The first question is; "Just what is LAMBDA"?
Well Lambda is the symbol used for the wave length. Thus if I
want the system to operate best at a wavelength of 10cm or
3,300RPM the length of the inlet pipe would be 10cm. But that is
too close to the stall speed of the T/C and I really want it to
operate better at the higher revs. The peak power I have decided
will be at 8,000RPM (or 5,000RPM above stall speed for the T/C).
This gives the value of Lambda to be 4cm.
Well I will admit I have been busy with other things
and now at the end of the Financial Year -I can return to
playing with ideas!!! The main problem with this loco has been
that fact that the motor will be moving at thousands of revs
while at idle and several more while at power. If this was a
"nitro racer" then this would not really be a problem -however
my loco has to start and stop very gently this means that it
will require gearing down. Not I hasten to add by the gearbox of
the original -but by a far more simpler 3 stage reduction.
The "theory" goes like this...
I feed the output from the Motor through my homemade
magnetic clutch system and by varying the field strength I can
have from 0 to 100% transmission. The output shaft then feeds
the first stage reduction gear, this then feeds the second stage
reduction gear located centrally between the two power bogies.
The output from this feeds a central shaft to two Universal
Joints located directly at the pivot points of the bogies. Then
from here it feeds a worm and spur gear to the OUTER axles of
the power bogies. The INNER axles are then fed power via the
external fly cranks -in the exact reverse of the original
manner!!!
See Picture 6.
The "maths" runs
something like this
The First Stage reduction is 1:10
The Second Stage reduction is 1:10
The Worm and Spur Gear is 1:30
Therefore the TOTAL reduction is 10 x 10 x 30 = 3,000 : 1
This may seem to be pretty excessive -but remember that the
stall point of the torque convertor will be 3,000 RPM.
Now, if we assume that we are running at 100% transmission and
the driving wheel diameter is 5cm then the revolutions of the
wheels will be 3,000 RPM.
The speed the loco will be travelling at will be 3,000 x Pi x
5cm per minute = 47,130 cm per minute. This rather alarming
figure drops down to 7.8 metres per second, still rather on
the high side, but this is a perfect case scenario!!! At best
I think my clutch will be 10% effective -this gives a far more
respectable 78cm per second. This equates to a "rough" 20MPH
at scale speeds.
Well -slowly but surely things and ideas are starting to come
together... Over the past few nights scribbled bits of paper
with dozens of rubber marks have come together to produce the
following "chassis mk1" drawing.
See Picture 7.
Everything can be welded together from 3mm plate stock(!)
I know that a few months have passed since I last made an entry
onto this thread -but I have been busy -with home and work. I
have decided to use the ENYA SS/50 Marine engine for my power
plant. This is "slightly" overkill for a Gauge '3' model -but it
is a very common IC engine and parts are easy to obtain.
Experiments with the magnetic clutch system were errrmmm...
"Interesting" as the amount of fluid and iron filings that were
sprayed when the sides of the containment vessel breached were
spectacular!!! I never even thought about the amount of force
that the spinning oil and iron filings would exert on the sides
of the clutch. I have come to a 3 plate clutch of the "wet
plate" variety operating in a fillable oil bath. This closely
resembles the "MekHydro" torque system, no I didn't crib the
idea! But the problem is how to pump the oil that provides the
frictive medium in and out of the clutch vessel.
I have found a source of wheels for this loco that are an exact
match -but where they are from is probably correct too... The
Driving Wheels are an exact fit for an LMS "Duchess" and the
Bogie Wheels are an exact match for a tender for a "Duchess". I
am not saying that BREL Derby had a few hanging around -but it
does fit the ethos and manner of the Litchurch Lane Works!!!
The final part of the assembly drawings is now complete. I know
how to make it reverse... This may not seem a major achievement
but the reversing mechanism is one that has bothered me for some
time. So, in my typical manner I decided to cheat. I have a
design for a reversing mechanism that goes forwards or backwards
whether it is going round -or not!!! I have decided to adapt the
standard method of dog bevels. These do not transmit very much
torque -but the system should only be going in reverse for short
periods... The drawing below shows the arrangement for it.
See
Picture 10.
Models are getting completed thus other models rise up the
construction ranks. This is what is happening with this model. I
have the majority of the parts figured out and some of the
prices. This is of course highly elastic as there have been
occasions when I have worked out parts and suppliers only to
find that, not only are the parts no longer made -but the
supplier is now defunct!!! As this is a diesel model and will
operate at a fairly high temperatures the bodywork is going to
have to be made from metal. This is a radical departure for me
as readers will know wood and plastic are more my medium. The
body is going to have to made from Aluminium sheet. This is
because it is light and very ductile. I have a small
“fabricator” which consists of a Guillotine, Folding Break and
Rolling Slips built into the one chassis. Whoever coined the
sales term “light weight and portable” obviously never had to
lift and shift this brute. I am not the puniest of people, but
humping 45Kg around my small work shop is very bad for my
back... Most of the time the fabricator lives on the floor on a
slide out shelf, (I wonder why???), but for some jobs it has to
sit on the top of the work bench.
Most of the body work is simple curves that can be rolled in the
slips and there are some compound curves where the corners meet.
But that is wooden dolly and sand bag work and can be formed by
hand (hitting it slowly) with a “Repousement” hammer and a
“Cloth and Hide” mallet. Joining the sheet metal sections is
going to be “fun”. The normal tricks of soldering , brazing ,
and welding are not going to work with Aluminium. What I am
going to have to learn is the magical art of “Aluminium
Welding”. I put the brackets around the last part as my best
friend refuses to call it welding -he is an instructor at “The
Welding Institute” and he prefers the term “Eutectic Solvent
Union”... Despite this I am going to use the words “Aluminium
Welding” as, (to me), this is easier!!!
I will use non fluxed rods -in short the “Techno Weld Process”
to create my shell that I will simply lift off and drop over the
loco.
September -the start of the new school year and the start of
this locomotive. Today is my son’s 11th birthday so the first
order for the parts has gone out to the steel stock holders.
Bogies:
3mm
bright mild steel flat
20mm x
50mm 2
rqd
20mm x
87.5mm 2 rqd
30mm x
115mm 4 rqd
Traction
Chassis:
3mm
bright mild steel flat
40mm x
300mm 1 rqd
50mm x
90mm 2
rqd
50mm x 30mm
2 rqd
50mm x
130mm 4 rqd
Power
Chassis:
5/8th" x 5/8th" x 16SWG wall
500mm
2 rqd
120mm
4 rqd
After the accountancy job with the “Krokodil” I am loathe to do
it again -because this is going to be the most expensive
locomotive I have built so far.... But on the other hand perhaps
it would make a good contrast to show what can be done on a
restricted budget of £10 per week.
Ok -so what has changed in the intervening months since the
design phase was “completed”?
Well, I have found that the manufacturers called “ASP and “Super
Custom” make a more suitable range of Marine engines for model
locos than “Enya”. These range in size from 0.12 cu in to 0.91
cu in. All the calculations are for the Enya SS/50 engine -but
the price comparison between the manufacturers is “interesting”
(!)
The 0.91 cu in motor is very tempting as a power supply as
it has the flattest torque curve and slowest idle of the group.
All the fluid calculations were also based on SAE 30 automotive
motor oil -which is for some reason is suddenly becoming very
hard to find(?) The reason I chose this was its characteristics
at all temperatures are well documented and there are no
“additives” that change the viscosity with temperature as in a
modern motor oil. Thus the viscosity at any temperature,
pressure and rotation can be read off the graph. The alternative
is to use good old fashioned Castor Oil as a medium for the
torque convertor -but this will undergo oxidation in quite a
short time interval... The other thing to consider is that the
oil will be returned from the torque convertor looking like, and
with the consistency of: dirty “whipped cream”. The pump will
force cooled oil into the convertor and a large bore pipe will
return the oil to the tank where it will separate and hopefully
it will disperse back into oil and air. One of the things I am
going to have to look out for is the danger of my model
“exploding” in a sea of oily froth.
If I go for the 0.91cu in motor, (as now seems likely), then the
cooling system will have to be adjusted to suit. The size of the
radiator is going to be doubled -to two rows and there will be
“tinselling” on the vanes to improve air to copper mixing. This
also means that an Ice and Water tank is now a Definite rather
than a Probable. Thus the direction of coolant air in the loco
has to be reversed. Instead of blowing air INTO the loco I now
have to blow air OUT of the loco. Initially I was going to blow
a draught of air along the length of the loco -but it seems that
there are going to have to be some auxiliary fans to do this
whilst the radiator fan sits in a sealed environment. The
electrical supply for them is going to have to be from a NiMH
bank of cells. I know that I normally use Sealed Lead Acid
batteries -but on this model I don’t think that is any room for
them and the environment could be a little “hot”....
There is actually a design “progression” in my models. From this
locomotive I learn about using a 2 stroke glow motor as a power
plant for a loco and the cooling problems inherent with it. The
next loco is a diesel electric in which I learn about the
problems of building an alternator and the power bogies for it.
The next loco is a gas turbine electric in which the only
problems I now have to crack are those of actually making a
working gas turbine since all that goes with it has been already
been built!!!
As they say around here “I may be crazy -but I am not daft”
I posted what I had done to a forum, (I know call it vanity if
you like -I prefer to call it explaining!), I got some useful
comments about sources for SAE 30 oil. The most common source of
it nowadays would seem to be garden machinery...
This is my reply.
“No, it has to be the
automotive stuff because the SAE Rating of your common or
garden(!) lawnmower stuff is not designed to operate at 0 deg
C. If you examine the drawing above you will see that the oil
feed to the torque convertor passes through the Ice and Water
tank. By fixing my oil temperature at 0 deg C I can then
accurately know what the viscosity of it is.
So, what I want is a thick
treacly oil that will get thinner and easier to flush out when
it is hot and not foam too much. I know that SAE 30 is usable
from -10 deg C to 100 deg C and the viscosity varies from 9.3
to 12.5 Centi-Stokes (mm sq per second) at 100 deg C.
Strangely enough if you
examine the sizes and viscosities of the design above -there
are lot of 0s and 1s on both sides of the equations.
I am an inherently lazy
mathematician... “
Silkolene in Belper is where I used to get most of the SAE 30
from -but they have stopped making it!!! The only real source is
Comma, who sell it for £14 per gallon -however there does
only appear to be two tins left in the entire UK???
"Gertrude" runs on MOTIL 1 -which is a great oil -but totally
useless for this as it retains the same viscosity whatever the
temperature...
It is not the lubrication aspect that is important here -it is
the viscosity at 0 deg C.
My home made gear pump is going to be made from Hostaform using
MOD 2 gears so that rules out any form of thixotropic wax, (i.e.
Automatic Transmission Fluid).
"Rob" has given me a quote for the start of the steel. This has
been duly paid for via "plastic" and I await delivery. I now
have to design my bodywork... The body is, in theory, two cubes
stuck onto the ends of a rectangle with an arch over it. But as
anyone with even a little knowledge of stereo geometry will tell
you "HA HA -no it it's not"...
So it looks like I will have to fight my wife for possession of
the left over cornflake boxes to make "pattens" from. She
normally uses them for Quilt Templates.
I have been researching my wheels over the weekend and come to
the conclusion that the "perfect wheels" are too damned
expensive. These are the tender and bogie wheels for an LMS
Duchess -however if I accept that they are 1 spoke too few then
the wheels from 4 foot and 3 foot wheel castings from Walsall
Model are a better bet. Having no experience of machining cast
iron I would rather "crack" a cheapish wheel of £4 than
one of £30. I think in this case it is an example of "all
the traffic will bear"...
The problem is that you CANNOT trust the original calculations
as provided in the original -because very few of us have gardens
capable of 6 chain and 8 chain (scale) curves. This means that
you MUST recalculate all the parameters for your model before
you begin construction.
If you take the dimensions as provided by BR you end up with
this...
See
Picture 11.
This CAN be made to work however quite a bit of cheating will be
required to get it around my std 2.2m radius curves. If you
examine the "nice" drawing above all the distances are regular
and the manner of its entry and exit to a curve are predictable
as it is all a segment of a polygon. What now has to be done is
to "ditch" the Bissell Bogie at either end and replace it with
an Adams Bogie on a control arm. The will enable the Adams Bogie
to rotate far more than a simple spring system would alone, and
also to traverse more easily. The result is that the Adams Bogie
does not steer the Loco as much. However if the Adams Bogie is
spring coupled not to the Power Chassis (red) but to the outer
part of the central bogie of the Traction Chassis this force
will act on it to steer the bogie (or more correctly pre-load
it) as it enters and leaves the corner. The Adams bogie is also
going to have to have independent turning wheels i.e. the wheels
will have to be able to turn on bearing fixed on the axles not
the axles themselves...
This will stop the flanges trying to "climb the rails" due to
the forces that would be transmitted through a connecting axle.
At the moment I am having "fun and games" designing the one
thing that I will admit that I am totally useless at, (other
than welding!), and that is gearbox design... I have something
on my drafting board that fulfils all the criteria -except
beauty. On the other hand this locomotive is infamous for its
gearbox -so I should take some heart in the matter. Whether the
locomotive will performs with "seamless acceleration" as the
original was supposed to do, (or as it was claimed), but just in
case I will keep close track of all screws locally!!!
I have a “number” of old DP sintered iron spur gears wheels
-these are basically metal powder pressed into a mould and
heated until it all fuses together. I think I have had them for
over 15 years now -so it is time for them to pay me a little
rent money... They seem to be the curious ratio of 38:16 (yes I
have counted them twice to be sure!) One of them will form the
output gear of the torque convertor and this will drive a
cascade of three of them giving an output of 1:13 (!) this then
feeds a 12 tooth 0.7 MOD gear to a 100 tooth 0.7 MOD gear that
drives the reversing dog bevels. The output shaft from this then
drives a second set of bevels to a worm and spur gear set up of
1:50.
So, at the end of this
1:13 from the torque convertor
12:100 from the reversing dog bevels
1:50 from the worm and spur gear
Which is the wild figure of 1:5,146.66 recurring...
Most of this is going to have to be transmitted through Iron and
Hostaform gears using 6mm shafts. These are going to have to be
mild steel or possibly brass. This is because the real
alternative is “Silver Steel”. Which would be marvellous -if I
could actually get a drill bit through it without it wandering
all over the place.
This locomotive relies on one thing -that the torque convertor
works!!! Having worked out; how the oil will interact with the
discs, the sizes of the holes in them, the expected heat
transfer due to friction and finally the force of the oil on the
sidewalls of the container(!)
The one problem that is still problematic is, how to put the
damned thing together...
The only way I think to do it is via the "Layer Cake" method. I
make some plywood discs and assemble first the impeller disc, a
sheet of plywood, a driven plate etc etc. The central shaft is
then silver soldered to the impeller disc and the next layer
added. Eventually after three iterations I have a stack of
discs, burnt plywood and plates. The position of the holes in
the discs does matter. The impeller discs are aligned so that
the four holes will overlap producing a spiral. The driven
plates are aligned so that their holes produce a spiral in the
reverse direction. This means that the impeller discs force the
oil right to left and the driven plates accept the thrust
generated from it.
The driven plates have to be larger than the impeller discs -so
that the torque from them is larger than that supplied by the
impeller discs. The oil viscosity is measured in mm sq per
second. This makes the maths simple, the torque increase is the
area of the driven plate divided by the area of the impeller
disc -provided we ignore the changes in the viscosity of the oil
due to the rise in temperature due to friction!
This gives me (see above) 8 faces of driven plates to 6 faces of
impeller discs, which works out to 8 x (2.54 cm sq)
divided by 6 x (1.27 cm sq) which equates to 16 / 6 = 2.66
recurring.
The interior of the torque convertor chamber is going to have to
be profiled to provide a profiled smooth “return” to the moving
fluid -this will be done on a rotary table and a ball ended
mill. The oil is going to have to enter the torque convertor at
the base of it and exit at the centre. This may seem strange but
remember that the fluid will be moving with the driven plates
and centrifugal actions will force it to the walls. Thus the
centre of the torque convertor IS the top!!!
The pump will have to empty and then fill the torque convertor
chamber for starting the IC engine and changing direction -thus
the pumps default mode of operation will be to EMPTY the torque
convertor chamber. Allied to this there will have to be some
form of “parking pawl” on the torque convertor output gear to
lock all the gearbox and transmission static once it has done
so, and to release once the torque convertor chamber starts to
fill...
How I am going to do this I am in two minds about. One option is
to use the level of oil in the chamber to block an IR LED -the
other is simply float something in the oil tank and close a
switch with it! The tank volume is going to be 500 cm3 (for ease
of maths!)
The pump controls the level of oil in the torque convertor and
since it is, (hopefully), at 273.16K the viscosity has to be
12.3 mm sq per second. The volume of oil pumped into the torque
convertor has to be Pi x R x R x H which comes out to 3.142 x
3.5 x 3.5 x 3 = 115.5 cm3. If we assume a fill time of 3 seconds
this means that the pump must be capable of roughly 40 cm3 per
second or 2.5 litres per minute. This is also the rate of flow
of oil through the Water and Ice intercooler... The question I
now have to ask is how fast do I have to turn my pump gears?
Since I am going to use MOD 2 gears they have pretty coarse
teeth. The smallest gear I think that I can get away with is a
12 tooth. This works out at a diameter of 2.8cm and a tooth
depth of 0.2cm. This means with two gears I have an effective
driven volume of (Pi R sq H) -(Pi r sq H) per revolution, (the
teeth occupy 50% of the volume of a gear wheel).
So: Volume = [(3.142 x 2.8 x 2.8 x
1.5) - (3.142 x 2.4 x 2.4 x 1.5)] per revolution
= 33 - 27
= 6 cm3 per revolution
Thus my pump must turn at at least 420 RPM to fill the chamber
in the allotted time. Now to work out how much torque the pump
motor has to deliver. This is (thankfully) simple mechanics.
The torque required is that to lift 6 cm3 of oil with an SG of
0.8 a distance of 2.8 cm with an arm of 1.4cm
Torque = (6 x 0.8 x 2.8 x 1.4 ) = 19 Gramme Centimetres
There are several motors of this rating available -but what we
need is one capable of generating this torque at the, (fairly
low), revs of 420 RPM. This will mean searching for a motor with
a wide but short armature -ideally one that is roughly the same
diameter as the gear wheel. This will mean driving the pump
motor from a PWM supply as this will deliver full torque at all
speeds -again the pulse frequency will have to be low to
encourage high torque in this situation. The MFA RE-360 is
probably overkill -but it will just keep going!!!
The heating effect of friction is the rate of rotation and the
area of the discs divided by the viscosity of the fluid. The
Specific Heat Capacity of the oil “should” be 2 Kilo Joules per
Kilo Gramme per deg Kelvin. So the mass of my 500 cm3 tank of
oil is 0.8 x 500 = 400 Grammes. This means that the amount of
heat that has to be used to raise it’s temperature is 800 Joules
per deg Kelvin.
Unlike Joule I had other things to do on my honeymoon than
measure the temperature at the top and bottom of water falls!!!
The manufacturers spec says that the IC engine develops 2.2 Kilo
Watts, (or more usefully here), 2,200 Joules per Second. So, if
my torque convertor is ZERO percent efficient then the oil
temperature would raise by [ ( 500 / 115.5) x (2,200 / 800) ] =
11.9 deg K per second!!!
So, what is the safe level of efficiency?
I am going to assume that my torque convertor is 75% efficient
-or rather that I have to dump 25% of the shaft power as waste
heat into the Water and Ice intercooler.
Ploughing through the maths gives an output temperature of
276.13K and the oil now has a viscosity of 11.73mm sq per
second. This gives me the amount of heat I have to "dump"
through the intercooler and the expected life of the ice in it.
If I construct a tank holding 1,000 grammes of Water with 1,000
grammes of Ice then the entire volume of oil (400 grammes @
276.13K) gets cycled roughly every 12 seconds.
This gives me a thermal input of 5 x (2000 / 400) x 2.97 Joules
= 75 Joules per second.
To melt 1,000 grammes of Ice to Water takes 334,000 Joules =
(334,000 / 75) seconds of running time or 4,453 seconds or 74
minutes (roughly).
The exact Specific Heat Capacity of the SAE 30 oil I will get is
as yet unknown. I think I will have to do the ‘O’ level Physics
experiment of boiling brass kitchen weights in water and
dropping them into measured volumes of oil, measuring the
temperature rise, and then calculating it from that.
Despite all this the one major source of heat is, the IC engine
that drives this loco...
Having decided on an Ice and Water cooled system -why do I
need a radiator for the IC engine? And the answer is that MOST
of the time -I won’t... However when the IC engine is really
“clogging it” then the temperature of the returned water will be
higher than ambient and it will need cooling. Thus the fan speed
with depend on the temperature of the returned water. This will
be a simple thermistor located in the return feed to the
radiator feeding it to a PWM driving the fan. I am a purest but
not even I am going to produce a “wax stat” for the radiator!
The pump motor will be the same as that of the torque convertor,
(I think it might be wise to make three in total)...
OK, so what are the cooling requirements for the IC engine?
Oddly enough -not that high... UNLIKE the torque convertor this
is an “open” system. The high temperature component comes from
the exhaust, which is dumped to atmosphere, so all that is
required is enough cooling to keep the cylinder head (and glow
plug) within parameters. The water feed intake will be at
273.16K and the outlet temperature at higher than 280K before
the fan triggers. The Specific Heat Capacity is slightly over
twice that for the oil -so it gets hot slower -but it also gets
colder slower too. I can dump most of the heat via the radiator
before it returns to melt the ice. The curious thing is that I
might be in the position of having to cut down flow through the
radiator to stop the ice cold melt water warming up from the
ambient air in high summer. One possible way out of this is
“sacrificial cooling” in which I squirt some of the ice water
onto the radiator, and the evaporation of it assists in the
cooling of the water. A similar thing was done on the Reed
Ramsey Turbine Electric. The limiting factor for this is of
course that you will run out of water to cool and pump around at
the same time...
The only way out of this is a method I don’t like.
I am going to have to link two independent PWM supplies to the
same thermistors in the water circuit from the outlet side of
the IC engine. This will have to control the PWM supply to BOTH
the water pump and the radiator fan. What I am afraid will
happen is “short term ripple”. The system begins to range
between too fast and too slow -it can get into a type of
positive feedback. I think that the trick is going to be to
attach the thermistor to a “damping block” of metal that will
heat and cool a lot slower than the water around it. I then take
the output of the thermistor and tune the two PWM supplies (pump
and fan) to match it.
This is my pump design. It uses two MOD 2 Hostaform gears
from Muffet Gears (part number S2.0 012H). They have an 8mm bore
-which is unfortunate! This means that I will have to fabricate
an adaptor to the 2.3mm shaft of the RE-360 motor. The easiest
way to do this is probably to just to connect the driven shaft
of the pump to the motor via gears. A push on MOD 1 gear to the
motor and a bored out to 8mm MOD 1 gear to the shaft. There is
plenty of “meat” on the boss of the MOD 2 gear to directly pin
the MOD 1 gear to it with four M2 nuts and bolts...
See Picture 12.
Below is my “standard” PWM circuit -but in this case(!) I have
replaced the 100K log potentiometer with a pair of 47K
thermistors. These are NTC (negative thermal coefficient) i.e.
the warmer they become the more that they conduct. What can
happen, (and I am ashamed to admit It has happened to me), is
that the current running through the thermistor actually heats
it up!!! The water passing over the thermistor should dump the
heat from it and (hopefully) simply allow the thermistor to
work.
See
Picture 13.
So, (in theory), this is what happens! The IC engine coolant
return water gets hot, this lowers the resistance of the two
thermistors which increases the pulse length in the PWM. The
motors pick up speed and then more water is pumped across the
cylinder head, cooling it down. This lowers the temperature of
the coolant return water and increases the resistance thus the
pump and fan slow down. The dangers of “short term ripple” will
mean that I will have to fiddle with the thermal mass (the metal
dampening block that the thermistors are bolted to) and fine
tune each PWM with a set of multi turn potentiometers.
Having nailed down the cooling of the cylinder head and the
cooling of the torque convertor the one great question remains
-the radiator. Fortunately I am something of a “petrol head” and
I am a “computer geek” as well. Both of these have now glorious
united in this design problem... My solution is one I have used
before and long before it became “trendy” I was using liquid
cooled computers. Modern methods are to cool the CPU directly
and dump the heat outside to a fan driven radiator via brightly
coloured coolant fluids. My solution is ugly -although it will
have green food colouring in the coolant to aid topping up(!)
