Excessive steam locomotive wheel slip.

jonsearles62

New member
I have a quick question for steam-engine file development. I apologize if this is the third thread I've posted, but the Content Creation thread was way down the list, and I was told that the "General Trainz" forum was less than ideal for my question.

I'm modeling a New York Central L-4 Mohawk 4-8-2, using the E-specs09 Excel sheet, first with standard settings, and then with the scalar, to raise the horsepower to a (supposedly) more accurate 5,400 horsepower. The first of these produced excessive wheelspin over (not under) 30 m.p.h., while the 5,400 horsepower version immediately runs out of steam and water, literally while standing still. The real locomotive reached peak power at 76 m.p.h., and topped out at around 94 m.p.h. The adhesion is set to 4.45, same as the real locomotive, and the weight is accurate. I've set the axles to 7 (after initially trying 4), which is also accurate, as the Trainz Wiki claims this tag is only for rolling drag. I've also tried a standard E-specs09, but with the steam-chest-max-flow raised to 234.7 (in line with the 5,400 horsepower version). This also failed to produce a normal result. I'm continuing to tweak it as of this hour, and so far my results have been even worse than the above.

What am I doing wrong, exactly? Note here that I'm not talking about how the drivers look. I mean that the locomotive is slipping, physically, so much that it gets stuck at 55 m.p.h. with a 10-car passenger train.
 
I have a quick question for steam-engine file development. I apologize if this is the third thread I've posted, but the Content Creation thread was way down the list, and I was told that the "General Trainz" forum was less than ideal for my question.

I'm modeling a New York Central L-4 Mohawk 4-8-2, using the E-specs09 Excel sheet, first with standard settings, and then with the scalar, to raise the horsepower to a (supposedly) more accurate 5,400 horsepower. The first of these produced excessive wheelspin over (not under) 30 m.p.h., while the 5,400 horsepower version immediately runs out of steam and water, literally while standing still. The real locomotive reached peak power at 76 m.p.h., and topped out at around 94 m.p.h. The adhesion is set to 4.45, same as the real locomotive, and the weight is accurate. I've set the axles to 7 (after initially trying 4), which is also accurate, as the Trainz Wiki claims this tag is only for rolling drag. I've also tried a standard E-specs09, but with the steam-chest-max-flow raised to 234.7 (in line with the 5,400 horsepower version). This also failed to produce a normal result. I'm continuing to tweak it as of this hour, and so far my results have been even worse than the above.

What am I doing wrong, exactly? Note here that I'm not talking about how the drivers look. I mean that the locomotive is slipping, physically, so much that it gets stuck at 55 m.p.h. with a 10-car passenger train.
Did I read right? A Mohawk?!?!? YAY!!!!' Is it private or will you share it?
 
Apologies for my absence. In the past few months, I have been far too busy to deal with this problem. To clarify, the Mohawk I'm using was originally built by Wgkandsk, but the engine file on the original doesn't work properly, so I have tried unsuccessfully to build a new one. I now have a video on YouTube with both a demonstration of the locomotive, and a scroll-through of the engine file. I hope this will clarify the problem for the people here.


As back in February, any assistance would be greatly appreciated. The problem is so frustrating that it's a distraction from the larger layout project that I want to use it with (up to 180 miles long at the moment).
 
Sorry jonsearles62 I had something written up yesterday and lost it - somewhere out in electron heaven now. I have a copy of Bill W's NYC L4 4-8-2 somewhere in the archives and I'd also like to dig it out and set up a new espec for it. I had several discussion with Bill Fock (developer of E-specs09.xls), before he passed, about the steam physics simulation in Trainz (or lack of it depending on our moods). If your using his spreadsheet there are a few things about it that might help you out. It is a simplified method that uses just a few input values to calculate the steam-engine epsec. I have my own spreadsheet that's a lot more complicate and requires significantly more input to setup the espec. I feel I can get a closer fit to an existing loco doing it my way but I guess we all have trouble with that aspect of Trainz.

I won't go into any detail in this post but Bill and I both agreed that the power output of steam locos in Trainz can be excessive. We both addressed the fault in a similar manner - namely reducing the cylinder volume to some fraction of the loco's original size. We differed in exactly how and how much we reduced them but we agreed on the reason. If you dig into his spread sheet you see the factor he uses is the reciprocal of the driver diameter in meters. The units are not consistent her only the value. For the L4 with 72 inch diameter drivers (1.829 m) the actual cylinder volume is multiplied by 1/1.829 which reduces the volume and therefore the work done by the steam in the cylinder to approximately 55% of the original ones. This reduces the TE produced and unfortunately the steam consumed by the same amount.

Before I go too far I have to ask if you are discussing wheelslip in DCC control or CAB since the video shows the loco under DCC control mode? Almost none of the steam container tags in the espec are used for the loco in DCC mode. Also I'm not aware that wheelslip is functional or considered in DCC mode.

If you are using DCC control then the appropriate tags to consider are in the motor container: max-accel and max-decel. There is another discussion on these in the Surveyor forum at the moment you might want to check.

