Coupled wheels - what if the wheels are slightly different sizes?

Please bear with me as I'm not an expert or a technical person, but what happens with a loco's coupled wheels if they are all not exactly the same circumference?

Even if they are only a couple of millimetres out, one (or more) wheel will complete a rotation faster (or slower) than the others and hence get out of sync. If they are are coupled, whether it be an 0-4-0 or a 2-10-0, how does this work? It's unlikely that all 10 driving wheels on 9F will be exactly the same circumference, so what happens to compensate this?

Presumably there is some slippage - but does this add the the stresses and strains of the axles & motion?

Sorry if this is a stupid question, but it's been troubling me!

 

Loco wheels are turned / profiled as a set so will to all intents and purposes be the same circumference. There will be slight differences due to engineering tollerances and these differences will be accomodated by slight slipping between rail and wheel.

Have you considered what happens on a curve. a pair of wheels is solidly mounted on an axle. the inside wheel will travel a shorter distance than the outer so some slippage between wheel and rail will occur.

 

The 1 in 20 taper on the tread compensates for the shorter distance on the inside rail, the wheel will find the correct position to more or less prevent slippage.

 

I guess that it all pretty academic. I would, however, refer you to the advertisement hopefully reproduced of the American solution which appeared in that country's 'Locomotive Cyclopaedia' 1947 edition which covered all aspects of the railway locomotive and runs to over 1300 pages. Whether the wheels given this treatment were of the same size, round or oval, is another matter. I believe it comes under the heading of 'The American Way'.

Or should this have appeared in the 'Have a laugh at silly things' thread in NGC?

 

The wheels are never quite the same size: see photos of locos after having been moved around with their coupling rods removed. No two crank pins will be in the same orientation. The coupling rods maintain the wheels' relative positions at the expense of a certain amount of slipping between the wheels and rail head. This in turn results in a higher running resistance. It was this which led to the retention of singles until into the twentieth century; while train weights were low enough for the limited adhesion to cope, the reduced rolling resistance allowed higher speeds.

 

MT 276 and subsequent standards decree that coupled wheel diameters should be within 0.010" of each other.

For the less numerically open minded; that's about 1/4 of a millimetre!

 

http://i.imgur.com/o5sVz.jpg

 

Only up to a point. After which (i.e. on sharper curves) you get lots of squealing and start to need check rails and flange lubricators.

 

Very true! to degree the taper compensates but it cannot cope with tighter radius's.

 

Its a remarkably complicated subject.
There are a whole raft of complicated interactions going on.
As well as the effects of coning, as highlighted above, there's also "creep", which is a very odd and hard concept to get your head round. Just before true slipping occurs the two surfaces can be said to be moving very slightly in relation to each other, and this is where maxiumum grip occurs.
Then there's also stick/slip effects, and then at the end of it true slipping.
Given a set of coupled wheels on a curve then all this things are happening at the same time in a different manner on different wheels..

 

I seem to remember that in the 80s there was reported rough riding on Mallard or Scotsman, or a n other loco and it was found that the drivers on one side were larger than on the other..... Or am I just suffering from a bad memory?

 

I have heard tales of 46521 having wheels of differing sizes when first returned to traffic on the SVR and how it used to seem to skip, wasn't a very good ride and had trouble with adhesion. I don't know if there is any veracity to the stories.

 

Certainly something that I know nothing of and I was heavily involved with its initial restoration and fired and drove it on many occasions in the 70s and 80s. Always a strong engine and a well riding engine in my experience.

 

If road vehicles are anything to go by, odd sized wheels on the same side shouldn't be a problem.
The Scammell Pioneer had an unusual arrangement of using a single axle and differential with pivoting gear cases to drive two wheels on each end, rather than the more usual method of two separate axles and three differentials. This means both wheels on the same side will always turn at the same speed, like a railway locomotive, except that it uses gears, not coupling rods.
The best illustration I could find of this is a photo of a model upside down. http://www.swannysmodels.com/images/Scammell/chassis8.jpg
Ideally, tyres of identical diameter should be used, but if they are not, the bigger one will always wear down to match the smaller one, the smaller one never wears down even smaller.
Assuming the same is true on locomotives, coupled wheels on the same side would always naturally wear evenly. Can anyone confirm that this happens?
The difference, of course, is the differential, which compensates for uneven wear between opposite sides. It sounds like, from the examples above, that the fixed axles and tapered wheels of a locomotive will constantly push it across the track towards the side with the smaller wheels and this problem doesn't correct itself.

 

A.J. Powell wrote about trouble on one occasion with 44686 giving a truly wild ride when freshly overhauled, after which it was discovered that one driving wheel had been machined 5/8" larger than the other...

