Realistic Physics and train handling.

[bavli]

Under MSTS Locomotive design forum,especially the New Method For Overspeed Cutout thread, a number of Physics issues have come up. In many ways MSTS does not properly support realistic locomotive physics,especially for electric and steam locos. I would like to start with some suggestions as to how to fix some problems. The people at MSTSX are listening so please add useful suggestions so that we get something where the Physics is as good as the graphics.

Since the 60s some EMUs have had a form of speed control-this then spread to locomotives as well and seems now to be common with AC motored trains, particularly in Europe. Basically the driver sets the speed and, perhaps a maximum current. The train or loc than maintains the set speed,unless the gradient causes the current to rise to the limit set by the driver or locomotive's protective system, at which point the speed does fall to keep within the current limit. This enables the driver to keep focused on what is ahead him and not worry about the speedometer.

I think MSTX should support this-will anyone back me up on this?They do listen-here's your chance to be heard.Dear reader,what say you?

I have other ideas ,but this should do for now.

[TrainMan_112]

Well, since MSTSX won't ship with any EMUs so I don't think the developers are concerned about that.

You'll have to wait for some third party developers.

[bavli]

If the piece of kit you are uploading has speed control ,the engine file should support this.

[mestevet]

Well, since MSTSX won't ship with any EMUs so I don't think the developers are concerned about that.

You'll have to wait for some third party developers.

They may not be, but it was pointed out in the thread bavli is referring to, that some modern diesels have similar features (sort of like cruise control) as well. Remember that a diesel is only an electric locomotive with the power generating plant onboard.

If they want 3rd party add ons, they can't neglect EMUs, Electrics and the like, or a large segment of railfans/Simmers (particularly outside the US, but many of us in the US who are interested in heavy electrics, as well as traction, subways, etc) will be mighty disappointed.

Steve

[jtr1962]

A couple of other things which MSTS-X should support:

1) Ability to customize tractive effort curves via a table of some sort.

2) #1 should include the ability to specify the tractive effort curve in each throttle position.

3) Blended braking that works properly (i.e. brake cylinder pressure starts to rise only when the dynamics are fading).

4) Braking friction variable with speed, fully customizable via a table (in the real world brakes don't stop as fast at higher speeds).

5) Maximum wheel-rail adhesion also variable with speed (particularly relevant with high-speed trains as there is considerably less adhesion at 200 mph compared to very slow speeds).

6) Friction of each individual railcar fully customizable via a table. While the traditional Davis A, B, and C coefficients work well over most of the speed range via the equation friction = A + B*velocity + C*velocity², it has been observed that in the real world friction increases markedly when speed drops from 5 mph to zero. Some way to duplicate this would make the sim that much more realistic.

[muskokaandtahoe]

Not that my thoughts on the matter would get serious consideration, but I've long advocated the ideal would be to make all of the physics values and formula open source. That gives the experts in the community the means to fine tune (and more likely, fix) the vocational content, leaving to MS the full control of what they do best: the graphics display and machine interaction.

[TrainMan_112]

They may not be, but it was pointed out in the thread bavli is referring to, that some modern diesels have similar features (sort of like cruise control) as well. Remember that a diesel is only an electric locomotive with the power generating plant onboard.

If they want 3rd party add ons, they can't neglect EMUs, Electrics and the like, or a large segment of railfans/Simmers (particularly outside the US, but many of us in the US who are interested in heavy electrics, as well as traction, subways, etc) will be mighty disappointed.

Steve

Which modern diesels have cruise control? Freight locos sure don't.

For me loading would be a big thing. In MSTS all locos are basically the same because they gain power just as fast as the others.

Realistic slack action would also be nice.

[bavli]

Custom TE curve might be:example
0.mph-74,000lbs
3.mph-71.000
3.1mph-57,000
41mph-57,000
45mph-48,000
61mph-48,000
100mph-21,000
There is no reason I can see why the computer couldn't connect the points in this matrix linearly and give you your custom TE curve. Would this demand more from the computer than the current algorithms? I'd be surprised if it did.One thing though-no one could say "my locomotive doesn't perform realistically because the algorithm won't accommodate it". They could leave default settings for those who want it and everyone could be happy.

[mestevet]

Which modern diesels have cruise control? Freight locos sure don't.

For me loading would be a big thing. In MSTS all locos are basically the same because they gain power just as fast as the others.

Realistic slack action would also be nice.

This is what I was referring to, I've heard other descriptions of systems elsewhere, but I'd have to go search them down too (might've been on passenger equipment though):

http://forums.flightsim.com/vbts/showpost.php?p=1434730&postcount=46

I say cruise control, but that's not really the correct description, maybe "load control" or something like that.

