Wheelslip is it accurate ?

[Torque55]

It's been a long time since I posted something, please be patient with me. I am not very technically minded and perhaps I'm not fully aware of the proper train jargon.
Anyway, I still hope someone is able to answer my question and I do hope it's not a silly question I'm asking.

My question is: Is the way that wheelslip works in TS2013 accurate, I don't mean the graphic side (like slipping the wrong wheels etc), but I mean on the tecnical side of TS.
In my opinion wheelslip feels applied too soon or I'm stuck in first notch to get up to speed.

Some details about what I'm driving in TS2013:
The new class 66 v2.0 (with the freightliner livery).
I do think that the new class 66 or the v2 version is very well designed and looks perfect to me.
But my main concern is still the wheelslip.
When observing real life trains and particularly the class 66 it seems that they can apply a reasonable high amount of throttle right from the start without slipping one wheel.
But in TS2013 it is sometimes impossible to apply high throttle right from the beginning and even with sanding on, it just will not get good grip.
I'm driving in the summer with dry weather.

Aside from me operating the train incorrectly (if this is the case), do you guys think that wheelslip is working as intended within TS2013 (besides the graphical respresentation flaws) ?
Is it possible for me to change the wheelslip numbers or is it advised to keep everything at default values ?
I know where to change the wheelslip numbers and how to edit files with converting them first with "serz.exe".

Anyway, thank you for reading!
Cheers,
Jan.

[AndiS]

Unfortunately, wheelslip is considered broken. A search for this term here in this forum will lead to several desperate threads. Basically, it seems to depend on the number of wagons which is nuts. There would be numbers in the engine .bin file controlling adhesion, but they seem to be ignored by the game.

[bdy26]

Its affected by the starting state of the consist - if you add wagons to a loco only consist it will behave differently.

Physics not the strong suit of TS13.

B

[ttjph]

The .bin file numbers (at least the 'dry' one) seem to work OK provided that you load the scenario light engine (and changing them certainly has an effect).

I haven't done a comprehensive test, but the relationship between max tractive effort, max current, weight and friction coefficient is somewhere near, ignoring the consist length insanity.

[Torque55]

Ah I see, I had my suspicions indeed and thank you very much for confirming that the wheelslip or adhesion is not entirely as it should be or broken.
Thank you AndiS, bdy26 and ttjph for your comments ! Highly appreciate it.

I still consider myself not a very advanced user, but beginning to understand all the 'technical' issues that arises when trying to get as close as possible to the real life trains and their behaviour.

AndiS, thank you for pointing out that there were muliple topics about wheelslip, I should have searched better, but I was a bit impatient to be honest.
At least there is a topic now about wheelslip/adhesion in the TS2013 section of the simulator subforum on UKTS.

Also I'd like to thank all of you guys and many others who are still trying to script or manage to change certain files to create an as close as possible simulation of the real thing.
I also downloaded the class 70 diesel mod or patch files and they seem or at least feel to me that the operating of the train is better, although it would be very subtle to notice but still, any improvement is fantastic.
Anyway, I'm highly appreciative of all the efforts and high expertise you guys have and your regurlar posts on the forum.

edit:
Btw, the file I was referring to for wheelslip or adhesion numbers seem to be in this place (I take the class 70 diesel as example):
"F:\steam\steamapps\common\railworks\Assets\RSC\Class70Pack01\RailVehicles\Diesel\Class70\Default\class70_default.bin"
I am able to change these numbers and seem to take effect ! :

Code: Select all

<DryFriction d:type="sFloat32" d:alt_encoding="000000000000F03F" d:precision="string">1</DryFriction>
<WetFriction d:type="sFloat32" d:alt_encoding="000000000000E83F" d:precision="string">0.75</WetFriction>
<SnowFriction d:type="sFloat32" d:alt_encoding="000000403333E33F" d:precision="string">0.6</SnowFriction>

if I increase the numbers I can pull very heavy trains without slipping as much as before, although I realise that I don't have any scientific backup in how I should choose the appropiate numbers.
Anyway, off to driving again !
At least the changes seem to work in a free roam environment, have no idea how it will work within a scenario or other kind of driving experience.

Thanks,
Torque55 or Jan.

[ttjph]

Yep, those are the numbers.

There's one scenario supplied with the 70 where it's raining and you start off with a heavy train - it seems to slip quite appropriately in that particular setup, and given the consist length nonsense I haven't messed with it otherwise.

If you did want to play, I believe that they do equate to coefficient of friction when you start light engine. A 70 weighs 129 metric tonnes, so 1265.5 kN, and has a max tractive effort of 534 kN. Therefore, it can produce a coefficient of friction of at least 0.422 on dry rails. Set the dry friction to less than this and it should slip before reaching max amps (8250 IIRC). Set it to half of this and it should slip at 4125 amps...

Good luck!

[black8]

Hi ttjph

Here's another Dutchman

I have been reading these comments on the questions from Jan and I have seen you and AndiS writing things like "...it seems to depend on the number of wagons which is nuts." (AndiS), "...and given the consist length nonsense" , and "...ignoring the consist length insanity." (both from you)

What do you mean by that? The length i.e. weight of the train you are pulling does have an effect on slipping, right? Or am I totally off here and do you mean to say something else?

Thanks
Jos

[AndiS]

The number or wagons is irrelevant. The total weight (mass) of the wagons is what counts. One wagon at 100 t ought to cause as much wheel slip as 10 wagons at 10 t each.

