Steam Locomotive Performance: Topic 2 for discussion

The Trainsim modeling of steam locomotive cylinder performance is flawed in a couple of different ways. Fixes for a couple of these flaws have already been published in this forum: making the steam usage reflect a reciprocating piston and modeling changes in steam volume at different pressures.
Here are two problems for which I have seen no work-arounds and cannot think of any.
(1) The sim models only the admission and expansion phases of the Rankine cycle for the piston, giving a cylinder pressure (MEP or "mean effective pressure") that is too high. The cycle should also include and exhaust and compression phase, which reduce the force created during admission and expansion. A careful look at the timing of cylinder events reveals that expansion is never carried out to the end of the piston sweep: the exhaust port always opens before the end to allow a reduction in pressure before the end of the stroke and to help reduce back pressure on the return stroke. The shorter the cutoff, the earlier the exhaust port opens. As a consequence, the real MEP achieved at shorter cutoffs is well below the sim's reported cylinder pressure, while the relationship is much better at longer cutoffs.
(2) Piston speed plays an important role in determining the final MEP. As piiston speed increases, not only is the entering steam pressure reduced, but exhaust is inhibited, raising the backpressure and increasing the compression pressure, both of which reduce MEP. None of this is modeled in the sim.
To illustrate my points, here are a few MEP readings taken during some locomotive tests in the first decade of the 20th century. The locomotives were apparently fitted with slide valves and mechanical indicators were used to calculate the MEP. In a perfectly efficient cylinder, the MEP at a given cutoff should not vary with velocity, but look what happens in real life. All were taken at 180lb psi. Because of the way the forums compress spaces, I'll rearrange the data somewhat.
20% cutoff: at 60 revolutions per min: 62lbs; at 180 revs:42.2lbs; at 300 revs:22.6lbs.
40% cutoff at 60 revolutions per min:108lbs;at 180 revs:78lbs; at 300 revs: 47.5lbs.
60% cutoff at 60 revs: 130lbs; at 180 revs:97lbs; at 300 revs:64lbs.
The effects of piston speed on MEP were well known at the time. The use of superheated steam improved the results somewhat, but the greatest improvements came with the use of inside admission piston valves which cut the throttling effects in half in test cases.
In the above data, the throttling is worse the shorter the cutoff for obvious reasons: the shorter the cutoff, the greater percentage of the admission phase during which the entrance port would be partially covered by the valve.
In my own simulator, I came up with a somewhat complicated formula for handling the throttling effects that works with the indicator card data I have to work with.
Since I cannot think of any way to get around these limits, my real purpose in pointing them out is to hope that someone at MSTS notices and that version 2 does a better job at these matters. I've had great fun trying to "impair" the basic MEP calculations to match the data I have to work with. Someone proficient in working with steam boilers could probably do a much better job. Any volunteers?Waiting for Godot (as usual),

Bill Hobbs

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

RE: Steam Locomotive Performance: Topic 2 for discussio, DRGWK36, Feb-11-04, 07:42 PM, (1) RE: Steam Locomotive Performance: Topic 2 for discussio, Bill Hobbs, Feb-11-04, 09:21 PM, (2) RE: Steam Locomotive Performance: Topic 2 for discussio, CumbresEngineer, Feb-12-04, 01:17 PM, (3) RE: Steam Locomotive Performance: Topic 2 for discussio, Bill Hobbs, Feb-12-04, 05:20 PM, (4) Advantages of Superheat, Bill Hobbs, Feb-12-04, 09:36 PM, (5) RE: Advantages of Superheat, CumbresEngineer, Feb-12-04, 11:56 PM, (6) Sources on steam locomotive performance, Bill Hobbs, Feb-13-04, 09:26 PM, (7)

Conferences | Topics | Previous Topic | Next Topic

Bill - if I were a brave soul, I'd jump right in. By rights, I should, since I studied steam long enough in school. For the time being, though, I'm going to avoid the temptation to get sidetracked on this, but if you keep posting interesting analyses, I may not be able to stay away!By the way, you may want to watch out for the automatic smilies - they seem to have eaten a couple of your combos.Andy

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Please do jump in. I've been working on my own simulator and could use some competent help from time to time.