In a manner of thinking, automatic speed control was possible back in the 1930s when coded track signals came into use for cab signals. I recall reading an article about trials on automatically maintaining speed as a safety feature in that timeframe (before WWII), but I believe the issue wasn't that it was technically impossible so much as that it was too costly to outfit enough locomotives with the equipment to be worthwhile.

A number of different types of electrics, led by EMUs have what can best be described as Cruise Control. The original Metroliners have what was known as "Speed Maintaining" (looking at my copy of the operator's manual right now). The control handle was moved to a point on it's travel proportional to speed (marked out next to the handle) and the selected speed maintained to within an error of 2-3mph. I've heard this worked (while other things on the Metroliners didn't). There are others as well.

I know there's a segment of rail enthusiasts who have no interest or use for electrics of any sort. However, the worlds most powerful and fastest trains are electric. Electrics have moved large amounts of freight and passengers throughout the world on a daily basis since before diesels came into wide use. Just because 75% of the US has never seen one, doesn't mean they aren't vital, useful, interesting pieces of equipment that deserve to be modeled as much as the latest smoke belcher from GE or elsewhere. I mean geez, make a train simulator that's incapable of accurately modeling a GG1? Marketing suicide if you ask me.

I like the sound of the open source suggestion for handling the physics parameters. Leave it to those who understand the physics best (like Joe R. and others).

Steve

[bavli]

If it is faster than 200 km/h, is in revenue service, and runs on rails, it is electric.

[plainsman]

A few other points as well.
Please allow for more appropriate modeling of modern AC locomotives. The electrical operation of these traction motors is very different than DC traction motors, both in performance and in the way they are controled.
Another thing is to allow direct input of transmission efficiency, or better still a table function, as this is not constant with speed. The variation is quite large between very old DC designs (about 80%) and the best modern DC designs (about 90%), and the best modern AC designs (about 95%).
Also, throttle position is not linear with power.
Note couplers are not all the same. They vary in both strength and operation. An important part of a rail simulator is for traction motors to overheat and fail, and for couplers to break under the correct conditions (not at node points!!!). These failures need to happen under the appropriate conditions for that equipment.

[westerngy]

From what I understand Bob is that in the physics your going to be able to set up so that you can have a mottor or two with TM's cut out.

[bavli]

Hundreds of railcars,with mechanical or hydraulic transmissions, are available for MSTS ,but these are not properly supported,if you want realistic performance. The good news is that it should be quite easy MSTSX to get right.
The problem is that for any gear the TE, or torque,is much lower at low speed and increases linearly with speed.For example,the Class 180 Adelante (UKTS 7009) displays a TE of 4000 at 0 MPH and 16,000 at 45 MPH at which point the gear shifts automatically.Any train using the gear-box will have this southwest to northeast straight line for any particular gear. I have found this in all the railcars I have tested . In the RW the torque would be close to constant except near the top RPM it would decrease somewhat. Why not more specific? Because it depends on the particular prime mover.Why not allow the uploader the option in posting #8 here? If the TE or torque curves are unavailable, the uploader could just opt for a horizontal line ,which would be better than what we now have.

Hydraulic transmissions are treated like mechanical gear boxes except there are ony two or, sometimes, three speeds.That is not accurate, but the system could support it very well if modified. Many uploaders now use diesel electric files for this.That is not accurate either.

Typical hydraulic TE curve
0km/h-81kn
20--50kn
40--30
60--20
69--18
70--15 1/2
110--14
For Diesel Hydraulic railcars the "first gear" is a torque converter.This can produce up to four times the torque input at low speed.Usually at about 60% maximum speed, this cuts out automatically and a more efficient fluid coupler takes over; this behaves more like direct drive,except there is a reduction gear to the wheels.The transition is very smooth with no loss of power. Railcars meant for 160 km/h or better usually have two fluid couplers-like having a 3 speed gearbox but more flexible at low end. Allowing the custom TE curve would make a realistic sim in terms of physics.

In the RW diesel hydraulics usually have hydrodynamic braking-akin to dynamics except you are heating hydraulic fluid instead of steel grids.This would be good to have too.Supposedly in MSTS the Kiha 31 has an engine brake-press P. If it has ever worked for anybody-what's your secret?

For gear boxes, in the RW you have to cut power while changing gear.This doesn't apply to hydraulics. I think this should be an option for the uploader.MSTS allows you to cheat, and I must admit I do.

Trains with gearboxes are fun to drive-they should be realistically supported.Make your views known.

Anyone wanting further material on Hydraulics-e mail me.