[Jonatan]

What Brandon implies is that there is no inertia of mass in the wheels themselves which causes the instant and limitless runaway when traction is lost, and the sudden stop when traction is regained. In the videos the wheels begin to spool up and then also take some time to spool down because they act as flywheels as long as the traction is lost and friction builds with slowdown, they only fully stop once friction is high enough to overcome any momentum. Note also that in many cases they do not reach escape velocity even when left unchecked because of inherent friction between wheel, rail and moving parts. MSTS inadvertently simulated this correctly as there was a hard-coded upper limit to the wheel rotation speed during wheelslip.

I've had to tune my models so they physically can't slip. I rather have no slip at all than going from no slip to BRRRRRRRRRRRRRRRRRRRRRR the moment the adhesion treshold is reached. This is with the adhesion modifiers set to default in the options menu AND with my best attempts at tuning the eng files in the best way I know how.

There is the Ctrl+A command that completely closes the throttle, although it's not intuitive and therefore not used.

The adhesion models (Davis A B C D etc) are difficult to set up and not everyone has the patience or knowledge/software to do this so the "correct" and "realistic" measures are omitted. In my experience I've found the use of them does little if nothing at all for my adhesion.
Perhaps if the casual OR user had been given every concievable tool to competently redesign their files things would be different.
A comprehensive eng/wag file optimizer as part of the OR package would be desirable to allow the most accurate use of the parameters.

Maybe it's worth looking at the requirement of a ReciprocatingMass parameter to simulate the total mass of inertia in the rotating parts themselves?

Remember: we ask if you can do it because we don't know how to do it.

[Weter]

@RR1
In fact, third number within throttle controller's definition refers the "speed" of regulator's movement...
However, that is NORMAL operation mode's speed.
@Jonatan
Unfortunately, Ctrl-A uses exactly thAt rate of regulator decrease now.

[scottb613]

Hi Folks,

Yeah - I've brought this up a number of times.

I've spent as much time as anyone trying to perfect steam locomotive performance with what's available in ORTS. I can't get anything that resembles what I've watched in RW videos of wheel slippage.

As Jonatan stated - all we seem able to achieve is zero to a million revs per second - there is no middle ground.

I'd really like to see some work done in this area - as managing wheel slippage seems to be a big part of steam locomotive operation - that we're currently missing out on.

Peter - if you have a locomotive tuned to represent realistic wheel slippage - please let us know - I'd be more than happy to take a look. Thanks for all your effforts.

Regards,
Scott

[darwins]

Davis A, B, C and D are I think only used for the resistance to motion of the whole train. (Though I imagine that Davis A and Davis B together with the bearing type could contribute to calculating the mass and inertia of the wheels and motion rotating when slipping.) There are separate parameters for adhesion, but I think the Curtius-Kniffler stuff is only valid for making modern diesel locos stick to the rails.

Something I would like to see in the future for adhesion is having some local variations, perhaps through route editor being able to add the greasy bits where locos stand at platforms, or the odd wet or icy patch in a cutting.

In the old days, in this country, they used to cut all the undergrowth on cuttings and embankments every spring. In the years since they have stopped doing that, trees have caused a big problem every autumn. The juice from squashed leaves is far more slippery than rain or ice on the rails. This is the report of a serious accident caused by autumn leaf fall - Salisbury_Tunnel_Junction.pdf

[Weter]

Yes, this is more painful for trams, who can lose electric connection with rails.

[ATSF3751]

I have to agree with you Scott. It is truly an art to try and make a steam locomotive not slip and keep a handle on it. We truly are missing out on this aspect of steam locomotives. Even a steam locomotive running light can slip if given the right conditions and it is just not possible in Open Rails currently. I do think this needs to be looked at more closely and figure out a solution. I know this is not Peter's fault and I am truly grateful for everything Peter has done for steam locomotives in Open Rails but I feel as if the Open Rails team puts Steam Locomotives on the back burner and only focus on diesel and electric locomotives a lot of the time.

