Atomic Systems

Firstly – if this is old hat, apologies (also for the length of the thread, but i thought it might be of use to someone...)

Following on from the thread http://forums.uktrainsim.com/viewtopic.php?t=62800 I was trying to see how the regulator coding affected the Steam Chest Pressure and thereby slipping.

The TechDocs give the following details:

Parameter – Default Setting – Notes
RegulatorValveType - Twin_port - “single_port” or “twin_port”.
RegulatorSecondValveStartPosition – 0.5 - The degree through which the regulator must be moved in order to start the second valve operating.
RegulatorPilotValveExponent – 1 - The curvature of the control setting to effect curve for pilot or single port regulator operations.
RegulatorSecondValveExponent – 1 - The curvature of the control setting to effect curve for second valve regulator operations.
RegulatorPilotValveFullOpenning - 0.4 - The proportion of the maximum effect when the pilot valve is fully open.
RegulatorMainValveInitialOpenning - 0.5 - The proportion of the maximum effect when the main, or second valve is first opened.

The following code was used for the tests:
RegulatorValveType( Twin_port )
RegulatorSecondValveStartPosition( 0.6 )
RegulatorPilotValveExponent( 1 )
RegulatorSecondValveExponent( 1 )
RegulatorPilotValveFullOpenning( 1 )
RegulatorMainValveInitialOpenning( 0.6 )

1) Steam Chest Pressure - Varying the regulator:

This was not what I’d expected. From the code I assumed that there’d be a change at 0.6 (ie 60%) as the second valve “opened” rather than a straight line relationship throughout.

2) Steam Chest Pressure - Varying the cut off:
a) Regulator at 100%
b) Regulator at 50%

Again, straight line relationship, unaffected by the Regulator Valve parameters.

3) Steam Cylinder Rate - Varying the regulator:
Using the Extended HUD available with MSTSbin found the following, which matches the coding exactly, changing at 0.6 (60%):

The Steam Usage rate seen in the standard full HUD, appears to be linked to other parameters as well as this as it varies unpredictably during the tests; sometimes very close to the Steam Cylinder rate, sometimes thousands of lb/hr above it.

4) Steam Cylinder Rate - Varying the cut off:
a) Regulator at 100%
b) Regulator at 50%

As expected, the Steam Cylinder Rate at 100% regulator is more than twice that at 50% due to the Second Valve opening.

Before starting this I assumed that the Steam Cylinder Rate and the Steam Chest Pressure would be directly related – the higher the pressure the more steam went into the cylinders, but they’re two very different curves. Does anyone know how prototypical these behaviours are?

This could get to be quite a big topic

asharte wrote:
Before starting this I assumed that the Steam Cylinder Rate and the Steam Chest Pressure would be directly related – the higher the pressure the more steam went into the cylinders, but they’re two very different curves. Does anyone know how prototypical these behaviours are?

I think 'not very' is the short answer! I've done tests similar to your tests 1,2 and 4 (hadn't tried no. 3, v interesting result) in the past and found similar results. How did you keep the speed of the loco and the boiler pressure constant during the tests btw? As far as I can tell, steam use in the cylinders is a straight line relationship with speed in MSTS, assuming everything else is constant. Boiler pressure also affects steam usage.

I don't have MSTS bin, so haven't experienced the extended HUD. I would assume that the steam cylinder rate shows the steam just used in the cylinders, while the standard steam usage shows all steam used by the loco (so adds in steam used for injectors and ejector, or lost via safety valves).

What MSTS calls 'steam chest pressure' is actually more like what's known in the real world as the 'mean effective pressure' (MEP). This is the average pressure acting on the piston throughout its stroke and is affected by both the regulator opening and cut-off. The MEP on a real loco can be worked out from indicator diagrams, brief explanation of the latter here: http://www.greatwestern.org.uk/basic8.htm

The MEP is more or less directly related to the tractive effort the loco is generating at any one time.

To answer your original query a little more fully, as you reduce the cut-off on a MSTS loco the amount of steam used falls too quickly and the amount of tractive effort the loco is generating falls too slowly. This means that you can only set the steam consumption on an MSTS loco to be correct at 1 particular cut-off value. If you set the consumption to be correct at full cut-off, the loco will become much too efficient at short cut-offs. If you set the consumption to be correct at short cut-offs, the loco will use far to much steam at long cut-offs.

