BR Standard 9F Tyre Profiles

[don9f]

Hi, from time to time the subject of the 9F tyre profiles crops up....particularly the profile of the "Flangeless", centre drivers. When I built my 5 inch gauge Les Warnett 9F model back in the 1980's, I just machined the flangeless profile as per the drawings and this simply followed the same coned profile of all the other wheels.
Part way through the build of this model, I became involved with the restoration of the "real" 92214 and I eventually learned that, whilst when first built, the 9F's did have the flangeless profile coned like the other wheels, this was modified at some point to a different "cylindrical" sort of profile.

For anyone interested, the following profile drawings for the 9F are taken from the old "MT 276" that was one of the Maintenance Documents that was relevant to Private Owner vehicles running on British Rail infrastructure. (This document is available on-line).

So, in steam days, the pony truck, leading drivers, trailing drivers and the tender tyres all had the "A" profile, which had the normal 1 in 20 coning (2.86 degrees).
For many years and still today, the equivalent to this is the "P1" profile.



The leading intermediate and trailing intermediate drivers had the "E" profile which was the same general shape as the "A", but had a slightly thinner flange....including about 1/8" skimmed off the back face. This would affect the way these wheels interacted with check rails on curves and through pointwork etc. This feature, coupled with greater sideplay on these two axles, enabled the 9Fs to negotiate a minimum 4.5 chains radius curve (99 yards or 90.5 metres radius).
The equivalent to this today is the P9 profile, although it's not actually exactly the same.



Finally the much talked about "Flangeless" centre drivers. These eventually had the "X" profile as related above. This axle actually had very little sideplay but of course it didn't need it and note that the width of the tyre is a bit greater than the others, although the "back to back" dimension is the same, so the extra width is on the outside.
There is no modern equivalent to this profile, so it's regarded as a "Special". I assume that anyone re-profiling 9F centre drivers has to either make use of a special CNC program, or template (if the lathe uses such a thing), or create the profile manually?
During our ownership of 92214, we once explored the possibility of having the drivers re-profiled at the Ilford Car Shed's Ground Wheel Lathe, for which someone there was creating a special CNC program for the "X" profile, but the engine was sold before we ever really needed the work doing!



In the other "Wheels video" thread , there were references to wheels "slipping" on curves, either due to the coning not being enough to cope with the differential rail lengths, or in the case of the flangless wheels not having coning at all!, but another factor touched on that is particularly relevant here, is that the wheelsets of a 9F are not properly radial to the centre of a curve....due to the length of the wheelbase etc., so some slipping is bound to occur anyway as the wheelsets adopt a slight "crabbing" effect.
From my own experiences, when negotiating a sharpish curve on a dry day, 92214 would make this phenomenon very "audible", whereas on a wet day, it would just glide round the same curves almost silently!

Hope this has been of interest....

Cheers   Don

[lesstoneuk]

Excellent explanation Don, I remember when Evening Star moved around the yard at Didcot the creaks, bangs and squeals as the axles equalised on the curves. Beautiful loco, racehorse lines, no wonder they did so well on the S & D.

[92220]

Just as a matter of interest, in general it is considered that scale wheels are a no-no for a working loco. I remember when I let it be known that I was going to make my wheels to scale, everybody said DON'T!! The middle wheels will drop inside the rails on curves. I'm afraid I was bolshie enough to ignore all the criticism, and did make my wheels to scale. When they were fitted on their axles, and in the frames, I took the chassis down to my model engineers club track - Cheltenham Model Engineers - to try pushing it around to see just how bad the problem might be, and then to remake the centre wheels if necessary. I needn't have worried. The chassis rolled around the track fine, and the rims came nowhere near dropping inside the curves, and Cheltenham track has some tightish curves!

Bob.

[don9f]

When my 9F was quite new and therefore bearings etc. were “unworn”, I took it to a ground level track near Derby where I was told that there was a particularly tight radius turnout from the steaming area, onto the main line.....IIRC it was 27ft radius (but couldn’t say if it had any gauge widening or not).
In 5 inch gauge, this is approximately equivalent to a “full size” radius of 102 yards, or 4.6 chains....very close to the minimum for a 9F! Anticipating trouble with the drivers on this turnout, it surprised me that they were fine and it was the pony truck that derailed, due to insufficient sideways movement. I modified the pony truck to gain a little extra movement, but never went to this track again and probably never will now! (Couldn’t drive on a ground level track anymore).

Cheers Don

[Gary L]

Jan 9, 2022 16:18:59 GMT 92220 said:

Just as a matter of interest, in general it is considered that scale wheels are a no-no for a working loco. I remember when I let it be known that I was going to make my wheels to scale, everybody said DON'T!! The middle wheels will drop inside the rails on curves. I'm afraid I was bolshie enough to ignore all the criticism, and did make my wheels to scale. When they were fitted on their axles, and in the frames, I took the chassis down to my model engineers club track - Cheltenham Model Engineers - to try pushing it around to see just how bad the problem might be, and then to remake the centre wheels if necessary. I needn't have worried. The chassis rolled around the track fine, and the rims came nowhere near dropping inside the curves, and Cheltenham track has some tightish curves!