I had a question about the value for adhesiveRatio tag you used but I'll have to ask that when I post back later. Oh and where did you get the 5400 hp at 76 mph? I did some calcs a long time ago and thought 4000 IHP at somewhere between 55 and 60 would be closer to what they produced.

Bob Pearson
 
Last edited:
I won't go into any detail in this post but Bill and I both agreed that the power output of steam locos in Trainz can be excessive. We both addressed the fault in a similar manner - namely reducing the cylinder volume to some fraction of the loco's original size. We differed in exactly how and how much we reduced them but we agreed on the reason. If you dig into his spread sheet you see the factor he uses is the reciprocal of the driver diameter in meters. The units are not consistent her only the value. For the L4 with 72 inch diameter drivers (1.829 m) the actual cylinder volume is multiplied by 1/1.829 which reduces the volume and therefore the work done by the steam in the cylinder to approximately 55% of the original ones. This reduces the TE produced and unfortunately the steam consumed by the same amount.

Bob Pearson

This is very useful advice. The key figure seems to be 55%, as opposed to the 60% figured used as a scalar by Fock's original Excel sheet. Although the horsepower rating comes to 3160, which is technically inaccurate, the locomotive performs far better in Trainz. The excessive wheelspin over 30 m.p.h. is still there, but oddly it doesn't seem to affect performance.

A 10-car passenger train can be run at top speed (94 m.p.h.), a 1000-ton freight can reach 60 m.p.h., and a 1900-ton freight can hit 53 m.p.h. It's possible a real L-4 Mohawk would have more brute force, but indeed steam-era freight trains were generally shorter than modern ones, so 50-70 cars seems reasonable to me.

The only remaining bugs, therefore, are in principle the wheelslip, and the cutoff, which has to be set at the maximum, but this is a problem common to many steam locomotives in Trainz 2009. I'm not aware of whether this has been addressed in 2010, 12, and/or TANE.

In any case, thank you very much for your help!! :-)
 
Don't confuse the R parameter in Bill's spreadsheet with the factor he uses to reduce the cylinder volume. They are different. The reduction in the cylinder vol is to fix the problem he found thru testing a lot of steam locos in Trainz. If everything was simulated correctly in the game he wouldn't need to resort to this. I've done extensive testing in the game and I do the same thing when I develop my especs - but not everyone making espec does.

The R param in Bill's spreadsheet is actually the ratio of the speed at max hp to the max speed. Bill uses the old rule of thumb that the maximum speed of a steam loco in mph is numerically equal to the diameter (D) of the drive wheels measured in inches. He calculates the speed (in mph) at HPmax as R x D. To calculate the HP produced he assumes the tractive force at this speed to be equal to 1/2 times the TEmax at starting (the usual rated maximum TE). At any speed v, HP(v) = TE(v) x v / 375 with TE in lbf and v in mph and 1 HP = 550 ft-lbf/s, - the constant 375 converts the units used to HP. Bill uses a slightly different version of this formula so in the spreadsheet you see HPmax = 1/2 x TEmax x R x D x 88 / (550 x 60) which gives the same result.

Anyhow that's the math behind it. The TE at HPmax probably isn't equal to 1/2 the TEmax. The max speed might be closer to 1.25 x D for fast passenger locos. This is just an approximation of the maximum HP and simplifies the input for a lot of Trainz users who want to make their own especs. In practice you can adjust the R param to change the HP to match a known value for a specific loco if you have the info available.

In your case if you want to use 5400 hp as the correct value for the L4 then you have to input an R value of 0.94. That looks a bit high to me and as I note above I don't think they could develop 5400 IHP. Personally I'd probably use a value around 0.7 which gives about 4000 hp which I can justify based on calcs of evaporation rates for the L4's boiler.

You'll note that Bill sets the speed tag for the design condition in the steam container equal to R x D (using 0.447 to convert units to m/s). Since in practice the R param can be used to correct for both speed and TE at HPmax, I don't know if it makes sense to use this speed value or if a more accurate value might make sense here. You also have to set the cutoff and efficiency for this design point. Bill uses 1.0 for the efficiencies and 0.55 for the cutoff so there's probably some room for adjustments here too. But one of his goals was to simplify the whole procedure and come up with an espec that works in the sense that it provides a reasonable approx. of the TE produced.

For those with the necessary knowledge, abilities and data, you can use this as a start and modified as you see fit to come up with a more accurate simulation of the loco's performance in the game. I developed my own spreadsheet that's a lot more complicated and seems to get modified each time I use it. I try to resort more to 1st principles in the mechanics and thermo aspects of the problem and include a simplified steam table. In cases where the performance data is not know I use another program I wrote to estimate it.

Bob Pearson
 
Here's the latest incarnation of the Mohawk physics file...


It isn't perfect, but it will do for now. It actually ran better when I did the playthrough than it had during testing on perfectly flat track. Go figure. It might be that the track type on the Auran Trainz version of the Harlem Line is stickier (I don't know if you can do that to a track spline).
 
Back
Top