OO Planet - Development

[What About The Bee]

When last updated, Planet's chassis and footplate was complete.

I indicated that the next phase would be the gear train, followed immediately by the oscillating handles.  

In prior posts, the tender, with a Hanazono motor boggie, proved to have fantastic tractive effort. Planet will have the Hanazono motor as well, because a Hanazono motor will match an identical Hanazono motor! So two motors will not only let Planet have all the steam engine motion, Planet and Tender together will be a fantastic puller.

Herein is the first issue. The Hanazono motor boggie has 10.46 mm diameter wheels at the drum. Planet, with a properly scaled OO rear driving wheel, has 20 mm diameter wheels at the drum. I cannot simply place the 14 tooth worm gear that is driven by the Hanazono motor worm on Planet's axle. That would be entirely too fast. The Hanazono wheels at 10.46 mm diameter are 32.861 mm in circumference, Planet's wheels at 20 mm diameter are 62.831 mm in circumference. If the two axles spin at the same rotational speed, Planet will be 1.912 times faster than the tender, on linear rails.

After a tremendous amount of playing around with gear trains, I have found a gear train which is very close indeed. In terms of one motor rotation, the Hanazono boggie turns the wheel axle 25.71429°, providing 2.347219 mm of linear travel. In terms of one motor rotation, my gear train turns the wheel axle 13.46939°, providing 2.350852 mm of linear travel. Planet will advance 0.003633 mm more for one motor turn than the Hanazono motor boggie tender. Planet will be pulling the tender, the tender will not be pushing Planet.

If there is any interest, I will gladly explain the gear train, an intensely mathematical experience. I will say that I did enjoy it, quite the problem to crack!

With the gear train mathematics resolved, I set about drawing all the gears and axles required. Here they are, presented in linear fashion.

Yet without a Planet shell, there will not be a way to confirm that the gears do in fact fit inside the shell. Here is the crude shell, noting that a detailed shell will come much later.

Of course, I must also include the Hanazono motor and worm. Presented here with the shell.

The electric motor is simply too large to fit in the boiler. Yet standing up in the firebox is a very near fit. I could just extend the firebox by 1 mm to accomodate.

In order to make them fit inside a shell, the gear train must fold up upon itself. Trigonometry to the rescue, each gear mesh pair is a polar coordinate system, with R fixed (radius gear a + radius gear b) and theta the only independent variable. Therefore, each polar coordinate system stacks on the previous polar coordinate system.  

And then the gears and axles are folded up, using the polar / cartesian coordinate transformations.

The gears do modestly interfere with the top of the boiler shell. Lifting the shell by 1 mm eliminates the interference

The gears sticking out of the bottom will be covered up by the inside frame members. The eccentrics will be mounted to the red axle, and the undercarriage eccentric rods should be quite visible to the casual observer. These will eventually drive the oscillating handles on the backhead. So the steam engine motion will be presented!

Some self criticism.

1) A nominal involute gear mesh will transmit ~98% of the input torque.  Yet the gear train illustrated here has 7 involute gear meshes. So 14% is the expected loss for a ball bearing system for the axles. Sleeved bushes are envisioned here. Expect double to triple losses, just to overcome the sliding friction. Perhaps as much as 42% loss in torque, just for the gear train. Hopefully less!

2) The oscillating handles and other pretty bits of the motion will add further frictional loses. Whilst unlikely, there is a non-zero chance that the entire system will present too much friction for the motor. Low probability, but not zero. If it stalls the motor, it can overheat, melt coil insulation and fail. I am reminded of the fantastic tractive effort of the Hanazono boggie, so fingers crossed on this topic. I have no way to calculate this. It will either work. Or not.

3) When the tooth count on an involute gear falls below 17, independent of the module size, the tooth becomes undercut and is weakened. 6 of my 11 gears have a tooth count below 17. This weakness is mitigated by the tiny torque output of the motor and my intention (for now) to use brass gears. Not only will metal gears present greater strength, but also add weight directly over the driving axle