So all in all, roll on KRS! And I hope Kuju have done their homework...

Cheers

Rich

I have noticed some interesting issues here while performance testing:

1) Better steam utilisation comes when the TE is set at 100% BP and the Regulator is limited to 85% maximum opening.

2) Steam consumption from the HUD is usually about 5-10% higher than the equivilant water consumption from the HUD

3) I have set the cut off in uneven notches to better balance the curve as you have noted.

Ian

Go read this line...........

SteamSmallestCutoff( 3/4/5/6/7//8/9/10 ) whatever, the slash = or

Problem is the link between Speed and Steam Usage is wrong.

MEP seems to give a answer but it does not bear out.

Regards, Rick

Added

(Steam locomotives only).
The lowest cutoff (another term for the regulator) that the engine can use. For example if this is set to 10 and the cutoff is set to 5% direction the game a cutoff of 10% will be used by the model. This is typically the lead in the engines valve gear.

Think we may be in the realms of MSTS doing different things on different PCs, and different things at various times on the same PC! I have tried altering the SteamSmallestCutoff before now, but it made no difference to steam consumption in my tests. I'll have to have another play when I get the chance.

Quite agree that the link in MSTS between speed and steam usage is wrong, doubling the speed simply doubles steam consumption (everything else being equal) in MSTS. As I understand it, in real life this doesn't happen as at higher speeds the ports aren't open for long enough to fill the cylinder with as much steam as at lower speeds.

Not sure what you mean about the MEP? I'm just saying that it seems a better description (though not perfect) for what MSTS calls the Steam Chest Pressure, as the following formula seems to apply in MSTS:

Tractive effort at any one moment = (Steam chest pressure/Max boiler pressure)*Max force

as posted in the slipping thread. I think the MEP idea might have been suggested before somewhere, but I can't find the post at the moment.

Regarding the regulator parameters, the implication to me is that Kuju were trying to replicate the pilot valve/main valve characteristics of many steam locos, but only applied the code to steam use and forgot to apply it to the 'steam chest pressure' as well. This means the relationship between 'steam chest pressure' and steam use changes dramatically at 60% regulator (in this case). I think supergoods has done work on this before, see here: http://forums.uktrainsim.com/viewtopic.php?t=25040

cheers

Rich

MEP seems to give a answer but it does not bear out.

Rich I said this above as there was no way MS/Kuju would have got
into this, you only have to read the description for the cylinder volume
to recognise what the core programmer was employed on afore he was
asked to tackle steam output and consumption.

The program as I have said afore is clever but still falls short of being
a perfect simulation of a steam locomotive.

Regards, Rick

1crick14a wrote:MEP seems to give a answer but it does not bear out.

Rich I said this above as there was no way MS/Kuju would have got
into this, you only have to read the description for the cylinder volume
to recognise what the core programmer was employed on afore he was
asked to tackle steam output and consumption.

The program as I have said afore is clever but still falls short of being
a perfect simulation of a steam locomotive.

Regards, Rick

Completely agree with that . I guess we could maybe call the HUD 'steam chest pressure' : "Something Kuju happened to come up with that's similar to the MEP, but probably not on purpose"? That's a bit of a long name though . Whatever else the HUD 'steam chest pressure' is, I think we agree its definitely not the steam chest pressure?!

Cheers

Rich

First to answer some queries:

1crick14a wrote:Go read this line...........

SteamSmallestCutoff( 3/4/5/6/7//8/9/10 ) whatever, the slash =

Here's the value - SteamSmallestCutoff( 8 )

portbury wrote:How did you keep the speed of the loco and the boiler pressure constant during the tests btw? As far as I can tell, steam use in the cylinders is a straight line relationship with speed in MSTS, assuming everything else is constant. Boiler pressure also affects steam usage.