The temperature of the output from a 91 series Marine IC engine
cylinder head has been measured, (Thank You “Lady in Blue”!),
the flow put through it however was not possible to measure as
it was fed from a scoop that the boat pushed through the
water. However I do have an input temperature (12C = 285.16K)
and an outlet temperature (26C = 299.16k).
The fuel used was a 15% Nitro Methane, 20% Castor Oil, 65%
Methanol blend set to “rich” with a “cold” glow plug. This is
going to be more thermodynamically powerful than the stuff I
intend to burn which is a simple 20% Castor Oil to 80% Methanol.
So, if I take the temperature differential as being 14 degrees I
can plug away at the maths...
Ok -lets assume that the pump normally operates at the same rate
as the torque convertor pump i.e. 6 cm3 per rev. The coolant
enters the cylinder head at 273.16K and leaves the cylinder head
at 287.16K this means that the radiator design has to dump 6 cm3
x 14 degrees = 84 Joules per revolution. But it gets worse...
The pump rotates at 420 RPM this means that the throughput is 6
x 14 x 420 = 35,280 Joules per minute or more usefully here 588
Watts - yes about the same as a small domestic fan heater...
The main problem is that, despite what it might feel like,
temperatures of 0C or 273.16K and below in the UK climate are
actually fairly infrequent(!) The normal ambient temperature for
easy running is probably between 10C or 283.16K and 20C or
303.16K. This means that the lowest temperature that I can cool
my return coolant to is from 303.16 to 283.16K. By simple
equation the amount of dry air that I need to pass through my
radiator to do this is 20 Litres per Second.
From an engineering perspective, a radiator varies from an ideal
black body by a factor, ε, called the emissivity, which is a
spectrum-dependent property of any material. Commonly, a fluid
thermal mass, containing the heat to be rejected, is pumped from
the heat source to the radiator, where it conducts to the
surface and radiates into the surrounding cooler medium. The
rate of heat flow depends on the fluid properties, flow rate,
conductance to the surface, and the surface area of the
radiator. Watts per square metre are the SI units used for
radiant emittance. If the system is not limited by the heat
capacity of the fluid, or the thermal conductivity to the
surface, then emittance, M is found by a fourth-power relation
to the absolute temperature at the surface. The Stefan-Boltzmann
constant is used to calculate it, as M = εσT4. Since heat may be
absorbed as well as emitted, a radiator's ability to reject heat
will depend on the difference in temperature between the surface
and the surrounding environment. For particular operating
temperatures, a system's overall heat flow may be given in
thermal watts, abbreviated Wt.
I have been working out my radiator... Classically there is only
one metal to use -and that is Copper. However Copper is
increasingly expensive and has no real alternatives, other than
Gold or Silver -for conductivity. There are two major designs
for a radiator, the one most modellers are familiar with is the
“square spiral”. This is simple to produce and works well. It is
a length of pipe soft soldered to a strip of metal and the whole
lot wound in a spiral like shape. The “automotive” multiple tube
type is the one that I am going for. This is because it will be
the most difficult to produce and hopefully once it is complete
other people will see how to produce one for themselves. The
material I am going to use to build it is BRASS... I can hear
the screams of “De Zincification!” from here. However since most
of the time the radiator is going to be empty the Zinc problem
will not be that bad, (hopefully). Added to this I will be
using distilled water and ice made from distilled water this
means that the ion exchange that leeches the Zinc out of the
brass should be slowed down. But if the plumbing does spring a
leak then I have only myself to blame for it!
The top and bottom tanks will be made from Brass “L” section
soldered together to produce a square section tube -the ends
will then be plugged in the normal manner with a piece of Brass
sheet (this also gives me the mounting holes...) The tubes will
be made of 3.2mm 24SWG brass tube and they will have their ends
cut at 45 degrees to ensure that if the do slip whilst I am
soldering them into position water will still get through them.
There will be 60 tubes each effectively 8cm long giving me a
surface area of 482.5 sq cm. This should work as a bottom feed
system, i.e. the radiator fills from the bottom upwards -thus
flushing out any air in it. The cooled water then returns to the
top of the tank and moves the ice around ensuring good thermal
mixing!
See
Picture 14.
Well having paralysed everyone with the raw maths perhaps I
should break it down into simpler chunks?
The bore of each tube is 0.3 cm making the cross sectional area
0.706 cm2.
The pumps supplies coolant at 2,500 cm3 per minute across the
radiator.
Each of the 60 tubes passes coolant at the rate of 0.7 cm3 per
second.
The fan passes 333 cm3 of dry air over each tube per second.
Thus each pipe has to pass 9.8 Joules of heat from the coolant
to the air in 1 second.
This gives a radiator rating of 9.8 Joules per 0.566 cm3 per
second or in SI 17,647 Watts per sq metre.
So, as further proof of my lazy mathematics -hands up those who
noticed why the radiator design has SIXTY tubes....
Having worked out some of the “mechanicals” I now have to work
out the bodywork. This is as I have said going to be made from
Aluminium sheet with the sections welded together. This will
involve some work with a sand bag and a “Repousement hammer” and
I foresee quite a few “words” as the metal refuses to go in the
right direction... The thickest metal that I can cut and roll
with my “lightweight fabricator” is 0.1cm thick and 30 cm wide.
I do have a 0.1cm “joggle” set for my joggle pliers. These are
strangely named things which simply produce a stepped edge when
the two jaws are brought together. The plan is that I cut strips
and then roll them in the slips part of it to the correct
curvature and “joggle” the edges to get the overlaps for
welding. This will leave me with a section seam that (hopefully)
I can fill and polish smooth. The formers that I am going to
have to make to “bash” the metal to shape over I will make from
plywood. That and handy lengths of steel bar and a lump hammer
hitting a sand bag over the two will produce the correct high
curvature that is required for some of the parts. I admit when I
first saw this technique I was horrified! But the sand
transfers the impact evenly and forces the metal to move across
the former.
The bonnets at the ends of the loco are going to cause some “fun
and games” as well. The ends are square these then rise to form
a curve at the windscreen... So, rather than forming section
through a frustum, (which would be so easy!!!), they instead
will have to be radially step folded and the folds bashed
smooth(ish). The rest of the bodywork is (thankfully) not so
wild -but it does have some tight curves at the corners that
will have to given the sand bag and lump hammer treatment.
See
Picture 15.
The next step is to “crack” the problem of the windscreens
themselves... These should be three simple panes of plastic
-however the shape of them in somewhat “unique”. The curve
of the roof is a a pure arc of 7cm radius and the top of the
bonnet forms a pure arc of 13.2cm. So far so good. The problem
begins as the windscreen is inclined at an angle that is not a
“normal one”... The angle is 1:2 or horribly enough 26 degrees
43 minutes from the vertical. I will not ask why they did not
make it 30 degrees -as this would have been too easy!!!
Flipping through the Sine,Tangent and Log tables section of
“Zeus”... The central windscreen presents a frontal dimension of
4cm x 2cm whilst the vertical dimension now works out to be
2.2cm and the base of the side windows work out to be 1.3cm.
If you consult the drawing above you will be able to see the
rough position of the panels. The longest thing that I can roll
is 30cm -thus the roof is rolled as one complete section
and then the flanges “joggled” to it. This is going to have to
be a monocoque structure and I am not quite sure if I simply
design it to lift off as a whole -or to have a roof access
hatch. The former is going to be the stronger whilst the latter
is probably going to be the easiest to operate with(?) This is
not a simple question. Starting a 0.91 series Marine motor is
not something that can be done with a flipped finger on a aero
propellor. Smaller engines have “pull starters” whilst the
larger ones have external battery electric starters operated via
captive rubber belt. One of the “radiators” on the nose is
probably going to have to be an access panel for the captive
rubber belt. This will not be a problem with the diesel electric
locomotive as the starter motor is the generator.
I could laugh! The motor dimensions are all metric -except the
screw thread coupling from the flywheel which is a 5/16ths UNF
thread... I of course, have no UNF taps.
According to the e-mail from “Rob” the nice man should deliver
my steel sometime tomorrow. I am starting to “itch” quite
badly!!! As promised the nice man did turn up on the doorstep at
just before 10:00am. and I duly unwrapped the very well wrapped
pieces of steel. Here you can see them on the side of the cooker
-yes next to the kitchen sink....
See
Picture 16.
A quick ID Parade is as follows.
On the far right there is the length of the chassis plate and
the longitudinal sq tube for the power chassis.
On the top left there are the B plates for the power chassis and
their top spacer plate.
Below them are the 4 transverse sq section tube that for the
rungs of the power chassis.
On their left are the bogie spacer plates.
And finally on the bottom are the side plates, arms for the
front and rear bogies and the bottom spacer plates for the B
plates.
Having looked at the drawings above you might think that I have
cheated (which is normal for me!) and I have two central
pivoting bogies. But technically this is not true... What I have
used in this loco can be classed as a Beugnoit lever system. The
Beugnoit lever (Beugnoit-Hebel) is a mechanical device used on a
number of locomotives to improve curve running. It was named
after its inventor Edouard Beugnoit. Around 1860, when Beugnoit
was the chief engineer at the firm of Köchlin, he developed
a system whereby wheel sets are housed in pairs in the
locomotive frame, with side-play, and connected by a lever.
These levers are fixed to the frame in the centre and thus
enable the sideways movement of the connected axles in opposite
directions. In this way, instead of being fixed in the frame,
the axles are able to move sideways rather like a bogie, but
clearly nowhere near as much. On locomotives with a side rod
drive, the axle side-play is balanced using longer coupling pins
(Kuppelzapfen) on which the coupling rods are also able to move
sideways.
On running round a bend, the first axle is pushed sideways by
the curve of the rails and so moves the second axle parallel to
it in the opposite direction, until the wheel flanges of both
axles align with the rails. This distributes the guide forces
between the two axles which reduces wear and tear on the wheel
flanges. By enabling this transverse movement of the wheel sets,
locomotives with rigid frames do not have to use the thinner
wheel flanges etc. normally needed to facilitate smooth running
through points, bends and tightly curved sections of track. The
'guide length' of the locomotive is formed by the distance
between the two fixed Beugnoit lever pivot points.
See
Picture 17.
In Germany, Beugnoit levers were used mainly in the middle
of the 20th century. The best known examples of locomotives that
use this type of lever are the MaK side-rod drive locomotives,
the Class 105 and 106 engines in East Germany (DDR) as well as
steam locomotives like the DB Class 82.
Since this locomotive is “steam era” design technology then it
does rather make sense to use this. The “Adams style” bogies are
another “steam era” device that I am using. These will
“pre-load” the chassis when cornering, whilst putting them on a
floating arm will enable me to control them in a more
predictable manner.
I am now in the first stages of turning my “primitives” into the
shapes that they will be on the finished loco. The first things
to do are to produce the Bogie side plates. This is a simple
process of chain drilling and (ugh!) lots of grinding and filing
to produce the half dog bone shape that they require. It may not
be prototypically correct -but I have opted for a simple 1cm
radius from the centre of the axle for my dog bone curve. The
Bogie wheels are going to present me with a problem... The last
set I made were mounted as a floating wheel between two
sets of ball races. These will have the no journals so I am
going to have to do some ”brainstorming” with some friends as to
how I will connect them to the axles. The favourite method at
the moment is to drill out the centre of the bogie wheels and
insert the bearings on an allen headed bolt into the axle -which
I have bored and tapped to that thread. Once the thread lock has
set solid (a short pause for the oven...) then I can grind off
the end of the allen bolt to a flush end. This is all very well
and good -but until I actually get the wheel castings in my
fingers this is all theory. However on the plus side of the
argument it does seem that it can be made to work! If it does
work than this will give me a free rolling bogie with no
cornering problems due to the wheels rotating at differing
speeds. Ideally the coning on the wheel should give me the
differing effective rates of rotation that I need. However I
really am that cautious!!!
I have often said the British do business in the pub. Yesterday
I went to a beer festival and supped some nice stuff (and one
VILE concoction -not unlike trying to drink battery acid!!!)
There amongst the books at the beer festival (held at The Great
Central Railway in Loughborough) was a book to die for. It has
the dire name “British Railways Design,Theory and Operation of
Diesel Locomotive Traction Volume 1”. They wanted £8 for
it. I did explain that I had come to the beer festival to drink
beer and not really to buy books. They said I could have it for
£5. I offered them £3 and they willingly took it. I
placed the money in the tin, walked out of the “shop” (a
converted horse box), turned the corner, and then ran like a
thief!!!
It is PACKED with drawings, formulae, tables and installation
instructions plus (oh thank you!) worked examples. I now can
very accurately reproduce BR transmission fluid for a “Western”
and DERV of a Cetane rating to run a “Deltic” with... If I am
right(!) the hydrostatic formulae in the book point to me
needing a smaller engine than the .91 series that I was thinking
of. This means that it will be cheaper to buy, create less heat,
(and thus cooling), whilst producing a higher torque from the
convertor. Given the tables in the book I can now successfully
“dope” my SAE 30 oil with additional additives that will give it
a higher thixotropic factor than it would have simply as a cold
liquid. All the substances mentioned are commonly available and
I have decided to use Aniline, Canubra wax and Flax Oil in a 1%
5% and 10% mix with the SAE 30 oil. This will give a smoother
transition from a “liquid” to the “solid” as the rotational
speed difference between the driven plates and the impeller
discs alters.
The common sources for Canubra wax are shoe polishes -but the
raw stuff is available from artists and cosmetic suppliers. Flax
Oil I got from my local Health Food Shoppe. The assistant then
explained how it would be good for my skin and muscles and how
to consume it with “normal foods”. I then explained that I
wanted it for putting into the torque convertor of my model to
alter the thixotropic co-efficient (Rho) whilst still
maintaining the di-oilic paraphase constant (Phi) at the point
of solidus (Kappa) in a Standard Temperature and Pressure
environment. At which point she asked me to enter my PIN into
the EPOS terminal...
Anyway -this is how it works(!)
The motor is idling at 2,000 RPM there is no fluid in the
chamber. I pump fluid in and the drag begins to turn the driven
plates. The fluid now fills the chamber and the drag between the
impeller discs and the driven plates causes the fluid to slow
down between the plates -this makes it behave more like a solid.
The impeller discs and the driven plates now move at the same
speed. The fluid now acts as a solid between them despite the
fact that the whole thing is going around at several thousand
RPM...
Only the fluid at the edges of the chamber behaves like a fluid,
whilst it is between the discs and plates, it behaves like a
solid.
I have begun to accurately machine the primitives to size. The
first things to go into the jaws of the mill were the two
longitudinal square section tubes. These need to be cut to
exactly square ends -otherwise nothing will line up.... The
problem I have at the moment is that there is “Dobbin” the
rotorvator stashed in the workshop. It seems to take up 50% of
the available area whilst only occupying 5% of the floor! Dobbin
is providing muscle to produce the lawn that the railway will
run around -so it has to be handy for when the weather permits
him to function. The soil has to be just right for the tines to
cut through the clay -otherwise he just stalls. Too wet and it
just churns it to slurry... Anyway as you will gather some fun
and games were had getting the mill , its vice, and the sticking
out lengths of steel in a position where the operator could use
it with out recourse to advanced Yoga positions. I loaded the
EM32 collet with a 10mm HSS cutter and began to nibble at it.
After a few passes I had a nice square edge and some tape
holding everything together to flip and do the other end. The
transverse pieces are going to have to wait until I have been at
them with the hacksaw, as I need to produce a flap edge for the
ends. When I start welding them into position I can safely weld
an end plus both of the laterals. The last end will have to have
a bolt arrangement as this will carry the motor and the torque
convertor on it. This will make it easy to remove as a unit for
working on.
ALWAYS design your models to be easy to work on!!!
After some delay caused by domestic priorities, (my son has to
choose his Yr 7 school and I have been ferrying the assembled to
the various “Open Evenings” at schools that are available). I
can now return to building a loco. This Wednesday I plan to
order my wheels from Walsall Model Industries. They are not the
“correct” wheels for the loco -but they are a fraction of the
price of the “correct” wheels.
The ones I am going to use are GWR ones....
The bogie wheels are going to be the 3 feet 3 inch bogie wheel
part number: T1818 and can be machined between 39mm and 44mm. As
the size I need is 40.5mm this is well within par. The main
driving wheels will be the T1841 casting the size of this is
between 49mm and 55mm and is thus a little “shy” of the required
size at 57mm, but to be honest it would cause the model to sit
1mm lower, have 2 spokes less per wheel, and at a saving of
£60 -I can live with it!!!
I have never worked with cast iron so this is going to be a
“first” for me. All the books say to take a heavy first pass to
get rid of entombed grains of casting sand and the toughened
skin that forms as the shock cooling of the molten iron occurs.
It is also in the books that this is going to be noisy, so I
have alerted the Domestic Goddess to go out to the Summer House
and quilt etc -whilst I am busy at the lathe.
Well the wheels have been ordered and then steel is starting to
be machined to size. The Mill has however done its usual trick
of blowing the fuse holder -not the fuse.... (I will buy another
one from Maplins tomorrow!)As I have written elsewhere both the
Mill and the Lathe are very nice pieces of equipment -but the
quality of just a few of the components could do with some
examining. To replace the fuse holder is not a difficult job -it
is however very fiddly and to be honest it not something I feel
like doing at the moment. SO, the pieces of steel that have been
cut to size are the bogie side plates and just one side of the
lower cross member for the central power chassis. The vertical
worm shaft will be held in position by the upper and lower cross
members. At the top of the vertical shaft will sit the reversing
dog bevel system. This part of the drive system is at the “I
know how it works -but how do I get it to work?” stage of
operations. In theory all I have to do is slide a shaft to
connect with a forward facing bevel and then slide it back to
connect with the reverse bevel. Thus the shaft still rotates in
the same direction but the take off is in one of two directions.
The shaft will have to “lock” in a Forward Neutral and Reverse
manner. This is going to have to work from a sort of “notched”
pyramid shape. The central notch will have to hold the slide
shaft centrally clear of both bevels and then lock it solidly in
either fwd or rev. All this now has to be interlinked to the
pump from the torque convertor so that the pump fills the
convertor in the Fwd/Rev whilst it empties in the chamber in the
Neutral. Time for some serious brainstorming and an ample
supply of tea!!!
Believe it or not less than 24 hours after having placed my
order (via my wife) over the telephone my castings arrived on
the front door step held by a very pained looking Postie... The
castings are of very high quality and errrm -they weigh a
tonne!!! Well perhaps not that but the larger ones come in at
450 Grammes and the smaller ones at just under 350 Grammes.
Since this is a 2-D-2 loco the poor Postie must have had a
hernia carrying the crate from his van up the garden path.
See
Picture 18.
See
Picture 19.
I have the new fuse holder from Maplins -but it is going to have
to wait until Thursday Afternoon at the earliest because of
domestic commitments. I will have to move “Dobbin” out of
the shed and then cart the Mill to the kitchen -where there is
plenty of light. Fortunately this is not too far but at 35
Kilogrammes weight it will probably feel as far as the distance
from the front gate to my door for the Postie.... Then it should
be a simple case of unscrewing the control panel and then
fighting through the tangles of cabling until I can get at the
back of the fuse holder to unscrew it and solder a new one into
position.
Having played “Mr Fixit” with the Mill I can now get back to
work....
Well over the past couple of days I have been busy, some of it
to do with this loco -but quite a lot of it to do with my
fettling my sons bicycle.... The next shot shows the Mill with
its vice clamping the work piece safely. (D.Ennis Esq please
note). The ends of the plates are now “square to themselves” and
as far as I can tell exactly square as well!!! The stack
of squared off primitives now await the scriber and rule for
their holes.
See
Picture 20.
While this has been going on I have been having a “practice”
with one of the wheels. I have never “done” cast iron before and
I thought it an opportune time to have a go with it. Having
spoken to “them as does” to quote a Southern Derbyshire
friend -I changed the jaws of the lathe over to external and
then fitted the wheel casting in -nose first(?)
See Picture 21.
The cutter then passes over the rear of the casting to produce a
flat face. This is a vile operation as all the black iron dust
clings to any slightly magnetic object and soon the lathe
resembles a magnetic iron filings demonstration in 3D.... A few
flicks of the trusty paintbrush and things are again visible.
The now flat face is needed to clamp it against the face plate,
(to be used later). The next operation is to centre bore the
casting. The next shot shows the Slocombe fitted to the chuck
about to start boring. I have always found the double ended
Slocombe drill to be a very strange thing to look at!!!
See
Picture 22.
Monday Evening. All of the Driver wheels have been faced and
drilled to M6, there are 3 Bogie wheels still to do and to be
honest my “eyes” have gone!!! However I am not displeased with
the evenings work and the stink of hot iron chippings has almost
certainly killed all the bacteria that may lurk in my shed... I
have found cast iron to be remarkably easy to work on -bar the
smell and chippings that seem to somehow get everywhere. Do I
prefer cast iron wheels to the ones I normally make from steel?
Well the answer really has to be no.
For ease of machining and gripping in the jaws of the lathe
steel wins hands down every time. I worry that pieces of the
castings are going to fly off whilst I am machining them. One
actually did on Sunday Afternoon and although it was not very
large, it left quite an impression on the safety shield...
Wednesday evening. All of the wheels have been faced and the
flat faces bolted together. Some “fun and games” were had as I
machined off the casting “buttons”. I had originally intended
simply to pass a cutter around the end of the casting and thus
cut through the button leaving a ring of cast iron to be
disposed of. This worked -sort of... It did in the end prove to
be simpler and quicker to knock off the spare cast iron with
a cold chisel and then turn the complete wheel as a pair
of ends. Something to remember for the future.
The next problem that faces me is the axles for these wheels. I
think that I have some silver steel rod of the correct diameter
but I have always found silver steel to be very difficult to
machine and in this case I think I will have to order some mild
steel rod to make them from. The bogie wheels will have to have
an allen cap head bolt fitted to the non rotating axle shaft as
this is what the floating bearings will be mounted onto (see the
NER EE-1 for more explanation). The drive axles will be 8mm
turned down to 6mm for the wheels and the bearing for these will
be in the horn blocks. Fixing the drive wheels to the drive axle
is going to be “fun” as well. My best option is the drill and
tap an M3 hole in the wheel boss and then use a grub screw to
secure it to the axle.
Thursday Evening. I have begun to prep up the wheels for their
finest moment. The first step is the face the fronts of the
wheels. The next shot shows a driver wheel in the jaws of the
lathe.
See
Picture 24.
The next shot shows a bogie wheel still with its button casting.
The axle hole will be enlarged by 0.5mm and then an end
mill will be used to cut the “blind hole” for the ball race to
fit into.
See
Picture 25.
This shot shows the tread for the bogie wheel being cut on the
lathe. This is actually very easy to do as the button allows me
plenty of room to get my cutters into position.
See
Picture 26.
Unfortunately the same thing cannot be said for the driver
wheels... These have had to be mounted on a coach bolt and then
“nutted up” in the jaws of the lathe. This does work reasonably
well -despite the chattering of the cutter. But the main problem
is that the jaws have to be tightened up so much that it
destroys the thread on the bolt....
See
Picture 27.
Well a few days later and all the wheels have been “roughed out”
on the lathe. They are now the correct size for loco the bogie
wheels have come out at 48mm and the driver wheels have (just!)
come out at 60mm. I will admit to cheating like mad and using
part of the button for the flange for the drivers -however I
have (mostly) got away with it. There will be a couple of
“dings” in the flange edge -but this would have happened anyway
with wear.... (he says....) So, at the moment all the wheels are
set for machining the flange angle and the coning on them. This
means using the one piece of equipment I have for my lathe that
I really HATE!!!
This is the Compound Slide.
It is a very simple piece of equipment but for some reason it
seems to delight in; fouling the saddle, bashing against the
tail stock, or being in such a position that it is impossible to
fit the cutter to the work piece, or, (its favourite trick) of
being just right to foul the chuck/faceplate at almost the end
of the cut...
Here you can see the slide set up for cutting the back angle of
the flanges for the Driver wheels.
See
Picture 28.
I have been going through the suppliers catalogues for suitable
bearings for my Bogie wheels and there is only one real
choice... This is a bearing of the tiny type! It is 8mm diameter
with a bore of 4mm and a width of 3mm -as is my norm I intend to
get it from Technobots. They stock it as part number 4255-122 at
a not unreasonable cost of 65p inc VAT! The load is a little
light for my liking at only 395 Newtons -but spread over four
bearings per axle it should take the strain. I know the STATIC
load would be over 40 Kilogrammes but the DYNAMIC load is far
higher. This will mean boring “blind holes” in the Bogie wheels
to take the bearings and some “Loctite Green Sleeve Retaining
Compound” to hold them in place. I have no idea what that stuff
is made from -but it has held together things that I would have
never hoped to stay together for durations that I would have
called impossible!!!
