Tread angle 2 deg, 3 deg or even flat ?

[GWR 101]

Hi I am sure this has been discussed before but I can't find the thread. I have seen standards that mention both 2 and 3 degrees, assume there should be some tread angle but does it make any difference.  At the moment I am looking at 3 1/2" running on Aluminium rail. Regards Paul

[stevep]

This is a contentious issue, so I'll be the first to put my head over the parapet.

In full size, wheels are coned, and the track is canted in at the same angle - so you get the dual advantage of self-centering and also maximum surface contact area.

Most (if not all) model engineering tracks are set with the rails (aluminium or steel), vertical. So I turn my driving wheels parallel, to get maximum contact area. If you turn the wheels tapered, the contact is on a narrow piece of the wheel just outside the flange, and you will wear a groove in your wheels. (I have seen it).

I do, however, turn carrying wheels (pony trucks and bogies - which carry far less weight) tapered to achieve the self centering, which 'guides' the engine into curves.

[ettingtonliam]

I know you are talking about 3 1/2" gauge, but the wheel standards for the 7 1/4" Gauge Society, which I am using for Locomotion, specify a tread coning angle of 3 degrees (max), and show a coned wheel on a flat top rail.

It is, as Steve says, a contentious area.

I'm curious to know, when track has been in use for several years, how flat the top is then.

[ettingtonliam]

May 28, 2021 12:45:28 GMT ettingtonliam said:

I know you are talking about 3 1/2" gauge, but the wheel standards for the 7 1/4" Gauge Society, which I am using for Locomotion, specify a tread coning angle of 3 degrees (max), and show a coned wheel on a flat top rail.

It is, as Steve says, a contentious area.

I'm curious to know, when track has been in use for several years, how flat the top is then. I haven't been to our club track (which is steel) for over a year, otherwise I'd check up on it.
It does occur to me that steel tracks tend to be black hot rolled bar set on edge, and the edges of such bars can be quite rounded in profile, not dead flqt like cold drawn bar would be.

[Deleted]

GL5 wheel profile standards are also 3 degrees

Pete

[GWR 101]

Thanks all for your replies, yes I also tend towards 3 degrees for everything. I have over the last couple of years machined around 100 wheels mainly on club carriages (3 1/2" and 7 1/4") and have always aimed for 3 degrees. The majority of these have been on normal lathes but a few on a CNC programmed lathe. The reason for raising this again is that I am about to machine wheels for my Jubilee (6 main and 6 pony /bogie wheels) and I have been shown a set of standards that indicate 2 degrees is the standard, they also mention 0.375 wheel thickness for 3 1/2", where our club recommends 0.406"plus and 3 degrees. As previously mentioned our elevated track is aluminium and we have noticed rail wear on the inside of the outside rail on bends. Regards Paul.

[mbrown]

Only a hunch, but I wonder if the wear on the inside face of the outer rail is caused more by locos with an inadequate radius between the flange and the tread, rather than coned or unconed treads. Even with a parallel tread, a good radius should prevent flange cutting on the rail.

Malcolm

[Gary L]

May 28, 2021 11:44:28 GMT stevep said:

This is a contentious issue, so I'll be the first to put my head over the parapet.

In full size, wheels are coned, and the track is canted in at the same angle - so you get the dual advantage of self-centering and also maximum surface contact area.

Most (if not all) model engineering tracks are set with the rails (aluminium or steel), vertical. So I turn my driving wheels parallel, to get maximum contact area. If you turn the wheels tapered, the contact is on a narrow piece of the wheel just outside the flange, and you will wear a groove in your wheels. (I have seen it).

I do, however, turn carrying wheels (pony trucks and bogies - which carry far less weight) tapered to achieve the self centering, which 'guides' the engine into curves.