Bob.

This story reminds me of Scalefour (and Scaleseven). Everybody said it was impossible, and then ‘everybody’ had egg on their faces because it manifestly did work. Admittedly the proponents also put a lot of effort into their trackwork, and the vehicles are generally sprung (or at least equalised) throughout. In ME sizes, the former is not always to be relied on entirely but the latter is always true, so why not? Full marks to Bob for making the commitment.

In another thread, I suggested people make their wheels to the latest standard for their gauge, and ignore what the (model) designer drew. It seems to me that this is the most practical way forward for the ‘average’ Model Engineer, but Bob is not in that category. I would add that the various Standards are all a compromise, usually specifying a deeper than scale flange. There are good technical reasons why Bob’s approach was unlikely to fail:
(1) because on track with decent scale curves the flange should never contact the rails at all, except for checkrails on points
(2) the flange depth compromise means that the flange angle is also compromised (you can’t have one without the other, it’s impossible), You can't have an angle as slack as the prototype if you want the deeper flange. This means that the danger of 'flange climbing' derailment sets in at a somewhat larger radius than with Bob’s scale flanges, the opposite to what you might expect.
(3) I think I can guarantee (!) that Bob will spend more time than most of us will on optimising the suspension, which is the other ingredient to keeping the tyres securely on the rails.

I can’t emphasise enough how important that flange angle (and the root radius) is. Lack of it is the reason for the well known phenomenon of alloy rails shedding “needles” on the outside of sharp curves. (Steel rail would do the same if it was softer; instead it grinds away the cast iron flanges instead!) Our weights are generally out of scale, which is why flange climbing can be converted into this rail abrasion; whereas a lighter vehicle might just flange-climb and derail.

Great stuff Bob: Respect!

Gary

[ettingtonliam]

I'm curious to know why it is thought necessary to have a special CNC program for the X profile tyre. For goodness sake, its cylindrical with 4 well defined chamfers, and it isn't as if theres going to be a production run of hundreds of the things is it?
A CNC program for the A and modified E profile, I might understand, to form the fairly complex shape of the flange.

[lesstoneuk]

Jan 10, 2022 0:33:51 GMT ettingtonliam said:

I'm curious to know why it is thought necessary to have a special CNC program for the X profile tyre. For goodness sake, its cylindrical with 4 well defined chamfers, and it isn't as if theres going to be a production run of hundreds of the things is it?
A CNC program for the A and modified E profile, I might understand, to form the fairly complex shape of the flange.

The beauty of CNC, is that once set, its a doddle producing identical items.
Plus its easy to modify the cross section with a few mouse clicks. For a 9F you'll need 4 to leading/trailing cross section.... 4 to intermediate cross section and 2 to flangeless. CNC is only a tool and a means to an end. If I was making a Riddles 9F and had CNC capability, I'd use it in a heartbeat.

[Gary L]

Jan 10, 2022 2:13:22 GMT lesstoneuk said:

Jan 10, 2022 0:33:51 GMT ettingtonliam said:

I'm curious to know why it is thought necessary to have a special CNC program for the X profile tyre. For goodness sake, its cylindrical with 4 well defined chamfers, and it isn't as if theres going to be a production run of hundreds of the things is it?
A CNC program for the A and modified E profile, I might understand, to form the fairly complex shape of the flange.

The beauty of CNC, is that once set, its a doddle producing identical items.
Plus its easy to modify the cross section with a few mouse clicks. For a 9F you'll need 4 to leading/trailing cross section.... 4 to intermediate cross section and 2 to flangeless. CNC is only a tool and a means to an end. If I was making a Riddles 9F and had CNC capability, I'd use it in a heartbeat.

It's horses for courses isn't it Les, but I agree with your point. I don't have CNC and despite Roger's inspiring example, a lot of us are unlikely to be able to invest in a CNC lathe. But for wheels, there is a lot to be said for CNC, because to get the flange root radius in particular, form tools have their limitations, especially on the relatively light equipment that most of us are blessed with.

At a slight tangent (bear with me), until recently I was in charge of the 7¼" rolling stock at Bristol. All of the stock is mounted on self-steering bogies, which are very good indeed, but their workings are not easily understood, so I made it my business to research the subject in depth. The wheels are effectively scaled up from the David Hudson profile, which is in turn a slightly simplified version of the BR P6 Heuman profile. The system has been proved in service with us for nearly 20 years, and it gives a markedly lower rolling resistance, amongst other benefits. The tread profile is an important part of the system, and consists of a 3mm flange root radius, blended into a 38mm concave radius which forms the bulk of the tread width, running out to a 3 degree cone at the outermost part of the tread. Still with me?