The bearings for the Driver wheels (part number 4255-182) that
will have to fit into the horn blocks are of a similar loading
at 1,252 Newtons. Alas they are 16mm diameter which means that I
am going to have to unship the 3 jaw and hernia on the 4 jaw and
take out the 12mm hole to 16mm with a boring bar. I like the 4
jaw but its icy cold 2.5 Kilogrammes is slightly awkward to “nut
up” whilst held up by two fingers and a thumb...
Tuesday evening. The work on the wheels progresses and now I
have (finally) 8 Driver wheels ready for final mangling... The
off set inclined hole for the grub screw has yet to be done and
I will admit that I am not looking forward to doing it. I have
used this time to experiment and apply my own personal
theories.... Well it is my model so I am allowed to. I accept
that the standard coning is 3 degrees and all but four of my
wheels are cut to this. The outer drivers are cut to 5 degrees
as this (I have worked out) will give me “just that bit more”
cornering ability -something that I need on my tight curves
The Bogie wheels have come along in leaps and bounds and all
that they require doing now is the coning and the flanging
angles cutting on them. I have bored the “blind” holes for the
bearing on each side of the Bogie wheel. This was a truly nasty
operation as the Slocombe insisted on clogging with iron dust
and and having to be withdrawn from the work piece and blown
clean -thus covering everything (and me) with fine iron dust...
Once I had got a lip of about 1mm I could then change over to
the end mill and bore down the hole to provide a flat bottom to
it. This was very quick easy and (I am told) unbearably noisy!!!
I bored in 5mm deep on each side of the Bogie wheel and I
decided to take out the centre hole to 6mm while I was there.
This will provide a good oil reservoir when I assemble the
bearing.
The next shot shows the Bogie wheel being bored with the end
mill. If you look at the two bogie wheels on the saddle you can
see the bore holes on either face.
See
Picture 29.
The same procedure is (I think) going to have to be done when I
start producing all the carriages and wagons that I will need.
Despite, “certain persons”, claiming that my locos could pull a
house around my garden, I think that the use of fully
floating bearing into my rolling stock will produce a more
nicely behaved item. This will mean that instead of relying on
the coning to provide the difference in radii (and hence linear
speed) The wheels are free to move at differing rotations. This
will place more load on the flanges as the steering effect will
mostly be down to them -rather than the continuous “sliding down
the coning” that provides the side thrust to the axle.
Now that the wheels are almost finished and the shed is getting
rapidly colder.... I have been sat near the cooker in the
kitchen working out things from my old text books. I have an old
book from the company PIPER FM ltd that has all the formulae in
it, (called not unnaturally the PIPER TUNING MANUAL). I will
confess that I am not sure if the company still exists(?)
however it does explain the reflection supercharging of a
2-stroke exhaust system -probably better than I am about to!
When the exhaust port opens a shock wave travels down the
exhaust pipe. As the shockwave expands it slows down and draws
the exhaust gasses with it and pulls clean air and fuel with it
-hence the first cone section. Then it travels at a speed along
the straight section (the length of which is time dependant)
when it is compressed by the second cone section and exits the
exhaust system. As it exists the exhaust system a secondary
shockwave travels towards the exhaust port of the cylinder and
pushes BACK any clean air and fuel as the exhaust port closes.
This then adds to any shock wave charging from the inlet side of
the cylinder...
I hope the diagram below helps to explain!!!
See
Picture 30.
The length of the timing cylinder is dependant on 2 things; the
speed of sound and the contents of the exhaust gasses. The speed
of sound through the gasses increases as the temperature
increases -but (Ha Ha!) not in a linear fashion... The content
of the exhaust gasses depends on what is being burnt...
Fortunately the book contains three sets of tables from which I
can “perm” my exhaust gasses. One table contains Nitro Methane
mixes, another Petrol mixes and finally (thankfully) Methanol. I
know that what I am going to burn is NOT “Straight Methanol” but
this is the only data table that I have -so I will have to use
it (and abuse it!)
Based on my design for the convertor I have to pump in figures
that will force the system to supercharge at my desired rotation
-8,000 RPM I end up with an impossibly big exhaust system with
an infinitely thin timing cylinder 48mm wide and the system is
478mm long!!! For obvious reasons something is going to have to
“mangled” -seriously. The best bet is to fit the exhaust system
with some form of cooling fins to lower the temperature of the
gasses and then to force cool the fins. The draught from the
radiator fan should do nicely and the fins will provide
structural strength, rigidity and something to bolt it all into
position with...
If I can cool my gasses using the coolant water then things
become manageable -but the heat still has to be dumped via the
radiator. The best bet would seem to use the coolant water from
the cylinder head to cool the gasses as they enter the expansion
chamber. This would mean using a co-axial tube for the “Header”
section from the exhaust port to the start of the expansion
chamber. This cools and (more importantly) lowers the speed of
sound through the gasses. Curiously enough this does not affect
the length of the system -but it does cut down on the width of
the pipework and I can now use more “delicate” 15mm pipes.
The Resonator chamber is going to have to be build in sections
-or built as a core with end flanges -I haven’t decided yet.
Each “chamber” of the Resonator has to be 1.4 times the volume
of the previous one -this has to do with the adiabatic modulus
of the air and the pipes are very “flute like” in their
construction. There have to be holes drilled in the walls
between each part of the chamber that are 35% of the surface
area and the whole lot internally “lagged” with some form of
dampening medium (probably glass fibre!)
A couple of evening at the calculator and I will admit I had to
unearth a teenage “A” level Physics text book that (errrmmm....)
never seemed to have got returned to my old grammar school at
the end of my upper 6th form(!) At the end of the maths and the
brain creaking with node and anti-node plus reflection and
constructive and destructive interference I have emerged from
the trauma with the following rather large full scale drawing.
See
Picture 31.
Now, this is an ideal drawing- and not a real world practical
one. I cannot see how I could possibly fit that lot into the
loco shell along with the tanks and pumps etc that are already
having to be shoe horned into it!!! The first part (going left
to right) is the expansion chamber. This is a “Brooklands” type
which gets its name from the race track where they were first
used. They are very simple to work out, having a 45 degree
expansion cone and a 45 degree compression cone. Being a lazy
mathematician -guess which angle has a tangent equal to 1.00?
Being of a “flat” design rather than conical -this removes Pi as
well.
The specifications for a Brooklands Silencer are pretty easy…
The exhaust gasses are lead to a chamber that is:
"Of not less than SIX times
the swept volume of one cylinder. The diameter D (if
cylindrical) being not less than ONE FOURTH of the length L ,
the tail pipe shall have an internal diameter not more than
ONE HALF of the equivalent diameter of the silencer"
(Don’t you just LOVE the 1920’s English!!!)
The inlet pipe has perforations along its complete length as
well as being open at the end. The shock wave then hits the 45
degree end and gets bounced back on itself. The exhaust pipe
from the chamber is similarly a length of perforated pipe open
at the end. This type of expansion chamber has one serious flaw
-it produces the “bark” like sound that is so characteristic of
old time racing cars. The resonator chamber is similarly a
classical design from a racing pedigree -this time Bugatti...
The inlet pipe has perforations on the top facing the curved
part whilst the exhaust pipe has perforations on the bottom
facing the curved section. Both pipes are open at the ends.
Whilst it is well known that Bugatti came from a background of
furniture making -he also liked music...
The five chambers damp out the frequencies generated by the
”Brooklands” expansion chamber. The effective length of each
being 1.4 times that of the last. Between each each pipe there
are loosely packed glass fibre strands to break up the shock
wave and render this to a pressure flow across the chamber.
So far so good. I now need to do some real world investigations
as to the exhaust temperature etc of my engine and the water
temperature that will result from water cooling the exhaust
manifold. Hopefully at the end of the experiments I will have
some hard data to plug into my dimensions and end up with a
somewhat smaller exhaust system!!! The design is based on a gas
temperature of 450 Kelvin and even if I can only knock down the
exhaust temperature by 20 Kelvin -I can shrink it by 25% due the
lowering of the speed of sound.
Some mangling with the Universal Gas equation, ((P1 x V1) / T1)
= ((P2 x V2) / T2) , produces the volume of exhaust gasses that
the motor produces based on its RPM and the temperature at the
start of the exhaust pulse. This is as yet “simply theory” -as I
have yet to push a pyrometer down an exhaust tube. Similarly it
produces the pressure at the point of exhaust thus the peak
shockwave pressure into the exhaust system.
Due to the Mill not behaving itself, the work has slowed down to
a crawl... I have had to unearth my pillar drill from it’s
hibernation. Admittedly this is something that I would have had
to have done anyway as the holes that are now required exceed
that of the chuck of the Mill. The next shot shows the four side
plates of the traction bogies aligned in the vice with a 3mm
drill to punch the initial pilot holes through the plates. The
stack is then taken to pieces and (hopefully) the pilot holes
will align the 8mm drill that is the next through them.
See
Picture 32.
After having produced the first set of plates and fettled the
holes to the a nice shape (de burr etc!) I had a trial run with
a couple of 8mm nuts and bolts then roughly gapped the wheels to
the correct back to back setting. I had a big grin on my face as
they rolled smoothly along the piece of track plonked on my
stool and then clonked onto the lino floor of the shed as it
fell off the end.
See Picture 33.
They moved so smoothly the question I now have to ask myself is
:”Do they need ball races?” The answer I am afraid has to be
“yes”. Even though they move well, they are moving without the
frictional weight of the loco on them, this may result in a
great deal of problems that will not occur if I use ball race
bearings... This does mean that my nice 8mm nuts and bolts are
not redundant until I get the 12mm drill out and bore them to
the correct (outer) diameter to take the flanged ball races.
Another thing that may alarm people is the the use of fully
floating compensated bogies -I intend to use these for my
traction bogies as well as my leading and trailing bogies. These
will be rather more “robust” than the more flimsy type commonly
associated with carriage bogies -but the principle will be the
same. I am not the first to use this in a mainline locomotive,
several Russian steam locos of the Stalinist pre WW2 era also
used this type of compensated steam bogies. But, this was
probably more to do with the appalling track conditions due to
winter and summer movements that would have snapped springs
etc. There will be torsional springs across the traction
bogie to force the wheels to follow the bumps, (hopefully not
that many!), that it will encounter.
There will be an 8mm threaded rod from one side of the bogie
sitting in a 8mm bore steel tube from the other side of the
bogie. The threaded end will be sunk into the tapped hole in the
plate and silver soldered there. The tube will be welded to the
other plate. I will bore an oil hole in the top of the tube to
admit some oil to the threaded bar and “keep it sweet” as we say
around here!!!
Due to the appalling weather I have been forced to spend a
morning (or two) in the shed (shame!!!) Armed with my sticks of
“B6” and “Easy Flo2” I started assembling the bogies (of both
types). Having unearthed my “hearth” from the shelf I then put
the 4mm bolts through the holes and nutted up the other sides
-being careful to apply plenty of pencil lead to the thread
before I did so... When silver soldering steel I tend to use
MAPP Gas -which I like -despite its evil and disgusting smell.
The ends of the bolts glowed a nice red and the B6 clung
lovingly to the steel and then flowed around the bolt head.
Before it could get much further I threw the flame to the next
one in line and continued along the line until they were all
done. The next step was to change over to Propane and then affix
the bolster between the two plates.
See
Picture 34.
The reason I switched gasses and solders has to do with the
differing temperatures of melting of the two solders. The cooler
Propane cannot reach the temperature needed to melt B6 -thus the
bolts remain firmly fixed and will not fall out... The bolster
was held in place by the simple method of clamping it in place
with a very long M4 nut and bolt (which also held the side
plates vertical-ish). I will be the first to admit that I do not
like Easy-Flo2, I find it far too liquid -but on the other hand
this extreme liquidity is just right for getting into the gaps
between the plates. You can see how it has just SPREAD
everywhere along the joint -but it has penetrated right through
the gap to the other side. So, typically enough for me -the
joint is ugly but strong!!!
The next shot shows the rough assembled bogie.
See
Picture 35.
Here you can see the bearings inserted into the left hand side
wheel -the stub axle nuts up nicely(!)
See
Picture 36.
The next step of operations is to start assembling the traction
bogies. The plates were bolted together as seen below. There
were I will admit a few panic stricken moments as I had to
really check that I had all the pieces in the right order -not
only are there a right and left there is also a mirror
positioning because of the fully floating aspect. (Check once
twice -maybe again???) In the sure and certain knowledge that if
I got this wrong, I screwed it up royally -I hit it with the
MAPP Gas and B6.
See
Picture 37.
The product of my labours -a slightly rough collection of wheels
and plates.. However I am not displeased with it as you can now
see the basics of the locomotive. This is a first for me as my
normal method is to build the bodywork first and then cram
everything into it. It feels very strange to look at the naked
bogies!!!
The next problem on the horizon is to weld the frame that will
form the base for the bodywork. I have most of the square
section tube cut to size, (bar some filing!) -they are going to
have to be “self jigged” welded together. This is a strange,
almost “ritual” like procedure, in which opposing ends and sides
are welded and allowed to cool before the next stage commences.
I am going to do a “belt and braces” setup and join the corners
together with some “pop” rivets that will take some of the
stress off me as I just know that I will disturb the layout when
I fumble around in my welding helmet!!! Once I have a tack weld
in each of the joints than I know that I am “safe” -but it is
the bit before that....
Well as they say “these things are sent to try us”..... After
having drilled all the 5mm holes in the steel tube and located
my pack of 5mm pop rivets I duly cranked the gun into the hole
crunched up the rivet and “THWUNCCHHH”. “This”, I said to
myself, “is not a good sound”... Sure enough it wasn’t. The pin
had bent inside the rivet gun and jammed it solid. The next two
hours were spent persuading the rivet gun to come apart. However
by dint of brute strength, dire threats and main force I got the
thing apart and lo and behold -the spring had broken. Thus the
tension on the jaws was never removed when the barrel moved
forwards. I duly replaced the spring with a couple of new ones
from my “box of bits” (where would the world of modelling be
without a “box of bits”???) and then in under 5 minutes I had
the thing back together...
The frame is now ready for welding.The aluminium pop rivets
will, I know, melt when the steel welds -but by that time they
will have done their job. The next shot shows the rough
assembled frame. There are only two really critical measurements
that have to be held “true”. They are the centre lines for the
frame -the rest of it can be “about there” as most of it will be
hacked to fit the components and brackets that will hold tanks
and radiators etc.
See
Picture 39.
Another thing that I am starting to look at again -is the final
connection from the gearbox to the drive axle...
I had intended this to be a vertical worm wheel to a spur gear
on the axle -but this is starting to look increasingly
vulnerable to; dirt, grit and the odd piece of herbaceous border
plant with a name always seemingly beginning with an “H”... My
new option is to drive the axle from a toothed belt from above.
This does have the added benefit that it keeps all the
vulnerable parts of the drive chain above the dirt line and the
toothed belt is fairly coarse so muck & “H” plants etc will
not really hamper its operation. This will entail producing a
floating platform that will move in the same axis as the bogie,
(this is not much of a problem!), and a flexible joint that
connects the drive from the reversing gear system to the
platform which will also be fully floating along with the bogie,
(this may require some thinking...)
Well after a couple of nights sleep, (I have also not been too
well -cough cough!), I think that I have most of the new drive
system worked out. Everything remains as it was in the original
setup -just that the worm dive is now inverted and drives a spur
gear above the chassis plate. This is going to require some 4mm
pierced plate, (it looks like sheets of “Mecanno”!), to be
assembled and soldered into place.
Having returned from the shed covered in Paxolin dust from
sawing the base board for the loco I return to my thread. Why
Paxolin??? It is oil and water proof, cuts with some difficulty
and it is rather strong -plus the fact that I happen to have a
large sheet of it in the shed left over from building the
Krokodil... The 3mm sheet was duly chopped to 12cm by 64cm and
most of the saw marks filed off it. It does not have to look too
pretty as its primary function is to be: drilled into, glued to,
screwed to and generally abused -during the life of the loco.
Its secondary function is: to stop things falling onto the
tracks(!)
I have scored a centre line along its length, and now that the
base board is there I can doodle onto it the positions of all
the bits and pieces -that have to fit there. Some of the shapes
for the tanks that I am going to have to build are “not pretty”
and I have resigned myself to the fact that both the oil and
water pumps are going to have to sit at the bottom of the tanks
and be driven by long shafts from overhead motors. This makes
the construction of them a little awkward as to examine what the
pumps are doing will mean that either I have to make the pump
out of transparent material (PolyCarbonate) or I simply “trust
to luck” that things are going around as they should do...
People who know me know that “luck” is something that I devoutly
disbelieve in. However, the innate perversity of situations, the
vile nature of happenstance and Murphy’s Law are things foremost
in my credo system!!!!
The tanks for the oil and water are going to be a problem.
Although I have a 30cm “fabricator” (bend/cut/roll) I have never
really used it -other than to make square and rectangular plant
holders for the greenhouse. The tanks are going to have to made
from Brass or Copper sheet and this then silver soldered
together. I have a feeling that I will be fighting over the
cornflakes boxes for supplies to attempt my tank designs with.
Some of the “joints” are going to have to be hand beaten with a
dolly and planishing hammer to get the seams smooth enough for
the solder to creep along the joint.
Decision time for the motor is coming.... Do I buy a 0.91cu
(15cc) motor or a 0.46cu (7.5cc) motor? The maths says that a
0.46cu motor is perfectly adequate for the job. However I favour
the larger motor on the factors that: it being the larger motor
it will be less stressed throughout its life and thus never have
to get into high rev situations and (being the larger motor) it
is Guaranteed to be powerfully enough!!! The price difference is
not that much either, £80 buys me the big 0.91cu motor and
£65 the smaller 0.46cu motor. Ok it may be £15 that
I have wasted, but then I have not gambled £65 -and
lost!!!
Having ordered it over the net the winner was the ASP 0.91m
motor. It is big brutal and honestly weighs a lot. Now that this
final part of the design is in place -but where do I put it???
Delivery will be in the first week of December 2010 (3 weeks
time) and probably by that date the shed will be too cold to sit
in for any length of time and static engine tests will have to
be done in the lee of the kitchen door -despite pleas of the
cold draught coming through the open doorway....
After running in the motor the first job will be to accurately
measure the heat generated by the cylinder head by means of a
constant flow calorimeter experiment. This was one of the
“nightmare” experiments of my “A” level physics years. Measure
the temperature of the cold water going into the electrically
heated calorimeter and measure the temperature of the water
coming out of the calorimeter. Measure the flow by counting the
number of drips per minute. However adjusting the flow of the
water to get a constant drip rate across the device was the
nightmare. However with a home made pump the rate of flow can be
adjusted with rheostat(!)
The next week is going to be interesting... I get to weld up the
frame for the power chassis and stick the Paxolin to the frame.
This is going to mean boring 3mm holes in it to nut and bolt it
to the frame whilst the epoxy sets, which despite the name of
it, it always seems to set in 2+ hours! I will then bore through
the holes from the piece of flat plate that forms the traction
bar between the two traction bogies -thus giving me my Beugnoit
lever array and the two sets of 3cm rubber blanking grommets
that produce the fully floating suspension for them.
Having been shopping for welding supplies and almost broke my
hand carrying a “carton” of welding rods home. I had a look at
the previous “carton” and it weighed in at 500 Grammes not 2
Kilogrammes. I am not a wimp -but the tube that they are
supplied in is just too large for me to grasp easily whereas the
500 Grammes tube is an easy heft with the grip I can get on
it... I got a new welding hood -not because the other one was
defective -just bereft of modern frills. My son is (I know) want
to go and watch Daddy at work -so a “suitably modified” welding
hood will be used as a deterrent. The welding glass will be very
dark..... (Almost opaque in fact!!!) Boredom will soon do its
work. I had to get a new set of gauntlets -the previous set (a
fetching shade of red) were grabbed by my wife for gardening in.
She does have a point though, not even the most sharpest of rose
thorns was able to pierce them. However it does render them
useless for welding in...
So, all in all we are ready to roll as they say. The one problem
is the dinner party that I have to cook for tomorrow -thus it
will be Thursday at the earliest before I can play with my new
toys.
Due to the infamous English weather it is now the weekend and I
have still to set rod to steel... Somehow rain and arc welding
is not a good combination. I am always reminded of the scene in
the graphic novel “Watchmen” :) where Rorschach is
imprisoned and kills one of his attackers who is cutting through
the bars with a plasma torch by smashing the toilet bowl and
cistern. The attacker is then electrocuted by the flooding water
hitting the cables...
Quote :)
“Never
thought of disposing of human sewage with a toilet -obvious
approach really...”
Unquote :)
So I have been shopping at the Warley National Model Railway
Exhibition at the NEC for the gears and cogs that will form the
basis of the transmission side of the loco. I found them on the
Squires Stand -which is always the stand I visit last as it is
normally at this point I am the most poorest and I feel that I
really cannot buy much considering how much I have spent
already. Sometimes it even works(!)
See
Picture 40.
The small 12 tooth spur gear is the “output” from the front
reduction gear chain of the torque convertor. This then drives
the 60 tooth spur gear to which is coupled the first of the four
bevel gears. These form the fwd/rev mechanism and the output
from the bevel directly opposing the 60 tooth is then fed via
two U/Js (not as yet bought) to the worm gears. The worms then
turn the 45 tooth gears which drive the 10 tooth cogs and then
via chains drive the 20 tooth on the axles. The chain is of the
self assembly type and four sections are required to be
assembled to produce one “link” in the chain. As my wife said
“this is something for long cold winters evenings when there is
nothing else to do”.
The first bevel is bolted to the 60 tooth by four M2 nuts and
bolts and the bore hole is enlarged so that it sits on a length
of brass tube that slides on the output shaft -but does not grip
it. The “cruciform” bearing for the other two bevel is, (I
think), going to have to be made from some scrap Aluminium bar
that I have. The carrier for the outside of the bevels is going
to have to be machined from polycarbonate. This will have a gear
wheel bonded to it. This will form the ”reversing pawl” that
will lock the transmission solid when the loco is not moving
backwards. The “reversing pawl” will only engage when the
“parking pawl” has engaged locking the transmission. Similarly
the “reversing pawl” will have to disengage before the “parking
pawl” when the loco moves forwards. I have a sneaking feeling
that I am going to have to construct some “truth tables” to get
everything to operate in the correct sequence. Fortunately I do
have a few solenoids left over from the bulk buy I needed to do
for the NER EE-1...
OK. Having sat down with a pot of tea some serious brainstorming
has had to be done... As far as I can tell EVERYTHING has to be
linked to the main servo that controls the throttle. I need at
least two “takeoffs” from the same channel. The throttle linkage
has to operated via an eccentric thus regardless of the
direction that the servo turns the throttle cable is still
pulled. The next bit is I am afraid -simply weird.... I have a
duplicate servo that turns a series of cams to micro switches
and at the and of this there is a sector plate with contacts.
Hopefully the drawing below will explain some of the madness.
See
Picture 41.
Monday Morning... Well it dawned bright and dry and I was
hopeful of finally getting some welding done. I unlimbered my
workmate, clamped the frame into the vice and put my foot on the
trigger plate. Then the mains power in the house went out and
the burglar alarm sounded... After having reset all 16 of the
RCBs(!) I dug out my trusty black “smiley” model 8 AVOmeter and
put the probes across certain terminal taps. The windings in the
transformer core had broken and there were at least four taps
that shorted to the PE connection... My much loved green Oxford
was now 25 Kilogrammes of oil filled steel and copper scrap. I
should not really complain. It was after all older than my wife.
How it came out of the BREL works in Derby I have no idea, but
it was fixture of my childhood and we all learned to weld with
it -even my sisters...
So after a little research on the net and a nibbled credit card
I will have to get a new one -probably from “Machine Mart” as
they are practically next door to the school where my son has
his extra-curricular science lessons. So it will be dump the
child and then dump the welding plant into the back of the car
and then position it in the shed -just in time to collect the
child....
Monday Evening... Well I have a nice shiny new welding plant
which is going to have to be classed as my Xmas present to me!
It is a nice shade of red, which my wife finds visually more
pleasing than Hammerite green. But she is still not too sure of
the 400V connection plug! Tomorrow I will start (hopefully) to
use it with avengence. Some of the parts are riveted into place
and require seam welding and others will require “persuading” to
stay in the right place whilst I hit it with the rod. I do have
a few NIB magnets that can be used to hold the steel pieces in
position... These have the added advantage that they have a very
low Curie point of 80 degrees Celsius. At higher than this they
cease to be a magnet and will drop off the hot metal. I used
them this way some time ago and although it worked perfectly, I
did have to lever off the NIB magnets from the steel toe caps of
my protective boots...
