Do we need coned wheels?

[AndyID]

Thinking it might be an idea to do a bit of experimenting to see what happens with wheels that do not have coned treads, it occurred to me that this might be well known already, so I thought i would post it. I wasn't sure where to put this, but because it might be of interest to those who build their own turnouts, I stuck it here.

 

Railway wheels are usually "coned" as this helps to center the vehicle as it progresses along the track (usually, but I understand not always). This centering action is largely a consequence of gravity, and while that works well on full scale railways, I wonder if gravity has much effect at the scales many of us apply to our models.

 

At model scales, could it be that coning wheels actually creates more problems than it solves?

 

Unless you model to very accurately scaled dimensions (e.g. P4), wheels traversing the crossing in a turnout experience a significant "drop" as the supporting edge of the wing-rail moves towards the reduced radius of the wheel's tread. The wheel is essentially running "downhill". That continues until the tread encounters the "vee", and when that happens, the wheel is accelerated rapidly upwards. That rapid acceleration can lead to problems.

 

A simple way to prevent any upward acceleration is to make the treads parallel rather than coned. The wheel would be supported at the same elevation by the wing rails all the way through the crossing.

 

A big unknown is how parallel tread rolling-stock would behave while being pulled or propelled in other situations. My hunch is it would be little different, but if there have been experiments conducted to find out, please advise.

[martin_wynne]

Unless you model to very accurately scaled dimensions (e.g. P4), wheels traversing the crossing in a turnout experience a significant "drop" as the supporting edge of the wing-rail moves towards the reduced radius of the wheel's tread. The wheel is essentially running "downhill". That continues until the tread encounters the "vee", and when that happens, the wheel is accelerated rapidly upwards. That rapid acceleration can lead to problems.

 

Hi Andy,

 

That's why I recommend that you follow prototype practice and take a few thou off the top of the nose of the vee, so that it drops below the level of the adjacent wing rails. It would be better to ramp up the wing rails a little, but that's tricky to do.

 

The main reason for the coning is to act as a differential on curves. The outer wheel runs on a larger circumference than the inner wheel, to accommodate the longer length of the outer rail. Without this coning, one or other of the wheel would need to skid or slip, significantly increasing the train resistance on curves.

 

In our models we tend to use much sharper curves than the prototype, so arguably we need more coning rather than less. Fortunately our wheels do not have several tons loading on them, so they can slip quite easily. Some slipping also helps to keep the wheels clean.

 

regards,

 

Martin.

[AndyID]

Hi Martin,

 

I forgot about the differential action. Is it self-correcting at any speed, or is it to some extent dependent on centripetal force and therefore speed?

 

In any event, as you point out, we are talking about radii and masses that bear little resemblance to the real thing. We know that coning does present challenges at crossings. I'm wondering if it really buys us anything in other situations. It is more correct cosmetically, but does it serve any useful function at the scales we use?

 

I have a bad feeling the only way to answer the question will be to try it

 

Cheers!

Andy

 

EDIT: Come to think of it, if the differential action was that good, there would be no reason to put flanges on the wheels.

[martin_wynne]

I forgot about the differential action. Is it self-correcting at any speed, or is it to some extent dependent on centripetal force and therefore speed?

 

Hi Andy,

 

Geometrically of course the speed is irrelevant. The coning is intended to be self-correcting on high-speed running lines -- in conjunction with the canting. That's why they have gentle curves and transitions. Otherwise flange wear and rail side-cutting would be horrendous.

 

I very much doubt that it's self-correcting when shunting the sidings round the back of the gasworks.

 

On sharply curved running lines the wheel flanges tend to be hard against the outer rail, so in that case clearly the coning has failed, or is insufficient. Such cases are often fitted with flange lubricators, and subject to severe speed restrictions.

 

As far as models are concerned, I think the CLAG group have done a lot of work on such matters: http://clag.org.uk/

 

It would be worthwhile asking on the Scalefour Forum: http://scalefour.org/forum/

 

regards,

 

Martin.

[Andy Reichert]

The amount of UK prototype coning is 1:20. An “00” wheel is 2.75 mm wide, including its ~0.75 mm flange width. So the radius difference between inner and outer tread edges is only going to be 2.0/20 or 0.1 mm. That’s about a scale 1/3”drop in 4mm scale, or the thickness of a single sheet of the paper in your printer. All of which suggests any coning related drop is going to be difficult to notice visually, and likely even difficult to even hear as a faint click.

