[copperpen]

I have noticed that slow down for the last few psi, but not tracked down why yet.

[Genma Saotome]

An update: Question posted to the Steam Era Freight Car List (STMFC) at yahoogroups returned an answer that the rate that air is released from the brake cylinder is very fast... not balloon popping fast but vented in a second or two. This suggests to me that the correct value for MaxBrakeReleaseRate() for 20th century North American freight cars should be at least 15 psi/second, maybe higher.

Stuff I read suggested that the valves on passenger cars behaved differently and I've got nothing yet to say about them.

One caveat for freight cars: I read a lot of technical material ranging from late 1890's technical journals to a contemporary Norfolk Southern training manual for Locomotive Engineers (which I highly recommend -- heck... where is that url when I need it?). I learned the triple valve changed some over the course of the 20th century, always trying to deal with the problem of the front end cars reacting and changing faster than those at the end of the train. AFAIK both MSTS and OR do not account for any of those design differences. I'm not sure I understand it all well enough to say whether these differences would be worthwhile to implement in OR. I mention this only as a warning to others that the technical literature they might read on triple valves can be contradictory in some details if the publication dates are quite different.

One other item: In all cases I found the literature said the mechanical process was pretty simple: For safety reasons a spring is located in the valve device so when the air reservoir(s) pressure is higher than the train line the brake cylinder will not vent. When the air pressure on the brake pipe exceeds what is in the air reservoir(s) the resistance of the spring will be overcome and the brake cylinder will vent, releasing the brakes. The difference between train line and air reservoir to overcome this spring was always 1 to 2 PSI with 1.5 most common.

I'm pretty sure OR is releasing brakes at a couple of PSI lower than the pressure in the air reservoir(s). IOW the difference in PSI is about correct but the sign on that difference is clearly wrong. You want that as a formal bug?

[Genma Saotome]

Braking manual from an operator of historical equipment.

Norfolk Southern's Locomotive Engineer Training Manual.

Both provide considerable information about how brakes work, what the equipment actually does, and what the Engineer needs to know.

[R H Steele]

Thank you for those links Dave.

[Genma Saotome]

More info:

Quote

This is from a retired Santa Fe road foreman of engines:

On current equipment, each car takes about 3-5 seconds to vent to to the atmosphere once the control valve is activated to the release position by a rise in brake pipe pressure with the retainer valve set to the EXhaust position. The SD (slow direct) and HP (high pressure) adds a different twist. Older ABD equipment took 5-7 seconds and some times 10 seconds to release depending upon maintenance. Propagation figures are a bit more complex depending upon train length, temperature, humidity, altitude, etc.

and

Quote

I took a look through some of my industry reference materials looking for actual times, which are really hard to find. Wabco liked to talk about improvements over older equipments as percentages, without actually committing to times, which could vary with conditions. However I did find a promotional flyer from 1975 for the new ABDW control valve that graphs performance vs. the older (but still newer than 1960) ABD valve, which was itself an improvement over the original AB valve.

The graphs plot brake cylinder pressure against time for the 1st, 50th, 100th, 150th car in the train, and the average, under a full service reduction (application). I will give you the time in seconds for each of these cars to an average 30 psi. brake cylinder pressure (first figure) and the max 58 psi. cylinder pressure attainable with the 80 psi. brake pipe pressure they use in the illustration (second figure):

ABD equipment
1st car 16sec. 48sec.
50th car 45sec. 110sec.
100th car 62sec. 125sec.
150th car 66sec. 130sec.
Average 50sec. 125sec.

ABDW equipment
1st car 14sec. 33sec.
50th car 30sec. 67sec.
100th car 38sec. 69sec.
150th car 41sec. 78sec.
Average 32sec. 66sec.

I can't find any data on release times because, really, they aren't that important... the engineer will start pulling before the brakes are fully released anyway, and a standing train will be held with the independent brake while the train brakes are released... But for the sake of a simulator, if the "engineer" doesn't release the train brakes beforehand, a minute or two delay before motion ensues would not be unreasonable for a long train.

The first info comes from Mark Hemphill, a consultant to the railroad industry and the second from Dennis Storzek, the Founder/Partner of one of the largest Model Railroad Mfgr's in the US. There is, perhaps a 2 to 2.5x1 difference in seconds WRT what they say.


When you look at the the numbers in the longer quote, from car 1 to car 150, it seems pretty likely that the of the difference in time shown above is due to the difference in how long it takes to raise the air pressure at both ends of the train... and that probably also explains why venting 30psi goes more than twice as fast as venting 58PSI. It could also explain the difference in the two answers: perhaps the first reply is just the brake cylinder after the valve moves to a release position whereas perhaps the second from when the engineer moves his brake handle.

