Loco Physics, specifically Donner Pass locos

[BoostedFridge]

Hi guys,

Can anyone shed some light on the difference in physics between the Donner Pass DLC locos and seemingly all other RSC locos? The DP units seem to be much more realistic at what speeds they develop wheelslip at, and what tonnage of train is realistic to cause them to wheelslip.

With almost every other engine from RSC you can take a heavily loaded train from a standstill, go to 100% throttle, and the train will accelerate. With the DP units any more than 25% will cause wheelslip with a heavy train, necessitating you to ease in to the throttle realistically. The same goes for what speeds the locos will 'stall' out at on a hill. With a heavy drag on a grade the DP units will get down to about 10-15mph then spin, whereas stock RSC units like the -2, or ES44 will slowly wind down to as low as 0.2mph and maintain traction.

Did RSC/Dovetail do this to make the grades on the DP route seem more challenging? Or does it have something to do with the physics of the freight cars that come with that route? Is it something you can 'mod' the other locos to replicate?

[Brahiam]

Hi There!
You must fix the friction, it is way to low.
try these:
<DryFriction d:type="sFloat32" d:alt_encoding="000000403333D33F" d:precision="string">0.63</DryFriction>
<WetFriction d:type="sFloat32" d:alt_encoding="000000A09999C93F" d:precision="string">0.5</WetFriction>
<SnowFriction d:type="sFloat32" d:alt_encoding="000000A09999B93F" d:precision="string">0.31</SnowFriction>

[mrennie]

Hi There!
You must fix the friction, it is way to low.
try these:
<DryFriction d:type="sFloat32" d:alt_encoding="000000403333D33F" d:precision="string">0.63</DryFriction>
<WetFriction d:type="sFloat32" d:alt_encoding="000000A09999C93F" d:precision="string">0.5</WetFriction>
<SnowFriction d:type="sFloat32" d:alt_encoding="000000A09999B93F" d:precision="string">0.31</SnowFriction>

Those friction values are way, way too high!

There have been several studies on the friction between rails and steel wheels, even a PhD thesis that I found, and they all agree more or less on figures such as these:

-- Dry and clean 0.25–0.3
-- Dry with sand 0.25–0.33
-- Wet and clean 0.18–0.20
-- Wet with sand 0.22–0.25
-- Greasy 0.15–0.18
-- Moisture 0.09–0.15
-- Light snow 0.10
-- Light snow with sand 0.15
-- Wet leaves 0.07

(extracted from the LUA script for the FEF-3)

[Brahiam]

In the real world I could agree with you but don't forget that we are talking about Railworks. There is no need of a Phd to know that...

[BoostedFridge]

Thanks for the comments guys. Its good to know that its fairly easy to change. I had initially thought that these units were MORE realistic because of their propensity to wheelspin, but perhaps I'm wrong. Anybody have any experience as to how these compare to a real world loco?

[mrennie]

In the real world I could agree with you but don't forget that we are talking about Railworks. There is no need of a Phd to know that...

That's right, the wheelslip physics in RW are wrong anyway (they made it depend on the mass of the consist, whereas that should have no bearing at all other than causing the loco to have to use more of its tractive effort to get the train moving, which is what then leads to wheelslip), but using wrong values just makes it worse.

[Ericmopar]

Actually the Donner SD40s aren't anymore prone to wheelslip than any other SD40s.
The Dash 9s... I think they slip to easily at too high a speed, on clear days anyhow.
I've noticed when mountain railfanning, that most engines can get down to 10 - 15 mph and be in a very high power setting, without sand, on 2% grades. I've never witnessed them in snow however, or during the fall when there is lots of forest debris on the rail head.
I've noticed the Donner Dash 9s slip above notch 3, even when it's clear and dry. (in TS 2024)

Other engines don't really slip at all.

I believe the mass of the train itself to impact the wheelslip a lot. A light engine can accelerate quickly with no slip, and a heavy train has to take it's time getting going, for more than one reason.

[mrennie]

I believe the mass of the train itself to impact the wheelslip a lot. A light engine can accelerate quickly with no slip, and a heavy train has to take it's time getting going, for more than one reason.