Bill Hobbs

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

LAST EDITED ON Feb-12-04 AT 01:21 PM (EST)

Bill, theres no way im going to type all that I know, if you want, email me your phone number or SOMETHING if you want to hear what all that I know.I will say one thing that will make a big difference in this sim, make all engines that would be super heated in real life, superheated! Superheated steam does make a huge difference compared to a non-superheated locomotive. The D&RGW literally gave, under contract, the 462 to the RGS. Simply because it wasnt super heated and didnt do the job for them. Super heated steam is also more efficient. Although it can create more problems simply because theres more there to have problms with. The tubes, and it runs hotter. But heat is good! You want the cylinders to get hot, all of this went into consideration for POWER! I also wish we could actually factor in grate area. I don't know maybe you can. And what about throttle valves? Can someone tell me how to make the engine brake on Fozzy's K-36's grab the wheels harder?! The way it is right now you can run it all the way up to Cumbres with it on and not even notice. Stupid...Jon Davis does a good job with the engine brakes on his engines.
Also Jon, email here jemabush@tconl.com and I want to talk to you about makin' an engine.See ya'
Kevin Bush,

I will have my hand on the throttle of a K-36 very soon Mon Freunds!

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Kevin,
Its easy to fix the brake. If you like, I can forward you my .eng file for the K36 or you can send me yours since mine is very likely quite different in some areas.
Yes, superheating does make a big difference, but you have to be careful in how it is set up in the sim. More about that later. I'll email you from home since if you forward anything to me through the forums, it goes to the office.

Bill Hobbs

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

The primary advantages of superheated steam were:
(1) The volume of a pound of superheated steam was greater than the volume of a pound of saturated steam, thus requiring less steam to fill the cylinder.
(2) The heat loss within the steam chest and cylinder didn't cause the steam to condense because of the extra temp above the saturation point.
(3) The extra BTUs required to superheat the steam were being wasted as heat going up the stack. Adding the superheater tubes in the flues increated the efficiency of the evaporation process.
All three of these combined to reduce the amount of coal required to produce a given level of work.

Bill Hobbs

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

LAST EDITED ON Feb-12-04 AT 11:58 PM (EST)

Exactly Bill, it was all about power and cost. They save money, and make more money. They use less fuel, but probably more water. What are your sources Bill? My dad has probably a couple thousand books on steam. All were written in the old days. They are all written to educate the operational and mechanical side of railroading of the day. I have been reading alot of them. #####, I mean he has 3 books on the Worthington Type B feed water heater! And 3 books on a certain stoker or a book differnt injectors. Books on every single little neuonce on steam locomotives. Right now i'm reading a book by a guy named Angus Sinclair, a locomotive engineer for the NYC I think in the 30's. He wrote many books on experiences running big steam. He was always experimenting it seems. Like he found out that steam locomotives run in there absolute smoothest stroke at full forward gear. And he said sometimes with different engines and in most cases, if at low speeds, he brought the gear up too high, it would slip. Alot of guys don't get that, they'll start the train and bring the gear up half way towards center or more and the thing will slip and then they'll hit the sanders or back off the throttle. Well that wont work cause when they come back into it it'll slip again. Especially big drivered engines will do this. A K-36 won't do this at all unless you bring it up WAY to high, like two notches from center at two miles per hour. But from what I heard from a veteran C&TS engineer for 22 years, told me that all of the K's were and are timed high. There isn't much messin' around with the gear once up to about 10mph and full throttle. I mean droppin it one notch will certainly make a difference but not a whole lot. Ive heard that with the 463 once they hit the 4% grade they just leave it at 4 notches from center and put the throttle in the tank and just leave it that way. Or they will adjust the gear up or down from 4 notches as they go. I know why they time them high but for some reason its hard to explain on here. It sll about tourqe and simply because the locomotives aren't designed to go any 50mph. The things would start hoppin up and down and tearing up the railroad over 35mph! Thats what happens if you exceed any engines limits. It starts actually hop up and down on the railroad. The K-36's and every other D&RGW K class loco were less prone to slip because of how far out the side rods were from the center of the drivers. The farther out they are the more tourque your gonna have. Damnit I wish I could explain the reason for them being timed high on here...I can't find the words. Lets see, it's just because, well, I explained it above pretty much. Just think about that and you'll figure it out. It's really all common sense. If you have taken science classes in school and if your at all mechanically inclined, you'll get this stuff. Just know that the main reason for all of these really cool appliances and stuff on steam locomotives anywhere from about the mid 20's to the early 50's or now were all to save fuel and cost to operate. Thats what all of this stuff was built off of. With that one thing in mind...I could go on and on and on about this stuff!
Kevin Bush,

I will have my hand on the throttle of a K-36 very soon Mon Freunds!