Hi Steve

Not really. Modern wheels are not coned in the way you mean, and the reason they are not is because coning does not really result in self-centring except in a very approximate fashion. I know there are demonstrations that seem to prove the opposite, but there are two enormous cautions before taking them on trust:

1. What you will see in the demos is not so much self-centring as hunting- a process of moving along a track by taking wide swings from one side to the other and then back again.

2. The demos only apply to single axles that are free to align themselves radially to the curve. The reality is that is never the case with normal rolling stock, because two axles are necessary, and when held in a frame they cannot take up the theoretical optimum radial position.

So the truth is more complex:

The coning is only secondarily there for directional stability. Its most important function (by far) is to compensate for the greater distance that the outer tyre has to travel around a curve, as compared to the inner tyre. If you fit parallel treads (which are, I'm afraid to say, no longer contentious, just discredited) then what happens is slippage when negotiating curves. So far from promoting adhesion, with 'steam-roller' tyres you are actually promoting the opposite and encouraging one of the wheels on an axle to break away. If there is no slippage, (due to heavy loadings, like under locos or their drivers) then you get a wringing strain on the axle seat which in bad cases can cause the axle bond to loosen, especially if it is a Loctite rather than a shrink fit. Cases of this have been reported in these very pages!

It follows then that the degree of coning needed is related to the degree of curvature. The Japanese high-speed trains hunted badly at first, until the engineers reduced the conicity; the tracks have very easy curves so this suits them. (There is a YouTube video about this.) The 2 degree figure comes from old standard gauge practice, where curves are generally more modest than on miniature tracks. Our miniature lines nearly all have very tight curves by full-size standards, hence 3 degrees is a better value by general consent.

The part of the wheel profile that is most important for self-centring is the flange root radius, which prevents the flange from smacking the railhead with a bang. This should never be less than the accepted wheel standards, and there are few disadvantages in making it bigger.

The part of the profile that is responsible for the shavings of aluminium rail that are often seen, as Paul reports, is the coning of the outer face of the flange, which is much more important than many people realise. If this angle is too slight (i.e. nearer vertical) and especially if combined with too small a root radius, then on a curve, the flange of the outer wheel makes contact at its tip, not its root. Shaving needles off the rails is the least of the problems; much more serious is that the wheel tries to climb onto the railhead which is a sure recipe for derailment.

David Hudson published a series of articles in Model Engineer in 2003 (Self Steering Wheel Sets and Swing Link Suspension) which explain this much better than I can. Amongst other things he illustrates the various BR Heumann type standard profiles. These have a complex curvature in the tread, but the part that does most of the work is somewhat concave. They are coned certainly, but the coning is steeper as you get nearer the flange, and somewhat convex in the outermost part of the tread (which is much less important). The function of this non-linear coning is to prevent linear feedback, which is what causes the hunting. This profile is required for steam locos on mainline tracks too I believe, not just for bogies designed to be 'self-steering'.

The science is a great deal more complex than this, and 'hydrostatic distortion' means that concerns about contact patches are greatly oversimplified. Suffice to say that all of the rolling stock at Bristol SMEE (raised track and ground level) runs on Hudson-profile wheels (and on self-steering bogies too, but that takes this discussion way off-topic!) with many benefits and no evidence of grooving where the wheel profile does not match the inclination of the railheads. This is possibly because the wheel tyres are steel, not the comparatively soft (and not really suitable) cast iron that most of us use for loco wheels.

Hope this sheds a little light on a dark subject

Gary

[Roger]

You definitely need some angle, else it's always going to ricochet from one flange to the other. As Gary says, hunting will likely happen, but that's preferable to running into the flange. The angle obviously affects how strongly it tries to self centre. I remember watching a program where they said that the very high speed trains use a shallower taper to reduce hunting.

[ejparrott]

2 degrees, or 3 degrees, never parallel.

www.youtube.com/watch?v=OD1t4BedIhw&t=92s

[keith1500]

I agree with Gary that the most important part is the root radius but not only that, there is also the angle of the flange. Closely followed by the back to back. This will allow the wheel to negotiate pointwork successfully. I would suggest you follow GL5 standards and choose whether to use flat or tapered.

