Page 1 of 1

steam locomotive emulator

Posted: Thu Apr 18, 2019 8:36 am
by gregoryc
a member of the NJ Live Steamers suggested I post my questions on this forum -- "PhD's who have written and studied these subjects"

i'm an electric model railroader (HO gauge) working on an electronic controller having both a reverser lever with multiple notches for cutoff and a throttle that will more accurately emulate steam locomotive behavior. It should require the operator to properly use the throttle and reverser to gain speed without depleting the boiler of steam. (i'm an EE retired from Bell Labs and Qualcomm)

the attachment plot from a simulation shows speed (blue), boiler pressure (white), cutoff (orange)

the software model assumes saturated steam where temperature, pressure and density are directly related. Use rpm and cylinder volume to determine steam consumption (lbs) and equal amount of water to replenish in boiler and determine resulting temperature/pressure. The plot shows the decrease in steam pressure with speed and its recovery with the reduction (% valve open) of cutoff

looking for help to verify my understanding and calculations for drawbar force, max TE and condition for slipping. In particular an understanding of how the throttle affects cylinder pressure.

dealing with a more accurate model based on super-heated steam, pressure drop thru steam pipe and back pressure thru exhaust remain beyond my understanding but may be unnecessary.

using the following parameters (and a gross estimate of boiler water capacity) for a few Reading locomotives and an Australian locomotive

Code: Select all

#        DrvrDia   CylDim   PSI  Wt  Grate   BlrLen  BlrDia  Desc
#             in    inXin       ton  sq.ft       ft      in
  locoDesc  61.5    27 32   220 132     95     13.5      88  I-10sa
  locoDesc    80    25 28   200 150     95    19.00      80  G-1sas
  locoDesc  61.5  30.5 32   220 200    108    22.25     100  K-1
  locoDesc    70    27 32   240 200     95    13.50      88  T-1
  locoDesc    36    13 18   180  13      9       11      40  NA_2-6-2
appreciate any help that can be provided

greg - somerset, NJ

Re: steam locomotive emulator

Posted: Thu Apr 18, 2019 10:07 am
by NP317
A thought:
The throttle primarily effects the cylinder VALVE CHAMBER pressure.
The cylinder pressure is secondarily effected when the valve admits/exhausts steam to/from the cylinder for a portion of the stroke, which also varies the pressure.

An interesting project. Are you assuming a "perfect" fireman, so you never have to worry about boiler pressure?
I don't have mathematical answers readily available for your questions on drawbar forces, tractive efforts and adhesion coefficients.
Tractive effort (TE) and drawbar force are primarily locomotive design characteristics. Adhesion is a variable in track conditions, that effects TE.
So you might need to decide which variables you wish to model.

I trust other knowledgable people (that's not me) will provide input here.
An interesting project.

I've often wished there existed a good computer model of steam locomotive operations, that required both a Fireman and an Engineer to operate the model.
Understanding the required operators' relationship with the machine - and each other- gets to the heart of operating steam locomotives.
Something that is lacking in most of today's machines.
~RN

Re: steam locomotive emulator

Posted: Thu Apr 18, 2019 2:34 pm
by gregoryc
thanks

i'm guessing that throttle restricts the flow by some fraction depending on the consumption rate. The higher the consumption, the greater the restriction. At very low consumption the throttle could barely be opened and have any affect

and i assume any restriction in the flow results in a proportional decrease in pressure affecting drawbar force. (we ignore super-heating)

while we could change firing rate (not accounting for any delays), so far we've kept it constant. We also add feedwater to the boiler (lbs) at the same rate steam is consumed by the cylinders (cutoff, cu.Ft, density, rpm, 2 strokes/cycle, 2 cylinder). This is something that could also be delayed.

we're also grossly estimating boiler water capacity based on boiler diameter and length from smokebox to firebox. We could use a better estimate.