Alert | IP

Printer-friendly page | Edit | Reply | Reply With Quote | Top

Kevin,
Before I get into a discussion, if your dad does have a good collection of books on steam, I would appreciate any data you might be able to pass on about indicator card readings. I am particularly interested in those which give the MEP (mean effective pressure) along with info about the cutoff rate, throttle rate, boiler pressure, cylinder dimensions, etc. I have a couple of sets of data and would like more as a means of checking to see how well my "model" calculates the mep. Let me know offline.
Ok, for the sources I rely on. Since I am building my own simulator of a basic steam locomotive, I have had to go to a number of places to get data. Finding basic formulae isn't hard. A good engineering book covering thermodynamics will have the formula for the Rankine cycle of a steam reciprocating piston, as well as some steam table data. But that won't get you very far in building an accurate working model. Why not? Well, let's consider the MEP formula as it is usually given in the reference books. I've put the four pieces on separate lines to make explanation easier.
MEP = p1 * C (pressure during admission)
+ p1 *(C + Cl) * ln((R + Cl) / (C + Cl) (pressure during expansion)
- p2 * (1 - K) (less back pressure during exhaust)
-p2 (K + Cl) * ln(((K + Cl)) / Cl (less pressure during compression after closure of exhaust port)
Where ln = natural log; p1 = admission pressure in pounds absolute; p2 = exhaust pressure in pounds absolute; C = cutoff in %; Cl = clearance in %; K = Compression %; R = release %
So far this looks like a straight forward thermodynamics formula for adiabatic expansion and compression and computation of force given pressure. So tell me the values for C, Cl, K, and R. C is obvious: it is the percentage of the stoke during which steam is admitted to the cylinder. And how do we determine cylinder clearance? But how do we know when the exhaust port opens (R) or when the exhaust port closes (K)? And what values do we assign to p1 and p2?
This is where the reference material becomes important. For any given valve gear, the release and compression vary with the cutoff. Shorter cutoffs have earlier releases and earlier compressions. The percentage, by the way, refers to percentage of the swept volume of the cylinder. These formula were intended to be used with indicator cards that were drawn on by the cylinder indicator device. An engineer sitting at his desk could put a ruler to the card and determine the point of cutoff, release and compression. That can't be done in building a model: one has to be able to calculate these points. I made use of Charles Dockstader's free software programs for setting valve gear for live steamers (available online). With a lot of patience, one can sit and time the points of release and compression for a range of cutoffs (say 5% to 90%). I then used a curve fitting technique called "least squares" to give me an equation I could use to predict release and closure given cutoff. Right now, I am working with an equation for Stephenson inside admission gear only, but I can go back at some point and do the same for other valve gear as well.
For the admission and exhaust pressure, it quickly becomes obvious in reading through the literature on steam locos that the cylinders did not get full boiler pressure unless the cutoff was long and the rate of piston travel was very very low. Otherwise, the pressure was throttled by the travel of the valve over the admission ports (they were fully open only part of the admission process). One must model this to get a realistic representation.
Two books I rely heavily on for this kind of data are Alfred Bruce's The Steam Locomotive in America, and some pages of a locomotive manual I Photostatted in the Colorado RR museum last year, titled "Locomotive Design: Data and Formulae", author's name lost. Recently I've added D Wardale's book "The Red Devil" to my library, but by the time I got it, my basic work was done. These books show graphs of the decrease in pressure to the cylinders given rate of piston travel. One has to look at the shape of the curve and then design a formula that will approximate that shape.
Exhaust pressure is really what we know as backpressure and here again, the kind of information one gets tends to be limited. Most books will talk about a range of acceptable back pressures, but one has to have some actual indicator card data to work with to get an idea of the range of values that might really occur -- and of course those calculations would be subject to the accuracy of one's assumptions about the admission pressure....
And these are just the problems I encountered with calculating the cylinder pressure. There is a lot more to the model, the most difficult part being the exhaust system: capturing the relationship between the exhaust steam, draft, combustion and heat release, and the heat exchange and evaporation. Once again, one must piece together the complete picture from a number of different and fragmentary sources.
Fortunately my local library has a very good collection of books on steam locomotives, donated by a fellow who used to work for the MoPac locally during steam days. Here is summary of the books I have worked with in the last year:
The Steam Locomotive in America by Alfred Bruce;
The Locomotive Up-to-Date by Charles McShane;
The Red Devil by D Wardale;
Calculations, Delineations, Classifications by Baldwin Locomotive Works;
The Locomotive Engine: Running and Management by Angus Sinclair (you know him)
2 different manuals on the Progressive examinations for Engineers and Firemen
How Steam Locomotives Really Work (authors names lost, but a recent Oxford University Press publication)
Locomotive Boilers and Engines by Professor ? of Purdue University (available online at www.sdrm.org)
Valve Gear simulations by Charles Dockstatder (available online)
Articulated Locomotives by Lionel Weiner (Kalmbach)
Steam tables (available online )
various handbooks on thermodynamics for engineers. John Wiley press has these working with imperial units rather than metrics
Study materials for the boiler license exams
Applied Multiple Regression by Draper and Smith (remind me how to set up the curve fitting).

Bill Hobbs

Alert | IP

[ Home | MSTS Forum | MSTS FAQ | File Library ] Copyright © 2004 by Nels Anderson. All Rights Reserved.