I have been reviewing the reports from the Beta testers on the update to the .eng file calculator.
There were definate problems with the steam distribution I had used succesfully on maxinum weight trains when it came to the operation of lightweight high speed trains.
After eliminating other possible causes, investigation focused on the way the Kuju regulator works. No logical combination of the parameters gave realistic results, both from the driving point of view and from the fuel and water consumption point of view.
Yesterday I made a great discovery about the regulator as designed by Kuju
They describe the regulator as " RegulatorValveType ( Twin_port ) "
Everybody has assumed that what they were describing was a normal British twin port regulator with a pilot valve and a main valve, however I believe that what they actually designed was a twin port compound regulator where the first port admits high pressure steam to all cylinders to provide starting tractive effort and then on opening the regulator past a certain percentage, the first port closes and the engine goes into full compound working.
I first noticed the way some of the French compounds were set up which made the effect even more obvious and decided to see what would happen if I changed the description of the regulator to " RegulatorValveType ( Single_port ) " and made the first valve open for the full 100% and deleted all references to the second valve.
I tested this in an old French 0-6-0 on the LDM route initially and the results were remarkable
I now get an even steam distribution over the regulator arc and the steam chest pressure is responsive to changes in cut off and regulator.
Later tests on the 0-6-0 and also on the Stanier 8P Heavy test train revealed that the parameter "RegulatorValveExponent" now has a direct relationship to power developed and steam consumed and the locomotive can be properly driven using the regulator and the cut off as intended.
In addition the whole locomotive is much more powerful than before, although wether this is may be that I can run with a full regulator and more expansive working will require more than the few minutes testing that I was able to carry out before work.
The theory of the compound regulator is confirmed by the fact that the Golsdorf engine which is a compound has the same regulator as the Scotsman which is not. Presumably Kuju did the Golsdorf first and did not have time to change the inner workings of the programme to reflect a simple expansion locomotive.
To incorporate this into the calculator is going to require some more testing.
Regrettably the British twin port regulator may be a myth in the simulator as I would love to see the lower steam flow on the pilot valve compared to the higher flow with both ports uncovered.
At present the settings below seem to give good results although the changes in the steam flow also demonstrate a need to modify the coal burnage and water consumption.
RegulatorValveType( Single_port )
RegulatorPilotValveExponent( 0.9 )
RegulatorPilotValveInitialOpening( 0.01 )
RegulatorPilotValveFullOpening( 1 )
Kuju Regulators
Moderator: Moderators
-
supergoods
- Very Active Forum Member
- Posts: 2752
- Joined: Fri Jun 07, 2002 12:00 am
- Location: Sugar Land, Texas
Kuju Regulators
Ian
-
mgreenland
- Been on the forums for a while
- Posts: 123
- Joined: Tue Jun 04, 2002 12:00 am
- Location: Co. Durham
Hi Ian,
I have tried the single port regulator on both the A4 and the FR Alco and they both work much better with it. You have made a major advance, a touch of genius, I think.
I used just two lines in the eng file. Type = single port and exponent = 0.7. As you say, the locomotives seem more responsive but I think it is more to do with the fact that (especially) at short cut-offs they use less steam. Of course, this means they are using less water but I think this is correct for the high speed lightweight trains. On the Alco, your settings and the original regulator used 276 gallons to TYB which is correct to a few percent. Not surprising because she is at >80% regulator and between 25% and 35% cut-off for much of the time, which exactly how you designed the calculator. The point is she shouldn’t be at that long a cut-off. Assuming the gradients on the route to be more or less correct (and they seem to be) she should be at 20% to 25% (when I was firing her she had the original boiler which had too large diameter tubes and the fire therefore heated the smokebox as much as it did the water. She was therefore limited to 5 cars and the driver drove to the available steam. The fireman’s job was just to produce as much of it as you possibly could. The final point of this little tale is that she was usually worked on the last notch of the cut-off lever sector which must have been about 15%. The driver even showed me once how she performed in mid-gear so I have practical experience of 0% cut-off! The Ladies in my day generally took 7 and used 20% to 25% cut-off so the new Alco, with the Lady boiler, should be about the same. Derek Evans and I took 10 with Blanche once (in 1972) and then it was just as with the Alco – give the driver as much steam as you could. There were no problems but we climbed at 10mph instead of the required 14mph and lost time all the way up). In the sim with the new regulator the Alco definitely seems to perform but it remains to be seen how it affects the water consumption (I havn’t done a full route yet).