So rather than any model wheel drop due to coning, we are really talking model wheel coning "drop" being less than the error in the construction of using any hand made track itself.

Martin’s suggestion of filing down the vee point of the crossing to compensate for the (really) paper thin coning drop is not so much unnecessary, but in fact rather risky. It takes a very skilled model metal worker to take the vee tip down accurately as little as 0.004" and not to extend the drop at all past the point where the vee swells to a scale 1/2" wide.  That’s only around 1 mm along or a scale 3 inches on a #6 where the wing rails stop supporting the wheel anyway. File away beyond that point and you are instead causing significant wing rail wheel drop, rather than fixing insignificant coning drop.

On model trackwork, arithmetic (and a little geometry) is your friend. It has no hard set opinions or axes to grind. It just gives unbiased truthful helpful answers. And that can save a lot of time for modelling, rather than lots of trial and error experimenting, just repeating to find out what has been solidly known for the past 150 years.

For example, from the above, you can easily see that reducing the crossing flange way from 1.25 mm to 1 mm only reduces the already insignificant coning  drop by 0.1/20 mm. So “better running” due to having coning run over slightly narrower flange ways is a complete fallacy.  By arithmetic, not opinion

Andy


 

[martin_wynne]

So “better running” due to having coning run over slightly narrower flange ways is a complete fallacy.  By arithmetic, not opinion

 

I don't recall anyone ever claiming better running for this reason. Ever.

 

Here's an excellent pic from Mick Nicholson showing an almost new bullhead crossing. I have added yellow lines so that you can see how the vee nose has been blunted back from the intersection, and also profiled down below the level of the wing rails to match the coning angle on the wheels.

 

 

Having built many hundreds of model crossings in my time, I can assure you that it is well worth taking a few thou off the top of the vee nose. The improvement in running is easily demonstrated. A rub with wet-or-dry abrasive paper is usually sufficient.

 

regards,

 

Martin.

[AndyID]

The amount of UK prototype coning is 1:20. An “00” wheel is 2.75 mm wide, including its ~0.75 mm flange width. So the radius difference between inner and outer tread edges is only going to be 2.0/20 or 0.1 mm. That’s about a scale 1/3”drop in 4mm scale, or the thickness of a single sheet of the paper in your printer. All of which suggests any coning related drop is going to be difficult to notice visually, and likely even difficult to even hear as a faint click.

 

Andy,

 

I agree it is not much, but I also think it's dangerous to assume you can scale the step function back to full scale, although, if you did, I suspect a 0.3 inch step on a real railway might have rather dire consequences in some situations. Also, I'm willing to bet the step (even without the tapering described by Martin above) on a real crossing would be nowhere near 0.3 inches.

 

In the model situation we are dealing with an entirely different set of variables in terms of inertia, elasticity and damping that could easily exacerbate (or even exasperate) the dynamics. I suppose we could create a very clever dynamic simulation to model these effects, but I suspect this is a case where obtaining the empirical evidence would be a helluva lot quicker.

 

That's why I asked the question. I was interested to know if anyone had actually tried parallel treads on a model.

 

Cheers!

Andy 

[AndyID]

I don't recall anyone ever claiming better running for this reason. Ever.

 

 

Hi Martin,

 

Then let me be the first to make that claim

 

Admittedly the difference may not be that great, but there is a relationship between the flangeway gap and the amount a wheel actually drops in the crossing.

 

Leaving the math as an exercise for the reader, I would assume that a 10% reduction in the flangeway width would result in a 10% reduction in the drop - but I'm willing to be corrected.

 

Cheers!

Andy

[Andy Reichert]

To save anyone going back to see the diagrams, I've re-posted that info here:

 

Posted 07 July 2015 - 12:16

 

 

First I want to give the back-up data I worked through back in 2011, to find out the extent of the problem of using too narrow wheels when passing over a crossing (frog) of a standard that required full width wheels.

 

 

 

This 2011 drawing shows the length of the frog gap that is wider than too narrow wheels (0.088 wide instead of 0.110 wide)  when using the NMRA HO standard of 0.050" flange ways. In this case this closely matches the DOGA 00 Interim standard, when using DOGA Fine wheels instead, so is appropriate for 00 too. In the case of using PECO's wider flange way, the length of the gap will be slightly longer, but not significantly so.  Sorry about some of the my small dimension texts.  But only the results matter.

