Compensating an eight coupled chassis

[RedgateModels]

In a job lot of stuff that came my way on Friday was Ian Rice's book on kit/scratchbuilding. Chapter 6 fell open and the subject of compensation was read. Now, the 2-8-2 Mikado is on the bench as yet unstarted.

 

Should I add some form of compensation/springing and if so how do I do it? as I understand from Mr Rice you should have 3 points of support which is fine for a 6 coupled chassis but what about 8?

 

Sketches and pics would be appreciated as well as words of encouragement/dissuasion

 

I will be scratchbuilding the frames, so a steep learning curve may ensue .....

[Penrhos1920]

In a job lot of stuff that came my way on Friday was Ian Rice's book on kit/scratchbuilding. Chapter 6 fell open and the subject of compensation was read. Now, the 2-8-2 Mikado is on the bench as yet unstarted.

 

Should I add some form of compensation/springing and if so how do I do it? as I understand from Mr Rice you should have 3 points of support which is fine for a 6 coupled chassis but what about 8?

 

Sketches and pics would be appreciated as well as words of encouragement/dissuasion

 

I will be scratchbuilding the frames, so a steep learning curve may ensue .....

 

3 point or beam compensation is a bit old hat now. Yes many people still use it but it has been surpassed by Continuous Springy Beams. A single piece of guitar string down each side of the frames supports all of the driving wheel hornblocks. It is mounted on fulcrums (usually handrail knobs) between each axle plus one at each end. It makes fitting compensation a lot easier as there are less parts to fit and adjust. My chassis below is for a 0-4-2T but I've sprung it like a 0-6-0T. The location of the handrail knobs used for the fulcrums are circled blue. Three of the handrail knobs that connect the guitar string to the hornblocks are circled red. The spring is free to pivot in the handrail knobs.

 

 

If the chassis runs too low a thicker guitar string is used and visa versa. All very easy to build and set up. And possible more important, the hauling power is supposed to be better, but I can't vouch for that.

 

The CLAG website says it all a lot better than I have and gives many worked examples. Halfway down the page is a Stanier 8F example.

[34theletterbetweenB&D]

For OO purposes, Bachmann's two sprung axles on the coupled wheel base of their WD 2-8-0 performs so well in pick up and adhesion, that I would suggest that anything more is overkill. With the models adjusted for the same weight on the coupled wheels it comfortably outperforms the rigid chassis on the otherwise very similar Hornby Stanier 2-8-0.

[craigwelsh]

Here are some options http://clag.org.uk/41-0rev.html#section20.2 . A rigid-beam version of c is included in my Finney 28xx kit to clear both the inside motion on the front axles and the gearbox on the rear.

 

You may find moving away from rigid for a first time on something with that many wheels a challenge or it could give you a big improvement.

[bertiedog]

 

 

The above drawings, from 63 years ago, nothings new, show a CSB system on the Varney Super Pacific and 8 coupled Super chassis and in particular the 8 coupled version.

 

The actual layout can be totally different, they used frame spacers each side of the bearings, which were sleeve bearings not hornblocks, but the same principles apply totally.

 

The wire I use is 14 thou guitar steel,the spring wound from the same.

 

The advantage over all other systems is it's adjustable, just re-tension the wires to suit the final weight of the model.

 

The "buttons" can be a simple bolt head with an undersized washer, or none, screwed into tapped holes, or soldered on to thinner frames.

 

Any modern motor and gearbox would fit the system, it floats with the axle, attached at a single pivot point to allow slight movement up and down, but not fore and arf.

Simply the best arrangement ever designed, fully sprung and fully compensated and fully adjustable.

 

I stress no commercial connection, but I feel the late Gordon Varney deserves all the credit for the design from the 1940's.

 

 

For ref:- this is the full chassis with ballraced 7 pole star wound skewed slot motor, all sprung, all compensated, all adjustable, and far beyond the rubbish made these days...remember this was 63 years ago.

 

Stephen.

[Flying Pig]

As an aside, please can we not perpetuate the dreadful modellerism of calling axle bearings "hornblocks". The hornblocks are the guides fixed to the frames within which the axleboxes slide on the prototype. There's really no need for this inversion.

