Fixing the brakes on the London and Port Stanley Railway

[Quacking Duck]

I have been trying to use the London and Port Stanley Railway mini-route[1] in Open Rails. I soon noticed, however, that the brakes on many of the rail vehicles in that package do not work. As a representative example of this problem, the following observations were made using the consist "No 2", which contains a single electric passenger railcar.

Upon applying the brake, the brake cylinder pressure rises to 1 psi but no further. As the brake is applied further, the brake pipe pressure decreases further, but the brake cylinder pressure remains at 1 psi. This remains true up to and including when an emergency brake application is commanded.

A search of this forum appears to show that I am not the first person to have experienced such problems. For example, this[2] forum post claims to have solved it in some way involving the following lines in the locomotive file:

Comment ( MaxBrakeForce set to 0.6kN/t )
MaxBrakeForce ( 0.282kN )

I have found these lines in the relevant file, but am not sure what should be changed about this to fix the problem, or even what about this causes the problem. Is a line which specifies a maximum brake force the cause of what appears to be a limit on brake cylinder pressure?

Beyond obtaining working brakes, what should be done to restore the braking behaviour of these rail vehicles to something resembling the original design intent of the developers?

[1] file "lpsv3-mini.zip", file number 89028
[2] https://www.trainsim.com/forums/foru...10#post2322210

[pschlik]

So, the actual problem is a combination of OR not behaving the way MSTS does and the setup of some stock in the LPS route. This stock has BrakeCylinderPressureForMaxBrakeBrakeForce set to 0.50 (psi). Open Rails interprets this number as a limiter on the brake cylinder pressure, preventing the brake cylinder pressure from going above half a psi (rounded up to 1 psi when shown on HUD). Meanwhile, MSTS interpreted this number as, like it says on the tin, the brake cylinder pressure at which MaxBrakeForce occurs-no mention of it being a limiter. The pressure can then be lower or higher than what's given, with the resulting force scaled accordingly.

So now we have lots of content made for MSTS that assumes there's no pressure limit and lots of content made for OR that assumes there is a pressure limit. This happens quite regularly as MSTS's exact functionality has never been very obvious and often deviates from what MSTS documentation suggests the functionality should be, so we end up with an implementation in OR that's not quite backwards compatible yet also can't be changed because the behavior is vital for OR-specific content. Best that can be done there is adding new features in that support MSTS methods better...but require editing the old stuff.

In this case, on testing versions of OR and New Year MG, BrakeCylinderPressureForMaxBrakeBrakeForce can be deleted and replaced with ORTSBrakeForceReferencePressure to provide functionality equivalent to MSTS, allowing this rolling stock to actually provide brake force. There are 'Find and Replace' programs out there which can search for every time "BrakeCylinderPressureForMaxBrakeBrakeForce" is used in the installation and replace it with "ORTSBrakeForceReferencePressure" in one click, so this doesn't need to be time consuming.

Still no idea why a brake cylinder pressure for max brake force of 0.5 psi was used on the LPS route. Mathematically it works, but most people would prefer to work with bigger numbers to avoid mistakes and to replicate actual brake force tests which need to be done at higher pressures.

[Quacking Duck]

I updated to the testing version of Open Rails from within the launcher. Then, using Notepad++, I performed the find-and-replace operation you described.

I am happy to report that the affected rail vehicles now have working brakes. Thank you quite seriously.

Could it be desirable to publish a version of the London and Port Stanley mini-route with this fix applied so that future users will not have to discover and perform the same fix on their own?

What is "New Year MG"?

[jdgremsjr]

What is "New Year MG"?

It is an fork of open rails available at http://www.interazioni-educative.it/...NewYear_MG.zip It uses the Mono Game engine

[baldwin]

I think that all OR versions now use the Monogame engine. The ORMG version to my point of view now acts as a cutting edge development fork because it contains things that are not found in the other versions.

[Train-a-Mania]

Thank you for providing this fix. I'm finding myself nostalgic and returning to Wayne's wonderful work.

I'm still feeling like the brakes are a bit sluggish compared to how they performed in MSTS. What other values should I tweak to improve this?

Unrelated to the brakes, but I'm also noticing the whistle clip while looping inside the cabs. Outside it seems to be fine. Has anyone solved this issue?

