A few quick thoughts

[SCLJim]

Thanks for the comments everyone. If I understand correctly there is also a discrepancy on the brake force of cars that can be both loaded and empty? If that is the case do you think it is best to model separate versions of loaded and empty cars in order to have completely accurate numbers?

[GreatNortherner]

Hi,

Yes I think this is correct. RW calculates the brake force by using a Percentage of Vehicle Weight setting. This might have been a simple way for the original coders to develop a braking system that would model the higher brake forces on loaded cars, but it can't produce anywhere near realistic numbers as far as I'm aware of.

And besides, on all default cars this percentage setting is also very high, giving unrealistically strong (or let's call them "arcade mode") brake forces on all cars.

Kali and a few others once did a lot of groundbreaking work in this department, and I think Kali also came up with a new "brake force percentage" setting that gave a good average brake force for LDs and MTs so that at least mixed manifests would brake reasonably prototypically. Unfortunately I couldn't find the thread when I searched for it now, does somebody remember where to find it?

However, coming to think of it, I think your assumption is absolutely right, to get really accurate brake forces for MTs and LDs one would most likely need to set up individual car blueprints for both just like we did in MSTS. So we would have to sacrifice the in-game (un)loading features, but I could easily live with this if I can get my mile-long braking distances back.

Cheers,
Michael

[philmoberg]

... And besides, on all default cars this percentage setting is also very high, giving unrealistically strong (or let's call them "arcade mode") brake forces on all cars. ...

That raises an interesting point, Michael, particularly with respect to the sorts of rolling stock that were common in the early and middle part of the last century. As a general rule, we never made more than a 15 lb. reduction with freight equipment, and we tended to make that reduction slowly. This was because freight cars tended to dump the air with little provocation. Some of the more skillful engineers could make a reduction as low as 3 to 5 lbs. and take it down in 2 to 3 lb. bites after that. Freight trains in those days ran with 60 lbs. in the train line.

Passenger trains ran with 90 lbs. in the train line, and brake valves for passenger equipment (as well as "passenger equipped" freight equipment) tended to be more sophisticated and forgiving. Passenger engineers would often make 25 -35 lb. reductions fairly quickly with no problems. That said, there could problems with a wide variety of passenger brake valve schedules in a single consist. I personally observed this a few times, almost ending up on the floor as a result of one of them.

My impression of RW3 is that the brakes tend to be more consistent with current standards for passenger equipment, on the one hand; while on the other hand, it is possible to make unrealistically high brake applications on older freight equipment without flattening wheels, breaking knuckles or (in rare cases) putting the train in the ditch. IIRC we had the same problem with MSTS, and worked out a series of refinements in a fairly short span of time. The balancing act - at least from the standpoint of developer of a platform such as RW3 - is how realistic the braking (and other) forces can become before the more casual users become frustrated and quit. I'm wondering whether it would [ultimately] be possible to incorporate levels of increasingly realistic physics into the realism levels currently available for controls, or whether there would be an optional realism setting for the physics somewhere else (the settings menu, for instance). FS9 has this sort of a feature, with respect to an increasingly realistic flying model, for example. It would seem RW3's architecture would be more amenable to this than MSTS was.

[SCLJim]

However, coming to think of it, I think your assumption is absolutely right, to get really accurate brake forces for MTs and LDs one would most likely need to set up individual car blueprints for both just like we did in MSTS. So we would have to sacrifice the in-game (un)loading features, but I could easily live with this if I can get my mile-long braking distances back.

Cheers,
Michael

Thanks Michael! It's something I have been contemplating for a while. Every time I think about it, I keep finding myself leaning to the more realistic side of things. I love the challenge of having to think ahead on your moves as far as throttle and braking. In my book that is a huge huge part of what running a train is about, thinking ahead, as opposed to being able to arbitrarily do as you please. I want to know that if I don't set my brakes up at a certain point and a certain reduction, I could very well get by that red board. Right now when I run in RW, I find myself stopping way early and having to adjust. Now that I'm getting ready to release my chip hoppers I'm thinking I will, indeed, go with separate blueprints for loads and empties. As you mentioned it does sacrifice the unloading features but that is honestly something that has never really been very high on my list.