4) The smoke box is entirely empty. This is an excellent place for a keep alive and / or a Bluetooth decoder.

~~~~

Close enough to consider it viable. Onward to the oscillating handles!

Bee 

[Rana Temporia]

Do you intend to electrically link the loco and tender? It would help to make sure both motors run more smoothly and avoid excess strain on each if power was cut intermittently to one or the other.

Also, will you be 3D printing the gearbox frame to ensure accuracy in the gear meshing?

[What About The Bee]

Hello Rana

Excellent questions. One of them I am still studying, the other I have a direct answer.

As to a 3D printed gear box frame, that is a definite yes. Once all the components are resolved and I have a firm, complete mechanical solution, the gear box frame will be resolved. I took extra care in the development of the gears and axles to insure that they didn't interfere with each other. I do want sleeved, metal on metal bearing surfaces for a low Coefficient of Friction.

When it comes to the electrical subsystem, I am far from decided.

I will be using Rumford type axles and wheels. The axles are metal, as are the wheel rims. The body of the wheel is plastic, non conductive. The Rumford system has conductors you can add to bring power from the rim to the axle, but then only one wheel per axle can be in circuit. Two wheels with conductors on an axle is a dead short. So in essence one wheel on the forward axle and one wheel on the rear axle will pick up power. Yikes!

The Hanazono motor boggie picks power from all four wheels, something I need to think about for Planet. I could bypass the Rumford system duplicating the Hanazono system.

Linking them together electrically, like modern Hornby Rocket, Lion and or Tiger makes the world of sense.

But all of this is just a jumble for now. It gets worse. I am currently using analog DC control. This is why I took extraordinary effort to speed match these two mechanically. DCC would permit a decoder, but lacking experience, I am uncertain if I can run two motors from one decoder or I need separate decoders with a unified pickup system. Why bother with a gear box speed match if I can run Planet and Tender as a two locomotive consist, speed matched electrically.

Far from decided Rana. I'm still puzzling over it. Hopeful of some guidance from the more knowledgeable and clearly more experienced lads here. I have time before I need commit so I'm not in a panic.

Bee

[81F]

Given that planet would not need much pulling power, have you considered using a drive band rather than gears?

A traction tyre from a Steam loco might make a suitable drive band?

[What About The Bee]

Hello 81F

That isn't a bad idea at all. The size and quantity of gears is concerning. I could create pulley sizes for a perfect speed match using a belt.

The common complaint I hear is that the chuffs from the sound decoder do not match the wheel rotation. I have seen this myself in videos and it is disconcerting.

Therefore, I want the handles to beat in perfect synchronization to the wheel rotation for my Planet. It is going to be a slow locomotive after all.

The wheel axle and oscillating rod axle have coupled identical gears. Eccentrics on the oscillating rod axle will cause the handles to move. Keeping them tightly coupled is a must.

But the velocity matching drive system need not be gears. Something to consider when I try merging all these major assemblies.

Bee

[What About The Bee]

Another major sub-assembly is now sketched. It comprises the valve gear Robert Stephenson designed for Planet.  As this design is a first generation, all of the valve gear that we know so well is not present. If you ever would have liked to see how Planet’s valve gear works, then you are in for a treat.

Eccentric Rods

The eccentric rods are driven directly from Planet’s main axle. The stick out from the front of the smoke box and are shaped oddly. Note the through hole and the lower notch.  Above them are two axles, mounted co-linearly. These axles operate independently of each other. Three supports are present for the two axles. The center support provides support for both. The eccentric rods are shown properly quartered. In the image the eccentric rod on the right is further in towards the face of the smokebox.  

Clutch

The eccentric rods are supported at the front through holes by a pair of rods. They are connected to a pair of axles at the top of the smokebox (purple) and thence to the footplate. The engineman can manipulate each rod, independently. When the engineman pulls the lever, the eccentric rod is lifted. When the engine man pushes the lever forward, the eccentric rod is lowered. If you look closely at the image, you will observe that the rods are not parallel. This is because they swing from the top, at the purple transfer mechanism. The two blue rods on the front of the smoke box will appear to “walk”.

Eccentric Rod Follower

Mounted to each axle is a lever, which I term the eccentric rod follower. These will follow the eccentric rods, but only when the clutch lever is pushed forward, lowering the eccentric rod onto the rod follower. Note that the eccentric rod cannot come out to the left or right. Nor can the eccentric rod retract far enough to ever fall out of the rod follower. The eccentric rod is always held captive by the follower. It can be engaged or disengaged from the eccentric rods, only.