Tuesday Morning... Well everything worked bar the expected
problems of the learning curve for the new plant. The dial
indicator is only a “judge” rather than the “specific” taps of
the Oxford. So I started off with around 60 Amperes and worked
up to what the indicator on the slug said was 80 Amperes. At
this point the arc held true and steady as it would have done on
the 60 Amperes tap on the Oxford. The welding is rather messy,
even by my standards -but it does hold everything together.
The central strap for the traction bogies has ben unified with
the transverse pieces of square section tube steel and then the
hole lot joined to the rectangular frame. I did not in the end
bother welding up the end pieces as they seem to be quite happy
with the rivets. The next step is to dig out the angle grinder
and make things look pretty prior to paintwork. Yes -it is
going to be Hammerite green. Well I have to use it up somehow
now....
Tuesday Evening... The rolling chassis is now complete -all I
have to do is power it!!! The leading and trailing bogies have
been positioned with their M4 pins and arms -but the transverse
tensioning with springs to force them into “guiding” the loco
has not yet been fitted (or even worked out!) The traction
bogies have had the M6 bolts that are their pivots welded into
place at the bottom of the U. Slight “adjustment” will have to
be made to ensure that they sit perfectly vertical, (pass the
hammer). The sheet of Paxolin has been drilled and bolted to the
frame -it awaits daylight for the rubber grommet treatment, (my
eyes have gone!) The chassis has been tested on the sharpest
curve that I have and nothing binds. The tensioning spring
between the two traction bogies may be a little weak for the job
(?) but it provides enough tension to force the traction bogies
back into alignment after exiting the curve. One of the most
annoying aspects of the day was the discovery that my nice
beautifully machined cast iron wheels where now starting to to
show slight but definite signs of RUST. Time to move the chassis
out of the shed into the warmth of the kitchen...
See
Picture 42.
Wednesday Morning.... I should have moved the model into the
kitchen last night(!) The temperature in the shed fell over
night to minus 2 degrees C -thus the steel of the model was
“sticky” to my fingers and quite frankly painful to touch!!! So,
after having disassembled the rolling chassis from its curve to
the piece of board on the cooker to warm up. I dutifully covered
everything splashable with pages from The Express, I dug out the
first pot of primer that I reached. Typically enough it had this
nice crust on it -of ice... Time for more tea! Once the paint
had thawed I stirred it and slapped it on. It does look rather
more attractive in the photo than real life. The nice red colour
dries to a mud brown. But since the last loco that had to be
primed prior to painting ended up a fetching shade of beige when
it dried -I don’t think this really matters as it will all be
dark green soon.
See
Picture 43.
The first parts of the frame have been painted green and are
drying over night on the cooker (which has been well stoked for
the coming cold night!)
As the motor is due sometime in the next week, I have sat down
and begun playing with pieces of cardboard to find out where
everything will sit. Breakfast food cartons are cheap(ish) and
they take several rubbings out before I wear a hole in them(!)
What the cardboard cutouts have shown me is that the area that I
have to play with is even smaller than I thought. Some of the
gears are going to have to be mounted directly onto the sheet of
Paxolin and I plan to use the 4mm perforated steel strapping to
provide the spacings for all the gears. Once they have been
secured then I simply bore through the holes in it into the
Paxolin and everything is rigidly mounted (hopefully!!!)
The nasty that has arisen is the expected problem of changing
gear...
I had hoped to be able to do this with just one set of dog
bevels but no matter how hard I fiddle with the layout I end up
needing two sets... Likewise I was originally using one solenoid
-but now I have to have two.
But, all is not doom and gloom as it might appear. This does
make the solenoid a simple pull-pull fed by a DPDT relay. The
gear shift will resemble a slotted plate that is pulled forward
and backward by the two solenoids -not unlike the standard set
of points motor used by my son’s Hornby.
The cardboard also reveals the positions of the tanks -and it is
not a good one. The shapes that they are going to have to have
are dictated by the volume that I need to store in them and the
pump that is going to have to sit at the bottom of each one. The
tanks are going to have to sit towards the radiator end of the
loco. This was rather expected -but it does produce plumbing
problems. The torque convertor will require a drain back to the
tank -this means that the drain will have to exit above the
surface of the oil -it will do this via a “Swan Neck”. The pump
will pressurise (slightly) the torque convertor and there is the
“environmental problem” of an oil leak. This also means that I
am going to have to figure out some form of oil seals for it.
“How do I start the motor?” is another problem starting to rear
its head. I have a plan in which I extend the motor drive shaft
out past the end of the torque convertor to a hex end. This
connects to a battery drill fitted with a over run coupling. The
idea is that I power the drill and the motor starts. The Glow
Plug will I feel be of the “permanently live” type, in that it
will have a current running through it all the time.
Normally Sunday Morning and Afternoons are times when I am very
productive -however due to the incredible cold snap that we are
having this means that I am keeping warm in the kitchen -so work
will slow down -but it will be nice and warm....
This particular Sunday Afternoon was spent in the kitchen on the
floor mat -near the cooker...
I did some trials with the positions of gears and shafts. The
first shot shows the initial trial on the piece of Paxolin.
See
Picture 44.
The next shot shows the provisional working arrangement. This is
not as “elegant” as I would like -in fact it could be quite
difficult to service. However, I feel that this is the correct
arrangement for the system as it can be manufactured in modules
and then “layered” together. Moving the worms towards the centre
of the model allows me to fit a larger gears to the “input”
shaft.
See
Picture 45.
Most of the day has been spent sitting at my drafting table and
abusing sheets of graph paper... It is amazing how thing
conspire to clash against each other when you start to draft up
the arrangement. I was sure that I had enough room and yet,
despite my best efforts, I had to screw up two drawings -the sum
of three hours work.
I have also found that my self assembly chain may be the only
length that I ever buy. Tedious is not the word!!! I have
assembled 24cm of it -enough to go around one gear and that is
enough for one day(!) As with the “Krokodil” there will be a
chain link between the power axles and the conrods will be
purely decorative.
At the end of the day I do have a drawing with most of the
things in the correct position and good idea how the reversing
mechanism will function and where it will fit, (this time it
will not clash if it knows what is good for it...)
The locking system for the reverse/forward gear is very simple
-every filing cabinet has one that works in the same manner. It
consists of a plate with two parallel slots with a diagonal with
a captive pin in a transverse slot. Moving the plate across the
pin moves the pin from one channel to the next and it cannot
move back without the plate sliding. The plate will be pulled by
two solenoids acting in the same direction and a return spring.
This means that if the power fails the default direction will
always be forwards. The plate will act on a bar at the end of
the two transverse shafts sitting between two thrust bearings
and this will push/pull the shaft across the dog bevel to select
direction.
Although the gears are nylon I am going to sweat a piece of
steel to the shafts and then pin the gears to this with nuts and
bolts. Thus the torque will be taken through the nuts and bolts
and not through the grub screws on the gear wheels. I still
remember the grub screw that cut a neat thread on the shaft when
I was testing the gearbox for the NER EE-1...
The arrival of the motor is in sight and I have to give some
serious thought as to how I get the power shaft from the front
of the torque convertor to the central gears. I had originally
intended to use some old powder metal gears -but I now believe
that this is the wrong way to do it. So I am thinking of driving
things from toothed belts. This will damp down any vibration and
hopefully give me some quieter running.
Wednesday 1st December 2010.... Possibly the most successful day
ever with this model!!! Although at least 25cm of snow had
fallen over night and the outdoors was in the less than balmy
temperature I set to work on the kitchen floor. The previous
night I had completed drawing №4 of the gearbox and miraculously
it seemed to fit al the parts into the right places and nothing
clashed (too badly). Braving the cold I stepped out into the
white wilderness and retrieved my set of marking instruments and
made some hot tea!!! Well this shot shows the the №4 drawing and
the transcribed positions on the sheet of Paxolin bolted to the
frame. Drawing in 16 pretty colours is normal for me -a 0.3mm
felt tip is more transportable than a Unix workstation running
AutoDesk...
See
Picture 46.
The next shot shows the progress after a couple of frustrating
hours. I had to chain drill the slots with a 3mm drill and then
“pick” the chains apart with a floor board saw until there was
enough room to get a proper saw in there. I had to smooth up the
shape first of all with a Rasp, then a Bastard file and finally
a half round to smooth it all off with. Paxolin dust of course
went everywhere and “Henry” spent quite some time sucking up the
mess... Cutting the final slot for the central gear was I admit
risky. I nearly cracked the sheet twice -I am going to
re-enforce it with some transverse hard wood strips later.
I drilled out the gears to 4.5mm and inserted a CA’ed length of
K&S brass tube -giving me a 4mm bore to the grafted gear
made from a bevel and a 45 tooth spur. Whilst the CA sets hard
over night the grafted gear will then be drilled and pinned in
the morning. These stand on M4 bolts that fasten through the
steel frame and emerge onto the Paxolin sheet. The central 60
tooth spur gear was fitted to its 4mm shaft and the two worm
gears duly bashed along the shaft until the hit the right
position, There will have to be a further bracket and ball race
besides the gear as the only supports are at the ends at the
moment.- and you can just bet the thing will wobble madly
without it... All the raw brackets are going to be drilled to
take M10/4mm bore ball races.
See
Picture 47.
The next two shots show the rough assembled final drive and
reversing system.
The blue self assembly chain on the small sprocket is obvious as
it vanishes through the slot to the large sprocket on the outer
drive axle. The dog bevels are on the shaft and (hopefully) you
can see how moving them along connects first with one side and
then the other.
See
Picture 48.
This shot shows how the input drive from the torque convertor
reaches the central gear. The grafted gear JUST misses the shaft
by 2mm and the 20 tooth gear drives the 60 tooth gear cleanly.
See
Picture 49.
After some discussion via “chat” I have taken the advice of a
friend and purchased the 0.46M motor as; “Thats too big an
engine for a toy train”, so we will see... The big motor 0.91M
is a special order -but the 0.46M is an off the shelf part. I
wonder which will arrive first?
If you examine the gearbox drawing at the top of the page then
examine the Picture 49 -you will see that the one that I have
built is upside down compared to the original. However like the
original gearbox it should be pretty smooth in operation.
Well the snow still falls so I spent the day in the kitchen.
First job of the day was to strip off all the test stuff from
the Paxolin sheet and dig out my bow saw and start sawing up the
hard wood framework for the gear box. “Why Wood”? I hear you ask
-because it is nice to work with and quick and easy to change
things when you have an oops moment!!!
I drilled out the brackets to take the bearing -rather painfully
having to “wind out” the hole from 5mm to 10mm a millimetre at a
time... I have found it is better to be bored than to twist the
bracket. The next shot shows the worm gear with a bracket and
bearing. the basic framework for the reversing “Ballista”(!) can
be seen in the top right.
See
Picture 50.
This shows it all far better. The throw arms sit either side of
the central gear and the whole thing is assembled with “Croyd
Aero” and then bolted and glued to the Paxolin with epoxy.
After 2 hours I could start again. It may be me but I have found
the slower 2 hour epoxies to be far stronger then the 5 minute
ones(?)
See
Picture 51.
The next shot shows the first assembly of the transverse shafts.
The brackets are bolted in place and the dog bevels “tapped”
into distance along the shaft. This is tooth plus 4mm -so the
“shot” from the Ballista is 6mm in total.
See
Picture 52.
Ok -over head shot. You can see the third bearing on the central
gear -this should remove any wobble from it (please!)
See
Picture 54.
The next shot shows the end of the days work (and my eyes have
gone too) So sorry -I have had to use AutoFocus... I need to
ream out the small sprocket to 4mm and start trimming everything
to size. I will admit that it has been a very rapid couple of
days progress!!!
See
Picture 55.
It is still snowing -so I am back to work in the kitchen, (what
a shame!!!) The first thing to do today was to construct the
gear shift arrangement. This, as I have said before, uses the
same mechanism as a set of filing cabinets. The slide is made
from a 50mm M3 bolt, (suitably filed), and a length of U section
brass. These were then silver soldered together -they will never
come apart, (which is how I want them). These now form the pin
slide that transfers the longitudinal movement to transverse
movement.
See
Picture 56.
The next shot shows the pin slide with ABS sheet CA’ed to it to
act as a lubricated channel. The pin slides on a I beam section
of ABS. A little touch of sand paper ensured that there was very
little rattle. The top of the slide is now flush with the arms
of the ballista. You can see the 4mm collets that will take the
thrust bearings for the transverse shafts.
See Picture 57.
This is the base plate -made from a 3cm length of 1 inch 64
thous K&S strip. The shape of the slot does not really
matter as all we need is a 3mm wide line between the extremes of
shift. The function of the base plate is to hold the pin slide
in its groove as it is pushed and pulled 6mm. A little CA holds
it in place whilst I drill down through the arms of the ballista
to hold the M3 nuts and bolts.
See
Picture 58.
This is the top plate. The nice sections of L brass provide the
smooth direction for the movement of the slide pin -regardless
of the dire nature of the hole that they are covering... They
are held together with a little silver solder and an M3 bolt is
used to provide an accurate line between them.
See
Picture 59.
The completed top plate. The slot can now be trimmed up with a
needle file until the hole at the bottom of the plate matches
the slot provided by the two L sections(!) A quick once through
with a 3mm drill clears out the holes in the top plate that have
been covered with the lengths of L section.
See
Picture 60.
The reach rod is at the moment rough cobbled from a few lengths
of pine strip wood, (see above!), it has a 60 degree slot and
captures the pin thus pulling and pushing it across... The final
version will have a short length of brass tube around the pin to
provide a rolling surface. The rough version tends to catch on
the threads of the M3 bolt.
See
Picture 61.
This shows the completed ballista gear change system. The pin
has been fitted with a cross arm, a 1/4 inch sq K&S tube, to
push and pull the shafts. This will be re-enforced with brass
plate in finally assembly. The reach rod is held flush to the
back of the gear change with a simple spring loaded shoe, (a
length of bent brass strip). This also pushes the slide pin into
the angled capture slot. In the final version I am going to have
to make the reach rod in two thicknesses and machine out the
capture slot. One of the quirks I have noticed is the fact that
the gear change is “upside down” -i.e. moving the reach rod
forwards engages reverse... The Mk 2 version of the reach rod
will have the slots in the other direction. The pencil marks on
the reach rod indicate the positions of Fwd and Rev. The
distance between them is 11mm -this is well within the 18mm jerk
range of the solenoids. The main problem now is making sure that
the solenoids have sufficient power to over come the strength of
the return spring and the effort of moving the shafts across.
See Picture 62.
The last day of building before the seasonal madness... I
trimmed a spring or two to get the correct rating to hold the
pin against the reach rod and then sufficiently strong enough to
press the pin into the capture slot. Too strong it made the pin
act like a brake, making the reach rod hard to move -too weak
and it didn’t force the pin into the slot. So, after taking a
too strong spring and slowly snipping off a quarter spiral at a
time -I got one with the correct rating to work!!! The screw and
swing flap will be replaced in the final version...
The next shot shows the completed ballista with the two
“selector forks” I suppose I should call them(?) they are simply
U channel lengths of brass with an M3 bolt silver soldered on to
them. These sit on the shafts and push/pull the collets. They
will be terminated with a couple of M4 washers to act as capture
rings & thrust bearing surfaces so they do not slip off the
shafts during movement.
See Picture 63.
Ok -now the glossy shots!!! Pulling the reach rod pushes the pin
out, the ballista pulls the shafts across, locks it in the out
position and the left hand bevels engage with the grafted gear.
The pin out position will be the forward position as the pin in
position does force the bottom of the grafted gear away from the
worm gear, (slightly), whilst the pin out -does the reverse(!)
See Picture 64.
The final shot shows the reach rod in the reverse direction. The
custom tuned spring forces the pin and shafts across and the pin
is captured by the slot and rammed home into the locked
position. The right hand bevel gears now engage on the grafted
gear and the loco reverses...
See
Picture 65.
Q.E.D. (!)
The next problem is to transfer the filed and twiddled length of
wood that is the reach rod into something made of brass. There
are, (alas), no suitable K&S sections that can be fiddled to
fit -so it is down to the scraps box to supply the parts... I
have decided to ditch the idea of a solenoid and return spring
and simply have two opposing solenoids switched through a relay.
Now that everything works it all has to be taken to pieces and
fettled into something that could be called a reliable working
device...
After suitable repast I am thinking of ditching the solenoids
now!!! The reason is that they are just too fast acting. There
is a chance that the pin will not be forced into the capture
slot and the whole lot will jam solid. The only other thing is a
slow moving linear actuator. This is a fancy name for a nut on a
thread!! I plan to drive a 30 tooth Mod 1 gear from a 10 tooth
Mod 1 on a small electric motor. The shaft of the 30 tooth gear
is, (surprise), a mangled of M6 roofing bolt... Silver soldering
a couple of M6 bolts to a length of L section brass strip gives
me my arm to the actuator. I will use this to slowly, (but
powerfully), move the reach rod through the gear selection
system.
OK -any idea where I fit it???
Time for more breakfast cereal packet cardboard...
Sunday afternoon produced the required linear actuator -pretty
much as described above. I used one of my many £0.80p each
SME motors (eventually I will run out of the box of them?)
Opening up the 5mm hole in the bracket to the required size to
fit the head of the motor through, required some patience with a
broach -typically enough the size is Imperial but the rest
Italian motor has metric dimensions. The nylon adaptor to fit
the 3.2mm shaft to the 4mm gear was delicately inserted with a G
cramp and the shaft of the motor pressed into the adaptor by
crushing the lot in a vice... Don’t knock it -it works!!!!
I lopped off a length of brass and filed a flat square edge on a
handy M6 hex nut. A few minutes with a cooks blow torch and they
were silver soldered together. A suitable off cut of ABS was
pressed into service and the brackets CA’ed to it. Tests with a
few kitchen weights proved that the linear actuator is capable
of lifting 500 grammes with 12 Volts feeding it.
The acid test was: “Could it change gear?” The hardest part is
going from Reverse to Forward selection -so this was the test.
The test was done and the answer (thankfully) was: “YES!!”
The place you see it in the following shot is probably the best
place for it. But it is going to make the design of the reach
rod “interesting” to say the least.
See
Picture 66.
Now that the actuator works I have to set the limits of how far
it will travel. There are certain points motors that work on the
“stall” principle -but I do not like this approach. What I
intend to do is to use DPDT relay to provide Fwd and Rev to the
electric motor -no surprise. BUT the trick is that I have a Push
to Break switch in the feed to the cables to the motor at either
end. This means that I have two pairs of cables feeding the
motor terminals. The motor turns in the Fwd direction until it
breaks the contact and then stops. The opposite direction now
has continuity, as the actuator closes the contact as it moves
away -so when the relay flips the motor has power to move in the
required direction...
I have been playing with card board and the designs that have
emerged for the tanks -do not look pretty... I am resigned to
having submerged pumps for both coolant and oil. I have decided
to go for pure glycol rather than water for my coolant as this
is inert and non-conductive. This does mean a bigger tank
though. However the tanks can be made from simple tin plate and
soft soldered together as the contents will not rust them. I
will confess to looking through the IKEA and LAKELAND ranges of
plastic food boxes to see if there are ones that are suitable!!!
To paraphrase a famous game... “Colonel Fell”, in the Kitchen,
with the X-Acto knife!!! Now that I have a general idea where
things are going and nothing is going to be delivered in this
weather. The only thing for me to do is to sit by the cooker,
make sure it is well stoked, and make my bodywork for the loco.
As usual it is a plywood carcass that will be laminated with ABS
sheet. I did honestly intend having a “go” with Aluminium sheet
-but I feel that this is a little risky for a first attempt -so
I am going to have a go with the tanks first...
I also had a “go” with Titebond” for this carcass instead of my
usual Croyd Aero -as I have nearly run out. “Do I like it?” Well
to be honest -I am not sure... I have several American friends
who swear by this stuff -but it doesn’t seem to suit my methods
or style of construction. So, all the flat plywood sections are
done with Titebond and the laths are all stuck with my remaining
Croyd Aero.
See
Picture 67.
The strange cross sectional problems reared their heads almost
immediately. The windscreen section with the infamous angle was
the first victim -it became 30 degrees!!! This made the problem
of producing the cross section through the hyperbola that was
the roof -a lot simpler... The square to parabola that is
the top of each of the noses I simply gave up on a mathematical
calculation and did it by drawing. The laths that form the cross
section where simply aligned with the side of the carcass and
the angles traced onto them. Some of them are a bit wonky -but
since they are just the ribs I am not bothered.
See
Picture 68.
After having attacked the collection of breakfast food packets
in the pantry I have a template for the bonnets at either end of
the loco. The template resembles a cubist version of a cup cake
drawing. This was done by measuring the lengths from the curve
on the windscreen to the edge at 0.5cm intervals and
transcribing this to a sheet of cardboard. The final act was to
bend, (somehow), a ruler across all the points and drawing a
line through them. The template was then tested and fettled
against the curve and I have a template that both fits and
works.
The next job was to “roll” the roof. I used a rolling pin, (no
seriously!) to wrap the 60 thous ABS sheet around and kept
tightening the curve until I got something that wanted to sit on
the roof rafters. Then the hard work began... Applying bucket
loads of CA to the central rafter stuck the sheet to it. Then by
means of brute force I had to bend the ABS to the curve and then
tape it to that position. After about 20 lengths of packing tape
I had all the edges flush. Pumping in another bucket load of CA
filled the gaps at the windscreen ends, (inside and out). Then
it was a simple job to mix up some 2 hour epoxy and apply it to
the gaps between the top of the side and the ABS sheet. I use
about 1/2 of a “double syringe” of epoxy doing this. OK it might
be too much, in fact it probably is -but it will hold it firm
while I molest it with the sand paper to get a “true” edge to
the side. The assembled nastiness is currently “fast curing” by
the radiator in the kitchen.
See
Picture 69.
In the morning we will find out if I have got a roof and two
bonnets -or had an interesting waste of time...
Well the mummification went well and the epoxy had oozed through
the gaps and united the roof to the sides. The last thing I did
at night was to turn the model upside down and dobbed spot welds
of CA along each of the rafters to lock the roof to them. I do
not know if this helped(?) but in the morning the model was
unwrapped and despite a few dribbles of epoxy I was very happy
with the result. There was of course now some “delicate” work
with garnet paper to remove the dribbles from the ply wood at
one point I will admit to using a rasp! After about an hours
abrasive surgery I had smooth sides and square edges. Now it was
time for the ABS sheet to come in from the cold -shed...
Needless to say they took some time to regain room temperature
and their flexibility returned. I have found that it is easier
to “crack and snap” warm ABS sheet than cold. When it is cold
the sheet is too brittle then the crack can wander from the
initial score and you end up with a wasted piece. I used thick
CA (which I rescued from my wife!!!) and then trimmed the
laminated ABS to fit. At the end of each side I trimmed and
sanded each corner square -but there were still the inevitable
gaps that had to be filled. I normally use epoxy putty
(Milliput) to pack and fill the gaps with. This is very strong
and takes a nice finish. The final dent and surface
filling is normally done with Squadron Green cellulose putty.
This Sunday afternoon has been devoted to the cooling side of
the locomotive. I scoured the junk draw and extracted an XP1700
heat sink and fan. I know that I saidI was going to construct a
tube one -but this is a quickie. I am proficient at liquid
cooling computers -so why not use the same techniques to cool my
loco? The next shot shows that fan that has been grafted to the
Paxolin sheet.
See
Picture 70.
More words were uttered as I had to cut a circular hole through
the Paxolin by chain drilling it. But it fits nicely. The next
step was to measure where the “exhaust” from the heat sink/heat
exchanger would vent and then to cut a suitable hole in the
correct position. This was then layered with Dacel Plait to act
as a grill. The loco is symmetrical and there are similar aero
vents at the other end -this is more for show than anything
else. The central hole is there because it is there on the
original(!)
See
Picture 71.
The modifications to the heat sink are as follows: I drilled
3.5mm holes through the base and then drilled 3.0mm holes cross
ways to form internal pipework. Into this I epoxied a length of
3.5mm copper tube. The hole lot is then topped with a 2mm thick
sheet of ABS this too is epoxied to the heat sink. I now have my
thin film radiator. The next step of the operation is to drill
and tap M3 into the heat sink through the ABS sheet. I then slap
on top of this a sheet of polycarb -this will enable me to see
the liquid flowing, (or not as the case may be....)
See
Picture 72.
Monday Morning. And as I step out the door I find a collection
of packets left there!!! One of them was obviously for me so I
opened it and found myself to be the proud owner of a 46 size
marine glow motor. I have not studied the motor in detail -just
given it the once over and had a look at the instruction sheet.
