Will L wrote
At to the truth of the assertion that equal distribution maximises haulage power, we have had this discussion on the forum before, although on the last occasion it was because I was questioned in my assertion that the pulling power of a coupled chassis is limited by the most lightly loaded driving wheel
Yes, I understood all that (and was part of the discussion). The problem is, I haven’t managed to demonstrate it in practice.
Here as an example is one of my best achievers of a high coefficient of friction (defined by me as drawbar pull / adhesive weight):
The loco is an 0-4-0T ‘Pug’, built in the 1960s. It has a rocking front axle and fixed rear driven axle and, incidentally, split frames, double-reduction gearing, a flywheel and a clear cab. It has a 24mm wheelbase.
It weighs 87gm, with 52gm on the front and 35gm on the rear axle. Stability is lousy, and it tips with a little pressure on a front dumb buffer. I suspect it would not be happy at pushing wagons if there was an asymmetric buffer load - so all in all a rubbish design.
On my drawbar pull test rig (consisting of a thread over a wheel hanging down below the baseboard to a little aluminium tray that I can put weights in), it produced the following results:
Going forwards - 15gm lifted without slipping; 20gm with slipping; 25gm with frantic slipping
Going backwards - 15gm lifted without slipping; 20gm with a lot of slipping; 25gm no progress
To test the equal weight distribution theory, allowance needs to be made for weight transfer. The coupling is approx. 14mm from the rails. With a 15gm drawbar pull and a 24mm wheelbase, the weight transfer is 14/24 x 15 = 9gm.
Going forward, the effective axleloads are 52 - 9 = 43gm (front) and 35 + 9 = 44gm (rear)
Going backwards, the effective axleloads are 35 - 9 = 26gm (front) and 52 + 9 = 61gm (rear)
This does give a little support for the equal distribution theory in that the performance, once slipping, is a little better with the more equal effective axleloads - but I’m clutching at straws! Basically, there is not much difference between equal weights and weights in the ratio 2 : 1.
The remarkable thing about these results is how good they are. An achieved coefficient of friction of 15/87 = 0.17 is usefully higher than the 0.15 I've come to think as what you can generally hope for and, with some slipping, it rises to around 0.23. That then brings the awkward question of why I get better results with slipping than the best I can get without slipping? Control, incidentally is through a variable voltage regulator.
I think the problem with the theory may be that it is unrealistic to expect our locos to operate in the narrow band between static and sliding friction - the band exploited by sophisticated diesel locos like Classes 59 and 60 - and no doubt many more since.
Regards Andrew