 

Back on message - can't wait to see the 2-8-2 begin to emerge, Redgate, regardless of how you choose to suspend it. One loco Riddles *should* have built.

[DLT]

In a job lot of stuff that came my way on Friday was Ian Rice's book on kit/scratchbuilding. Chapter 6 fell open and the subject of compensation was read. Now, the 2-8-2 Mikado is on the bench as yet unstarted.

 

Should I add some form of compensation/springing and if so how do I do it? as I understand from Mr Rice you should have 3 points of support which is fine for a 6 coupled chassis but what about 8?

 

Sketches and pics would be appreciated as well as words of encouragement/dissuasion

 

I will be scratchbuilding the frames, so a steep learning curve may ensue .....

 

Hi There,

 

If you can get hold of it, Edition No.23 of Model Railway Journal (from 1988) might help; in it Iain Rice describes the building of a Somerset & Dorset 2-8-0 in EM gauge. Its written in his usual style, so its an entertaining as well as highly informative article.

 

Best of luck; I will be following closely as I have an 8-coupled (in 00) to build.

 

All the best,

Dave.T

[Miss Prism]

Craig's right; it doesn't get simpler or better than this (Figure 77c):

 

http://clag.org.uk/41-0rev.html#figure77c

[Scale7JB]

As an aside, please can we not perpetuate the dreadful modellerism of calling axle bearings "hornblocks". The hornblocks are the guides fixed to the frames within which the axleboxes slide on the prototype. There's really no need for this inversion.

 

Are you sure? I was under the fairly educated impression that the 'guides' attached to the frames, were know as hornguides, and the axle bearings are indeed hornblocks...

 

JB.

[RedgateModels]

Thanks everybody, certainly food for thought, I'll see how the comet 9F rods go together - I'll be using them with one "link" removed and the distance between centres altered. Depends how much articulation I can get as to what I decide on.

 

I would like to make it a good runner though, even if it's OO

 

 

[craigwelsh]

Craig's right; it doesn't get simpler or better than this:

Well I was more giving a link to some options than recommending as I haven't had one working yet..

 

Are you sure? I was under the fairly educated impression that the 'guides' attached to the frames, were know as hornguides, and the axle bearings are indeed hornblocks...

JB.

I thought that as well due to suppliers descriptions but this Websters definition suggests otherwise http://www.websters-online-dictionary.org/Ho/Horn+block.html as does the original MRSG P4 notes http://www.scalefour.org/history/Feb76/feb76.htm

[Buhar]

This is a really interesting topic, thanks folks, I do try to keep up but, apart from Stephen's notes about the Varney chassis posted just before The Great RMW Silence, I had never heard of this method. I read the CLAG page (and then re-sat my Physics O Level) and am now a bit confused between "springy equalising beams" and using the wire along the length of the wheel-base as a form of independantly springing each axle bearing. I would have thought springy equalising beams would need a bit more thought about weight distribution.

 

On the full-length type presumably the weight of the loco and the number of load-bearing wheels have direct relationships with the spacing of the fixings (handrail knobs) and the type of guitar string or piano wire. Can anyone enlighten me?

 

Also can this method cure the 4-4-0 problem and cope with a weighted tender?

[bertiedog]

Clag's site is the main source for CSB calculations, the method does work, but most examples I have examined run on a bit of a wing and a prayer, there is no adjustment apart from the steel wire diameter, which at 14 thou or so, is a limited selection.

 

The Varney wire method overcomes this by having at the end a coil spring, and you can pull tight to adjust, or add a screw adjuster to each side. The Varney also works with nylon cord, and stainless steel woven cable, which is made at these sizes. I tried the nylon cord but found it stretched in the long term. Mylar cord can be used, but in the end simple 14 thou guitar steel is the easiest to source.

 

There are other sources of fine spring steel wire, but most are bulk suppliers.

 

I should add that there is another secret to the Varney system, and that with the spring slack, (or removed!!) all will work 100% anyway.... Now this sounds insane, but think about it, the spring is only there to apply a force that at max will deflect the axle boxes down about .5mm or so.