I am happy with the size and look of the assembled bits -so the
other motor is going to power “Ixion”.
See
Picture 73.
As you can see I have marked the picture to show the various
connection points for the plumbing. One of the unfortunate facts
of life is that the plumbing points are always in the wrong
place -this is no exception.Ideally I would like the positions
of the coolant IN and coolant OUT to be reversed... This will
make “bleeding” the coolant line harder -and I will probably
have to bleed the lines with the engine out of the loco before
fitting it. The only position, and one that I planned for it -is
inclined to an angle of 45 degrees. This brings the silencer to
the vertical and it sits over the gearbox. It is a free silencer
and if it works to the satisfaction of the assembled family it
may well stay. Failing that I build my own.
The specs for the motor say that it is “timed” for a 5%
nitromethane mix and should not be run on less than 18% oil. I
personally do not like nitromethane. The combustion products
(i.e. NITRIC ACID!) are to me unacceptable. As I said above I am
therefore going to use a 20% caster oil and 80% methanol mix.
The motor is supplied without glow plug -so I am basically going
to pick someones brain as to what would be a suitable starting
point.
Having had the motor now for a couple of days I have spent most
of them simply looking at it. Not out of admiration for its
sheer beauty, but what shape it is and measuring it to within a
millimetre of its life... This plus more than a few drawings
have shown me that I have to move the centre line of the motor
and torque convertor to sit 2cm to PORT. This and the fact that
if I change the angle of inclination to 30 degrees gives me a
nice corridor to get everything up and down. This way I can
squeeze everything into a rectangle 12cm by 14cm. I have yet to
fire up my motor, but I have been twisting it every now and then
to free it up, whilst slow feeding it castor oil from the same
bottle that I use to oil my record player with. It now moves
with a polite hiccup at Top Dead Centre and a small length of
knotted string has successfully pulled it over at cranking
speeds. Things are nearly ready to go!
I have my temporary water pump ready (a Lucas windscreen washer
pump) and I need a fuel tank and fuel... The latter is causing
me some grief(!) there is now only one supplier of “straight
castor” and that is Model Technics. But this is very nearly
unobtainable... The Model Technics fuels “GX-5” and “GN-5” both
have the correct ratio of castor to fuel but have 5%
nitromethane -as does Flair “Yellow”. I might have to explain
why the “H” plants are not feeling too well along the
border... Why do I call them “H” plants? for some reason my wife
has taken a great fancy to a large range of plants which all
seem to have names beginning with “H” viz: Hosta, Heuchara,
Heutunia, Heather, Heath, Hebe, Hydranga -you get the idea!!!
I do at the moment have absolutely no idea of how large a fuel
tank I need. I would like a run time of 10 minutes so if I play
with the figures -based on what I know of combustion I can
perhaps get some approximation(?)
The perfect air to fuel ratio is 7:1 for Methanol...
I sent an enquiry e-mail to Model Technics to see which fuel
they recommended and this is their reply:
Hi Ralph
I will answer your
questions not in any particular order.
1 Todays engines run
much better and more reliable with a small amount of
nitromethane and 5% is not unreasonable.
2 All castor fuels in
todays engines is not good. Tolerances and castor's nature to
varnish causes problems in reliability, running etc due to the
varnish build up. I would suggest therefore Sport 5 which has
a small amount of castor as an additive. The rest of the oil
being a highly regarded synthetic from Klotz.
Total oil content is 18%.
3 As for the acidity
of the exhaust, well the exhaust is acidic due to the nature
of the combustion process. The small amount of nitromethane
will not increase the overall acidity of the exhaust as we
have found in tests. So your wife’s roses and plants beginning
with H are safe.
4 The main components of
nitromethane combustion are members of the nitric acid family
but nitrogen, water and the oxides of carbon.
I hope this helps but if
still unsure please phone.
So there we are. Any problems with a “Key Lime Pie” or
“Obsidian” or “Creme Freshe” variety of Heuchera is NOT MY
FAULT!!!! The Hostas will still have to battle the slugs -but
not the passing loco...
Now that the seasonal madness is in full swing I can divert my
time into applied insanity(!) One of the things that
distinguishes this loco from all others is of course “The Fell
System”. This is not only the infamous gearbox -but also the
fact that the engines operated from a supercharged source of
air. This was supplied by two “Rootes” style displacement
compressors powered by two diesel engines running at constant
speed. Thus the main power engines were fed a high degree of
supercharge at low revs and this fell off at higher revs. To me
it seems that I have to duplicate this in some manner. I do not
have the technical ability or tools to produce a “Rootes” blower
but I can produce fairly easily a centrifugal blower. There are
two main methods of supercharging. Draw Through The Carb [DTTC]
and compress the mixture in the supercharger, (the system used
by Bentley). Or, Blow Through The Carb [BTTC] and modify the
fuel system accordingly, (the system used by Mecedes-Benz).
There are pluses and minuses to both systems. In the draw
through you have a large volume of fuel and air churning and the
heated output behaves more as a perfect gas than a fuel air
mixture. In the blow through the hot compressed air is cooled by
the evaporation of the fuel droplets inside the engine and there
is less backfire risk. On the balance of ease I have selected
the blow through the carb system. I have to design some method
of pressurising the fuel tank other than by using the exhaust
gasses from the silencer -as I doubt that they will have
sufficient pressure to power the carb(?)
What I intend to use is technically speaking an axial flow
compressor -or to be blunt a few computer cooling fans stacked
together!!! I can stack three 25mm fans together to produce one
fan. I don’t know how much “supercharging ” I will achieve
-maybe a few millimetres of water. Each fan is rated at 155
Litres per minute, thus it should provide “supercharged air” for
the engine up to 20,000 RPM -but as I said above the amount of
supercharging may not be very high. However the simple fact that
it IS running supercharged captures the spirit of the loco
rather than being just a model of it!!!
On the second day of the New Year I was able to spend some time
in my shed at last. Most of it was spent rebuilding the gearbox
to a far stronger standard. I drilled down the grafted gear
sides and united them with 2mm steel bolts. These were CA’ed
into place along their length and then trimmed to size with a
junior hacksaw. The bevel gears were introduced to my new(ish)
Vee Blocks and I delicately and carefully drilled through the
boss at the back with a 2mm drill and CA’ed a pin through it. I
find that I have almost run out of 2mm roll pins -so I will have
to use bolts only soon. The M4 steel washers that form the end
loops for the “gear selectors” for the ballista were just as
evil to fix as I imagined. I polished up everything nice and
bright. As my former Brother in Law comments -”you have to be a
SAINT to solder”, and applied the heat from my chefs torch. A
little flux and then I hit it with the electrical solder, (good
old Savbit). After some, “thinking about it”, it did bite into
both the brass and the galvanised steel, then flow around the
complete brass U section. The L brackets have been re-enforced
with some offcuts of teak so the bevel shafts have something to
rattle against.
The next thing to do is to investigate the final connection
between the gearbox and the drive axles. In short -can we figure
out how to assemble the last pieces of “chain” in situ???
The big 20 tooth sprocket that will take the drive from the
gearbox above will have to be “custom cut” into the bogie as it
is over 4cm wide. This means some “delicate” work with an angle
grinder and a hacksaw to cut a slot for it to fit into. Looking
at the sprocket again I think that I bite the bullet and wait
until the budget can afford some 10 tooth sprockets and use
them. This will mean an order for a further 6 of them from
“Squires” -but that is not really a problem. The shopping list
is starting to grow again!
Well the New Year has sprung and now two weeks into it I can
return to work on my loco. After some self disgust I hacked away
at the noses of the loco at each end -until I had something that
looked “true”. The next stage was to apply basic auto repair
systems!!! In short P40 then P38 and Squadron green putty... Now
a couple of days after I began the “rhinoplasty” for the loco I
am much happier how it looks. The problem was that the ABS sheet
that I was using was too thin to take a true corner from. But if
I made it thicker -then I could not bend it. Hacking away a
section and then filling it in has provided a more smoother look
to the nose. The shot below shows the result of wet sanding and
filling. The section cut out can just be seen from the end of
the side radiator to the windscreen, (as will be!)
See
Picture 74.
Contrast the nice square to round shape with the raw shape found
in Picture 71.
I am now busy starting to paint the things that I will not be
able to paint once they are in position, and if they get
scratched -then “tough”!!! The B0-B0 look of the loco means that
it post 1954 -and it will be “Diesel Green”. What exactly this
colour is -is debatable... But for the sake of accuracy,
(meaning that the paint will be the same throughout my Green
Diesels) -I declare this colour to be Rover Green 17. Since this
is the colour that they use at Barrow Hill to paint the restored
Deltics with -I have no quarms in cribbing it’s use.
The flycranks et al will have to be signal red. however I am
going to depart from the accepted colour scheme for this and
paint my wheels in light green with white rims. The orange
stripe and the Lion will of course be there -but where I get my
Lion from I am not yet sure(?)
Tomorrow I have to get my knees X-Rayed and I will grab this
opportunity to visit my local model shop -it is just up from the
DRI. There I plan to raid the owners for information as to
the price of Glow Fuel, and what they stock. I am still not sure
if I require a fuel pump -or the exhaust pressure is going to be
high enough to feed fuel to the carb(?)
I will admit to spending most of today simply looking at my
model. This has nothing to do with personal vanity -but a lot to
do with having to see whether I have the essential feel of the
model... This normally involves looking at my drawings and then
pasting / taping pieces of cutout to the model. Sometimes even
when you have the piece in exactly the right position according
to the drawing -you find a picture that contradicts it... Thus I
admit I played shuffle board with my bits and pieces all over
the roof of the model. The roof slots were made simply by
sticking the lengths of ABS strip to the roof and then cutting
through the now strengthened roof section with a cutting needle.
The two central inspection hatches are one piece of ABS curved
and then grooved -the two offset hatches are done in the same
way. The doors and windows will have to be “black areas” as in
order to cut these through I would have to cut through the main
support brackets inside -this as they say is: “not a good idea!”
See
Picture 75.
The one thing that I have to admit COMPLETE FAILURE with -is the
twin exhaust system. I would dearly love to have the fumes from
the glow engine expel from them. BUT the exhaust outlet from the
silencer is in the wrong position and even if I used a length of
silicone pipe to join it all up it would still not be “safe”.
So, what I am going to do is to use a length of aluminium tube
and vent the exhaust underneath the model. This will keep small
fingers safe from hot exhaust gasses -but the ants and spiders
might need to run rapidly(!)
The next thing to do is to make sure all the pieces that I have
laboriously cobbled together actually interlock together. This
may mean silver soldering more brackets to the gear box and then
getting everything tapped and screwed to within an inch of its
life -but with the G3 AGM at the end of next month I want to
have something that can be seen to be a, “work in progress”,
rather than a collection of modules...
Well today has been fairly productive(!) I have finished off
building the radiator. This required cutting out a piece of poly
carb, (always a messy job), and then clamping it onto the top of
the radiator (XP1700 CPU heat sink as was...), and drilling down
through the poly carb and ABS into the Aluminium below.
See Picture 76.
The holes in the heat sink were then drilled and tapped to M4.
Some gasket paper and red gasket compound were suitably cut and
smeared and the lot bolted together. The bolts were cropped to
length once the compound had “set”.
See
Picture 77.
Simple tests with a tap and a length of pipe have shown that
although ugly -it is water tight.
The replacement PCB for my Milling machine arrived on Saturday
and this morning was the first chance that I had to fit it. It
was a Monday Morning and as usual things decided to “play up”
after the weekend off... I fitted my PCB and -nothing happened.
Basic tests with my “smiley” AVO showed things to be where they
should be and I continuity in the right places dependant on the
position of selector switches. Still it refused to go round....
I telephoned the Manufacturers -which being deaf is very hard
work for me (!) and discovered that I had to short a link on the
PCB to get it to function(?) After having unshipped and fitted
the shorting link I applied power and both LEDS on the board lit
up. Turning the pot knob and the motor growled, a little more
and we had rotation!!!
Now that I have after 2 months plus of waiting a functioning
Mill I can press ahead with things that have been waiting for it
to work....
The first thing to do is to paint the wheels -this might sound a
very strange thing to do but it does require a Mill to cut fly
cranks etc with and once I have them I will not be able to paint
the wheel spokes. A short squirt of spray primer and a some
green paint from the pot and I have wheels for my loco.
The time is drawing near to start assembling my series of
modules into a locomotive. I have ordered the bearing for the
bogie wheels, the grub screws for the driver wheels and the
bearing for the driver axles. The next thing on the shopping
list was the 20 tooth gears for the drive axles and the
additional “chain” for them. The body has had its first layer of
grey primer and when this has hardened off (48 hours) then I
will wet sand it flat and apply the next layer. Normally two
layers is all that are required -then it is on with the green. I
have to mark out and cut the windscreen holes in the front of
the loco -this could be a rather delicate operation... Once the
holes have been punched through they have to be “edged” and then
the whole think primed again. Why did I not do this before the
initial spraying? the reason is simple. The edges are very
delicate and if I sprayed two coats of primer on then they would
disappear under the layers.
A little further on in the week and have progressed a bit
more. The bearings for the front and rear bogies are now firmly
installed and they scoot along the floor nicely. I decided to
build a dedicated oil cooler for the torque convertor rather
than having a cooling spiral in the water tank. This doubles the
amount of cooling in the loco and lowers the strain on the water
cooler. The oil cooler is made pretty much in the same manner as
the water cooler except that I have used a heat sink from an
Intel Pentium 2 series 3 cpu. This fits nicely across the front
of the loco.
See
Picture 78.
See
Picture 79.
Well a busy and productive Saturday in the shed! Despite the
finger numbing cold I did get some of the main targets that I
wanted done. The first thing on the list was to drill and tap to
M3. The next shot shows a driver wheel on the rotary table with
a suitably modified centring peg!!!
See
Picture 80.
The next task on the list was to chain drill the chassis to take
the chains (!) Here is where my beautiful day hit the skid
pan... I had purchased from a TV Shopping Channel a set of
Cryogenically hardened Cobolt Steel drills. The demonstration
showed how they drilled the same drill to drill through glass, a
file, and concrete plus they had a 10 year Warranty. I was
sufficiently impressed to order a set there and then. They came
the following Wednesday. I admit they looked very pretty and
functionally rugged for the job. So, it was with high hopes I
inserted the 3mm drill into my pillar drill and started to punch
holes through the steel following the scored line.
See
Picture 81.
On the third hole the drill broke....
See
Picture 82.
If you look at the picture you can see where the hardened tip
has simply come away from the drill proper. Needless to say “I
was not impressed!!!!” I then fitted a standard TiN drill bit
and continued to punch the rest of the holes. A slight attack
with a hack saw and a set of mole grips and the resultant cactus
edge is now visible. A quick once over with an angle grinder
soon tidied it up.
See
Picture 83.
It was now time to say “goodbye” to the M8 coach bolts that had
provided the loco with axles for nearly four months now.
The axles I cut some 6mm bar into four 10 mm lengths squared
them up on the lathe and then bored 3mm diameter holes for
5mm into the ends of them .This will provide the holes to hold
the lengths of M3 threaded bar that will hold the fly cranks to
the ends of the axles.
See
Picture 84.
The wheels will have aluminium tube shoes in them this will act
as a crumple layer for the grub screws to bite into. The next
step was to mangle the sprockets to fit the axles. The bore of
the nylon sprockets is 4mm and they sit on a brass insert bush
cast in. They had to go... I ripped out the bush on the lathe
with a 6mm drill.
See
Picture 85.
I now had to make a new bush for fixing for the sprocket. A
length of Aluminium rod provided the new bushing. Oh so slowly I
drilled down the length of the rod and out the other side. I was
lucky if I could cut 3mm down it before the bit clogged up and I
seemed to spend HOURS simply winding the bit in and out of the
hole to brush the swarf off the end.
See
Picture 86.
Once I had my “tube” I then clamped it up and drilled the 3mm
holes down its length giving me what my son called a “small
flute”.
See
Picture 87.
I put it back into the lathe and used the parting tool to cut
myself some slices from the tube. Bolting the slice to a
sprocket and then drilling 2.5mm holes in the two provides
union. There was a slight panic stricken moment when I had to
count up to see if I actually HAD enough 2.5mm nuts and bolts to
complete this. The double sprockets have been fitted with three
2.5mm nuts and bolts as these will have to take rather more
strain that then the end ones.
See
Picture 88.
The following shot shows the double sprocket and the single
sprocket (unfortunately still on M6 coach bolts....) When it
comes to actually fitting the sprockets I intend to drill down
the already present 3mm holes and through the steel axle shaft
and then pin the lot together.
See
Picture 89.
The outer of the two sprockets connect to the gearbox above. In
the original this would have been done via; an idler gear, to a
quill drive, to the driven axles -but to be honest given the
reduction and the torque convertor I do not think that this
particular loco will need one as it will have no “cold snatch”
from electric motors. The next stage of operations will be
to section and slot the power bogies to take the modified
sprockets. The chain return between them is very likely going to
have to sit in an external “conduit” to keep it clear of muck
and leaves etc.
See
Picture 90.
After a respite to deal with a slight case of flu!!! The chains
have been assembled and the fore and aft chains fitted to the
output sprockets. The single sprocket is the outer and the
double sprocket the inner set. The transmission between the
inner and outer axles will be via conrods and the link between
the two bogies via the chains. This is because chain systems
only work by pulling -and it gives a flexibility not found in a
rigid system. Chains react differently depending on which
direction they are tensioned thus the requirement for chain
tensioners to keep the forces equal regardless of rotation. I
will have to screw together some form of “slipper” tensioner as
there isn’t enough room to produce a tensioning sprocket.
However initial tests with a wheel brace connected to the
“input” shaft of the gear box show that it will propel the loco
forwards -but without the tensioners the “take up” of the chains
is very slow.
The postman has been busy and now have a fuel tank for my loco
along with a mounting bracket (nylon) for my engine with, (at
the insistence of my wife), a nice bright shiny purple anodised
fuel filter. I was simply going to buy a plain aluminium one
-but she liked the colour.... I did however sneak in the plain
grey plastic exhaust tubing -despite all the “nice colours” that
were on offer(!) The plastic plumbing tubing is clear -the
fluids that will flow through them will be colourful enough...
The outer shell has finally been given it coat of green paint.
This went better than I thought it would as well the loco has a
very “sad” expression on its face -I cannot explain it any other
way. After two coats of high build primer some soap and wet
sanding I had a surface that was “flat” and all the runs and
sags had been rubbed down. The sun duly came out and the wind
dropped. After some applied “rattling” I sprayed the body with
BL Rover Green 17 and I was initially horrified!!! It was FAR
too light... After a deep heavy sigh I put it back in the shed
to set overnight. I put a “post it” on the fridge which read
“buy paint stripper” The nighttime worked some kind of magic as
the morning brought a shell the was dark glossy green and almost
a perfect match for the colour I wanted it to be. The body now
needs its roof coating with “Dark Grey”.
See
Picture 91.
After a not so brief spell digging holes in my garden (more next
week...) I have some time again to play with my loco -yes it is
throwing it down again! I ordered some more bits from Technobots
-these are the 16 tooth sprockets that I will mangle to produce
the flycranks and the aluminium pulleys that will connect the
engine to the torque convertor and the torque convertor to the
gear box. Why Pulleys? These have self damping connections -in
short big rubber bands. This will enable me to have a flexible
connection between the engine and the torque convertor, direct
alignment will not be required -it makes it simpler and if all
else fails it is a easily removable module.
See
PIcture 92.
In the shot above you can see the main pulley on the engine
-which will have the rubber band to the pulley on the torque
convertor. The pencil line to the left of it shows the centre
line of the torque convertor. On the left of the line is the
linear actuator that will be used to change gears.
The nice people at “Just Engines” who supplied me with the ASP
46M engine also supplied me with the fuel, water and oil tanks.
You may remember that I was looking for something to cobble up a
water and oil tank system from, well I had to mangle part of
their “economy” range of fuel tanks and I have something that
not only looks as if was designed to fit there -it should work
too. The system has two L shaped pipes -one of which has
suffered at my hands and is now the “return” to the tank. The
feed still has it’s “clunk” at the end of it. The fuel tank is
as supplied at 8oz and the water / oil tanks are 20oz. The
dimensions for these are given in imperial and I had to beg an
imperial ruler from my wife, (Quilting is still quite firmly
Imperial!!!)
See
Picture 93.
Yes, I know it looks messy in the photo -but it will look a lot
neater once the cradles are wired in. The “returns’ are clearly
visible as are the air vent pipes inside the tanks. However they
fit nicely when taped into position and the next instalment for
the loco awaits -the pumps.
After much thought I decided to use Lucas style windscreen
washer pumps and then mangle them for my purpose. They are
cheap, (reasonably reliable), and very common. I will have to
remove the casing, (delicate work with a hacksaw...), and then
extract the vitals.
Well another wet weekend (and it is now snowing...) However on
the plus side the pumps have arrived from the Channel Isles and
I have been playing with them. The rate of flow was the first
thing I had to find out. So, after duly filling a large milk
bottle (3 Llitres ) I connected the pump up and times how long
it took to pump out 1 litre. The answer was 48 Seconds. This
gives me a flow rate of 1.25 litres per minute. The tanks on the
loco are 634 cm3 thus the entire contents of the tank is pumped
twice per minute. Don't you just love easy maths!!! The next
experiment was rather more messy -as I had to find out at what
"pressure" my pump was delivering its liquid. This was found by
simply lifting the out let pipe until no more water flowed out
of it -the slightly soggy answer was 2.35 metres. Translating
this it becomes 0.21 BAR of pressure. Once the primary
experiments were out of the way it was time to cut open the
pump....
There is a suitable gap between the motor and the plastic wall
of the pump body. It should be possible to cut some vent holes
to help cool the motor as it will be running throughout the
progress of the loco.
The next thing on the shopping list are some solder on "nipples"
for all the 1/8th inch plumbing -these will probably come from
MaccModels. The pressure from the pump is enough to force the
connections from the rubber pipe -and yes it was very funny!!!
I now have the 3mm bearings for the conrods and the brass strip
that will form them. The next step will be to produce the
flycranks from the sprokets. This is done by sawing off a
section of the sprocket and then grinding off the teeth of the
sproket. I know this sounds very wasteful -but it is guarenteed
to work. The centre of the, (now), flycrank is epoxied to the
3mm threaded insert and nutted up. This not only holds it while
it sets -but gives more mechanical "grip" to the flycrank. A
little dressing up with a few suitable sections of ABS scrap and
I have something that not only looks like -but actually will
-work!
As with the "Krokodil" the main problem will be getting the
flycranks all synchronised with the rotation. The original had
two huge long ones, then split into 4 shorter ones -due to
resonances at high speed. I wonder if I should bite the bullet
and produce a "4 Quarters" layout in which non of the conrods
are in alignment and they occupy the 90 degree sections of the
circle. I know that I have phrased this VERY badly but let me
see if I can explain it better(?)
front lhs at 00 degrees.
front rhs at 90 degrees.
rear lhs at 180 degrees.
rear rhs at 270 degrees.
This SHOULD even out the vibrational modes of the system and
hopefully give me a smoother running loco.
This weekend (26th of Feb 2011) is the Gauge '3' AGM. I have
been busy building a trainsporter. I intend to get the Fell and
the Krokodil to Biggleswade and back... Maplins supplied the
foldable trolley and my son supplied -his toy box!. I looked at
it and then dropped the Fell chassis in it -it fit exactly. I
them dropped the Krokodil body into the box -that fitted it
exactly too. Never in a million years would I have expected this
to happen and I confess that I did laugh for a few minutes at
the irony of it. If all else fails we will have a seat on the
train -by sitting on the trains...
Well nearly one week after the AGM -how did it go? I have to say
that the AGM itself went very well indeed and I could have spent
rather more than I did at the traders.... Fortunately I ran out
of money and non of them would take "plastic" -otherwise I would
have been in hock for many a month!!! I did come home with 8
sets of Carriage wheels -which will go with my 4 bogie sets and
slowly the bits for a carriage amass -more later! Getting to
Biggleswade was nasty -there is no other word for it. Since my
son wanted to do some "trainspotting" I studied the time tables
and worked out my route. If it was not subject to any delays it
would be an easy run. We got up at 3:30 am had a decent
"college" breakfast and made our way to Derby Midland station
for the 6:01 train to London. We got on board the "Meridian"
which then only paused at Leicester before hammering it to St
Pancras International -total time on train 93 minutes and the
coffee shops were open for more food for son (and an
espresso or three for daddy!). He got to see Javelins and
Eurostars and more Meridians. Even though it has been several
years -I still expect
to see the "Gasometers" on the final approach to St Pancras...