Brandon

[Jonatan]

ATSF3751, on 12 March 2022 - 08:11 AM, said:

I feel as if the Open Rails team puts Steam Locomotives on the back burner and only focus on diesel and electric locomotives a lot of the time.

This might be the case that the team (asuming they're mostly young people raised in our modern times) are more interested in contemporary forms of traction than steam and may resort to online data for the latter in leu of hands-on knowledge, which may not give an accurate picture of how it truly works as the data usually covers a specific part rather than the whole.

Very few people today truly know how to properly run a steam locomotive or how it should perform. My best bet would be to consult as many old timers as possible for input. A true steam engineer being shown the results in form of a video and review what is accurate and what isn't would be a great source of feedback for tuning the simulation.

Steam operation and wheelslip are governed by an infinte number of variables. Something as miniscule as the air moisture can affect the tractive effort.

[Traindude]

One other side effect of wheel slip worth noting is that on solid-fuel (wood or coal) burning locomotives, there is a tendency for the momentary excessive draft during the wheel slip to "tear holes" in the fire bed. This is particularly problematic on mechanically-stoked locomotives since the fire bed is usually 10 inches or less in depth.

This could be potentially simulated by the fire mass rapidly dropping while the driving wheels slip, and the fireman (or the player when the AI fireman is shut off) will have to rebuild the fire mass back up to the ideal level.

...Not sure if this is worth implementing or not, but once we get the wheel slip itself figured out, then this could potentially be simulated.

[steamer_ctn]

Jonatan, on 12 March 2022 - 06:06 AM, said:

What Brandon implies is that there is no inertia of mass in the wheels themselves which causes the instant and limitless runaway when traction is lost, and the sudden stop when traction is regained. In the videos the wheels begin to spool up and then also take some time to spool down because they act as flywheels as long as the traction is lost and friction builds with slowdown, they only fully stop once friction is high enough to overcome any momentum. Note also that in many cases they do not reach escape velocity even when left unchecked because of inherent friction between wheel, rail and moving parts. MSTS inadvertently simulated this correctly as there was a hard-coded upper limit to the wheel rotation speed during wheelslip.

The OR steam locomotive slip model is based upon the physics describe in this book.

The model does have inertia of mass included in it to calculate the angular acceleration (ie the speed of increase of the wheels). The above book is also used as a guide in calculating the inertia.

As always, I am happy to review this model if we can find some good physics references that describe how to accurately calculate the angular acceleration once uncontrolled slip has occurred, and some "worked" example calculations that we can use to compare the OR model to.

Jonatan, on 12 March 2022 - 06:06 AM, said:

The adhesion models (Davis A B C D etc) are difficult to set up and not everyone has the patience or knowledge/software to do this so the "correct" and "realistic" measures are omitted. In my experience I've found the use of them does little if nothing at all for my adhesion.

The Davis values are used to calculate the train resistance to movement, and play no part in adhesion.

The Curtius - Kniffler parameter is used to adjust the adhesion under "normal (non-slip)" conditions.

[ATSF3751]

Here is another book that could come in handy for you Peter if you have not come across it already! Its the Locomotive Cyclopedia of American Practice: Volume 6 from 1922. Not sure if it will help you with wheel slip concept but it could help you out in other areas of steam locomotives. Going back to wheel slip, when I look at American Steam Locomotives VS European it seems that European locomotives have more tendency to slip then North American steam locomotives. A few factors I have thought about is most European locomotives are lighter then North American locomotives and the fact that on European locomotives have the cylinder cock smoke blasting out the front of the locomotive while starting up vs North American locomotives have it blasting out of the sides. I would think this would cause more water to get on the rails in front of the locomotive for European locomotives as that is the whole reason to use the cylinder cocks in the first place to drain water out of the cylinders after sitting for awhile so you don't cause a cylinder to blow under pressure.


Brandon

https://play.google....g=GBS.PP8&hl=en