Wheelslip ought to depend on just three things - the current coefficient of friction between rails and wheels, the weight on the drivers (which comes entirely from the locomotive's mass on top of those drivers), and the current amount of tractive effort being applied. Wheelslip will occur if current tractive effort > current friction coefficient * weight on drivers. The only reason the mass of the consist has any effect is because more tractive effort (a higher notch in a diesel) is needed to overcome the inertia of a heavy consist and get it moving, and if that t.e. overcomes the product of coefficient of friction times weight on drivers, you get wheelslip. A light engine can accelerate quickly with no wheelslip because it can use less t.e. to overcome just its own inertia. It'll still suffer wheelslip if you whack it straight into notch 8, just like when you floor the gas pedal in a car as you attempt to speed away from the traffic lights. I've put these calculations into the scripted wheel motion in the FEF-3, to get extremely accurate wheelslip. It hardly ever coincides with the indication of wheelslip that appears in the F5 HUD (wheelslip calculated incorrectly by the core code). In fact, I put a value of 10 for all the friction coefficients in the engine blueprint, just to stop that wheelslip indication from ever appearing. The point I'm making is that wheelslip in the core code is borked, so it doesn't really matter what values you put in the engine blueprint, the wheelslip is going to be wrong. Unless you script the wheel motion (like I've done).

[Ericmopar]

I believe the mass of the train itself to impact the wheelslip a lot. A light engine can accelerate quickly with no slip, and a heavy train has to take it's time getting going, for more than one reason.

Wheelslip ought to depend on just three things - the current coefficient of friction between rails and wheels, the weight on the drivers (which comes entirely from the locomotive's mass on top of those drivers), and the current amount of tractive effort being applied. Wheelslip will occur if current tractive effort > current friction coefficient * weight on drivers. The only reason the mass of the consist has any effect is because more tractive effort (a higher notch in a diesel) is needed to overcome the inertia of a heavy consist and get it moving, and if that t.e. overcomes the product of coefficient of friction times weight on drivers, you get wheelslip. A light engine can accelerate quickly with no wheelslip because it can use less t.e. to overcome just its own inertia. It'll still suffer wheelslip if you whack it straight into notch 8, just like when you floor the gas pedal in a car as you attempt to speed away from the traffic lights. I've put these calculations into the scripted wheel motion in the FEF-3, to get extremely accurate wheelslip. It hardly ever coincides with the indication of wheelslip that appears in the F5 HUD (wheelslip calculated incorrectly by the core code). In fact, I put a value of 10 for all the friction coefficients in the engine blueprint, just to stop that wheelslip indication from ever appearing. The point I'm making is that wheelslip in the core code is borked, so it doesn't really matter what values you put in the engine blueprint, the wheelslip is going to be wrong. Unless you script the wheel motion (like I've done).

I understand what you are saying Mike.
I however think train weight still plays a role.
Lets say a light engine move can get going if you slam it to notch 4 without slip. It's far more likely to slip in that same notch 4 if there is a heavy train behind. Increased mass of the train increases the moment of inertia.
A practical example would be a heavy trailer behind my father's gardening truck years ago, vs a truck with no trailer. A person had to remember you couldn't take off without the tires spinning into heavy traffic, the same way as when the truck wasn't pulling the trailer.
It's like holding a dog by it's tail. If you leave it alone, it's like they have built in 4 wheel drive, and they can take off very quickly. If you hold the dog by it's tail, it's feet will quickly break traction.
Adding cars to a train, is like holding the dog by it's tail. (not that I've done that)

[Chacal]

It's called tractive effort in mrennie's post above.
Which combines weight of consist, rolling resistance, grade, etc.