Slide Valve Levers

We have two axles, now driven into oscillation by the eccentric rod followers. Stephenson adds a lever to each axle, so as to drive the slide valves. The boxes on the slide valve rods are effectively a Scottish Yoke.

Therefore, as Planet’s crank axle drives the wheels, the eccentric rods are driven. If the clutch rods are down, the eccentric rod followers will do just that, follow the eccentric rods. This causes the axles to rotate, which drive the slide valve levers. This motion is transferred to the slide valve rods via a Scottish Yoke. The admission of steam will cause the pistons to move, driving the crank axle.

But where are the levers on the footplate?!?!

Note that the axles on the front of the smoke box protrude just a bit beyond what seems necessary. Stephenson adds manual control levers on those ends

Stephenson adds four control rods, which lead to the footplate

The rods connect to corresponding levers, which cause the oscillating levers to move. When the engineman requires, the clutch control rods are pulled, lifting the eccentric rods from the eccentric rod followers. The oscillating levers on the footplate can be moved which drive the corresponding axle, which in turn drives the corresponding slide valve lever and then slide valve. Viola!

And now the Tension in the Model

The firebox assembly was tweaked, permitting it to house the motor. In doing so, the clearances to the rear wheel were present, but only just so. When the blue manual control rods are added, an overlap is present.

Note that the lower of the two manual rods goes THROUGH the firebox. Yes, that is not going to work!!!

Two Possible Solutions

The first is to retain the motor in the firebox, yet rotate it 90 degrees around a vertical axis. This shrinks the width of the firebox, but lengthens it considerably. In my view, it appears visually jarring. If you have ever observed the reproduction at the museum, or in images, you will see it.

The other solution is to place the motor in the smoke box, but instead of a vertical arrangement, lay it down horizontally. This provides us with the appropriate firebox length, yet it comes with an issue. The bottom corners of the motor will poke out from the boiler. In this image, light red. I can disguise this with the underframe and enclose it in a black box. The carriage, springs & etc will further disguise it, yet, it will be there.

The model continues!

I’ve been at this since early February, with the first animations of the oscillating levers presented here on 19 Feb. Up until now, each subassembly has been approached as if each was the only assembly, with little thought as to how they are integrated. The next major step will be the integration of the motor, the involute gears, the Stephenson Valve Gear and the shell. The artistic, non-functional elements like the steam dome will be added once a functional assembly is complete.

Fabrication

And yet one more dilemma! Some of these parts are ridiculously tiny. I am not yet sure what part of the Stephenson Valve Gear will be operative, and which parts merely static display. An example is the purple clutch, that clearly can be static. But the eccentric rod follower? If printed, it will be entirely too fragile. I could use the 3d print to cast resin objects. Resin cast would be strong enough. Alternatively, what about etched copper plates? So many questions and internal conflicts!

Onwards!

Bee

[Rana Temporia]

If this can be made to work it will be fantastic. Are there any ready made metal parts from other sources (model ships, aircraft, clocks?) that can be substituted for prints at the high stress points in the mechanism? Or would you be happy with something slightly overscale to add strength where it’s needed? Please keep us updated.

[What About The Bee]

Hello Rana

It may have slipped by. I certainly didn't mention it. The smoke box door isn't a door! It is simply a bolted on plate.

You need not worry, I will continue to update, even as I encounter difficulties or problems. I find the LMR and Planet quite interesting. This model has revealed details I would not have known otherwise.

At the moment, I am leaning towards etched copper plate for the tiny bits. But frankly, that is far from certain.

Bee

[bill7437]

Is this taking model railways TOO FAR??? I admire tenacity but I think this

is mildly ridiculous.

[ModelerXYZ]

Great work Bee.

bill, there is no such thing as taking model railways too far. My own interest is in running trains and building lights. I also like permanant way engineering. I have little interest in building my own locos or waggons. I am happy with using ready to run models. Bee clearly enjoys working out how the LMR locos worked, so he can do that.

I will support bee if I can, but my interest is in other aspects of railway modelling.

XYZ

[Rana Temporia]

The moving parts on this small loco are a vital part of the appearance when in motion and thus need to be modelled to get the full effect. It would b like modelling locomotion with the motion on top of the boiler static.

[What About The Bee]

Hello Bill

I always attempt to explain, in detail, what is presented. The usual result is a large body of text. It is always possible for a reader to miss a critical statement. Here is one you may have missed: 