Keith

[andyhigham]

Are the flangeless wheels on 9Fs coned or flat?

[don9f]

From memory, as I don’t have the proper documents these days, when first introduced the flangeless wheel profile was “coned” the same as what was the “A” profile of the leading & trailing drivers, obviously without the flange! At some point, probably early in their existence, this was changed to a flat, ie cylindrical profile over the centre section of the tread, with two different coning angles repeated on either side of this....what became known as the “Special X” profile. The diameter of the cylindrical part was turned to the same diameter as those drivers with the ”A” profile, measured at their reference point, which is (nowadays) 70mm towards the tyre’s outer face from the back of the flange.
The flangeless tyres on a 9F are slightly wider than the other tyres and the intermediate driver’s tyres have thinner flanges than the “A” profile. (Nowadays the “A” profile is called “P1” and a thin flange profile is “P9”).
When I built my model, I didn’t know about all this and just followed the Les Warnett drawings and made all the driving wheels with the same tyre width and coned the flangeless ones at 2 degrees, like all the others and didn’t machine any thin flanges! After hundreds of miles, the cast iron wheel treads of its drivers all show some hollow wear, but the flanges still look ok.

Cheers Don

[dscott]

I have always done 2 degrees and have had no trouble.
Lily spent part of the day lugging some EN3B down to the Workshop for some outer tires.
The inner bits will come back as smaller trolley wheels.
Berkshire Metals got in some 6" diameter and we go and get some cut every so often.

We have now done quite a few miles with the Electric Beast so will take a photo of the steel
wheels on that.

Having been involved in the building of track panels for our ground level.
I can report that the profiled rail is superb compared to the older steel section which is next to replace.
One day.

David and Lily.

[stevep]

Gary,

I didn't 'quote' your post, as it would be just repetition, but thank you for the comments.

I totally agree, that whether the wheels are parallel or coned, both the root radius and the tapering of the flange are critical.

I can also understand the argument of coning to allow the wheels to travel around a curve without slip. That is a very good point, and as far as I am concerned, probably the best argument for coning that I have heard.

However, I also stand by my assertion that if you have a coned wheel, especially in cast iron, running on the corner of a vertical positioned bar steel rail, a groove will get worn on the wheel where the contact with the rail is made. I have seen wheels like this - probably most prevalent where there has been slipping. Aluminium rail is normally a more prototypical profile, and has a rounded corner, so there is no point contact.

Maybe we should start a campaign to get all ME societies to either tip their rails in at 2 degrees, or grind off the top at that angle and provide a good radius on the inside corner!

[GWR 101]

All, many thanks most informative. It's going to be 2 degrees for me on the loco wheels and a root radius. We wondered if the wear on our track which is on the inside of the outer rail and extends down for about a 1/4" was caused by insufficient tip radius as we found some axle sets which were lacking in this area. We assumed that this would produce a "chisel edge" and hence be more prone to machining the rail profile in this particular area. No one has mentioned this, is it because it's not considered an issue, or is it that it's considered as obvious that a radius is always required. Regards Paul

[Gary L]

May 29, 2021 8:42:12 GMT GWR 101 said:

All, many thanks most informative. It's going to be 2 degrees for me on the loco wheels and a root radius. We wondered if the wear on our track which is on the inside of the outer rail and extends down for about a 1/4" was caused by insufficient tip radius as we found some axle sets which were lacking in this area. We assumed that this would produce a "chisel edge" and hence be more prone to machining the rail profile in this particular area. No one has mentioned this, is it because it's not considered an issue, or is it that it's considered as obvious that a radius is always required. Regards Paul

Hi Paul

The tip radius is not responsible for your rail wear, though it is certainly logical that an absence of tip radius will make it worse if the flange angle is too small. It is absolutely prime that the the flange tip never contacts the running rail, and the flange angle must be great enough to ensure this is the case. If your Society enforces the relevant wheel standards and does not have unusually sharp curves there is not likely to be a problem. (Working out the geometry of this is a complex problem, because it involves conical sections in 3 dimensions- you can't do it from a plan view alone, because this only shows the flange profile at centre height, never at rail height.)