our simulation currently runs for some number of one second iterations. Between each set, we can change parameters such as cutoff. eventually we will add ~200 lines of code to a model railroad controller and update each iteration using throttle and reverser inputs

greg

Re: steam locomotive emulator

Posted: Thu Apr 18, 2019 4:26 pm
by Asteamhead
Hello gregoryc,

What do you think about the good old Train Simulator? Maybe this will be only working with ancient WIN 2000 or Expert on your labtop yet?
Years ago I checked out, if the calculations used with this programme were true regarding the real thing. Found out that both drawbar power and -force were quite right calculating velocities, loads and line profiles!
If you are running the "Settle and Carlisle" in GB f.e., you may run a train very stable at 15 +mph on steady 1 % upgrade.
Runniing there a German class 44 on the point of a 1.400 T (1.500 tons) train will run a steady 15 mph, which is just right with the performance of the prototype.
All speeds downhill seem to me a bit optimistic yet, but that' s British history :wink: (look at A4's downhill record run!).
Had a lot of fun working different locos on different profiles! :)
To simulate like the real thing you are to use a lot of calculations or tables of performance, formula of train resistance and so on.
One graph, which may show the typical performance of a steam locomotive, regarding speed or better rpm, drawbar power and cutoff is shown.

Best regards
Asteamhead

Re: steam locomotive emulator

Posted: Thu Apr 18, 2019 6:49 pm
by Cary Stewart
I suspect that to fully understand the details of a steam locomotive's operating characteristics you will have to delve into thermal dynamics, Ranquine cycles, etc. It is a very extensive and deep subject that most live steam model operators do not bother with. We learn how to build, mnatain and operated or models. The act of firing a loco economically is the key to a good run. I have only known one man that truly studied and understood the technical details of how heat is generated in the firebox to create steam to drive the pistons and thus the drivers. He was the only man I have met that could read technical books on these subjects in French, German and Italian and truly understand what was written. There were many texts that were never translated into English but were very important to the study. If you are willing to do the study go for it.
Cary

Re: steam locomotive emulator

Posted: Fri Apr 19, 2019 11:02 am
by gregoryc
you mean Rankine.

yea, i'd like to talk to a guy like that. But I just need a basic understanding

Re: steam locomotive emulator

Posted: Thu Apr 25, 2019 10:56 pm
by k4kfh
gregoryc wrote: Thu Apr 18, 2019 8:36 am a member of the NJ Live Steamers suggested I post my questions on this forum -- "PhD's who have written and studied these subjects"

i'm an electric model railroader (HO gauge) working on an electronic controller having both a reverser lever with multiple notches for cutoff and a throttle that will more accurately emulate steam locomotive behavior. It should require the operator to properly use the throttle and reverser to gain speed without depleting the boiler of steam. (i'm an EE retired from Bell Labs and Qualcomm)

the attachment plot from a simulation shows speed (blue), boiler pressure (white), cutoff (orange)

the software model assumes saturated steam where temperature, pressure and density are directly related. Use rpm and cylinder volume to determine steam consumption (lbs) and equal amount of water to replenish in boiler and determine resulting temperature/pressure. The plot shows the decrease in steam pressure with speed and its recovery with the reduction (% valve open) of cutoff

looking for help to verify my understanding and calculations for drawbar force, max TE and condition for slipping. In particular an understanding of how the throttle affects cylinder pressure.

dealing with a more accurate model based on super-heated steam, pressure drop thru steam pipe and back pressure thru exhaust remain beyond my understanding but may be unnecessary.

using the following parameters (and a gross estimate of boiler water capacity) for a few Reading locomotives and an Australian locomotive

Code: Select all

#        DrvrDia   CylDim   PSI  Wt  Grate   BlrLen  BlrDia  Desc
#             in    inXin       ton  sq.ft       ft      in
  locoDesc  61.5    27 32   220 132     95     13.5      88  I-10sa
  locoDesc    80    25 28   200 150     95    19.00      80  G-1sas
  locoDesc  61.5  30.5 32   220 200    108    22.25     100  K-1
  locoDesc    70    27 32   240 200     95    13.50      88  T-1
  locoDesc    36    13 18   180  13      9       11      40  NA_2-6-2
appreciate any help that can be provided

greg - somerset, NJ
Hi, I don't have a whole lot to offer regarding simulating a steam locomotive. But your situation struck me as a funny coincidence. I also model in HO scale, and I am an EE student right now, and a few years ago I got a very similar idea for diesel locomotives after riding in an F7 cab and realizing how much more complicated a real locomotive is than an electric model. So I ended up spending a couple years putting together a simulation software which works with any diesel locomotive, and runs in a web browser so it works with any device (including tablets). If you get this worked out, I'd love to integrate whatever simulation you come up with into my software (I open sourced my code and it integrates with JMRI so it's compatible with nearly every DCC system on the planet).

If you are interested, here is a demo: https://www.youtube.com/watch?v=GrTLCuW7GQw