David Hudson's articles (ME 21.2.03 onwards) include a profiling tool to turn this rather intricate shape, but in 7¼" it was hard to get acceptable results from it due to chatter. It is very important to the self-steering function that the wheels in a wheelset are identical, both in form and diameter, and highly desirable that all the wheels in each bogie are likewise. So for this application there really was no realistic alternative to CNC, and that in turn meant they had to go out to contract.

So these wheels are at the opposite extreme to the 9F centre drivers, and the break point must lie somewhere in between. But for any flanged wheel, you should not compromise on the standard root radius, because it is not there for appearance or even stress-relief, it is there because it is crucial in keeping the wheels on the track by keeping the flanges clear of the railhead. It can be turned with a form tool, particularly in the smaller scales, but even as a non-CNC user I can very much see the advantages of CNC for this job in particular. There are alternatives (like coordinate turning) but, in quantity, CNC would always get my vote, even if it meant paying someone else to do it!

Gary

[Neale]

Speaking from personal experience (although I'm sure that someone has done it this way before - described as "manual numerical control") there is a "middle way" to turn accurate profiles, between form tools and full CNC. I've just finished 4 bogie wheels this way, and turned a set of 6 tender wheels a while back the same way. Essentially, it comes down to generating a toolpath to follow (in my case using a 5mm circular carbide tip) and then slowly moving the tool to each of the coordinate pairs extracted from the toolpath file, but by manually turning the lathe controls and using the DRO for positioning. Of course, you could use dials but a DRO is much easier and less error-prone! I must go back to the original Hudson articles but I cannot remember now if he used a similar technique or a form tool. In fact, I just cut the flange down and just past the root this way, using the topslide set at an angle to produce the coning, but obviously if you have a profiled "cone" section, you could use the same technique. Just takes a bit longer! It sounds like a very slow process but in practice with a set of numbers printed out and a marker to keep your place in the list, it's quicker than you might think. Maybe if I were producing wheels commercially it's a non-starter - but then, it would justify a CNC lathe.

Generating the actual toolpath was easy. I used Fusion 360 to replicate the wanted profile (using the GL5 recommendations). I then used F360 CAM to pretend to cut this profile as if it were a flat piece of work and using a 5mm endmill. That gave a gcode file with a whole series of G1 commands and it was an easy job to extract the actual coordinate pairs which went into a spreadsheet so that I could tweak to get the starting coordinates correct, etc. I can hear some readers already throwing their hands up in horror - "I could have made a form tool and got the job done in the time you spent playing on a PC!" - but we each take our own path through the field of model engineeering and I'm a damn sight better with a PC than a lathe! In addition, of course, this technique (failing access to a CNC lathe) works for almost any profile on even a smaller lathe where a form tool is likely to chatter.

In due course, I shall be turning my D&C wheels the same way.

[don9f]

Jan 10, 2022 0:33:51 GMT ettingtonliam said:

I'm curious to know why it is thought necessary to have a special CNC program for the X profile tyre. For goodness sake, its cylindrical with 4 well defined chamfers, and it isn't as if theres going to be a production run of hundreds of the things is it?
A CNC program for the A and modified E profile, I might understand, to form the fairly complex shape of the flange.

I was writing about CNC programs for the flangeless wheels in the context of re-profiling “real” 9F wheelsets in ground wheel lathes.

Modern wheel lathes at Maintenance Depots etc. are CNC controlled and do the re-profiling with the wheelsets still in position....usually supported on rollers and axleboxes clamped down. The wheels / axles then revolve in their own axleboxes and the result is very accurate and suitable for high speed running. I doubt whether it’s possible to control them manually, but I was never a wheel lathe operator!

Where I last worked, the lathe there could re-profile individual wheelsets, holding them between centres, but would not have been able to accommodate a complete steam loco like a 9F, as there was no way to clamp the axleboxes down. Either way, the Management wouldn’t entertain the thought....so that was the end of that! At the time, Ilford Depot were prepared to take the job on, but then I think there was a sudden policy change and that avenue got closed off as well!

Cheers   Don

[steamer5]

I’ve built 2 x 8 wheel riding trolleys to David Hudson’s design……well almost, mine are able to be either 5 or 7 1/4, raised or ground level.
Anyway David provide the design for the tool post mounted profiling tool using a 5 mm round carbide bit. He also described how to use it to get the profile correct. Basically rough the wheel out using normal tools, install the profiling one & away you go! The tool cuts the curve & the using the top slide cuts the taper, reset to zero & repeat!

I tried a single axel on our track from the hi point I gave it a light push, it was interesting to watch as the axel moved back & forth until it centered itself gathering speed as it went,

Cheers Kerrin