If you try to juggle the angles in a standard wheel profile, you will soon discover why the standards should not be tinkered with. In fact it is only the conicity of the tread that is susceptible to alteration without having a deleterious knock-on effect on something else; they are not just random choices. Since I started to study this, I have formed the firm conclusion that 3 degrees is the most suitable angle for miniature railways, and there is simply no point in using anything else. (With cast iron wheels, you do need to expect a degree of wear over a loco's lifetime, and if for no other reason, 3 degrees gives a margin for wear before bad things start to happen.)

Going back to the tip radius, in full size this is very important, and it must never be less than a certain value. (In the P series of BR standard profiles it is a compound curve with a distinctly 'blunt' appearance). If the flange tip is too thin/sharp it can damage points and cause serious accidents. (In miniature, damage to points is not unknown either, but the consequences are not usually so grave.) The point about flanges, which many model engineers don't appreciate, is that they are only there to provide a safe transit through points. They should never touch the rails in ordinary service, and in full size, when they do (on check-railed curves for example) they generate a distinctive and unpleasant squealing that nobody enjoys.

Some MESs specify extra deep flanges on passenger stock, but I believe this is misguided. If track is so bad that the 'reserve' of a deep flange is needed, then the track needs relaying! The geometry of extra deep flanges (given that flangeways on points are generally fixed) means they can only be accommodated by slackening the flange angle, which brings us back to flange tip contact and shaving needles off aluminium rail.

Of course, this aspect is only relevant to ground level tracks. If you only run on raised tracks, which don't have check rails or points, your flanges can be as thick or as deep as you like! You still need the correct flange angle and root radius though.

Gary

EDIT
PS. I agree with Steve about the corners on bar rail, but by shopping around it is possible to track down bar with quite rounded corners (we have some at West Huntspill), where material from a different mill might be quite square, and we would all agree that is a Bad Thing for trackwork. However I am agnostic about whether matching the rail inclination to the wheel conicity is of any benefit in small sizes. I'm sure that scale effects and the huge masses involved mean that in full size, the wheels will force the track to conform even if there were a small difference in the actual angles. I'm not convinced this happens in our scales though, certainly WRT locos in 5" and below. Even with Hydrostatic Displacement, the likelihood is that a tiny (and inevitable) difference (fractions of a degree) between wheel angle and rail angle will produce a very small contact patch whatever happens, so the practical benefit of matching the angles 'nominally' will be small. Passenger vehicles, and heavyweight locos in 7.25" with greater axle loads might be different.

[David]

May 29, 2021 9:44:35 GMT Gary L said:

The point about flanges, which many model engineers don't appreciate, is that they are only there to provide a safe transit through points.

Nicely done on the beginning and end of that sentence, you've made some very interesting points here.

Seriously though, good arguments for coned wheels and I hadn't thought about how a straight flange would want to climb the rail and lead to trouble.

[steamer5]

Hi Gary,
Great discussion!
Ive built a couple Of David Hudsons trolleys, took me a bit to get my head around how his profiling tool is used / worked. Once I figured it out it make a great job of putting the profile on. I tried one axel on my club track its was fascinating watching it self center. Once I had the trolleys working with no seat I went to a friends private track, sitting on a box as a seat watching the wheels work was interesting!
On the deep flange issue on ground level, on one of my away trips one of the guys had 5 out of 6 flange have bits smashed out of his wheels...cause the local council had put a concrete path across the track & had managed not put the concrete down enough between the tracks, mind you the gent also had deep flanges...

Cheers Kerrin

[millman]

Kerrin, any chance you could provide details of David Hudson’s profiling tool and how it works, I have 24 passenger trolley wheels to machine in a few weeks time.