[pschlik]

Heya Nick,

Brakes in OR over the last couple years have become a bit more complicated, but more accurate, than MSTS could have dreamt of. That probably means things need tweaking to feel correct. I'd recommend looking at the "Brake Shoe Force" section of the Open Rails Manual. Most real-life brake tests would give you a "brake ratio" at 50 psi brake cylinder pressure. The "brake ratio" being the force on the brake shoes as a percentage of the weight of the train car (in the 1920s, this would be measured with the car empty, in the modern day both the loaded and empty weight are considered). The actual brake force at the wheel would then depend on the coefficient of friction of the brake shoes, and that's what the brake shoe force feature is all about. MSTS never thought about any of this, it just let you define the brake force at the wheels and didn't think about any of this 'brake shoe' or 'coefficient of friction' stuff, so it's less accurate.

For the 1920s era of the L&PS rolling stock, you'd be dealing with cast iron brake shoes. Iron brake shoes bite hard at low speed but do very little at high speed-MSTS doesn't simulate that one bit (brake force is constant with speed), which means in MSTS most historical trains brake too hardly at speed but too softly when coming to a stop, this is going to be the case on L&PS.
To set cast iron brake shoes in OR, use ORTSBrakeShoeType ( Cast_Iron_P6 ) in the Wagon ( section of the engine/wagon. Why "P6"? Well, there is also ORTSBrakeShoeType ( Cast_Iron_P10 ) which would be "hi-phos" (high phosphorous, hence "P10") brake shoes, but that wasn't in use at the time of L&PS.

The brake shoe friction code also needs to know the number of brake shoes, but that's as simple as ORTSNumberCarBrakeShoes ( X ). It's not hard to tell from models and photos if something has one brake shoe per wheel or two brake shoes per wheel. A normal freight car will be ORTSNumberCarBrakeShoes ( 8 ) (8 wheels, one shoe per wheel) but I suspect some of the locomotives and passenger equipment would be ORTSNumberCarBrakeShoes ( 16 ) (8 wheels, two shoes per wheel). The number of brake shoes affects the distribution of brake force, which affects the actual brake friction coefficient developed (more brake shoes is better, as this lowers the force per brake shoe which lowers the temperature change on each brake shoe, though temperature change isn't actually simulated yet).

And then the important piece is ORTSMaxBrakeShoeForce ( Xlbf ). Determining the brake force is the most important part, though we also need to specify how to convert brake cylinder pressure into brake force, which is why you delete BrakeCylinderPressureForMaxBrakeBrakeForce, and use ORTSBrakeForceReferencePressure ( 50psi ). Why 50 psi? Well, most brake testing in that era would have told you the brake force at 50 psi, so using that as the reference pressure removes guesswork.
Anyway, back to the brake shoe force, this is where the "brake ratio" comes in. You find (or in some cases, guess) the brake ratio and then multiply that by the empty weight (in pounds) of the locomotive/train car to get the ORTSMaxBrakeShoeForce in pounds. For the 1920s freight rolling stock with cast iron brakes, the empty brake ratio is supposed to be at most 70% to avoid skidding ("supposed to" being the key word, you could see anything between 50%-70%).

For a beefy 50-ton capacity freight car of the time, the "typical" empty weight of the freight car would be 35 tons, or 70,000 pounds. 70,000 lb * 0.70 = ORTSMaxBrakeShoeForce ( 49100lbf ) at most. Of course, not all freight cars would actually weigh that much (most included with the route are lighter) and this is the maximum so brakes should be weaker by a random amount, it varies. Note that this would usually be the same brake force on the loaded freight car, as very few things in this era would have load/empty brakes (if they did, it would be manually set). The brake shoe force should ideally be below 70% of the empty car weight and above 15% of the loaded car weight simultaneously. (If the two can't be achieved simultaneously, that probably means the car has some sort of load/empty lever.) If we look at our 35 ton empty, 85 ton (170,000 pound) loaded, train car, the max allowed brake force of 49,100 lbf gives a braking ratio of: 49,100lbf / 170,000lb = 28.9%, so this car would have more than enough brake force when loaded as well (this maximum value is arguably too much, so we could be more conservative and give it a 50% empty brake ratio, which would provide a 20.6% loaded brake ratio).

Locomotives and passenger cars are different. As their weight doesn't change as much, they can afford to use much higher brake ratios all the time. Passenger equipment should have a brake ratio less than 90%, with 70% being totally reasonable (though interurbans likely ended up on the higher end of this scale), and locomotives kept below 75%, though 50% would again be reasonable. For those in the audience, this is specifically 1920s era stuff, these numbers changed dramatically (usually, by getting smaller) over the years as things got heavier and brake shoes got stickier.