~~~~~

"I am not yet sure what part of the Stephenson Valve Gear will be operative, and which parts merely static display." - Bee

~~~~~

I've no issue with valid criticism of a model. I do take on board your comment, the absurdity of the situation is not lost on me.

The tiny motion bits will produce correspondingly tiny motions. For example, the total stroke of the slide valve rods is ~1.6 mm, roughly 0.06" ( ¹/₁₆"). Will that be observed?  After all the expense and effort, will that tiny wiggle be worth it?

Yet the oscillating handles on the footplate will rotate +/- 10°, clearly and readily apparent. Head on over to youTube, and watch the Planet replica move. The most obvious part of the steam engine motion are those handles. Please do go watch the video. What do you see?

Of course, it is entirely possible I missed why you think this "ridiculous".  Would you care to expand on your views?

Bee

[bill7437]

As I said I admire your tenacity but fear this is really NOT for the AVERAGE model railway enthusiast who looks at this forum.

[threelink]

I'm not sure what an "average" modeller is but I for one find this whole thread fascinating so please keep posting, Bee.

[What About The Bee]

Hi Three Link

An average modeler? That is a person far and away better than me. I am still a novice!

I look around here, and I am continually amazed at the level of prowess on full display. It is inspiring to see and something to strive to achieve. A goal, tantalizingly just out of reach.

Bill, may I suggest that a rising tide lifts all boats. 3D printing and CAD modeling represent the future. I would dare say that many here have produced excellent models using this, including Hornby themselves!

You need not worry Three Link, my intention is to continue with my development and posts about it.

Bee

 

[What About The Bee]

Whilst this component will absolutely NOT appear on my OO Planet, I do feel it is necessary to complete the description of how the Stephenson Valve Gear works. Why Robert Stephenson did not patent this is beyond me. It is extremely clever, permitting a locomotive to move either forward or reverse, at the whim of the enginemen.

Rocket had to be rolled about manually until the admission of steam would cause Rocket to move in the desired direction. Yet Planet simply required footplate controls.

If you have been following along, I have described in this thread the eccentric rods, and the manual control levers. You may also recall some chatter about a pedal on the footplate and 3 positions. As a brief review, the footplate pedal selected forward, reverse and neutral.

In the image, we see the green levers which connect to that pedal on the footplate. Pressing the pedal moves the eccentrics from left to right, along the main axle. The red eccentrics are essentially a hollow sheave wrapped around the main axle. When the eccentrics are pressed against Detent 1, the admission of steam into the ports at the cylinders is in the order 1,2,3 and 4. When the eccentrics are pressed against Detent 2, the admission of steam into the ports at the cylinders is in the order 4,3,2,1. Obviously, when the green control lever holds the red eccentrics in the middle, neither detent is engaged. Thus the admission of steam into ports is manually controlled by the oscillating levers, and the order of steam admission is completely at the discretion of the enginemen.

Bee

[DavidBigcheeseplant]

I find this thread interesting both on early locomotives and CAD design. Instead of putting the motor in the loco it could be placed in the tender and drive the loco via a cardan shaft to the loco, allowing more weight to be placed in the loco. Have you thoght of using High Level gearboxes.

[What About The Bee]

Hi Big Cheese

Having a motor in the tender is a good idea, and one very well underway. The Hanazono motor boggie has been tested for tractive effort and current draw, all whilst under a 3D printed tender.

https://uk.hornby.com/community/forum/planning-planet-343735#post-344650#post-344466

Thus, the motor in OO Planet need only drive the mechanism. Any contribution to tractive effort will be a bonus. Nice to have, but not necessary.

The Hanazono motor boggie is also providing requirements to the gearing seen earlier in this thread, as I wish to speed match OO Planet to that tender, under DC.

Bee

[threelink]

HI Bee,

Glad to hear of your intention to keep posting.

For those of us without such skills in the cad department, or in my case in most other modelling departments, I should mention that it is as easy as falling off a log to put a motor like the old Triang X04 or an hacked about Hornby or Lima power bogie into a Hornby LMR carriage and use it to push dummy locos about. It saves an awful lot of high precision work on gear trains and such. There is oodles of room for weights in the carriage, so long trains can be hauled even if the drive is only to one wheelset, so long as power pick up is from both. It's a crude and horrible cheat but it works.