(You may argue that the uneven force of resistance of 10 small freight wagons with loose chain couplings causes wheel slip earlier than a single modern 12-wheeler. But that are the finer bits. Then you also need to take into account if it is a tank wagon or not. But the game does not even consider whether the wagon is loaded at all or not.

[AndyUK]

Sorry to disagree but in reality, rather than in TS2013's broken wheelslip model, what counts is the force applied by the loco to the rail head. If that force exceeds the frictional force generated by the loco's mass and the coefficient of friction between wheel and rail then slip will occur. Granted that to move a heavy train the driver is likely to apply more power and therefore more likely to get slip than with a lighter train but at the end of the day it's the force applied by the loco that causes slip, not the trailing load.

Note that for simplicity I've ignored weight transfer and systems such as creep control in the above but I doubt very much that TS2013 attempts to model those.

Andy L

[AndiS]

Of course, I meant all of the above under the assumption that the engine is the same. The point is solely that the wagon count is irrelevant (in reality but not in RW) while the wagon mass is relevant (together with rolling resistance, acceleration and grade; and air drag, for higher speed only).

[nschichan]

The point is solely that the wagon count is irrelevant (in reality but not in RW) while the wagon mass is relevant (together with rolling resistance, acceleration and grade; and air drag, for higher speed only).

Hi,

My understanding of wheel slip is that only the weight on the powered axle is relevant.

Taken from http://en.wikipedia.org/wiki/Rail_adhesion :

f_max = friction_ coefficient * weight_on_wheels

f_max being the maximum force transmitted to the rail before wheel slip occurs.
friction_ coefficient (I also refer to it as "rail adhesion") is dependent on the time of the year (leave fall making the rails slippy) and the weather conditions (rain, snow, ...), and also more often than not industrial pollution.

If we take the class 66 (not the /6 subclass) as an example we have:

- weight: 1271.38 kN
- maximum (starting) tractive effort: 409 kN
-> this requires then a rail adhesion coefficient of at least 0.321 (weight / maximum_starting_te) if you want to perform a notch 8 start without slipping.

Now considering the class 66/6:

- weight: 1271.38 kN (assuming no changes)
- maximum (starting) tractive effort: 467 kN
-> this now requires a rail adhesion coefficient of at least 0.367

Typical rail adhesion under favorable weather condition, is 0.3 but I would guess that the supercreep technology fitted on the class 66 could allow a slightly higher rail adhesion.

Also, the weight on wheel varies with the gradient (not a big problem on a falling gradient, but that's another story if you have to start an heavy train on a rising gradient).

My uneducated guess on that is that the weight on wheel should be multiplied by cos(gradient_angle).

On an 1in37 grade (I have seen somewhere this is the maximum gradient on the network rail tracks) this should be negligible (the angle is 1.548 degrees and the associated cosine value is 0.99963).

To fix the simulation, It would be probably possible to script it, but I have not the slightest idea how fast the wheel would slide once the maximum tractive effort has been attained and how this would impact the power transmitted to the rail.

Nonetheless, It would be great to have this longstanding issue fixed once and for all. This would allow for some interesting scripting of the speedometer needles (brutal rise when applying power, brutal fall if wheelslip occurs during braking). Soundwise it would be interesting too (I recall of a youtube (sound only) video of a 319 slipping where the engine pitch would rise slightly each time a slip would occur).

But until it is fixed I would personnally not really bother with wheel slip. I would just put high values DryFriction/WetFriction/SnowFriction and be done with it.

Regards,

[AndiS]

You summed up the part I left out. We were just discussing the other half of the equation, i.e., the train part. Of course, on the engine side, all the things are valid that you list. However, the engine side is not so bad in RW. Is far as I know, entering the adhesion figures and trying it out with the light engine on a grade works ok. It is the way in which train length (instead of weight) is taken into account which causes the disappointment.

[Torque55]

Thank you all who commented, interesting discussion.
I think that I'm a bit over my head with this one.
I think the safest way to deal with this for me personally is to leave the original numbers intact and only change them when I'm absolutely certain that change is needed.
Also I like to do this only when I create a scenario by myself (I kept all of the original files of course).
So I'm tinkering only with files when I'm in my own scenario (free roam/standard).
When I'm driving in someone elses scenario or the ones that came with the sim I revert back to the original files or released patches on UKTS who are made by people who know what they're doing.

I am not going to dig any deeper into this subject, for me it is a bit too technical or scientific.
Anyway, I do read every comment and I seem to understand this adhesion/friction better but I'm rather novice with science and engineering.
But what you all have mentioned in the comments makes sense, even for me.

I will keep an eye on the physics subforum and will only post a question if I'm certain that I know a bit what I'm talking about

edit:
If I understand correctly is that the RW wheelslip problem is that the total train length is taken in account when wheelslip is applied and whether that particular consist is heavy or not, that is not measured.
Sorry if I can't explain it in better means or words, I am dutch afterall but I'm trying (I mean that English is not my native language). Anyway, I think I do understand it even when my written explanation is a bit wrongly worded. my apologies for that.

Cheers,
Jan.

[AndyUK]

...It is the way in which train length (instead of weight) is taken into account which causes the disappointment.

AndiS you seem to have missed the point that I and nschichan are making, the disappointment is that nothing about the train that the loco is hauling should to be taken into account. The only weight that matters is the weight of the loco on the powered axles.

Andy L

[black8]

....the disappointment is that nothing about the train that the loco is hauling should to be taken into account. The only weight that matters is the weight of the loco on the powered axles.
Andy L

OK this is interesting information. I think I understand; so you mean that in the event a loco tries to start a very heavy train (mass, not number of wagons) you would rather stall the engine than make the wheels slip, right?

Thanks
Jos