 

These are for #6 crossing, 0.073" long

These are for #8 crossing, 0.098" long

These are for #12 crossing, 0.144" long

 

Now looking at the effect of the gap on the standard US HO wagon wheels:

 

 

Here we have the larger circles representing the common US HO scale 33" dia (9.5 mm) wagon wheels trying to drop as low as possible into the frog gaps of the three frog sizes drawn above. (from L to R #6, #8 and #12). The mid size circles are the effect for close to minimum practical size 24" dia wheels (e.g. small trams) and the tiny circles are for a imaginary tiny wheel size the same diameter as the gap, which would actually fall in. The sigle LH and the other RH dimensions iare the measure of the maximum drop of the 33" wheel.

 

The respective rounded up max drops are 0.004" for #6, 0.006" for #8 and 0.014" for #12.

 

To put the #6 drop (,0.1 mm) into perspective, 0.003" is the thickness a single sheet of copy paper .

 

The larger size and scale of a 4mm scale wagon wheel, larger coach wheel and much larger steam loco wheel do of course reduce these drops even more. And using a SWAG, probably cancel out the impact of using PECO's larger flange ways, to come up with comparable or less drop numbers.

 

Again I should point out that these show the maximum drop at the mid point of the gap. The actual rate of decent is not instantaneous, but a portion of something close to a sine half wave falling from its peak, that occurs while traversing half the gap length. Then reverses back up in a mirror image of the first half.

 

It's pretty clear that for sharp frog angles (e.g PECO, and other RTR set track points), the drop is insignificant. Most plain rail joints are about as bumpy. But for those using relatively long turnouts, a fix is appropriate.

 

Andy

[28XX]

SNIP!

 

Cheers!

Andy

 

EDIT: Come to think of it, if the differential action was that good, there would be no reason to put flanges on the wheels.

On well laid prototype track a muted rumbling is the only sound heard from the wheels. When coning gives up and the flanges start to do their job, we hear the characteristic screaming grinding noise, essential ambiance for indifferently maintained sidings and goods branches.

[jamie92208]

Not wanting to get into some of the discussion above I wondered if some actual experience of using wheels that are not coned could be given.  I was involved in the restoration of a full size horse tram and we had to recreate the running gear from scratch.  research in 1890's engineering publications revealed that horse trams usually had  parallel tread wheels with one wheel on each axle free to rotate.  This provided the differential action mentioned above and also reduced rolling resistance when re starting from the loops that were very common on many tram routes at that time.   This helped reduce wear and tear on the horses.  Apparently the parallel wheels ran well.

 

As we wanted to be able to run the tram on other heritage tramways we consulted with them about wheel/rail interface standards and it became obvious that no one wanted us to run with non coned wheels though they did accept the 'loose wheel' concept.  We then had our wheels manufactured and machined with the standard 1 in 20 coning. We also had all the necessary collars and lubrication fitted to enable the 'loose' wheel to rotate freely.   We found that we had to be very careful handling the wheelsets before they were fitted as they had a tendency to rotate and try to drop off the track.

.

In practice the wheelsets ran very well, being nearly silent but we did struggle when pushing the tram on railway tracks due to differences in back to back standards compared with tram tracks.

 

Not related to the topic but I hope this is of interest.

 

Jamie

[AndyID]

Thanks for that Andy.

 

It will take me some time to understand what it all means and how it relates to the coning.

 

Meanwhile, the plot thickens. I've been looking at various wheel sets that I have and I have also modified a set so that they have parallel treads.

 

As we might expect, the "no cone" wheels glide through any crossing I have without the slightest bump. There is no apparent drop at all.

 

The other wheels are a mixed bag. Some of them do drop quite a bit, and others hardly drop at all. It turns out that there is quite a large variation in the cone angle across the sample I happened to pick.

 

For example, the cone angle on the wheels from a relatively new Bachmann mineral wagon have a very small cone angle. It's difficult to measure very precisely, but the maximum tread diameter is of the order 12.65 mm and the minimum 12.55 mm over a tread with of 1.85 mm.

 

That's an an inclination of 1.5 degrees (3 degrees inclusive cone angle). In fact, it's so small that it's not obvious there is any coning at all.

 

The next question is: If angles are small, do they serve a useful function, or are they merely cosmetic (as applied to models)?

 

That might be a bit more difficult to answer. However, I have some Lima coaches that need their flanges reduced, so I might as well "de-cone" them while I'm at it and see what happens.

 

Cheers!

Andy

[AndyID]

On well laid prototype track a muted rumbling is the only sound heard from the wheels. When coning gives up and the flanges start to do their job, we hear the characteristic screaming grinding noise, essential ambiance for indifferently maintained sidings and goods branches.