 

This movement is all that's required to maintain contact, so that when the spring is slack the axle boxes rest at the top of each hornway guide. In this condition it would work as an unsprung chassis.

 

As the spring is tightened up the loco will rise and a small .5mm clearance appears and allows the wheel to rise under force, or fall under combined spring/loco weight, and gravity.

 

As any individual wheel goes up it tightens the wire, restraining the others more, exactly the same as CSB.

 

But this is not, repeat not, an individual independent suspension system as on a car, which confuses some supporters of CSB. The axles lock pairs together, they are not independent as in a car.

 

On one side as the wheel lifts it also shifts the other side, as a lever, and although the other wheel does not rise it applies a leverage force back on the axle to stop the first rising, a complex set of dynamics going on.

 

This gets complex to work out, and is why the Varney idea works, as there is no need to work out anything first, apart from basics.

 

Stephen.

[Miss Prism]

I would have thought springy equalising beams would need a bit more thought about weight distribution.

They require as much thought about weight distribution as any other form of beams, whether they are rigid or non-rigid, continuous or non-continuous. The force impinging on each end of an equaliser (springy or otherwise), supported or pivotted at its midpoint, is half that of the force impinging at its midpoint. Learn "The Principle of Moments". (Plenty of good websites out there.) Springy equalisers do not require complicated mathematics, and are conceptually easy to grasp.

 

On the full-length type presumably the weight of the loco and the number of load-bearing wheels have direct relationships with the spacing of the fixings (handrail knobs) and the type of guitar string or piano wire. Can anyone enlighten me?

That is correct. For a 'CSB' (and I will have to call it a 'CLAG' CSB lest darling Bertiedog corrects me), the positioning of the chassis support points is derived from solving the second differential of the deflection curve. In models, our deflection curve is defined, and the boundary conditions at the ends of the beams are known, and thus we can work backwards through the maths to get to these support point longitudinal positions. The boundary conditions at the ends of a Varney CSB are significantly different. 'CSBs' exist in many forms.

 

Also can this method cure the 4-4-0 problem and cope with a weighted tender?

If by 'problem' you mean rigid 4-coupled chassis with bogies dangling or supporting the body in some non-precise way, the use of springy equalisers is in my view the easiest way of supporting both the 4-coupled part and the bogie. The only significant constructional problem with a springy equaliser is how to stop it slipping longitudinally along the chassis. (I think this danger is overrated personally, and one could say the same of CSBs, but there are simple solutions - how's about a bit of sleeving at both ends for example?)

 

On the weighted tender issue*, any form of springing will be able to cope with it provided the value of the weight of the tender impinging on the rear of the loco body is known, which can then be factored in to the support/pivot point positions and deflections of the springs. (The model axle weights of the loco would also need to be known of course.)

 

* I am not a fan, although my reservations are not for 'weight' reasons.

[Buhar]

The spring has sprung, the loco's riz,

I wonder where the forces is.

They say the force is on the track,

And Bertiedog says Varney's back!

 

By the "4-4-0 problem" I meant getting sufficient weight on the drivers for reasonable traction, solutions to which do indeed often lead to bogies wafting in the air.

 

I am going to have to have a bit more of a think before I try to tease some more out of Stephen and MP and it's too late for that now. Except that the CLAG CSB (2 axle type) seems to be not too different from a set of cantilevered springs. In fact I can't see how the item would pivot before the spring above an axle would deflect in response to that axle rising.

[Jol Wilkinson]

All this goes a long way to show why Mike Sharman's (and others) ideas on compensation using beams provides a simpler and in many cases, more suitable, answer.

 

One thing that doesn't "scale" is mass, and therefore, inertia. Springing needs to be relatively soft to cope with the mass involved in 4mm models. Perhaps that's why there is no universally acclaimed, commercially available, system available from the inventive designers/suppliers that populate the technical boundaries of our hobby.