Since I'm not very familiar with Railworks, I'm wondering how would the game react to the realistic brake forces with inappropriate operation? As phil mentioned, will you put them in the ditch easily? Thinking about this reminds me of an engineer we had that, I guess you could say, zoned out and forgot where he was. He wound up cresting a grade with a tonnage train and got a good bit over the speed limit. In panic and a very bad decision, he went from number 8 throttle to full dynamic in a matter of seconds. Needless to say, 35 cars took to the wind and as one engineer I worked with put it, "cars were stacked like cord wood." I didn't get to see the carnage since I was working a remote job but there was rumor that the cars were stacked high enough to almost touch some high tension power lines that crossed the tracks in the site of the derailment.

So browsing through the different threads on physics, I think the general consensus was 12-13% for a loadable/unloadable car, to average things out. So my question is, what percentage would you use for a flat out empty and load? I'm guessing the 12-13% still for the load but what would be a good figure for the empty version?

[GreatNortherner]

Hi,

Wow. That sounds like it was the mother of all derailments!

So browsing through the different threads on physics, I think the general consensus was 12-13% for a loadable/unloadable car, to average things out. So my question is, what percentage would you use for a flat out empty and load? I'm guessing the 12-13% still for the load but what would be a good figure for the empty version?

I think if you are going with different weights on MT and LD cars -- in which case I would suggest setting only one weight ("MASS") in the blueprint and leaving the cargo capacity weight setting at or near 0 -- you could use absolutely spot-on max braking force settings.

For example, you could pick Bill "TurboBill" Prieger's Pro Physics brake force settings from MSTS and use them as a starting point. Seeing that KRS and MSTS originated from the same software design studio, chances are that the underlying car physics calculations aren't all that different. So for example, a 27.5t car with cast iron brakes gets a 21.8kN max brake force with Bill's physics pack. For RW, that would get you a MaxBrakeForce% of 79%. A 100t car gets 39.4 kN max brake force, so that would be 39%.

In one of my old physics mod files from the wild MSTS days I also found some references that quote some AAR figures for maximum brake forces at 50 psi brake cylinder pressure: loaded cars mass*0.3 (0.39 for T/COFC) -- empties same as loaded, but not more than (empty mass * 1.2)

That said, there could problems with a wide variety of passenger brake valve schedules in a single consist.

This is fascinating stuff, Phil! I'm afraid Railworks is much too forgiving to simulate anything like this, I've been running long trains with mixed extreme brake settings strewn throughout the consist, and it didn't seem to have any bad effects on train handling. Probably because the air doesn't propagate along the train.

Cheers,
Michael

[mdurdan]

Hahaha, yes I've had this happen too me also. I was pulling a 140 car train with mixed power (SD70ACE oon the head end). Hit the dynammics after notching down from 5 at 60mph and boom goes the first 50-60 cars of my train, all over the place and piled up just as you discribed! Locos were perfectly fine and on the rails though!!!

[up_8677]

Yeah, the dynamics are terrible on most of the locos. Using the BritKits couplers helps some, but ultimately the dynamics are miles too powerful compared to the prototype. I think the ES44AC is the best driving of the American high HP units. Never had a derailment with it, that was caused by dynamics.

[mdurdan]

Same here, hope the new ones better yet

[philmoberg]

... I'm afraid Railworks is much too forgiving to simulate anything like this, I've been running long trains with mixed extreme brake settings strewn throughout the consist, and it didn't seem to have any bad effects on train handling. Probably because the air doesn't propagate along the train. ...

I think you're right about the air not propagating along train, but the slack sure does, and the effect seems a bit aggravated (not to mention aggravating) because of it. Just for the sake of trying it to see what would happen, I took an RS-3 and a short consist of freight cars from the same era, which consisted of two loads and eight empties. I noticed that the minimum brake application on the RS-3 is about 15 lbs., but it's a bit of a trick to get this consistently. When I did, I got results that were reasonably like the antiques I used to work with. That - at least in my opinion - isn't too shabby, and suggests 1.) that we have a lot to work with already, and therefore 2.) that it is reasonable to be optimistic about more realistic performance happening sooner rather than later. Of course, you have to think well ahead of the train top operate that way, but than most of the old-timers who taught me the business would grumble and say, "Of course you do!"

I was so impressed by this that I decided to try it with a passenger train consisting of two RS-3s and seven coaches. On a reasonably level route, I was able to make a 25 lb. application from a bit over a mile out, at about 65 mph., and stop the center of the consist about a car's length short of the platform. This is not inconsistent with what I've observed similar equipment do under similar circumstances on Class 1 railroads.

Is there room for improvement? Lots of it, to be sure; but it would seem you could make up for some of it with your approach to running the train.