 

I remember it well. I lived very close to the Canal line through Paisley near a particularly tortuous series of bends with continuous check rails. A Coronation rearranging the track was a sight and sound to behold.

[martin_wynne]

Then let me be the first to make that claim

 

I think the two Andy's have got in a muddle here.

 

"Wheel drop" for models calculations are based on the overall width of a wheel with respect to the flangeway gap. The coning angle is disregarded and doesn't come into it.

 

If the wheel is wide enough, it runs properly through the crossing fully supported on the wing rails. There may be a very slight drop due to coning effects, but it is negligible in relation to other variables. Its effect is mitigated by taking a few thou off the top of the nose of the vee, as indicated in my post above.

 

"If the wheel is wide enough", is the reason for using narrower flangeways for better running, not coning.

 

regards,

 

Martin.

[AndyID]

Not wanting to get into some of the discussion above I wondered if some actual experience of using wheels that are not coned could be given.  I was involved in the restoration of a full size horse tram and we had to recreate the running gear from scratch.  research in 1890's engineering publications revealed that horse trams usually had  parallel tread wheels with one wheel on each axle free to rotate.  This provided the differential action mentioned above and also reduced rolling resistance when re starting from the loops that were very common on many tram routes at that time.   This helped reduce wear and tear on the horses.  Apparently the parallel wheels ran well.

 

As we wanted to be able to run the tram on other heritage tramways we consulted with them about wheel/rail interface standards and it became obvious that no one wanted us to run with non coned wheels though they did accept the 'loose wheel' concept.  We then had our wheels manufactured and machined with the standard 1 in 20 coning. We also had all the necessary collars and lubrication fitted to enable the 'loose' wheel to rotate freely.   We found that we had to be very careful handling the wheelsets before they were fitted as they had a tendency to rotate and try to drop off the track.

.

In practice the wheelsets ran very well, being nearly silent but we did struggle when pushing the tram on railway tracks due to differences in back to back standards compared with tram tracks.

 

Not related to the topic but I hope this is of interest.

 

Jamie

 

That is interesting Jamie. I suppose they wanted the wheels to interface correctly with the rail surface (assuming it was inclined). Incidentally, the original Tri-ang wheels were free to rotate independently on the axles. Probably not such a bad idea when trying to negotiate 15 inch radius curves.

[AndyID]

I think the two Andy's have got in a muddle here.

 

"Wheel drop" for models calculations are based on the overall width of a wheel with respect to the flangeway gap. The coning angle is disregarded and doesn't come into it.

 

If the wheel is wide enough, it runs properly through the crossing fully supported on the wing rails. There may be a very slight drop due to coning effects, but it is negligible in relation to other variables. Its effect is mitigated by taking a few thou off the top of the nose of the vee, as indicated in my post above.

 

"If the wheel is wide enough", is the reason for using narrower flangeways for better running, not coning.

 

regards,

 

Martin.

 

HI Martin,

 

Well, as it says in my signature, I am a professional railway muddler, so what else would you expect?

 

Yes, I understand and fundamentally agree with your points, but that does not address my question, which is basically about the benefits (or lack thereof), of coning model wheels.

 

In other words, does anyone know if it really does the slightest bit of good, or is this simply a case of "the big railway does it that way, so we better do it too".

 

Cheers!

Andy

[t-b-g]

People can talk theories, maths and physics all they like but what it comes down to is "Does it work?"

 

There are enough people who have built model railways that run well over the years. there have been probably plenty more that do not run so well. I am talking about hand built pointwork here. Apart from filing things down or adding shims to check rails, there is little that we can do to alter RTR points.

 

As most of these all use the same components, we can be pretty sure that any problems in running are down to the care in construction.

 

I am part way through a project to build points for a biggish layout. Around 170 points and crossings (some highly complex) have been constructed and there are around another 80 to do.

 

The layout baseboards are not built yet, so all the points have been tested with a long wheelbase 4 wheeler van.

 

It is in EM gauge using a 1mm flangeway gap and there has been no filing down of crossing noses, just a very gentle rounding to remove a sharp point on the crossing nose. The test wagon rolls through every bit of it with no drop, no bumps or lurches. In fact it is all just as smooth as it is possible to get.

 

So trackwork built in a perfectly normal way, run with commercially coned wheels just as they come, can and does do the job exactly as required.