 

Spring mounting location is important, to provide consistent applied spring force (and deflection) as well as to avoid pitching (or porpoising). That's why the CLAG team have developed a downloadable spreadsheet to make the calculations easier. Even Bertiedog's beloved Varney system need correct positioning of the mounting points. Tightening the adjuster to raise ride height will also increase the effective spring tension and vice versa. Although the CLAG system needs different wire diameter or material to change the spring rate and static deflection, it's not too difficult to do. One etched kit manufacturer already supplies an etch with adaptors to fit their own and other hornblock bearings. This also includes a variety of mounting point fittings, including a screw adjustable version (albeit it is fiddly to make up).

 

Going back to the OP, I would compensate/spring an eight coupled chassis. My L&NWR C class has a fixed (4th) rear axle with the motor drive under the cab, the 3rd is sprung using .011" wire bearing on the axle bearings - which run in hornguides - and the front two axles are compensated by a single, central, beam. It has Sharman wheels, including the flangeless 3rd axle, which were fitted using a GH Wheel Press. The hornguides were set up using taper ended axle jigs and the coupling rods. It runs well and I would use the same system again, as it provides a simple but effective solution.

 

For 4-4-0s, a sprung (CSB type)bogie, with compensation or CSB's on the drivers would be my preferred route, so it's what I'm doing on a Brassmasters Precursor. Fortunately, London Road Models do a suitable sprung bogie (6' 3" wheelbase) from their LNWR Jubilee kit.

 

Jol Wilkinson

[buffalo]

Stephen (Bertiedog) and I had a long conversation about CSB on the old RMweb with others chipping in from time to time. The subject was a six-couple chassis, though may still be of interest. Although I've copied this to my blog (starting here), it's probably better to look at the old pages as they allow you to see Stephen's diagrams in line. In the end, I went for a simple CSB with no end restraints, but used the knife-edge fulcrums as suggested by Stephen. The end result can be seen in the first photo on this blog page.

 

Nick

[Miss Prism]

By the "4-4-0 problem" I meant getting sufficient weight on the drivers for reasonable traction, solutions to which do indeed often lead to bogies wafting in the air.

For overall loco stability in a 4-4-0, we need a proportion of weight on the bogie and a proportion of weight on the drivers. These static weight proportions will be given by where the centre of gravity is. (Taking into account any extra tender weight factor impinging on the back of the loco.) Additionally, in the dynamic situation, we want our wheels to stay on the rails over undulating (or superelevated etc) track. Springs or beams will satisfy this demand, but the bottom line is that traction will be determined by how much weight is on the drivers and how consistently those drivers will retain contact with the rails.

 

On a rigid, or essentially rigid, driver chassis on a 4-4-0, any height variation or setting of the bogie will adversely affect the consistency with which the drivers contact the rails, and in the extreme, only one driver could be contacting the rail. Hence the gross variation of haulage abilities reported, for example, on the Hornby T9.

 

If one wants to maximise the driver weight, the solution is to load the firebox with lead, put a gearbox on one of the drivers, and cardan the transmission from a tender-mounted motor. Complicated? Yes, to an extent, but there is no such thing as a free lunch.

 

I am going to have to have a bit more of a think before I try to tease some more out of Stephen and MP and it's too late for that now. Except that the CLAG CSB (2 axle type) seems to be not too different from a set of cantilevered springs. In fact I can't see how the item would pivot before the spring above an axle would deflect in response to that axle rising.

In terms of providing a springing function, that is correct to a large extent, although the value of the deflections (wire for wire) given by free cantilevers are of course different to the deflections given by a 3-fulcrum 2-axle CSB. Needless to say, for an 0-4-0, the latter is far preferable to a pivoted springy equaliser, which would provide an unrestrained and unacceptable pitch freedom. For the 4-4-0, the converse is true, because of the presence of the bogie, which constrains and helps the overall (loco) pitch stability. The slight snag with the 3-fulcrum 2-axle CSB for the drivers of a 4-4-0 (or 0-4-2 etc) is that the horizontality of the CSB fulcrum plane is critically dependent on the height setting of the bogie (and of course its height variation as it traverses undulating track). Pivoted equalisers (preferably sprung, but the case holds true, at least simplistically, for Sharman rigids) over the drivers are immune to this potential problem.