We then dragged the box across the road to Kings Cross -were
things began to fall apart -big time. I did not know that
"Kings" was being "restored" and that also -nothing worked. The
only functioning display board we could find was on the end of
Platform 8. When I say functioning I mean that it was lit
up -but that was all. When this photo was taken the time was
8:09. The roof leaked and (of course) it was raining.
See
Picture 94.
I wasn't until 8:18 am that we knew which platform the train was
going to arrive at. However what we did not know -was how to get
from Platform 8 to Platform 6. Fortunately a Porter gave a lift
on a towed "Brute" trolley and we made it just before the doors
closed.
We got to the AGM despite the map and unpacked. In case you are
wondering I do build large locomotives -even for Gauge '3'. Here
is a shot of the Fell with my sons loco the (as yet) unfinished
LMS 1831 with a GWR Pannier tank for comparison
See
Picture 95.
We left the AGM shortly after 1:30 pm and got home for just
before 6:20 pm. I managed to eat some Chinese from the takeaway
and my son promptly fell asleep after having munched his bag of
chips and he was out until 10:00 am the next day. I got him and
my wife upstairs and then shortly after 8:30mpm looked at my
wife and kissed her -and then somehow it was 8:30 am Sunday!!!
In short we had a great day out.
A couple of weeks have passed since the AGM and yesterday was St
Patricks day. So what have I been doing in the meantime -the
answer is getting very
wet! The only real way to find out if my design for a torque
convertor works well is to make a plastic sheet model and wind
it up to speed. My test fluid was tap water and my test plates
made of ABS sheet. Now there is a function of propellor design
called "rooster tailing" in which the waste energy of the
propellor flings up a stream of water in the familar "rooster
tail" of racing hydroplanes. So the amount of water that my
driver disc converts to a rooster tail shows the direct
efficiency of my design. In short I produced some magnificent
rooster tails until I got the shape that I needed.
In the original design I had nice evenly spaced holes, this did
S@D all.
I then tried slots. Better.
I then tried a series of small holes working out in a spiral, it
took me a day and half to produce, and did S@D all.
By mistake I drilled a hole that was obviously too close to the
centre shaft (groaned) and completed the series. This worked perfectly...
SO I have to have the holes for the discs close to the centre.
At the moment the only reason I can think of is that the centre
is rotating slower thus fluid has more time to penetrate the
cavity and thus create more friction between the plates(?)
What follows is a true rendition of my pet term "Kitchen Sink
Engineering"... This shows the test rig for my evaluation (!)
Yes I know it is really "Heath Robinson" but the fact is that it
does exactly what it says on the tin -or in this case -the
Muller Rice pot! The length of 4mm shaft is held in place with a
couple of notches sawn into the side and some CA. The driver
plate is driven by a power screwdriver separarted from the
driven by a spacer washer and the fluid is -used cooking
oil!!!
See Picture 96.
The next shot shows the driver plate being spun by the
screwdriver and the fluid forming the small rooster tail around
the driven plate and transferring the energy to the driven plate
and it spins with it. When the screwdriver stops then the oil
drains out from between the plates. I now understand perfectly
why there was a STEP UP
gear to the main convertor on the Voith system. As you can see
by the errm "excess oil" splatted around. The system works far
better at higher revs. The main problem being that oil is now
centrifuged off the plates -and of course goes everywhere....
The plates are made from the resident powder cast gears -it is
nice to see them actually at work rather than sitting in the
bottom of the draw.
See Picture 97.
I looked and thought -and looked and thought -and changed my
mind!!! Although I still like the idea of the Tesla disc system
I deciede to go with a paddle wheel design instead. Most of the
mayths that I didi for the tesla design still hold true for the
paddle wheel . So, after a good cup of strong tea I set down in
my shed and began the days labors.
The first thing to do was to fix the driven plate to the shaft.
Out with the carbon arc torch and dial 40 Amps on the welding
plant a couple of dabs with the stick of SIFbronze and the job
was done.
See
Picture 98.
The next step was to similarly braze the output shaft to its
gear wheel.
See
Picture 99.
The next question was how to fix the paddles to the plates? This
was simply done by using the rotary table and drilling a series
of 2mm holes around the edge. Some CA to fix the paddles
(actually 2cm lengths of brass U section), to the plates and
drill through them. The nuts and bolts then held everything in
position until I was ready to hit it with the MAPP gas and the
silver solder. By brazing the shafts before hand I was sure that
I would not disturb the joint!
See
Picture 100 (!)
The main shaft is held in a simple plate U frame with the driven
plate free to rotate on the same shaft. The output shaft is held
by a scrap piece of box section and this meshes with the central
gear on the driven plate.
The next twp shots show the assembled torque convertor.
See
Picture
101.
See
Picture
102.
The next problem is going to be a suitable oil leak proof
cladding for all this to go around in...
I know it has been couple of weeks since I last put
anything on this -but I assure you I have been busy! Now that
the torque convertor is finished -but still not "clad" I have
been surveying the market for what need to start it. I
bought myself a rechargable glow clip and a glow plug spanner.
The latter has useful nut sizes on it and it may get pressed
into service for other duties.
See Picture 103.
The next item in the shopping cart was a starter motor. I
originally designed the front oil cooler to swing out of the way
to get the starter motor to nose up to the front of the engine,
and this it does very well.
See Picture 104.
The problem is not one that I could have foreseen though.... The
problem is that my hands are really too big to fit comfortably
in the "grip space" of the starter. The starter has an oval push
pad to connect it to the battery pack and I can press this with
three of my fingers. The little finger I have to hold stiffly
out -emulating a famous species of "Invader"!!!
See Picture 105.
The other thing that really should have occured to me is that it
has to held in my left hand leaving my right hand to do the
business with the glow clip. This would not normally be a
problem for 90% of the population -however for the 10% of us of
the Sinister persuasion -doing things with the right hand
is just awful!!! YES I admit it, I own a pen with a bent nib for
writing left to right.
Another sunny afternoon spent in the safety of my shed! The
problem with the too small grip aperture has been solved with
the aid of a saw and some "polymorph" which is a resin that is
squishy when hot and rigid when cold -it is also reusable. I now
have a perfectly formed custom grip for my starter motor -no
more "stiff little fingered aliens" around here! The testing of
the starter motor to the engine has proved that I need what is
called a "cascade" battery supply. This is a set of batteries in
parallel to produce a very high impulse current. The eventual
system will use six 6 volt SLAs to produce three 12 volt
batteries in parallel. This will give me a cranking supply of
13.5Ah at 12 volts. The peak load from each chain is not really
that high but the sum of them is "enough" to turn the engine
over at more than idle speed I think(?)
Now that things are begining to form in the aether I can spend
some time on the wiring for my loco. There has to be two
independant supplies -one permanantly supplying power to the
fans and the pumps -the other supplying power to the RX
and the servos. The next problem on the list is laughable!!! The
amount of force that I need to press the rubber insert against
the coned flywheel of the engine -causes the engine to move
backwards on its mounting brackets. I have a friend coming who
will advise me on a suitable system to use -he will probably
weld the brackets up for me too!
Well the "consultation" went well and I have a good idea of how
to rebuild my support brackets for my engine. Once they have
been brazed together it would take an elephant to move it!!! I
have spent most of this afternoon playing with my coolant
system. I knew that the system had never been played with "in
anger" and it was time to see if it functioned correctly. The
first thing I had to do was to find and fix the leaks! These
were due to things "settling in" as the sealant moved during the
stresses of heating (and tighten the odd nut or two). I also had
to wire the ends of the piping as once the seals were
functioning they were the next thing that could (and did) move!
In order to check wether my seals were working I mixed blue food
colouring with the water. Since the system only functions when
it is full of water the only method of priming it is to suck
hard on one end of the plumbing... YES I have severe "blue
tongue plague". The pump gets a little warm when running -but I
think it is within parameters -I might drill a few coolant holes
in the side of the plastic casing to let a little air through
it?
The pump draws 2.5 Amperes when in full thrust thus the two
pumps will draw 5 Amperes when the system is running and the
loco is in motion. I plan for a running time of 10 minutes -but
this is dependant on the consumption of fuel -the fuel tank is
225ml or 8 Fluid ozs. I picked up my 1 litre bottle of glow fuel
and "After-Run-Oil" from the local model shop. The oil content
is slightly low at 16%, (ASP's like 18%), and the NitroMethane
more than a little high at 20%, (it is "timed" at 5%) -but it is
the only one I have so it will "do" for initial testing.
A sunny Sunday afternoon spent safely in the kitchen and shed. I
fabricated the new mounts for the engine -they are of 3mm thick
steel and use a steel re-enforcing bracket to hold it "true". I
have bolted it through the chassis members with 5mm bolts -it is
not going to move again! Now that the position of the engine
ifixed in 3 space I can get around to connecting the propshaft
from the torque convertor to the gearbox. This runs underneath
the engine and through a slot in the engine mount. There were
some very worrying moments as I twirled the propshaft and there
were loud "bangs" as the bevel gears jumped on their axles. The
reason turned out to be quite simple -but fatal for the insect
involved -some delicate brushing away of pieces of corpse then
re-oiling and the gearbox whired away niceley
This shot shows the new engine mount.
See Picture 106.
The next shot shows the test assembly of the torque convertor
and its propshaft. The belt takeoff from the engine to the input
pulley of the convertor and the output pulley of the convertor
to the propshaft. The propshaft will sit between two sets of
ball races in the position of the brackets. The brass union for
the propshaft and the input shaft to the gearbox can be seen
directly underneath the engine carburettor.
See Picture 107.
Today was the May Day Bank Holiday -so I spent it in my shed! I
had specified to the rest of the household that this was a "shed
day" so they ate their sandwiches for lunch in quiet meditation.
The first operation fo the day was to build the "cranking
handle" for the engine. When the torque convertor etc are
positioned there will not be enough room for the staret motor to
get into position -hence some form of reach rod was required. I
made this from 10mm silver steel bar -a little overkill -but it
was handy. The end I drilled out to 3,5mm on the lathe and then
tapped it to M4 on the lathe. This is the first time I have done
it this way and it was slightly scary the way the tap was
"eaten" by the steel bar. A length of M4 threaded rod was fitted
and then the free propellor shaft fitting that came with the
engine was added.
See Picture 108.
The next thing was to hit it with the silver solder and unify
all the parts. This is somewhat ugly looking result. The yellow
colour comes from the Zinc plating on the propellor shaft
fitting. If you are welding then galvanised stuff is a definate
NO -NO! but with soldering -it actually makes life easier.
See Picture 109.
I now come to the proudest part of the creation!!! The 10mm
steel bar is then punch fitted to a Hose Lock fitting and then a
3mm nut and bolt pinned the two together. The hose fitting
meshes perfectly with the yellow silicone cup on the starter.
See Picture 110.
Lunchtime and sandwiches saw the living room converted to my
drafting room. The next shot shows the designs for the torque
convertor housing. I settled on one using an asymetric roof
-because it made life easier to build. As you can see one side
of it has a radius of 4cm and the other 3cm.
See Picture 111.
Back in the shed I began to cut up the highly unwilling paxolin
sheet into the correct shape. The sawing of the straight sides
was the hard part the curves I simply ground out on the
grindstone. OK it smelt a bit -but it was quick! The ends are
made from two pieces of paxolin that wrap around the holes an
then lock together. There is a 3mm nut and bolt locking them to
the uprights on the torque convertor.
See Picture 112.
The internal space of the torque convertor is going to be filled
with shaped pieces of balsa wood. Once I seal everything with
epoxy the substance that I used to make the shape will not
matter -but the shape will...
See Picture 113.
Saturday is scheduled to be "Fell Fire Day"(!) The coolant
system has been bled to within an inch of its life and the
radiator fan has been whirring away like a top. All the power
bricks (SLAs) have been charged and everything awaits the dawn
of what promises to be a very noisy day.... I have been asked by
a few people to take a "video" of the device in action. Well as
a far as possible I will try to -but we might be a little
busy!!!
Well -I got it to fire...
Now it seems that I have to re-tighten every single nut and bolt
on the loco!!! The model is covered with oil -most of it from
the self undoing allen bolts on the exhaust manifold. A good
friend who was there to "hold my hand" commented "Is this what
is meant by self lubricating???" At the end of the first 2 minute test run due to
vibration I had "lost" 1 M4 nut and bolt and several M3 ones
-these were nicely rolling around on the sheet of ply on the
workmate . Unfortunately these were the ones holding the engine
to the bracket... The good news is that the coolant system
performed well, but then on the second test run -it blew up!!!
The fan died on me and with the coolant temperature rapidly
rising I put my thumb over the carb intake and everything
stopped. The temperature of the coolant rose and held steady at
30C during the first
test and the air temperature was a warm 23C. As per the instructions
we allowed the engine to cool -this allowed us time to retighten
all the loosened nuts. The engine had initially started with
"some words" but it started cleanly the second time. I will have
to modify my starter system though, the yellow rubber cup tends
to fly off when I pull the starter away from the running engine.
I think that I also need to investigate some form of "over run"
clutch for it(?)
The next thing on the shopping list is a shed load of lock nuts
and a large tube of thread lock!!!
The fuel clips didn't work to my satisfaction and weeped, so I
undid them and WIRED the hoses to the pipes -this held it all
leak proof. I also need to find some method of easily filling
the fuel tank when it is locked into position. At the moment I
am using a 6oz "squirty bottle" with a valve and filling the
tank that way. The engine does seem to like guzzling fuel though
-so perhaps at some later date I will investigate converting it
to petrol? The people that I bought the engine from do a petrol
conversion kit complete with "timing disc". This may be
something to think about as it would ease the types of fuel that
I would have to store -at the moment there is 2-stroke for the
rotavator, 4 star for the lawnmower and the glow fuel -not to
add the gallons of parafin for the greenhouse heaters.
The noise was as bad as everyone feared. I was the only one NOT wearing ear defenders,
and this was with the "Super Quiet" silencer...
I think that I terrified the neighbours though! A voice calls
out "The System is LIVE", there is a muted whirring from the
pumps, then "Cranking!", the starter screams and then the engine
squeals away at 1/4 throttle. The longest I have had it running
is for 6 minutes on the
second run, I have yet to follow through with the instructions
to run the engine at full trottle for 12 minutes. The crib sheet
says that "run in" will take around 30 to 60 minutes -now
everything will have to wait for a new fan and some lock
nuts/thread lock to come through the post.
Am I unhappy -NO. It was
a test run and as such it provides data on the problems to be
solved.
When I am ready to test things again I will have to do a time
temperature and voltage chart. This will give me some idea as to
the peak temperatures from the cooling system and the drain on
the batteries as time progresses. I will have redically
re-design some of the plumbing as the only way I could drain the
fuel from the tank was to blow down the silicone tube connected
to the exhaust silencer and then draining via siphon the fuel
from the tank out of the feed pipe to the carb. Similarly
fueling the loco will have to be arranged, the squeezy bottle
works -but it obviously a short term solution to the problem.
Well the fan has arrived and it seems pretty potent at 39CFM and it slotted into
the hole perfectly. I did some experiments to see if "blowing"
was better than "sucking" air through the radiator gills and
although there was no detectable difference in the amount of
heat that was extracted -I was told by "Her Pinkness" that
blowing was quieter -so this is the way it will operate.
The rest of the stuff arrived this morning, (fortunately the
weather was with me) -and it threw it down all day!!! I started
out by replacing all the nuts with nylock nuts these should
discourage the bolts from moving (hopefully!). The torque
convertor was the next victim on the list and I mixed up some
epoxy to try and seal all the edges of the t/c, once that had
set I then daubed P40 over all the places that looked as if
there was light coming through them, (and with what ever was
left I simply splatted on too...) It looks somewhat "hairy" with
all the loose strands sticking out of the resin but this is a
minor point. I got the rubber drive bands and fitted them. This
means that there is (in theory) just the one connection to go
between the engine and the gearbox -the fluid! The next shot
shows the days progress.
See Picture 114.
Famous Last Words...
The day dawned black wet and windy. Perfect weather for an
excuse to go and play in the shed with. I had been experimenting
with my pump for the torque convertor and had got it to work
both forwards (flood) and reverse (bail) quite naicely with the
oil mix that I was going to use. The only problem is that with
the viscosity of the oil the pump will not "self prime" as it
would with water -so there is going to have to be some form of
detcor to say when the T/C is nearly empty. The pump is of the
gear displacement type and once the oil has reached the gears it
squirts it quite effectively.
See Picture 115.
K&S provided the brass sheet that was to produce the wrapper
for the T/C and some slight persuasion with the slips formed the
correct curve. This was duly edged with some brass L section and
this soldered onto the curved sheet.
See Picture 116.
The wrapper was then wrapped around the body of the T/C and
secred with a handy vise while thegasket and screws were fitted
over the next two hours...
See Picture 117.
Nedless to say several "words" were uttered whilst trying to fit
the wrapper and the new scars from the raw edges dripped
nicely(!)
The "pressure test" as such consisted of a length of pipe
connected to the T/C and it filled with oil and it withstood 1
metre pressure of oil -this being the longest length ofsuitable
pipe that I had. The fiiled T/C was then connected to the chuck
of the lathe via long rubber band and it twirled nicely at both
ends. Everything was set for "GO!"
I drained the T/C fitted, it into position ,"dodged" the flying
yellow rubber cup thing (I really must do something about that!)
and the engine fired. Nothing seemed untoward so I hit the
switch to "flood" position -the cable to the T/C pump was 2
metres long and I was standing behind the corner! There are some
sounds that stick in your memory -before I lost my hearing there
was one sound that was as unique as it was lethal. This was the
sound of "Disc Dive Motor Head Bump" which is the sound of a set
of read heads on a disc platter bouncing from top to bottom and
knocking chunks off of the surface. I could describe it as the
sound of someone feeding a musical box mechanism into a
meat grinder.
IT was THAT sort of sound.
The engine stopped.
I peeked around the corner and there was nothing visibly wrong
-there was a gap where the T/C should be and it was visible on
the floor beside the loco. The rubber band was "somewhere"
and oil was EVERYWHERE!!!
I initially though that one of the driven plate turbine
blades had come off -but after examining the debris showed that
all of my turbine blades were still firmly fixed to the plate.
What had happened was as the oil filled the T/C the turbine
blades hit the oil and the shock of the impact snapped pieces
off of them. I had a good cry in my beer and did the only thing
possible -I went on the Internet and I asked for help!!!
Help arrived by 9pm the same day. The problem IS a type of Oleo
Kinetic Shock -but the shock is due to the foaming of the oil.
In short CAVITATION. The method advised by "The Person from
Glasgow" is to use a modern synthetic and to feed the oil from
the centre outwards. This spins the oil up to speed as it fills
the T/C. Draining it will involve some valve and plumbing work.
He is also going to provide some "suitable alloy" to make the
turbines from. Yes -the nymphomaniac metal... I have never tried
to work with it but I am told that it is "interesting" to
say the least.
I have been drafting some designes for the new T/C and I am
pleased to say that there are very few pieces of waste paper bin
fodder so far. Now that I have nothing to do whilst I sit and
design things I have turned my attention to the exhaust
system. This is a "Super Quiet" silencer supplied by the
nice people at Just Engines who also supplied the engine.
They do give some help in altering it to be quieter on their web
site and I e-mailed them to check out an idea that I was looking
at.
The "Super Quiet" is made in three sections, it is held together
by one huge long bolt and the end of part 3 is THREADED
-the lock nut at the end simply stops anything unwinding. It has
one baffle plate of six holes with vanes and is located in
position "B". They do suggest moving it to position "A" and you
can experiment with flattening the vanes. Another suggestion
they give is to partially obstruct the exhast vent in Part 3
with an M4 bolt. What I have done is to manufacture a second
baffle plate (in short a simple piece of brass sheet with seven
4mm holes!) Located this in position "B" alongside the supplied
plate in position "A" and also fitted a section of M4 threaded
bolt through the exhaust vent. This will no doubt RUIN the "top
end" of the engine but since it is never likely to hit that in
it life I feel quite confident that all will be well.
Well at the end of the second afternoon of test running I am
happy to say that the engine was "noticably" quieter. It still
sprayed oil over everything
but I did not lose too many nuts and bolts this time -but I did
lose some from different
places this time... This series of tests was designed to examine
how well my cooling system was now running so -for the first
time we had an "endurence" test. The first run was two minutes
long and the initial readings from startup were:
1: 24 C
2: 27 C
After a pause to tighten up the ehaust system and see where all
the new nuts and bolts had come from... I gave up on using a
gasket for the engine to exhaust system, (bits of it were just
splattered around anyway! ) and simply applied jointing compound
to the two cleaned faces and tightened everything up solid. This
seems to be the best method as they were still tight at the end
of the second run.
1: 32 C
2: 34 C
3: 36 C
4: 36 C
5: 36 C
6: 36 C
So, looking at the results it appears that my cooling system can
keep the motor at a constant temperature of 36 C. This gives me
something to work on. I now have a steady state temperature of
36 C and a rate of flow which is 1.2 litres per minute. This
means that the radiator is cooling the volume of liquid over it
at the same rate regardless of the volume -in short this means
that I can reduce the volume of the tank. In theory I could
dispense with the tank -but perhaps not yet!!!
Tomorrow (Saturday) is scheduled for another test run. I have
been busy unscrewing and dabbing "blue threadlok" and
"stiffnuts" over and in any threaded item that I can get at! The
aim of this series of tests is to see just how low a rev rate I
can get on my engine and (possibly more importantly) how lean a
mixture I can get it to run on. I do expect to have to play
"start rich" and "run lean" games for some time. I have put the
air filter on the engine for the first time. I confess that it
does look somewhat comical like a large yellow clowns nose
perched on the end of its rubber elbow. The reason for fitting
it at this time was that beforehand I could look down the
venturi and gauge (roughly) at what point I was at -min to full
throttle. Now the aim is to set the throttle arm and leave it.
The restriction through the air filter will alter the mixture to
be richer in any case.
For those interested in more information on the loco there has
just been published June 2011 a 4 page article in the magazine
"Traction" which you might find useful. Previously to this I
only knew of an article in the magazine "Back Tracks" from Feb
2009 -which is more detailed and goes into the design and
technical aspects of it more.
When I began my thread on "The Gauge '3' Society Forum" I used
the old schoolgirl skipping chant by way of introduction:
"I do not like thee Dr Fell,
The reason why -I cannot tell,
The only thing I know (and know
full well),
I do not like thee Dr Fell."
The article in "Traction" begins with the same refrain... (And I
thought I was being so
original!!!)
Looking at the pools of oil left after each series of test runs
-annoys me. There will have to be some form of "scrubber" on the
exhaust stack -or at the very least a drip tray... This does
raise the question -what do I do with all this oil? I am not au
fait with Klotz -but it does seem to be a very good lubricant
even when it has been through the engine. This does raise the
problem of oily rails so it does seem to be "needful thing" that
we have to contain the spent oil, (somehow).
Well possibly the worst series of stupid problems in the world
happened...
It began with the starter rod pin shearing.
Then the flywheel nut unscrewed.
Then the drive pulley on the starter motor unscrewed.
And then to cap it all -the drive belt fouled the motor and had
to be cut to pieces ....
I looked at Martyn, he looked at me and we both said; "Enough
-lets go make tea".
So we did.
After having had a good cup of tea and several days having
passed I have decided to abandon the starter system that I am
currently using. What the new starter system will be is as
follows, and to be honest it is a genuine cobbled together mish
mash. I saw off the prop shaft adaptor on the starter rod. I
then bore out a 14mm 1/4 inch drive socket to a bore of 10mm and
I braze it to the
starter rod. This gives me a rigid fixed connection. To the
other end (sawn off) I fit a "one way bearing" as it is
now called, but I still want to call it an "over run clutch"
-with an 8mm bore. This then provides a secure fitting to the
driven end of the engine. Once the engine fires the over run
clutch dis-engages and the starter motor can be switched off
with the engine still running then (hopefully) the 14mm socket
uncouples from the M14 nut without it flying off -somewhere!
The parts are on order and I await them in the post.
Tuesday morning the parts arrived in the post. I now have a nice
satin finish Chrome Vanadium 14mm socket and a 14mm hex nut "one
way bearing" for it to connect to. I hunted around the scraps
draws in the shed and extracted a length of 8mm rod, filed the
ends smooth(ish) and slipped on the bearing. As promised the
bearing went around only one way. So far so good -I thought. I
drilled a 3mm hole through the end of the rod and affixed it in
the spade jaws of the engine and nutted it up. I tried to turn
the engine via the bearing -and the nut came away in my
fingers... Needless to say -I was not best pleased!!! The nut is
at best a friction fit to the bearing and it could have just not
seated before it was engaged to the bearing. I whipped out the
CA and now it is VERY firmly fixed to the hex nut....