This is not an attempt to detract in any way from your exceptional work, Bee. It's just an idea for those of us with lower skill sets.

[DavidBigcheeseplant]

This is a link to the High Level kits gearboxes https://www.highlevelkits.co.uk/gearboxes

He sells a gearbox to fit in a tender and drops the drive to below the footplate to link with small gearboxes he sells for the driving axle. I would start with these and then design your planet round these.

When I designed my broad gauge locos I drew them in 4mm scale not full size, as I has to thicken various parts slightly to print.

[DavidBigcheeseplant]

I notice that you are designing the Planet full size then scaling it down to 4mm scale, I think if you do this then you will run in to a number of problems , probably the biggest being you are modelling it in OO gauge which is 16.5mm gauge if you draw it at prototypical size then scaling down exactly will gauge it at the correct 18.83mm gauge, so splashers and wheels will need moving in and clearances will be changed as you will be using larger than scale flanges.

David

[What About The Bee]

Hello David aka Big Cheese*

Excellent catch and a huge error on my part.

The sandwich frames, footplate, springs & etc were drawn full scale. Since FreeCAD is parametric, I installed a vertical, horizontal and longitudinal scale factor to each and every dimension. It was a miracle of scaling, functioning beautifully to tweak the shape. I was convinced I had hit upon a clever scaling method, but it functioned only so long as the model was simple. It did not work for a complex sketch. My sandwich frame model collapsed into a pile of 1s and 0s. Whoops!! I attribute this error to my learning curve in how FreeCAD functions. It was challenging enough to translate the Armengaud drawings into FreeCAD while simultaneously learning FreeCAD. Armengaud provides plan, elevation and some sectional views, without isometric views. Teasing information from the Armengaud drawings can be very challenging. I did not want to add in the complexity of the OO squish at the same time. My error. An error I will not repeat.

The gears, motor and boiler shell are drawn directly in OO, as a consequence of that catastrophe. I started with the Rumford axles and wheels. Those large yellow wheels you see are 20mm diameter with 14.3mm back to back. Gear design uses 0.4 module. The boiler shell follows the OO requirements and is NOT scaled directly from Armengaud. All those tiny Stephenson Valve Gear parts will be difficult to fabricate, but are drawn to OO.

So yeah, I get to draw the sandwich frame again. In OO. Weeee!

Bee

*Fantastic handle, please never change that!

[What About The Bee]

Hi Three Link

Look back at my first feeble efforts in FreeCAD. I wanted to draw Tiger's auxiliary buffer, so as to push Chaldrons about. The so called Banking Engine test. We discussed that extensively at the time.

The model was 4 rectangular blocks. 4 threaded rods. 8 threaded nuts. It took me forever to understand how to do this. How to draw it. How to arrange the pieces. I was so proud of myself. Yet that would only take a few moments to accomplish today.

As I continue on through OO Planet, my confidence and facility with the FreeCAD tool has increased. It is a marvelous tool.

I do encourage others to try their hand.Start small. Do simple designs. Learn the tool. Do not, however, expect to design a full scale working steam locomotive in your first week!! Hehehe! Maybe a few blocks to start. Perhaps a mile marker post, or simple gate.

You Can Do This!

Bee

[DavidBigcheeseplant]

I am not familiar with FreeCAD I use Fusion which in the hobbyist version that I use is free.

For the locomotives I have drawn, I start with the wheels, I have devised a parametric program in Fusion that allows me to put in wheel diameter, number of spokes, shape of spokes and all sorts of other variables, I can put in exactly the wheel I want and Fusion creates it. I can then export the wheel as a STL file and print it. inserting the wheel into a turned Alan Gibson steel tyre.

[Rana Temporia]

I have a lot of previous experience in using 2D CAD and originally found this programme difficult to draw with as my brain is thinking Autocad and there are some differences, especially as I can’t type quick commands such as poly, line, fillet etc. which is how I’m used to working. However, there are some excellent tutorials available free and I quickly learned how to do some things with it, such as producing hollow cones and other shapes.

It looks like I also have to option to import .dxf files which I can very easily and quickly produce in CAD for each axis and produce a 3D model from them. I’ve not had time to experiment yet due to work pressures but hopefully soon…..

For a freebie it is a very useful and unbelievably powerful application, I am impressed!

Just a note for anyone with an older computer, the 32bit version is no longer available, you need a 64bit machine and I would think plenty of RAM.