[renovater 1]

I have been following the discussions on CSB sprung locomotives closely, being into small narrow gauge battery electric locomotives i have limited space for CSB, so i am sticking for the moment with compensation. Having said that, i have a question, does a locomotive with CSB lift at the front when attached to a load ?

[t-b-g]

For OO and EM I have never been convinced that much is needed by way of springing or compensation. I have a fair degree of experience in such matters, having built upwards of 50 locos in OO,EM and even one P4! I have also been heavily involved with other folk, who between them have built hundreds of EM locos.

 

I can safely say is that there is no "right" or "wrong" approach and that anybody who tells you that you have to do something a particular way because nothing else is any good, is spouting nonsense! I have built sprung, rigid and beam compensated locos and they all work!

 

For an 8 coupled loco, I have seen the following variants in EM. "Solid" frames, no springing or compensation. "Sprung" with individual axleboxes. As the boxes rest against a solid "stop" I am not sure that these are actually truly "sprung". One axle fixed and the others sprung. Beam compensated, using a twin beam on one pair of axles and a single centre beam on the other pair. All of the above have worked and I could run them on a layout and ask somebody to pick which is which and they would not be able to tell.

 

The locos that have to do the most work are on Roy Jackson's "Retford". They do more mork than any other EM kit/scratchbuilt locos I have seen. As far as I am aware all the 8 coupled locos on Retford have "solid" frames and they plod on for miles and miles with no problems.

 

Back to the 4-4-0 matter. I have seen a 4-4-0 pulling 15 brass kitbuilt carriages arong Retford, with no problem at all. It was an old GEM Midland Compound kit, built 30+ years ago. It has a Triang motor and gears, the front bogie is just fitted loosely to a pivot arm and carries no loco weight and the tender hangs on the back and has a "free bogie" arrangement. It runs superbly! In my simple view of scientific matters, if you transfer ANY weight from the loco body onto the bogie, you take that weight OFF the driving wheels, which is where you want it.

 

Another friend has 2 almost identical NER 4-4-0s. One is rigid, weighted and has a dangly bogie. The other is fully sprung, including the bogie. Both run smoothly and without wobbling about but the sprung one struggles to pull 4 carriages and the rigid one pulls 8 easily.

 

In case I am sounding like a luddite (I was called that once by a very prominent modeller!) I will also say that I have seen some CSB locos in action, including those seen on the CLAG website. They are very, very good runners indeed! Enough to tempt me to have a go next time I build something!

[renovater 1]

Each to his own, i agree, but i would like to know from those who know the CSB system, if locomotives do lift up at the front when attached to a load ?

[billbedford]

I have been following the discussions on CSB sprung locomotives closely, being into small narrow gauge battery electric locomotives i have limited space for CSB, so i am sticking for the moment with compensation.

CBS supports and sprimgs take up less room than compensation beams.

 

Having said that, i have a question, does a locomotive with CSB lift at the front when attached to a load ?

No, why should it?

[Grovenor]

Quote

 

 

Having said that, i have a question, does a locomotive with CSB lift at the front when attached to a load ?

 

No, why should it?

 

Not when attached to a load, but when you accelerate the load then there will be load transfer caused by the traction forces that will increase the axleload at the rear and reduce it at the front, consequently the front may lift slightly. Same effect with any sprung design, CSB or not, Including the prototype.

Regards

Keith

[billbedford]

Quote

 

 

Having said that, i have a question, does a locomotive with CSB lift at the front when attached to a load ?

 

No, why should it?

 

Not when attached to a load, but when you accelerate the load then there will be load transfer caused by the traction forces that will increase the axleload at the rear and reduce it at the front, consequently the front may lift slightly. Same effect with any sprung design, CSB or not, Including the prototype.

Would you like to put a figure on that 'slightly'? will it be visible to the casual observer?

[John_Hughes]

The GWR's 4-6-0 designs, when starting with a heavy load, simply 'sat' on their rear drivers, rather than transferring the weight to a trailing bogie as they would on a Pacific; they were, as a result, allegedly less prone to slipping under these conditions.

 

So yes, the transfer of weight when starting is entirely detectable even on the prototype.