The next shot shows the hex bearing sitting on the length of 8mm
rod.
See Picture 119.
Having looked at the 14mm socket I am wondering if I could
simply cheat and drive it from my battery drill/driver. The
thing does have enough torque to turn over the engine and it is
a lot simpler than all the faffing around with milling and
brazing -plus at the end of it I have a 14mm socket that I can
use elsewhere. Hmmmm....
See Picture 120.
Having slept on the subject I decided that as the socket was
only £2.54p I could afford to buy another one -if I needed
one. So, the socket went to the lathe and then the fun and games
began. The first thing I tried was a 10mm TC end mill this cut a
nice groove through the curved edge of the socket hole and then
stopped... the next thing I tried was a 10mm Cobolt drill. This
cut a nice conical shape into the socket -and then stopped...
Hmmm... I seemed to be able to get a nice cut with the
edges of the cutter and then as the amount of cutting surface
increased the amount of pressure that I could get on the cutting
edge I hit a point of no return -in which the amount of force I
could apply with the screws was not enough to make the cutter
bite into the metal. I unlimbered my "ultimate weapon" as
regards holes.
I fitted an 8mm boring bar to my head and slowly opened it out a
few millos at a time. After the course of an hour I had removed
enough Chrome Vanadium steel to be able to get the entire cutter
width of the boring bar into the hole. NOW things began to
move!!! Now you will have noticed that using an 8mm boring bar
in a 10mm hole is "a bit tight" as they say -and I will be the
first to admit tht the head did bung up with chippings and stall
the lathe on far too many occassions.
See Picture 121.
After some polishing at very high speed I POURED Superglue into
the gap and left it in the jaws of the lathe to set over an hour
or so -time for lunch(!)
See Picture 122.
The next shot shows the completed starter socket and the lathed
to size stub end of the over run clutch fitted to the "business
end" of the engine.
See Picture 123.
So -what has been happening in the the weeks since the last
entry to this? Well actually quite a lot -but very little to
report on. The starter system works and the engine has done
endurence runs of 20 minutes. The peak temperature reached by
the coolant was 43C with an ambient air temp of 23C at that
time. The superglue did eventually give way and I was forced to
do something that I really should have done on the first day
-BRAZE the two parts together. The engine is now well "run in"
and I have been able to reduce the mixture down to something
that is both livable with and runable with! During the running
in process I was using a mixture so rich that it gulped down
250ml of fuel in 5 minutes.
Now that the engine side of the system seems to be ok and is "a
known unknown" I can get down to the business of building a
torque convertor. Rather than oil I have decided to use water.
There are two reasons for this -the stuff is cheap and easy to
get hold of... I did "jig" my torque convertor for SAE 30 oil
-but this is in short supply so Aqua Impura is what I am going
to use. Since I am using water I can simply use two propellers
back to back -non of this tedious mucking about with discs and
moduli of fluid surface tension. The simple fact is that I was
too ambitious and should have gone for the simple option
straight off -there learn from my mistakes!!! I ordered two nice
brass props from an internet supplier who promptly told me they
were out of stock and I would have to wait two weeks for them to
be shipped (ha!) from Holland. I then got two props from the
suppliers but with the wrong thread size. Needless to say this
did not make me happy and the sales person on the other end of
the line was made very aware of the fact. They did offer to
replace them if I sent them back and they would order another
pair from Holland and I could have them in another two weeks. (I
told them, "No thank you" -even if I did not use quite those words...)
One of the props will have to be bored to 5mm to take the input
shaft. The other will have to be bored to 6mm and then a "blind"
hole of 10mm cut in the front of the prop with an end mill for
the ball race to fit into. This "locks" the driven prop directly
in line with the driver prop from the engine. There will have to
be some return plumbing in the cavity to allow the water to
circulate around the props. Another advantage is that being
water the props are designed to cope with a low level of water
as they are lowered into the water. Thus there "should" be no
problems with cavitation. The top speed of the props is 4,500
RPM and I "think" that I have the engine running at below that.
So, nearly six weeks from the last real session on model
building -what have I been doing on this loco. The sad fact is
not a lot -but we do have a nice new kitchen, a new back door
and ground floor "facilities"...
Now that it is safe to begin work in the kitchen again(!)
This details the production of Torque Convertor mark 3.
Having aquired two propellors (see above) I then began to
mangle them to requirements. The first prop was brored out to
5mm and a 5mm bolt soft soldered to it. The bolt them went into
the lathe, centre bored with a Slocombe and a 3mm hole bored
20mm down it.
See Picture 124.
Into this hole a length of M3 threaded rod was inserted, this
does not connect at either end and simply acts as a floating
support and centres the two props to each other. There is a
spacer nut which is CA'ed to the thread to prevent contact
between the prop blades. The other prop was bored out to 6mm and
a M6 threaded bar inserted into it and then CA'ed solid in the
hole. The end of the hole then had a 3mm ball race fitted for
the 3mm threaded end from the other prop to slide into.
See Picture 125.
See Picture 126.
See Picture 127.
THIS version of the torque convertor is going to be made with
6mm thick polycarbonate. It is thick enough for bullet proof
glass -so I have hopes..... After the usual nasties of trying to
saw polycarb -i.e. keep the saw cold and play a stream of water
on it to make sure it doesn't think about getting warm. Polycarb
melts at around 80C which is easy to produce when sawing and
drilling it -sometimes the only way to extract your drill or saw
blade is smash apart the thing you are trying make with a hammer
and start again...
After due dictionaries of "words" and "pleas" I got my pieces
cut and rough CA'ed then together.
See Picture 128.
The next step was the drill, (carefully), the re-enforcement
holes and then tap them for the M3 bolts that would hold the
sides whilst more treatment was administered to the casing...
Tapping Polycarb is WORSE than tapping Copper. You forever live
with the perpetual fear that what you are tapping into will melt
and entomb your tap in it!!! I cut it 3mm at a time and "rested"
the tapping Tee tool on a bock of ice. I may have suffered third
degree frostbite -but it never stuck in any of the holes.
See Picture 129.
After having poured the best part of a bottle of thin CA around
all the joints on the inside to both seal and strengthen it, I
left it outside overnight to cure and it was ready for rough
treatment to come...
See Picture 130.
As you can see none of the sides is the same height -despite my
best efforts with a saw(!) I needed to produce a parallel face
to the open face -this meant putting it in the milling machine.
It is said that there are "Lathers" and "Millers" I am
definitely a "Lather" as me and the Mill do not get on at
all!!!!
After some improvisation in the clamping dept, it may look wild,
but nothing would move -which was the point.
See Picture 131.
The edges were trimmed to height with an 8mm end mill and filled
the shed with the evil polycarb snow that gets everywhere... A
little further along and the casing now sports M3 tapped holes
on the driven face and the driven prop has been test fitted to
see it is clangs against anything when twirled -it didn't
fortunately.
See Picture 132.
The final shots show the casing with the two props test fitted,
(the driven prop clangs slightly and will have to be trimmed to
diameter on the lathe).
See Picture 133.
See Picture 134.
The raw shafts will sit in ball races and have to have thrust
bearing fitted to them. The next stage is the plumbing part and
the problem of sealing the front plate. I do have gasket paper
but I think something a little more robust might be needed(?) I
keep looking at the pink silicone rubber pastry rolling mat but
it might be a step too far in the domestic scavenging dept...
A summer evening has been spent playing with water.... I
connected up the Torque Convertor and plumbed in the pump -which
fed from a large jug of water. The Torque convertor leaked all
over and to be honest I expected it to. BUT on the plus side it
leaked very little from the face plate and mostly from the
bearing holes.
See Picture 134.
The driven side is turned by the wheel brace clamped onto the
shaft and the system turns quite nicely thank you!
See Picture 135.
See Picture 136.
I did some "Classical Physics" experiments with the the T/C as
it was being pumped. To one side I affixed a 1 pound kitchen
weight and then to the other side a 1 ounce one. Both
weights were connected to their pulleys via lengths of "button
thread" and the 1 pound weight dropped over the side of the sink
onto the floor. The 1 ounce weight rose a few inches in the time
that the weight fell to Earth. Thus I can say that for every
pound I put into the T/C I get 1/16th out. This multiplied by
the gearbox gives me 90 ounces for every pound that I put into
the T/C. I think the loco will move...
This raised the question -exactly HOW much force do I have to
put into the gearbox to get the loco to move? After some due
experiments with weights and thread I can state that the gearbox
will move under 2 ounces of weight and the loco will move under
4 ounces of weight. Translating this into "real world" figures
that I can think with...
The pulley wheel is 2.5cm diameter thus 1.25cm radius.
The force on it is 4 ounces = 133 Grammes
Thus the torsional force is 133 / 1.25 Grames per Centimetre =
106 Grammes per Centimetre = 1.08
Newton Metres
Which is not a lot and it shows how well the ball races are
working.
Working back through the maths the loco will produce 1.08 / 5.8
* 90 = 16.76 Newton
Metres
I think this loco will pull a couple of carriages...
The next step will be to produce a cylindrical shape inside the
T/C as at the moment it is simply square. (I suspect that Joule
was right on the money here and there is a lot of rotational
enegy loss due to turbulence...) There being an extreme dearth
of correctly sized tubes available in the house (i.e. none) I am
going to have to cheat... There are two ways to do this; EITHER I take a short length
of drainpipe and saw it into four quarters and stick them on the
inside corners to produce a "cylinder like" shape for the water
to move around in. OR I
cut some sheet brass and roll a tube from it. The first option
is by far the easiest and the second will produce the most
accurate shape. Which ever option I choose the corners will
require "packing" with P40 glass fibre putty to hold them in
place and to stop the corners being distorted by the force of
the water on them.
After having "sat down with a cup of tea" I am going to use the
first option as it is by far the simplest.
The next thing I have to do is to "calibrate" my new T/C. The
only thing that I have that I can say with decimal point
accuracy in the rotation dept is the lathe. I do have an 80mm
chuck that I can bolt onto it, (or the far heavier 100mm chuck),
to provide me with a flywheel rotation at any speed from 100 RPM
to 2,000RPM. All that is needed is a large rubber band(!)
Delving into my collection of bits draw unearths an old belt
from my turntable, it is a genuine brown fuzzy felt covered
Dunlop SystemDek 3 that I have had since 1982 -and I wouldn't
swap it for anything. It
has undergone a few mods since I have had it such as; The Origin
Live mods to the RB300 Arm, The Origin Live DC motor, The
Goldring Elite cartridge (I used to have a Coral MC82
-unfortunately it just simply wore out....) The replacement
vinyl platter mat for the felt one and the matching record
clamp. This has added about 2Kg to the mass of the platter and
makes it just dreamy to
listen to!!! The only problem I have with it is the brown nextel
suede is wearing off and it looks more than slightly bald in
places (a bit like me!)
Anyway -back to the the T/C.
I know the rotation of the chuck to a decimal point per minute
-thus giving me a linear speed for the 80mm chuck.
I know the diameter of the pulley (25mm) and thus the
multiplication factor = Pi x 80/25 = 10.05
So it should be an easy task to ramp up the T/C to the full
4,500 RPM, (450RPM from the lathe chuck), that the props are
capable of, and hopefully nothing bursts!
Well I found a length of 60mm ABS pipe and split it into 4
quarters and then stuck the pieces into the corners of the the
T/C. This gives me a roughly cylindrical shape to play with. The
next stage is to pack the cavities behind them with P40
fibreglass putty. Once this whole lot has set solid it then goes
into the lathe and I very gently try to produce a cylinder from
it with the boring bar. I did spend some time this morning
looking at my loco and the shell that is going to have to go
over it. The idea is that I hinge one end (the one without the
engine!) and once the engine is running the whole shell then
flops forwards on its hinge and away we go...
I tried four times to get the shell to fit....
It seems that despite all the measuring the shell lips are "just
that bit" tight. So it is going to be out with the sanding block
to take off some of the edges on the sealing lips. Once on the
shell fits like a glove -but unlike the glove I seem to have a 3
fingers, a thumb and big toe on my hand!!! This is because I
tend to produce "exact fits", and once something moves -or has
to move in an arc -it just catches. If I engage reverse gear and
drop the shell absolutely vertically on the frame -it fits. What
is going to have to be done is that the lips of the shell are
going to have to be ground to take a hinged arc motion. (sigh).
The epoxy set and cured overnight and a slow slightly squeaky
time was had spinning the T/C on the lathe with the boring bar
"doing the business" on the inside of the unit. I smoothed out
all the lumps and bumps and took a smidgen off of the sides too
(oops!) The gasket paper did work and here you can see the sheet
stuck to the face of the the T/C -I punched the holes for the
allen bolts to go through later.
See Picture 137.
The modified T/C is now being tested with weights and thread
-and so far it has been quite an improvement. I knew that I was loosing some
power due to turbulence -but to be honest I did not think it
would be that much!
The first trial showed that the 1 pound weight was now capable
of lifting a 2 ounce weight at speed and then a 4 ounce weight
showed that it would be suspended at any height. This shows that
the now cylindrical interior of the T/C has improved the torque
to the ouput shaft between 100% and 400%.
I have worked out the amount of torque that I am getting out of
my T/C -now I have to work out how much POWER I am getting out
from it.
Converting everything to SI (!) shows that the driving weight
0.453Kg takes 2.3 seconds to fall 0.8m.
This gives me that the input power to the T/C is [0.453 x 9.81]
x [0.8 / 2.3] = 1.54 Watts
The output power from the T/C is [0.056 x 9.81] x [0.8 /2.3] = 0.19 Watts
Therefore lost to heat is [1.54 - 0.19] = 1.35 Watts
Pumping rate is 1.2 litres per minute thus the volume of water
per second is 20ml thus the water leaves the T/C raised in
temperature by [1.35 / 20] = 0.06
degrees Kelvin. Which may not sound a lot but I do now have some
idea of how much heat I have to get rid of through the T/C
cooling system. The modified surface cooler made from the
Pentium II heat sink plus its 40mm fan should be enough. The
problem is the volume of water that I have to pump into the T/C
to fill it and then pump out the T/C to drain it. Whilst this is
happening the loco is still being propelled!!! I think I have a
workable system using solenoids to hold on a parking brake which
draws its power from the "drain" setting of the pump and the
best place to engage the "parking brake" is I feel on the main
pulley from the T/C.
I am starting to run out of internal space rapidly....
What would be nice is to have a air bleed to the T/C so that
when it empties I only have to pump out that volume and when it
fills the same thing happens. I have an idea from the infamous
"automatic brake bleed valves" from late fathers Westminster...
I should be able to fabricate something similar from a couple of
springs and two viton rubber spheres(?) Once this is working I
then need to do some timing experiments on just how long it
takes the pump to fill and drain the system. A timer cobbled
from a NE555 should be able to do the drain time thing and then
(once empty) it could then allow you to trigger the fwd/rev gear
change. The fwd/rev would have to be a flip flop of some kind,
as there would be instances where you simply wanted the loco to
stand still -but still be in fwd gear.
Right. What you see below is my "doodle" of what the system
should do (I hope!)
See Picture 138.
I have ordered the bearings and relays from my suppliers and it
is up to "Her Pinkness" to cut felt washers to pack the sealing
glands that the input shaft and output shaft will have pass
through. I have no idea how leak proof greased felt is -but I
can devoutly hope that it is better than what I have got at the
moment. Yes there is a reason why the experiments are being
carried out on the drainer of the kitchen sink....
Having used a device called a "Rotary Evaporator" in my College
days I do know that centrifugal force will cause the fluid to
cling to the sides of the the T/C as it operates and the air
will be forced to the centre -and thus not leak (too much). I
actually wish I had a Rotary Evaporator nowadays, the main use
of it during my college days was to produce what would now be
called "Bio Diesel" from the used cooking oil from the chip shop
by the College. The fryers wondered what I wanted it for -the
other students wondered how I got such good milage. It had to
stop when the HOF thought that we were distilling moonshine and
he refused to believe that it was DERV that we were purifying!!!
I am still debating whether to use oil or water in the T/C. One
of the problems I think I might have with oil is that the props
will convert my oil to dirty whipped cream -this was a problem
that bedevilled early expreriments with turbochargers... I have
this horrible feeling that the only way to know -is to try it.
The bearing etc have arrived via post -and today was the day
that was scheduled for my Retinopathy examination. So, most of
the morning was spent in a dark room recovering from the
Tropicamine drops they put in my eyes. It stings like hell, and
for about 4 hours afterwards everything feels like you are
looking into a search light... But after lunch things had
returned to normal and I started work. I bored out the
shaft holes in the T/C to 10mm and inserted the roller races
then came the thrust bearings each side of the hole and
(finally) the grease that should seal everything (hopefully!)
The rather slimy result is shown below.
See Picture 139.
The next stage is to hack the hole in the chassis work of the
loco to fit the T/C. I have this feeling that I am going to have
to take a hacksaw to my chassis rails....
At the end of a Sunday "in the shed" I have fitted T/C version
3.... (And if this doesn't work I give up and go and cry in a
corner). To fit it I did have to saw through my chassis rails on
my loco -this does mean that one end of it is now errrmmm
suspended in mid air whilst I figure out how to re-enforce it. I
do have some wood that will probably be epoxied in place. It is
like its two predecessors a "sidewinder" installation -how much
this will affect the balance of the loco I do not know. You can
clearly see the sectioned chassis rail with the bottom of the
T/C now level with it.
See Picture 140.
This shot shows the front of the T/C. The output shaft is held
in place with an external ball race. The output from the T/C
pump is shown as the input side is now below the level of the
deck. Both of them are equi-radial from the centre and
(hopefully) once the fluid is spinning will ingress and outgress
smoothly without too much foaming…. Once the fluid has exited
the T/C it should then flow into the surface cooler and then
back to the holding tank.
See Picture 141.
Tomorrow evening will most likely be spent soldering the
electrical switching system for the pump and solenoid for the
the T/C. Last Saturday I bought a few metres of "hook up" wire
to begin building the loom for the loco. It is now time to take
apart the lash ups for the cooling system and start hard wiring
everything in place. The next step after that is to begin "hard
plumbing(?)" the pipework into the frame. The next series of
tests on the loco will concentrate on the T/C -should it survive
first filling.... There was a recent "Mythbusters" in which they
used Sodium Bicarbonate to break open a Gaol door -the pressure
was less than 5 pounds per square inch. I have no idea what the
pressure will be inside the T/C once it is running. It is made
of 6mm thick polycarbonate which from my experience will take a
Grade 2 lab explosion (100 Newtons impact) -my safety glasses
are made from 8mm thick polycarbonate!!!
The question is -how much pressure will 6mm polycarbonate take
with 3mm thick stainless steel bolts through it?
My son has asked me what would happen if I pumped a non
Newtonian liquid into the T/C -such as custard. I have to
admit that frankly I have no idea… The theory is fine as the
custard will act as a solid and still be liquid enough to be
pumped -but I think this is a non-starter. There are better uses
for custard -like with crumbles and pies!!!!
I will do the initial experiments with water as I have said and
then switch to oil.
The test control panel takes shape! It will sit above the gear
box and the silencer will sit behind the raised section of ply.
The breakout tags are fixed to the ply with M4 nuts and bolts
and ye olde piece of Vero Board has the relays mounted on it.
They are from left to right; the fwds relay (DPDT), the reverse
relay (DPDT) , the Fill T/C relay (DPDT) and the Drain T/C relay
(DPDT). The switches are; the gear lever (DPDT centre off), the
T/C pump control (DPDT centre off) and the main power switch
(DPDT).
See Picture 142.
How exactly I get the gearchange to function is at the moment a
puzzle... What the motor has to do is on the face of it is very
simple -turn clockwise and then turn anti clockwise. There could
be a simpe cheat (and I like cheats!) that uses a pick up track
split into three sections. The two small end sections have a
blocking diode on them arranged so as to face each other. A
travelling pickup runs along the pick up track and the motor
draws its current from that. Once the pickup has travelled into
the facing end section the current stops. Once the current is
reversed however... The diode now passes current to the pick up
and the motor moves into reverse. Hopefully the following doodle
will give you the idea.
Well I have had a busy day in front of the soldering iron and my
eyes ache from straining to see things at such short distances
-even with the magnifying glass from the set of "Helping Hands"
and my reading glasses. Anyway this is the progress that was
made on the Friday evening. You can see that the power feed for
the relays as started to work its way across the tag strip -the
main Positive feed for the board is on the lower right hand side
below the bolt.
See Picture 144.
This is the underside. On the extreme top right you see the fuse
holder -this is normally the thing I wire FIRST... The main on/off
switch is simply being used as a SPST -but by having a DPDT
witch in there to start with I can tag other functions onto the
same operation (if need to). The same is true of the Gear
Change switch -although all four outputs are wired in I only use
it as a SPST . The pump switch is always going to be SPST as all
it does is connect relays -I can extract other functions from
this switch via the tag strip connections.
See Picture 145.
The next shot shows the status at lunchtime. The gear change
Relay is wired in and so are both of the pump relays. The Gear
relay is wired BLUE and YELLOW the pump relays are wired in
GREEN and YELLOW. I know that I normally use Blue and Yellow for
DC of variable polarity but this simple colour change will
(hopefully) stop any cross connections. The solenoid side of the
"fill" relay and the gear select enable side of the "drain"
relay are not wired in yet.
See Picture 146.
By tea time I am the proud owner of technicolour squid... YES I
know the cables are far too long and I can easily shorten them
-the reverse is not true!!! There are a few connections still to
be made, such as the interlock for the gear change relay, but
since this will be a static test for the forseeable future -I am
not too fussed.
See Picture 147.
Well after another day at the front... I tested the newly built
control panel and found that, (somewhere), I have a major short circuit -Hey Ho!
So, in the spirit of "Oh
What The Hell" I unshipped the T/C
and began testing in earnest. Some Lucar
connectors and a length of twin later I
can report that the fill & drain system works and that
whilst under test the T/C did not
explode -although the seals on
the rotating parts will have to be rethought....
The pressure inside the T/C was only 1.3 BAR and whilst it
worked the amount of fluid loss was on the order of 75ml per
minute and the volume of the ATF fluid container is only 250ml.
This may not sound a lot but the jet of fluid was at least 4
metres high and quite painful when I tried to stem it with my
thumb. The tests had to be stopped at 3-4 minute intervals to
top up the fluid tank, (red coloured water). The bottom of the
garden now looked like an explosion in ketchup factory. Possibly
I think again on how I circuit the fluid -there was a drip
collector system on certain locos to deal with just this
problem. I have to ask myself would it be a good idea to have
the T/C simply vent to air and dribble down into a holding sump
and pump from that? (Hmmm....) On the face of it that is not a
bad idea and I will investigate that further over the comming
evenings. It would involve changing what the feed sequence is to
the T/C and cooling the fluid before
entry into the T/C -but on the other hand this does mean that I
can have a finned sump
which might be a better option for cooling(?)
I am going to have to replace the M3 cap heads with epoxied
lengths of M3 threaded bar and then produce a squishy rubber
gasket as the std gasket paper with sealant compound that you
use on the the car would not seem to be up to the job. This will
enable me to apply rather more torque to the nuts on the face of
the T/C than I can do at the moment. If I apply too much I will
simply end up with M3 sized holes as I completely strip the
threads there in the Polycarbonate.
The grease felt seals at the shaft holes do not seem to be up to
much "cop" either... Repacking them and tightening them has
reduced the dribble to a weep. What did make the most difference
was lowering the voltage to the feed pump from 12V to 6V -but
the amount of fluid that is going to have to go through the T/C
means that the higher voltage will
be needed. The tests with the kitchen timer showed that it took
8 seconds to fill the T/C -this gives me some sort of idea of
how long to run the pump in reverse to empty the T/C whilst
still having fluid in the pipes. On the shopping list for the
next dollop of pocket money is a simple NE555 timer kit.
All in All a good first test!!! THIS IS THE FIRST ONE THAT HAS NOT EXPLODED....
This morning was spent trying to find the "short" on my new home
made control board. I actually found it quite rapidly. That's
the problem... I unlimbered my much loved and ancient "Smiley"
AVO selected "Diode & Continuity" and applied the probes. I
got an instantly lit red neon "short" across any part of the
wiring. This I was expecting and some delicate work with a
soldering iron unsoldered three connections. Two minutes proved
that the short was not on the coolant feed relay system. A
further unsoldering removed the gear change system from the
list. The last thing on the board was the T/C pump system and I
was very puzzled at this... I poked at the terminals and got a
red light... I switched over to resistence mode and all became
clear.