I have spent all day playing with your settings on the A4 and the FR Alco and trying to understand how Kuju programmed the generation and usage values – I must admit without too much success. Without going into a long tale, everything seems pretty good except that they seem to have programmed the cut-off so that at 0% it gives zero average cylinder steam pressure (their ‘steam chest pressure’) and pulling capability (correct) but not BasicSteamConsumption + accessories and not a zero effect on steam generation (not incorrect, but the values of consumption and generation you end up with are far too high). I suppose they thought at the end of the day that as long as the usage and generation rates roughly balanced it was good enough. I don’t know for certain why they came up with this solution but suspect that it was to overcome a ‘divide by zero’ problem somewhere in the code. I can understand a ‘quick fix’ mentality but this really is a bodge from hell because it seems to progressively mess up the generation/usage as the cut-off is shortened ie it gives an exponent effect which we will never overcome with the constant variables (or should that be variable constants?) available to us in the eng file.
Anyway, more power to your elbow. It may be that the ‘linkage’ of the regulator effects and the cut-off effects on generation and usage in the sim mean that the new regulator can overcome the problems of the cut-off.
Best regards, Martin G.
I have tried the single port regulator on both the A4 and the FR Alco and they both work much better with it. You have made a major advance, a touch of genius, I think.
I used just two lines in the eng file. Type = single port and exponent = 0.7. As you say, the locomotives seem more responsive but I think it is more to do with the fact that (especially) at short cut-offs they use less steam. Of course, this means they are using less water but I think this is correct for the high speed lightweight trains. On the Alco, your settings and the original regulator used 276 gallons to TYB which is correct to a few percent. Not surprising because she is at >80% regulator and between 25% and 35% cut-off for much of the time, which exactly how you designed the calculator. The point is she shouldn’t be at that long a cut-off. Assuming the gradients on the route to be more or less correct (and they seem to be) she should be at 20% to 25% (when I was firing her she had the original boiler which had too large diameter tubes and the fire therefore heated the smokebox as much as it did the water. She was therefore limited to 5 cars and the driver drove to the available steam. The fireman’s job was just to produce as much of it as you possibly could. The final point of this little tale is that she was usually worked on the last notch of the cut-off lever sector which must have been about 15%. The driver even showed me once how she performed in mid-gear so I have practical experience of 0% cut-off! The Ladies in my day generally took 7 and used 20% to 25% cut-off so the new Alco, with the Lady boiler, should be about the same. Derek Evans and I took 10 with Blanche once (in 1972) and then it was just as with the Alco – give the driver as much steam as you could. There were no problems but we climbed at 10mph instead of the required 14mph and lost time all the way up). In the sim with the new regulator the Alco definitely seems to perform but it remains to be seen how it affects the water consumption (I havn’t done a full route yet).
I have spent all day playing with your settings on the A4 and the FR Alco and trying to understand how Kuju programmed the generation and usage values – I must admit without too much success. Without going into a long tale, everything seems pretty good except that they seem to have programmed the cut-off so that at 0% it gives zero average cylinder steam pressure (their ‘steam chest pressure’) and pulling capability (correct) but not BasicSteamConsumption + accessories and not a zero effect on steam generation (not incorrect, but the values of consumption and generation you end up with are far too high). I suppose they thought at the end of the day that as long as the usage and generation rates roughly balanced it was good enough. I don’t know for certain why they came up with this solution but suspect that it was to overcome a ‘divide by zero’ problem somewhere in the code. I can understand a ‘quick fix’ mentality but this really is a bodge from hell because it seems to progressively mess up the generation/usage as the cut-off is shortened ie it gives an exponent effect which we will never overcome with the constant variables (or should that be variable constants?) available to us in the eng file.
Anyway, more power to your elbow. It may be that the ‘linkage’ of the regulator effects and the cut-off effects on generation and usage in the sim mean that the new regulator can overcome the problems of the cut-off.
Best regards, Martin G.
-
saddletank
- Very Active Forum Member
- Posts: 14183
- Joined: Mon Dec 03, 2001 12:00 am
- Location: UK East Midlands
-
CaptainBazza
- Has a sign reading.. Its NOT the end of the world!
- Posts: 18852
- Joined: Tue May 13, 2003 10:21 am
- Location: Land of the Long White Cloud.
Fractions and frictions
I'm following this important discussion, both here and at trainsim.com, with great interest.
-
saddletank
- Very Active Forum Member
- Posts: 14183
- Joined: Mon Dec 03, 2001 12:00 am
- Location: UK East Midlands
-
CaptainBazza
- Has a sign reading.. Its NOT the end of the world!
- Posts: 18852
- Joined: Tue May 13, 2003 10:21 am
- Location: Land of the Long White Cloud.
More from across the way
This is the trainsim thread I was referring to.
http://www.train-sim.com/dcforum/DCForumID5/7428.html
http://www.train-sim.com/dcforum/DCForumID5/7428.html