 

I can fully understand and appreciate that somebody wants to query if coned wheels are necessary. It is an interesting question.

 

But such research is not going to lead to better running because running can be perfect with what we have now. If we are trying to recreate the real thing but smaller, then why should we even consider having flat treads when the real railway doesn't?

 

Tony

[AndyID]

People can talk theories, maths and physics all they like but what it comes down to is "Does it work?"

 

There are enough people who have built model railways that run well over the years. there have been probably plenty more that do not run so well. I am talking about hand built pointwork here. Apart from filing things down or adding shims to check rails, there is little that we can do to alter RTR points.

 

As most of these all use the same components, we can be pretty sure that any problems in running are down to the care in construction.

 

I am part way through a project to build points for a biggish layout. Around 170 points and crossings (some highly complex) have been constructed and there are around another 80 to do.

 

The layout baseboards are not built yet, so all the points have been tested with a long wheelbase 4 wheeler van.

 

It is in EM gauge using a 1mm flangeway gap and there has been no filing down of crossing noses, just a very gentle rounding to remove a sharp point on the crossing nose. The test wagon rolls through every bit of it with no drop, no bumps or lurches. In fact it is all just as smooth as it is possible to get.

 

So trackwork built in a perfectly normal way, run with commercially coned wheels just as they come, can and does do the job exactly as required.

 

I can fully understand and appreciate that somebody wants to query if coned wheels are necessary. It is an interesting question.

 

But such research is not going to lead to better running because running can be perfect with what we have now. If we are trying to recreate the real thing but smaller, then why should we even consider having flat treads when the real railway doesn't?

 

Tony

 

Thanks for that Tony.

 

Now, I'm not saying this will be the case, but what if it turned out "no-coned" wheels made it easier for modelers who are less skilled than yourself to construct reliable turnouts?  Would that be such a bad thing?

 

Cheers!

Andy

[martin_wynne]

Now, I'm not saying this will be the case, but what if it turned out "no-coned" wheels made it easier for modelers who are less skilled than yourself to construct reliable turnouts?  Would that be such a bad thing?

 

Hi Andy,

 

I don't know how it would affect model running. It would look a bit odd where wheels are exposed, such as on steam locomotives.

 

But I'm quite sure it wouldn't make the slightest difference to the ease or difficulty of building reliable model trackwork. The effects of the coning angle are negligible compared with the other variables.

 

regards,

 

Martin.

[Nimbus]

EDIT: Come to think of it, if the differential action was that good, there would be no reason to put flanges on the wheels.

In the dim past, when the P4 movement was just beginning to stir, the enquiring Dick Ganderton did run a species of 'Bubble Car' without flanges - and without derailment - on a model of Ashburton, to demonstrate that coning was more than mere hypothesis.

 

The Nim.

[Grovenor]

 

But I'm quite sure it wouldn't make the slightest difference to the ease or difficulty of building reliable model trackwork. The effects of the coning angle are negligible compared with the other variables.

However it does task this relatively unskilled modeller with modifying all his wheels to remove the coning, a process I think needs rather more skill than putting a slight slope on top of a crossing nose.

Regards

[Allegheny1600]

In my very humble opinion, all you need to "Prove" that railway wheels need coning - is to roll a ping pong ball along a length of '00' track as it is surprising what odd angles and gradients the ball will cope with without falling off.

Essentially the real thing replicates this "ball" effect with the use of coning.

But, I condone you for having an equiring mind, without that we may never have got beyond the oxen cart.

Cheers,

John E.

[Poor Old Bruce]

EDIT: Come to think of it, if the differential action was that good, there would be no reason to put flanges on the wheels.

 

IIRC, that was speculated by someone working on advance vehicles at BR Research at one time but it never got off the ground (so to speak) because the flanges really are necessary on sharper curves and to counter the effects of lateral track defects.

 

Geometrically of course the speed is irrelevant. The coning is intended to be self-correcting on high-speed running lines -- in conjunction with the canting. That's why they have gentle curves and transitions. Otherwise flange wear and rail side-cutting would be horrendous.

 

Cant is there for passenger comfort and to stop the tea spilling. It is not necessary to assist curving and sometimes can add to curving resistance.

 

The amount of UK prototype coning is 1:20. An “00” wheel is 2.75 mm wide, including its ~0.75 mm flange width. So the radius difference between inner and outer tread edges is only going to be 2.0/20 or 0.1 mm. That’s about a scale 1/3”drop in 4mm scale, or the thickness of a single sheet of the paper in your printer. All of which suggests any coning related drop is going to be difficult to notice visually, and likely even difficult to even hear as a faint click.