[mrennie]

I understand what you are saying Mike.
I however think train weight still plays a role.
Lets say a light engine move can get going if you slam it to notch 4 without slip. It's far more likely to slip in that same notch 4 if there is a heavy train behind. Increased mass of the train increases the moment of inertia.
A practical example would be a heavy trailer behind my father's gardening truck years ago, vs a truck with no trailer. A person had to remember you couldn't take off without the tires spinning into heavy traffic, the same way as when the truck wasn't pulling the trailer.
It's like holding a dog by it's tail. If you leave it alone, it's like they have built in 4 wheel drive, and they can take off very quickly. If you hold the dog by it's tail, it's feet will quickly break traction.
Adding cars to a train, is like holding the dog by it's tail. (not that I've done that)

Sorry Eric, but a locomotive isn't a dog. It's a fact that wheelslip depends only on the weight over the drivers, the friction between those drivers and the rails, and the force applied by the locomotive (the t.e. applied at a given moment in time) to overcome that friction. That's why locomotive builders put so much emphasis on the weight of a locomotive.
I'll see if I can dig up the technical explanation for it all - I'm sure there are people who've explained it better than I can

[slick204]

Eric posted this a while back and I thought it was a good explanation.

http://www.alkrug.vcn.com/rrfacts/hp_te.htm

[dtrainBNSF1]

How do I change the text? Can anyone give me a rough guide? Which file? .bin or a what?

As I have operated real life locomotives I believe the locomotive set up for Donner Pass is not realistic. Even 1 SD-70M couldn't lift 22 autoracks.

Please help. How do I change it?


Daniel

As I've found in experimenting in the scenario editor there can be a couple of problems. Either: 1) The physics of the locomotive are indeed incorrect or 2) The cars are not weighted properly. Maybe even a mixture of both. I've found a formula that can be used to calculate how much tonnage 1 locomotive can pull up a gradient at certain speeds. First you need to know the total horsepower of your consist and how much trailing tonnage is in the train (not including the weight of the locomotive(s) ). I haven't tampered with car weights so we'll focus on the power of your locomotive.

You took 22 autoracks up Donner Pass with 1 SD70M. Donner Pass has a ruling gradient of 2.4% if I'm correct. If the SD70M is powered correctly, then it should generate 4000hp. According the the consist editor the autoracks are weighted at 74t. each (check to see if I'm right). The weight of your train was 22x74=1628t. (assuming that the weight is 74t. even). We then can calculate the Horsepower per Ton by dividing 4000 by 1628=2.45HpT. Now multiply that by 12=29.4 and divide that product by 2.4 (the gradient) to end up with the quotient of 12.25. If the SD70M is powered correctly it should have crested Donner Pass at 12.25 mph. However because of your post obviously it didn't which proves that the locomotive is not powered correctly.

The extent of my experience with loco's simulation files is in steam locomotives, most recently the SP&S Northerns (with plenty of advice from Smokebox). I'm no expert. I've never worked with diesels but as they're easier to drive than steamers theoretically their simulation files should be easier to deal with. If you were to open the SD70M's simulation file with the serz application it creates an xml file. You can open that file with notepad. Once inside look for maxpower (horsepower) and maxforce (tractive effort). According to the developer docs the maxpower is represented in Kilowatts, where 1kW=1.341hp. Maxforce is represented in Kilonewtons where 1kN=225lbf. With some math you can figure out how off the original figures are and what they should be. Other figures inside the simulation file may also need to be changed because I know at least with steam simulation files it's a balancing act. Before you do anything backup the original simulation file so that in case something bad should happen you can still drive the loco.

The best thing I can tell you to do is download the Developer Docs from the file library here and read through the Diesel Simulation Blueprints section and check out sites like thedieselshop.us for other specs you may need to change.

Again I'm no expert but that's the only thing I can think of and I hope it helps. If anybody out there with more experience dealing with simulation files would like to jump in here please feel free.

[Kali]

I think the adhesion-based-on-consist-weight thing is fixed, btw. Dry adhesion around 0.2 for an older engine seems ok, can't just throw it in notch 8 & walk off. More modern designs will have better suspension & need a bit higher. Wagon weight ( and the lack of ) is something else...

Mmost of the simulation blueprint values don't do anything ( Maxforce just scales the ammeter I think ), so don't get too caught up there. As an example, for a straight electric literally the only thing that matters for the power side of things is the TEvsSpeed graph - the regulator control just sets a fraction of that. I wish the rest were like that!

[dtrainBNSF1]

Good to know Kali