As an example for why you might want to recalculate things with this new method, let's look at the 55 ton coal hoppers included in the route. These have an empty weight of 20.2 us tons, or 40,400 lb as written on the side of the car (I'm going to ignore the fact that the weight programmed into the .wag file is actually 25 metric tons, or 27.6 us tons). So let's see what the brake ratio is for this car from the MSTS brake force; we have 16.39kN (this is 3,685 lbf) (note: the loaded version of these cars specifies 27.31kN, which suspiciously is 1.66x as much as the empty brake force; this "empty force = 0.6 x loaded force" is a red flag for a 1920s route. That's definitely accurate on many modern things but is totally out of place in the 1920s where load/empty is very rare) and this force is at a max brake cylinder pressure of 69 psi (which is emergency). Converting this to the brake force at 50 psi we get: 3,685lbf / 69psi * 50psi = 2,670 lbf. Remember, MSTS brake force is measured at the wheels, so it needs to be converted to brake shoe force. If we are very generous and assume this brake force is taken at 50 mph, the cast iron coefficient of friction at 50 mph with that amount of brake force is roughly 14%. So, converting to brake shoe force: 2,670 lbf / 0.14 = 19,070 lbf, for a final brake ratio of 19,070 lbf / 40,400 lb = 47.2%. That's reasonable, if a smidge low.
However, that's at 50 mph. Recall that MSTS treats brake force as if it doesn't change with speed but cast iron brake shoes should get a lot stronger at low speed. If the same calculation is repeated at a complete stop (ie: the situation real life brake tests are done in), the iron shoe coefficient of friction OR calculates is now a staggering 51%! Repeating the previous calculation, 2,670 lbf / 0.51 = 5,235 lbf, and then 5,235 lbf / 40,400 lb = 13.0%. That is too low for a loaded car, and much too low for an empty car.


So long story short, at least this coal hopper (can't check everything at the moment) has brakes that are tuned to be reasonable at high speeds, but are unrealistically weak at low speeds because MSTS didn't model variable brake shoe friction, which OR now has the tools to properly model (this is only for testing, New Year, and upcoming 1.6 versions).

There are also a bunch of other things that can be done better with new features, but getting the brake force right using ORTSBrakeShoeType, ORTSNumberCarBrakeShoes, ORTSMaxBrakeShoeForce, and ORTSBrakeForceReferencePressure are most impactful for train handling. Users are likely to notice a car with cast iron brakes gripping hard at low speed. They aren't likely to notice a difference if you set the brake cylinder diameter to the correct size (which is a thing you can do now).

[superheatedsteam]

The number of brake shoes affects the distribution of brake force, which affects the actual brake friction coefficient developed (more brake shoes is better, as this lowers the force per brake shoe which lowers the temperature change on each brake shoe, though temperature change isn't actually simulated yet).

I don't understand this part. If you apply the brakes so that there is 50psi in the brake cylinders, how does adding another brake shoe to the system lower the force on the other brake shoes?

I was taught that the more brake shoes / brake area you have the more heat dissipation, therefore a wider range of braking effect.

I read the OR manual but the information on ORTSNumberCarBrakeShoes was not clear to me.

Cheers,

Marek.

[pschlik]

The total brake force developed depends on the brake cylinder pressure, size of brake cylinder, and brake rigging. Well-designed brake rigging will produce the same brake force at each brake shoe, so you take the force produced by the brake cylinder and brake rigging then divide that by the number of brake shoes to get the force per brake shoe. More brake shoes means splitting that same brake force up into more, smaller chunks so each brake shoe has an easier time. Changing the number of brake shoes won't change the total brake shoe force produced, and you only ever enter the total force into OR.

[Train-a-Mania]

Out of curiosity, is there anyone out there who has completely fixed the brakes on the equipment included with the L&PS Version 3?

For the 20th Anniversary of Wayne's V2, it would be great to have a drag-and-drop solution for other users. I would do this myself, but I lack the time (and some degree of understanding) to do this right now.

If you have the files, I would be happy to work with you on a method of releasing them.

[EvolutionMan]

This thread has some fixed values in the linked post. I applied it on my end and it works fine.

Nick: I'd be willing to contribute the updated ENG files if you ever commit to a re-release. I also have photos from inside the cab of interurban car #14 (taken from the motorman's perspective) which I can make into a working cab view, if that might be of any use.

[Train-a-Mania]

Hi EvolutionMan,

I'll absolutely take you up on the offer for both the updated ENG files as well as a new cab. I've sent you a PM.

[shanig]

thanks for the quality post, the was very informative good