I had ZERO ohms between N/O and N/C connections on BOTH relays.
I had in short (!) two duff brand new relays...
One failure I would accept, but two failures on two relays -in
the same mode??? I am not buying this brand of relay again!!!!
I re-soldered the red leads on the board to the coolant and gear
change systems and then tested them and (hoorah!) they worked as
expected.
After a quick trip to the local MAPLINS I have two relays that I
have replaced the duff ones with. The main problem is of course
that the leads are too short to fit them properly so I can see
Saturday Morning being spent -re-wiring the wiring... They
"clunnng" quite nicely and the flood / drain system now behaves
as expected. Tomorrow I am going to visit the local B&Q to
find some silicone sealant to see if that is the cure for the
errrm "slight leak" from the T/C. I will build a sump system for
the T/C as I think that this is my only real option -so at the
moment I am building a shopping list of bits for it.
The next step once I have got a working control board -is to
build the gear change system and its interlocks. This will
entail replicating the "hand tuned", (well it sounds a lot
better than being chewed to the right shape with a rasp),
wooden reach rod with one possibly made from metal. The linear
actuator, (GOD I could really get in trouble with Trades
Description Act from here on in), moves the reach rod through
the gap and the pin causes the ballista to move the bevel gears
across the graft gears. Well I gave in with being squirted at by
th the T/C and returned to the problems of wiring and changing
the gears. The gear change system has sat on the side of the
kitchen for a few weeks now and to be honest it is way past it
being fitted!
So, Saturday afternoon I sat down with a set of needle files an
slowly worked "the shape" into a length of deal which would be
the reach rod for the gear change "ballista" pin to be forced to
move against. I did try brass but it was far to unco-operative.
In its simplest form -it is a wedge... However due to the
requirements of moving and locking the pin the actual shape is a
delicately flattened 'S with a hook at the end of it. And if it
is not absolutely true -it BINDS. It took nearly two hours of
"Ho Hum" fitting and getting the reach rod through the gap until
at last it lifted the pin across and then latched into position
"like a good un". Now that I had a smoothly contoured shape it
was time to fit the gear change motor to the reach rod.
This sits at an angle
to the reach rod. As the motor winds the saddle down the thread
the angle between the pin on the saddle and the reach rod
becomes less as the reach rod is wound into the slot of the
"Ballista". This makes the reach rod move initially fairly
slowly and then at the end of its movement quite rapidly to lock
the pin into its new position. The reverse action uses the hook
and to grab and trap the pin against the slide thus slowly
forcing the bevel gears across the graft gear. The screw action
of the linear actuator provides a rigid termination to the
movement.
I know it sounds complex but it is actually very simple!!!
See Picture 148.
Some slightly painful experiments showed that the reach rod "in"
position is now designated as REVERSE and the motor has BLUE as
positive. The stall current on the motor is about 6 Amps and it
changes gear in about 0.8 seconds. There were a few ouch moments from hot cables
and the motor casing!!!
Now that I have a working gear change -despite the protests of
the motor. I then test fitted my control panel It is shown here
only supported by two of the bolts and it sits directly above
the gearbox -so I have to build some form of protective cage for
the gears.
Yes, the leads are too long, however this is easily fixed with
the soldering iron -and the may not be that much too long once
they have been threaded into the loom and thence to where they
have to go.
See Picture 149.
It is actually quite a simple board -the duff relays required
the entire pump drain system to be rebuilt (ugh!) and their
replacements are a little large in comparison -but they function
well and that is all that I require of them.
The rest of my weeks time is going to have to be finding some
way of making a seal on the power shafts of the T/C. Greased
felt does not seem to hold much above 1 BAR so I am going to try
to greased leather and we will see what happens.
After having had a good thought about it and slept on it I have
decided to go out and buy some commercial oil seals. Getting the
correct sizes proved to be far easier than I thought and the
only problem I have at the moment -is the fact that everyone is
on holiday. Since the same is true of the supplier that I will
have to use to get the finishing things for the T/C, I am not
overly concerned. The only seals that will fit are (typically)
far more expensive than the normal ones used in the car world.
They are both 16mm diameter and I am not sure if I have a mill
of that size -but I definately have a "spoon" bit of that size.
The polycarbonate is 6mm thick and the seals are 7mm thick. This
means that I can legally sandwich the seals between two sheets
of polycarbonate -I will have to cut an inset into the side of
the T/C and then construct a flange with an inset to sit over
the seal and then (somehow) stick the flange to the side of the
T/C. This as they say is going to be "fun"...
Polycarbonate is the least forgiving material I have ever worked
with -so I will have to make sure that everything that I use is
icy cold and wet. There is a good chance that I will get even
more wetter than I would get testing the T/C seals that I have
made... But on the plus side I do feel that I am on the homeward
stretch with this loco. I have gone through -"OMG I have had enough of
this thing!" and the "Whatever made me try this?" to the "Now I am
getting really bored
with it..." and finally into the "It might actually work?!?!?" frame of mind.
The last loco I built, (The Krokodil), I kept a track of all the
expenditure that I did and showed that it was possible to build
a 1 metre long Gauge '3' loco for under £200. This week I
will pass the £500 barrier, this makes it the most
expensive loco that I have ever
built and I think that I will be another £200 before I see
it running on the tracks...
Having sorted through my tubes of mills I do infact have a 16mm
end mill -still I am ashamed to say in its sealed and oily
cellophane wrapper... So it must be New Old Stock from 2008 when
I bought the mill(?) I now have to figure out how to do a "line
bore" on the T/C to produce the two 16mm wide and 2mm deep inset
holes on the sides. The easiest method is (I think) to fit a
length of 10mm rod to the chuck of the end stop head and align
the hole with the chuck. I then swap out the 3 jaw on the lathe
for the collet head, fit the 16mm collet and then gnaw the inset
into the polycarbonate -there is no other word for it! The outer
shells will be dead easy -but messy. I simply carve a piece of
polycarbonate sheet to the right size , true it up on the lathe,
and then bore the 6mm shaft hole and the 16mm inset hole.
Everyone in the family has been told "Tomorrow Daddy is in the Shed"
so apart from having to do the usual meal times I will not be
disturbed. It is going to be one of those operations where it
will take 2 hours to do
the set up and 20 seconds
to do the cut. Murpheys Law
states that the required 20 seconds will not be granted me!!!
Well "Shed Day" started at 8 am with a good breakfast inside me
and a couple of mugs of lethal coffee. When I went to Uni
I was given a kilogramme of coffee by my mother and told not to
use it all before I returned home in two weeks time -I didn't
manage it... After having delved through my draws and boxes,
(thank you IKEA!), I had all the parts infront of me to do the
transformation of the lathe. I unshipped the 80mm 3 jaw and
changed over to the 32mm collet chuck. Some mutterings later I
had the collet chuck aligned in the face plate. I make "patch
marks" on the chucks to allow me to align them easily, some
people dot them with a centre punch -I file V grooves. It
is easier to feel for them when you cannot see the location
holes... I have a mixture of 3 and 4 stud chucks to my name and
if you align the patch marks then everything slides home, the
next problem is getting my not too dainty fingers to the nuts at
the back of the chuck(!)
The next shot shows the vertical slide has been fitted along
with some very "custom" mounting studs to hold the work piece,
(i.e. the T/C), in the correct axial plane. This was established
by fitting a 10mm mill into the collet and then sliding the T/C
over it and then bolting the T/C to the vertical slide -thus the
position of the hole is perfect!!! As my late father used to say
-"You have to live in some bad houses before you get a home" and
yes POP this is definately "a bad house". The 10mm is then
swapped out of the collet. This he says so very glibly... The
entire sequence is: fit the bar to the chuck, fit the crowsfoot
spanner to the teeth on the collet cap. SQUEEZE both together to
undo the cap, slide the mill out of the collet. Now COMPLETELY
unscrew the cap from the chuck and then using "the knack" pop
out the collet and replace it with a 16mm one. Now reverse the sequence...
See Picture 150.
The whole thing is then wound towards the mill at rate of 1.25mm
per revolution of the crank until it touchs and then "backed
off" a turn. Once the lathe has spun up to speed (about 400RPM)
the T/C is then pushed into the end mill and after about 5
seconds -we have a 16mm concentric stepped hole for the oil seal
to sit in. Given that the oil seal is 7mm thick I opted to cut
4.3mm into the wall of the T/C and the rest into the sandwich
plate. This shot shows the stepped hole into the side of the
T/C.
See Picture 151.
A length of scrap polycarbonate gives me the two sandwich
plates. I would apologise for the repeat custom stud mounting,
but it works -so don't knock it!!! Here I bored throigh the
polycarbonate into the balsawood backing piece with an 8mm mill
and then swapped that over for a16mm one. The whole thing is
then slid across on the saddle and the process repeated.
See Picture 152.
A saw cut later and I have two plates to play with. A length of
M8 threaded bar and a couple of nuts gives me a shaft to spin
the scrap on and within the time it takes to boil a cup of water
I have two outer seal covers. I have to say that I HATE working
with polcarbonate -the one thing that redeems it in my eyes is
the beauty of the finished article. The smaller of the two fits
at the driven rear end and the larger at the front.
See Picture 153.
Ok a little shopping spree later at the local MAPLIN store.. I
now own five 1N1001 diodes for the gear change motor system and
a little NE555 timer kit for the not unresonable sum of
£7. Could I have sourced the parts and built the PCB
myself for that? I doubt it. So I am happy -as is the
exchequer!!! I am used the Vellemann kits and this is a kit from
a Uni here in the Uk. It lacks the detailed instructions of the
Vellemann kits -it seems to assume you know what you are
doing...
See Picture 154.
So it looks like another Saturday Morning at the soldering iron.
Saturday Evening has proved that it was correct! I now have a
working timer board set to 7 seconds, (or thereabouts). I had to
swap over the switch for the gear change to the one for the fill
drain system -not a great problem. The reason is that the switch
I was using for the gear change was a DPDT centre off and I only
needed a SPST centre off. The other pole on the newly installed
fill drain switch will connect to "Trigger" on the timer board.
The idea is this. The NC side of the power relay allows power to
the pump then the NC side of the relay on the timer then chops
it off. So that regardless of the direction of current, I only
get 7 seconds of power to the pump.
See Picture 155.
Sunday is occupied by "other things" as is Monday &
Tuesday -so it looks like Wednesday will be the next time I will
get to play....
Well having grabbed some "free time" whilst the children party
in the orchard... I have been devoting some thinking time to the
problem of the Radio Control of the loco. I need 4 channels to
cope with the amount of fiddling that it requires. I am going to
buy a new 2.4Ghz R/C system, (I already use one for the NER
EE-1), from the same people who sold me that one -the laugh is
it is the same device but at £20 is only 40% of the
original price... I need two sets of R/C controlled relays that
can be made to Latch. This is a requirement to prevent sore
thumb problems. The normal R/C flight layout are two joysticks
-the left hand one having a graduated north/south setting for
the throttle. I plan to use the right hand joystick to control
the actions of the Torque Convertor pumping system. Thus
east/west will be fill/drain and north/south will be gear change
fwd/rev. The joy of latching relays is that it takes TWO flicks
of the joystick to flip the state of the circuit. Thus I flip
the joystick (east) to fill the T/C and then flip it (east)
again to stop the pump and engage the solenoid parking brake.
Similarly the drain system (west) operates and stops via the
timer -the west position is set to non-latching. The north/south
option on the right hand joystick is latching thus locking the
gear into either fwd or rev. What I do with the east/west option
on the left hand joystick -I am not sure(?)
Today has been one of those days in which you look around for
someone or something to MURDER....
I found out why I seemed to be blowing fuses all the time. The
batch of 10 Amp fuses I had got from Maplins were a complete
waste of time. I tested the fuses BEFORE I fitted them and found
that I had 7 out of the 12 that were there -already "gone"
before I fitted them. I have spent nearly two days in careful
and fruitless circuit dissection trying to find the non-existant
short.
I then sneezed and yanked the daughter board with timer circuit
on it. Needless to say it hit the top of the main power
capacitor and the copper cladding on the PCB is no more.
Deep
Breath...
I packed everything away and went shopping in town for the bits
to replace what I had just scrapped. I will now replace the fuse
holder with a circuit breaker. Even if it trips it will reset
once the circuit is dead -or it has cooled again. So I might get
bored with the waiting but at least I will be safe(!) One
problem I have found with the timer board is that the "trigger"
is an on/off setting -so I will have to work out some method of
producing this once the drain option has been selected. (Is this
what I use the east/west option for?)
Well after a Saturday and Sunday in the shed I decided to try
out my new improved Torque Convertor. I did a static pressure
test up to 1 BAR and nothing leaked -I was amazed!!! I placed it
into the loco and wound up the system to speed with an electric
motor (after having removed the glow plug).
1,000 RPM holding pressure 1.4 BAR
2,000 RPM holding pressure 2.25 BAR -idle speed on engine
3,000 RPM holding pressure 3.7 BAR -Torque convertor lowest
"connection" speed
4,000 RPM -the system failed at
a new point!!!
There is no need to test beyond 4,500 RPM as this is the
maximum RPM on the prop design(!)
The plumbing system made of flexible polypropylene piping burst and I was thoroughly
sprayed with home made ATF.... I had decided that the leaks from
the front plate of the T/C were due to a lack of "depth" of seal
-so I CA'ed a wrapper made of strips of ABS sheet around the
edges of the T/C and then wiped the gasket seal around the new
lip I had created. Despite the fact that I had to use quite a
few tissues and paint brush cleaner to repmove the sealant from
the lip of the T/C (and me!) it seems to be a suitable solution
There is nothing new in that happening, but it does mean I am
one step closer to the actual running of the loco. If I can sort
the plumbing problem out over the next few days I will have a go
with a live test from the engine at the weekend. Closer
inspection of the ruptured pipe in the morning showed that it
failed at a point where it had been stored in a "kinked"
position i.e. the pipe had been folded back on itself. Thus I
presume it failed at a point where it had been pre-stressed (?)
Well I didn't get to fire the Fell over the weekend -I had other
things of a domestic nature to take care of (ugh!). But out of
it I have the parts to produce my parking brake system. This
consist of a couple of gear wheels -with the teeth lathed off
and some ply CA'ed to them to act as frictive material. This is
also secured by a couple of recessed 2.5mm bolts -so if the CA
gives way it will still hold together on the bolts.... The
solenoid sits alongside the output shaft from the T/C and will
pull on a "fork" fitting to push the two pieces of ply together.
The solenoid has a convenient pair of holes that I tapped to M3
-despite their unwillingness to do so. The fork fitting is
perhaps one of my finest pieces of self invention cum
scavenging! It is the modified spanner that came with the IKEA
light fitting that was put up in the kitchen on Saturday
afternoon. The solenoid has a spring from a large biro as the
return mechanism, (would I lie
to you???) The spanner will be hack sawed to length
later.
See Picture 156.
See Picture 157.
See Picture 158.
The last shot shows the ATF tank to the right hand side. This
and the Glycol coolant tank are going to have to be replaced in
the "running" engine.This is because I bought the biggest tank
that "Just Engines" sold for test puposes. Now that I know how
much coolant fluid and (soon) how much ATF I will require I can
buy smaller ones. People have asked me "Why Glycol?" It is inert and
doesn't have one of the more embarrasing problems of water -it
has a wide freezing to boiling point range and it actually has a
higher specific heat capacity than water. The mixture that I use
is a commercial one for liquid cooling CPU chips PC's. Ideally
the stuff that I would like to have pumping around my coolant
system would be a saturated solution of Sodium Nitrite which
used to be used for storage heaters -but with the amount of
aluminium and copper around I doubt that either of them would
last long!
I have decided that Sunday afternoon will be the time that
I test the torque convertor with the engine running and the wind
it up past the stall rev point into the solid phase point and
then over the other stall rev point. This could be very dangerous(!)
The amount of energy contained in the torque convertor is
frightening -in some respects it is like an angry captive genie
in well shaken bottle... Something that you need but dare not
open!!!!
The figures are simply nasty;
At 1,000 RPM The centrifugal acceleration is 33.54 G and the
centrifugal force is 42.11 Newtons.
At 2,000 RPM The centrifugal acceleration is 134.18 G and the
centrifugal force is 168.44 Newtons.
At 3,000 RPM The centrifugal acceleration is 301.96 G and the
centrifugal force is 378.99 Newtons.
At 4,000 RPM The centrifugal acceleration is 536.75 G and the
centrifugal force is 673.76 Newtons.
At 4,500 RPM The centrifugal acceleration is 679.33 G and the
centrifugal force is 852.73 Newtons.
By Sunday tea time I should know whether I have a runner or a wreck...
It is now Sunday tea time and I am proud to say -WE HAVE A RUNNER!!! OK -so
it might not move much, (the drive chains being disconnected
etc), but the engine did fire after some very dire threats and
the torque convertor did not leak too much. A forum member had
requested that I make a video of this titanic event and I had to
admit to myself that I had never used YOU TUBE in my life... So
I opened an account and readied my tripod and camera. Then it
threw it down and I had to rest everything in the corridor
between the sheds -Hollywood never had this problem. I did make
a small video sequence using my Mac and iMovie and uploaded it
(it was quite easy to do!) Anyway here is the sequence:
The first part shows the glycol coolant system running
-essential to any fire up -note that the pump and piping throb
and the fan flappers move in the draft from the fan. (They are
actually important -as it is very hard to tell wether the fans
are working when the engine fires...) The second part of the
sequence shows the torque convertor flooding and draining (note
the rising level of red liquid). The third part shows the engine
"on song" with my finger and thumb very painfully holding the
throttle at the point of solidus -when the torque convertor
begins to turn. Too much throttle and everything vanishes into
blurs and grease gets flung EVERYWHERE -too little and the thing
just sits there and laughs at you. Getting it right plus holding
the camera by hand was not achieved without dire words to all
parties involved. The first time I nearly got it right and then
the camera battery went flat...
See video
sequence fellpic159.
Well after dire threats I managed to drill a suitable expansion
hole in the torque convertor. The one problem that I expected
was heat.
And this I got in barrow loads. What I did not expect was HOT FOAM. The dissolved air
in the fluid was basically beaten out of the fluid and the
cutesy baby pink froth returned to the reservoir. I had expected
some "whipped cream" effect -but to be honest not a complete
bubble bath...
Since the foam is compressible, whilst the fluid is not, very little happened whilst
it was "foaming at the mouth" so to speak. So, what I have done
is to drill a port at 45 degrees from the front face and
inserted a length of 4mm pipe to as close to the centre of axis
of rotation as I can. Now, (hopefully), when the torque
convertor fills and froths, all the froth will rise to the axial
centre of the torque convertor, and thus be expelled by the
rising fluid. This does mean that I will have to block off the
drain from the torque convertor and thus force all the froth and
fluid through the central port until I get "clean" fluid coming
out of it. How this will be done during the test tomorrow
afternoon will be via applying by hand a pair of surgical clamps
on the tubes. I know how long it takes to completely fill the
torque convertor cavity -so it should be possible to set a timer
circuit alter a valve to change over from axial port to radial
port after the cavity is filled.
More Work (sigh)...
The loco uses a marine engine -thus the blue liquid is a glycol
water mix with "Fernox" corrosion and algae inhibitor. This is
pumped from the reservoir tank to the lower part of the cylinder
head and thence to the radiator (which is fan cooled). The
radiator consists of a XP1700 CPU cooler with a polycarbonate
plate bolted to it through which the coolant is passed via a
thin film. Not unlike the surface radiator system that cooled
the Supermarine Spitfire. The red liquid is automatic
transmission fluid (home made) and this is pumped into the
torque convertor to engage the gearbox and thus the amount of
fluid in the torque convertor varies the rotational speed of the
gearbox. The used ATF is then cooled via a modified Pentium 2
heat sink (as above) and returned to the reservoir tank. The
pinky yellow liquid is a home made glow fuel and is 20% castor
oil, 20% Nitromethane, and 60% Methanol. I will admit it is very
messy and I will rebuild the engine to use petrol and spark
ignition when I get around to track trials as it throws used
castor oil all over the place whilst running.
The next step of the development will be to fine tune the rate
of flow through the torque convertor and the fluidus to solidus
aspect of the ATF. This involves hot and cold viscosity tests.
The viscosity of the ATF is determined by dropping a ball
bearing down a glass tube and timing between two marks 30cm
apart at the set temperatures (293K and 333K). Too thick it will
not pump. Too thin it will not act as a solid at rotational
speeds. More canubra wax thickens it, more rape seed oil and
aniline thins it , (stir gently over heat...) The result from
the kitchen smells not unlike a strange mixture of "cherry" shoe
polish and frying chips(?)
I have three batches of fluid ready for testing tomorrow. YES
-The Goldilocks Principle...
One is as thick as I think I can pump it.
Two is a solid as can make it between the viscosity
temperatures.
Three is a 50/50 blende of One and Two...
Well at the end of a massively bad afternoon -here are the
results...
One -was too thick to pump -it may in fact have been this that broke the pump...
Two -was "neither here nor there" from the standard mix that I
have been using. Maybe it "came on" faster than the normal stuff
but is difficult to tell(?)
Three -was maybe better at pumping and it felt "softer" than Two
as it came up to solidus.
So, I now need a new pump for the torque convertor -this is not
unexpected as the poor thing has been sorely abused since its
purchase!!! It is is of the twin gear type and well it is
missing a few teeth -rather like me...
Right. E-Bay has sold me a new pump for £4 plus P+P. And
after the weekend I should be back in business.
I have to return to my plumbing problem, now that it works all
the silicone plastic piping has to be replaced with curved
copper with solder fittings -rather than "wiring them tight".
This is where things get to the level of "inspiration"!!! I will
have to cut away part of the re-enforcing sheet of plywood
underneath the curve of the bonnet and, (more than likely), cut
away part of the longitudinal spars that hold the bonnet shape.
I am going to have to remove the torque convertor and mark out
the positions of the return plumbing. Both end faces of the
convertor housing will then have to be cut into and a 13mm hole
bored into it. The problem is getting all this plumbing into the
confined space. The return is going to have to go from the LHS
of the front face to the RHS of the rear face. Thus the fluid
will move across the faces of the prop blades in the reverse of
what they would do. This sounds strange I know -but work it out.
The fluid will move around the cavity in an orbital motion from
the outside to the inside and everything moves at a steady
state. Which is all well and good -but what happens when we wish
to accelerate or deccelerate? The return allows a flow of fluid
to move directly across the blades -thus they now act as
propellors rather than turbines. Once the fluid is moving at the
same rotational speed, the return passes no more fluid, and the
system resumes steady state.
Assembling the return is going to "interesting"...
The parts that will be inserted into the cavity faces are going
to hasve to be epoxied into them. The first bend and its two
short sections of pipe will have to be silver soldered together.
The top connecting pipe and its two bends will have to be silver
soldered together. The two unions between all three parts will
have to be soft soldered together. This is not going to be "fun"
by any stretch of the imagination. I will have to somehow melt
plumbing solder near polycarbonate and not get it to melt... The
obvious ploy is the mother and father of all soldering irons
-the clamp on plumbing type.
Sorry for the delay in getting back to you... But there have
been some domestic developments that take priority!!! I am now
in the middle of converting the shed thus I cannot get any work
done on my loco. The shed is sealed tight and (hopefully) not
too much concrete and brick dust will penetrate into it. I
expect to have to clean and oil etc everything that is in there.
BUT at the end of the process I should have a shed with a nice
insulated pitched roof and a plumbed in RADIATOR inside to help
keep me warm. You know when you are getting old when you want to
work on your loco but the prospect of sitting in near zero
temperatures while you do so -makes you head for the fire and
throw some more coal on it!!!
So, have been doing some armchair research and bought a
couple fo books from "the deadliest catalogue in the world".
This is the one from Camden Miniatures. It contains all the
books you never knew you had to have.... Out of the current
issue I have bought Hydraulic versus Electric -which is a
study of the development of diesel traction in the UK.
Chapter Four is devoted the Fell loco. It contains the ONLY
picture I have ever seen of the Fell Gearbox in situ. The
inverted Y shape of the gearbox is plainly visible. My gearbox
is "sideways on" as for me this made it far more easier to see
that everything was working correctly. The chapter on the
"convertor wars" was very interesting from my viewpoint as it
clearly shows what they were using as a transmission fluid
-DIESEL FUEL OIL(?) There IS a logic to this as the amount of
fluid you need would be parasitic to the weight of the loco and
the diesel would burn better if pre-heated.
Note: This
page has been converted to raw HTML to help it load faster.
To Be Continued.