It should be noted that coupled wheelsets inevitably wear to a non-circular shape, with more wear at the maximum torque positions. Interestingly, L.D. Porta was an advocate of machining the driving wheelset larger than the other coupled wheelsets to compensate for the greater wear on that wheelset (3mm on the Rio Turbio 2-10-2 class, I believe). I think the logic goes something like:

• Power is only transmitted to the coupled axles via the coupling rods when the driving axle is trying to rotate faster, either by slippage or creep.
• As a result, the driving wheelset will transmit more power on average and therefore wear more, making it undersize compared to the coupled wheelsets.
• If a certain amount of size disparity is acceptable one way, then it should also be ok the other way.
• A greater service life will therefore ensue before the wheels need to be reprofiled because the driving tyres have been worn too far relative to the coupled wheelsets.

I instinctively recoil from this deliberate deviation from perfection as the initial state, but if the overspeed demanded of the driving wheelset is kept in the creep region rather than slippage it may not be so crazy. However, it's interesting to note from @twr12 that MT276 explicitly bans this short of thing.

 

Foden 6 and 8 wheelers up to the late 1970s with worm and worm wheel back axles didn't have an inter-axle differential. This meant that the tyres had to be 'matched' for diameter/circumference between the two sides otherwise 'wind-up' would occur between the two axles. Or to put it another way, one of the axles would be trying to go faster or slower than the other one. A friend had a 1970s 8-wheeler and wasn't aware of this and when I drove it on one occasion found that one didn't need to use the brakes for slowing down due to this. The diffs were so 'wound-up' that it would come to a stop all by itself! By not having an inter-axle differential this made Foden a popular choice for tipper operators especially when going off-road. With the 3rd diff if only one pair of tyres/wheels were to spin, that was it, you were stuck. The Fodens usually kept going.

 

I think that's more likely to be the resistance of the worm drive to being back driven, which is the reason model railway locomotives don't need a brake. A modern tipper lorry with locking differentials will still roll with all the diffs locked.

"Power is only transmitted to the coupled axles via the coupling rods when the driving axle is trying to rotate faster, either by slippage or creep"
This is one of those things that sounds like nonsense, yet perfectly reasonable, at the same time.
Each wheel has got a crankpin, with a coupling rod connecting all crankpins. Why should it matter which one of those crankpins the connecting rod drives?
On the other hand, there must be some clearance in each crankpin bearing. The driven wheel will be under load for a few thousandths of an inch twice every revolution as it takes up that clearance. Maybe, over time, that additional load will cause additional wear.

It would be interesting to know if, when wheelsets are removed for turning, they are all worn to the same diameter, or at least to within 0.010" of each other, or if the driven wheels are usually more worn than the others.

 

I am responsible for the steam locomotives at the 600 mm railway in Mariefred, Sweden.

https://en.wikipedia.org/wiki/Östra_Södermanlands_Järnväg

For the last 45 years we have had an eight coupled German HFB Brigadelok, no. 8 Emsfors in service. With the original Klien-Lindner movable axles front and aft, it has always had a quite bad running, swaying from side to side at anything above 18 km/h.

In the last few years the side rod bearings had started to warm at longer non-stop runnings, so last winter we lifted the loco and checked out bearings and lubrication. Nothing was really wrong enough to explain the problem and the wheel profiles were quite OK.

Finally we checked the wheel diameters, and the unbraked KL-axles were 3 mm bigger than the fixed, braked middle axles. We turned all axles to the same diameter, and heating of the bearings was gone, but most remarkable; the running was as steady as any of the other locomotives at up to 25 km/h!

Sometimes it takes some detective work to find out…

 

I think you may have misunderstood me, or I haven't made myself clear enough. If all the wheels on a rear bogie are exactly the same size/diameter/circumference then, in a steering straight ahead position, should roll freely. In the case I referred to, two tyres on the nearside of the leading axle of the bogie were of somewhat larger diameter etc., than all the rest, which resulted in 'wind up' between the two axles because the lead axle was going very slightly faster that the rear one.

Getting back to railway practice, and to throw some more into the conundrum, but apparently BR Standard 9Fs, the flangeless centre driving wheel set have a parallel tread whereas all the other wheel sets have a normal 1 in 20 taper profile.

 

A complicated subject indeed.

David Wardale was asked by his mentor, Livio Dante Porta, to measure tyre wear and by extension the wheel diameter of of SAR locomotives. These were two cylinder types and DW confirmed LDP's findings in that there were four areas of wear brought about by microslip to be found around the driving and coupled wheels. Porta went on to recommend that the driving wheels should be turned to a larger diameter than the coupled wheels since these were more effected by microslip wear. This microslip wear was felt to lead to axle unloading at certain rotational speeds and so regular light tyre turning to restore a uniform diameter was recommended.

Porta looked beyond how this particular phenomenon affected adhesion and so wanted and developed a high adhesion tyre profile, looked at improved rail head conditioning, recognised that cylinder drain discharge created problems. And went on search for other occurrences that effected the performance of the steam locomotive.