 

Prototype wheel profiles do not stay at 1:20. They wear! In extreme cases negative conicity can be generated. The actual conicity changes across the wheel, some profiles start with a bit of coning, then a parallel portion which runs into a transition towards the flange.

 

I don't recall anyone ever claiming better running for this reason. Ever.

 

Here's an excellent pic from Mick Nicholson showing an almost new bullhead crossing. I have added yellow lines so that you can see how the vee nose has been blunted back from the intersection, and also profiled down below the level of the wing rails to match the coning angle on the wheels.

 

 

Having built many hundreds of model crossings in my time, I can assure you that it is well worth taking a few thou off the top of the vee nose. The improvement in running is easily demonstrated. A rub with wet-or-dry abrasive paper is usually sufficient.

 

regards,

 

Martin.

 

As you say, Martin, an excellent picture. You can see where the contact of the outside of the wheel treads on the wing rails overlaps that on the crossing nose. If the flangeway was wider, the wheel would run out of wing rail support before reaching the nose and drop into the gap.

 

On well laid prototype track a muted rumbling is the only sound heard from the wheels. When coning gives up and the flanges start to do their job, we hear the characteristic screaming grinding noise, essential ambiance for indifferently maintained sidings and goods branches.

 

Also heard from Pacers when leaving places like Leeds and Newcastle when the inside wheels slide across the rail heads and the wheels resonate. Not necessarily applicable to "indifferently maintained sidings and goods branches" though, it is more associated with track curvature.

 

Not wanting to get into some of the discussion above I wondered if some actual experience of using wheels that are not coned could be given.  I was involved in the restoration of a full size horse tram and we had to recreate the running gear from scratch.  research in 1890's engineering publications revealed that horse trams usually had  parallel tread wheels with one wheel on each axle free to rotate.  This provided the differential action mentioned above and also reduced rolling resistance when re starting from the loops that were very common on many tram routes at that time.   This helped reduce wear and tear on the horses.  Apparently the parallel wheels ran well.

 

As we wanted to be able to run the tram on other heritage tramways we consulted with them about wheel/rail interface standards and it became obvious that no one wanted us to run with non coned wheels though they did accept the 'loose wheel' concept.  We then had our wheels manufactured and machined with the standard 1 in 20 coning. We also had all the necessary collars and lubrication fitted to enable the 'loose' wheel to rotate freely.   We found that we had to be very careful handling the wheelsets before they were fitted as they had a tendency to rotate and try to drop off the track.

.

In practice the wheelsets ran very well, being nearly silent but we did struggle when pushing the tram on railway tracks due to differences in back to back standards compared with tram tracks.

 

Not related to the topic but I hope this is of interest.

 

Jamie

 

So called 'Independent wheels' were tried on BR in the 1980s or 90s (IIRC) and a Sprinter was fitted with them and tried on the Blaenau Ffestiniog branch. It was not pursued as the lack of coupling (for want of a better word) between the wheels meant that there was no effective 'steering' by the wheelsets. May be OK at horse speeds but not deemed suitable for main line speeds.

[Andy Reichert]

I'm not in the least confused.

 

Coning works on the prototype to steer the wheels towards the centre of the track on straight track and their relatively large radius curves.

 

The mechanism that causes that is the larger working diameter (due to coning) on the wheel that is closer to its adjacent rail than the other. That larger diameter makes the wheel move further along the track as it rotates, than the wheel that is further from its adjacent rail. That steers the wheelset back towards the track centre.

 

As you can imagine, that doesn't work very well for sharp radius curves. I did the maths several years ago and found that for 00/HO the minimum radius needed for 1/20 or 1/40 model wheel coning to steer the wheels was around 30 ft!!!!.

 

So the quick answer is that is has no effect on any model railway curves, as far as wheel steering is concerned. But it will help centering on model straight track.

 

Andy

 

I believe the  (90 mph) North Shore Electric railway near Chicago, ran tests on unconed wheels in the 1940's, with TV cameras mounted under the cars to watch the effect. Apparently, IIRC,  wheel wear was rapid and dangerous "slotting" in the wheel tyres occurred .

[martin_wynne]

all you need to "Prove" that railway wheels need coning - is to roll a ping pong ball along a length of '00' track as it is surprising what odd angles and gradients the ball will cope with without falling off.

 

Yes, but where do you put the tail lamp?

 

Martin.