The other thing is it confirms my assumption there were at least a number of different triple valve designs in use in North America, from the H Series around the turn of the century, to the K, AB, ABD, and ABDW, and that's not addressing variants of each, like K1 and K2. So that's at least 5 different designs for Freight cars (passenger cars used different versions).

This post has been edited by Genma Saotome: 07 July 2015 - 08:23 AM
Reason for edit:: Disregard strike out text. Incorrect feedback.

[steamer_ctn]

Hi Carlo,

Csantucci, on 24 June 2015 - 12:01 AM, said:

I was testing a demo activity using as player train the standard consist 2 x gp38-2.con . I noticed that after some brake releases - brake applies the brake release didn't fully release the brakes even after considerable time.

I had a quick look at this consist (only 2 locos?) and didn't observe the same type of problem, but my tests may not have duplicated your process.

In NSW practice the Main Reservoir operated with a max pressure of 100psi. The brake pipe of freight trains would be set at 60 to 70 psi, and passenger trains at 75 to 80psi. I think that the restart pressure was around the 90psi mark. So 20 psi could be a bit too broad a gap.

The other aspect that maybe impacting on the operation of the brakes is the volume of the main reservoir and the charging rate of the compressor. Both of these will have an impact in OR, as I understand it.

For example, both the GP and Dash9 have the following main reservoir volume:

AirBrakesMainResVolume( 8 ) - in cubic feet according to MSTS documentation

This seems small to me for a modern diesel, as I have seen older steam locomotives with reservoirs over 10cu ft, and more modern units with volumes up to 30cu ft.

Reservoir Charging Rate is a "new" OR value, so I am not certain how these values were taken into account within MSTS.

Good driving practice was not to operate the brakes too often, as this could exhaust the air supply, and would result in the brakes taking a while to recover. This would be an even bigger issue on long trains, which need more air.

So perhaps an answer is to distinguish between "older" ENG files, and newer correctly set files. Thus, my preference would be to assume that if "ORTSMainResChargingRate" is present then it is highly likely that the brake parameters have been reviewed by somebody recently, and therefore can assumed to be correct, whereas if this value is missing from the ENG file, then the test that you propose could be implemented.

Thanks


EDIT: Just found this post that supports the view that 8cuft seems a bit small for the reservoir size - see last comment

[Lindsayts]

Sometime back I traveled on a train down to Melbourne (Victoria, Australia) on a train, the consist was an N class DE (2400BHP) hauling 5 N series pass cars, these latter where fitted with a twin pipe brake system coming from the sets built fro the Melbourne electric pass network. Both sets of vehicles were built mid 1980.s.
Now the last car of the set was fitted with a brake pressure gauge, unusually though this gauge was fitted to the brake line and not the main reservoir line.I was very interested as it gave a good indacation on how the brakes were used. The normal line pressure was just over 70psi, on applcation the pressure dropped t o around 60psi then down to around 56psi when the set stopped at the platform. The driver then released the brakes completely and I assume held the train stationary by the loco brake.
Now release timing, the pressure rose in two stages, immediatly on release it climbed to around 67psi, this took around 0.5 to 0.7 of a sec, the rise to line pressure of 70psi took another second. I assume this latter rise was due to the pressure drop in the main reserviour line while it was topping up all the trains Auxilary reservors.

Note, the observations were from the last car. I judged by the trains behaviour that the propergation delay from the loco to the rear car was nearly instantanious.

It would probably be worth remebering the brake system fitted to these cars came from a metro electric set, these usually were fitted with quite quick brakes.

Lindsay

[Lindsayts]

Post Script, The system fitted to the cars in the above post was Westinghouse twin pipe using Distributor valves, it was ____NOT___ an EP system.

Lindsay

[Genma Saotome]

steamer_ctn, on 09 July 2015 - 04:11 AM, said:

Good driving practice was not to operate the brakes too often, as this could exhaust the air supply, and would result in the brakes taking a while to recover. This would be an even bigger issue on long trains, which need more air.

Indeed.

This brings to mind a question about retainers. The other day I read retainers had just three positions: Open, which allowed the brake cylinder to vent completely, and two others (names forgotten already), which effectively forced the brake cylinder to retain (hence the name of this device) a certain amount of pressure within the brake cylinder even when the engineer had released his brakes. The two positions meant two different pressures were retained (I should find that again and get the names and pressures).

This must have been a brilliant solution that allowed the engineer to recharge his air w/o fear of picking up too much speed while doing so.

Does that description bear any similarity to what OR is doing w/ retainers?

[copperpen]

The three positions are EX = out to atmosphere. LP = low pressure and HP = high pressure. Far as I can tell, OR is using retainers in the correct way.

Back to the slow release of the final few PSI. I am becoming more convinced that it is a bug in the brake code that is hard to locate, which is why we have the Initialise Brakes key press.