It was a fudge - I wanted a quick & easy way to find the relationships, so made the the loco stationary (ie permanent wheelslip caused by reducing 'adheasion'). By stepping through the regulator / cut-off values quickly the pressure stayed almost constant at 200psi. Perhaps if the tests had been running ones the Steam Usage rate woul've matched the Steam Cylinder Rate more closely.

portbury wrote:Regarding the regulator parameters, the implication to me is that Kuju were trying to replicate the pilot valve/main valve characteristics of many steam locos, but only applied the code to steam use and forgot to apply it to the 'steam chest pressure' as well. This means the relationship between 'steam chest pressure' and steam use changes dramatically at 60% regulator (in this case). I think supergoods has done work on this before, see here: http://forums.uktrainsim.com/viewtopic.php?t=25040

The info on that link set me thinking about single port regulators, and here's the results from the following code:

RegulatorValveType ( Single_port )
RegulatorPilotValveExponent( x.x )

5) Effect of RegulatorPilotValveExponent on Steam Chest Pressure (at 75% Cut Off):
a) 0.5
b) 1
c) 2

Once again - no effect

6) Effect of RegulatorPilotValveExponent on Steam Cylinder Rate (at 75% Cut off):
a) 0.5
b) 1
c) 2

Surprised at this one - I thought that 1 would give the straight line...

The Steam Cylinder Rate was considerably less with single valve than twin. I suppose the 2 valve values must be added together in the twin valve setup.

PS loco rated as MaxBoilerOutput ( 20070lb/h )

Here's the value - SteamSmallestCutoff( 8 )

Why ( 8 ) ??

Regards, Rick

I assume everyone is using the regulator settings in the initial post. I'm impressed by the smoothness of the curves which i think are due to the settings of Pilot and Main valves which you haven't really highlighted. The critical point is the setting of the PilotValveFullOpening at 1: twice the value of the MainValveInitialOpening of 0.5 .

I think this avoids the unrealistic hump in steam usage which occured going from pilot to second valve in the original Scotsman Twin port regulator settings and their consequent unusability.

My own solution was to adopt a PilotValveFullOpening of 0.5 and set the MainValveInitialOpenning to the lower value of 0.3 which also works well.

I don't think it's possible to achieve more than a fudge with regulator settings but at least this approach to pilot and main openings gives some finer control and an illusion of reallity at lower outputs.

My personal opinion on this is that the twin port regulator in the sim is better used as a regulator for a compound loco.

The two ports can be modified to deliberately produce the hump in steam production found when a compound starts as a simple and changes over to a full compound after reaching about 20mph.

I use the single port version for simple expansion locos.

Ian

1crick14a wrote: Here's the value - SteamSmallestCutoff( 8 )

Why ( 8 ) ??

Regards, Rick

Rick - you have a PM

morice1 wrote: I'm impressed by the smoothness of the curves which i think are due to the settings of Pilot and Main valves which you haven't really highlighted ... unrealistic hump in steam usage which occured going from pilot to second valve in the original Scotsman Twin port regulator settings and their consequent unusability.

Morice1 - I didn't investigate the settings of Pilot and Main valves because I was trying to see how the regulator code affected the Steam Chest Pressure and thereby Tractive Effort. I've got my answer now: not in the least (!) - it's simply tied to % Regulator opening.
However, I see what you mean when you compare the curves of the test engine as the 2nd valve comes into play against those of the Scotsman (where the rate halves) -

Here's the code:
RegulatorValveType ( Twin_port )
RegulatorSecondValveStartPosition( 0.5 )
RegulatorPilotValveExponent( 1.7 )
RegulatorSecondValveExponent( 1.7 )
RegulatorPilotValveFullOpenning( 0.3 )
RegulatorMainValveInitialOpenning( 0.4 )

and here's the graph:
7) Default Scotsman - Effect of Regulator Opening on Steam Usage & Steam Cylinder Rate (at 75% Cut off):


It clearly shows the hump in steam usage. I've shown the Total Steam Usage this time (the figure that appears in the F5 x 2 HUD). It's only loosely affected by the Steam Cylinder Rate (but it's probably the test method).

supergoods wrote: I use the single port version for simple expansion locos.

8) Effect of Regulator Opening (RegulatorPilotValveExponent = 0.5) on Steam Usage & Steam Cylinder Rate (at 75% Cut off):


Ian - A couple of questions about Steam Usage:
a) How do you find it over a long period of time? In the graph it's erratic, but that's only a very short-lived test.
b) In the graph the Steam Cylinder Rate only approaches the coded MaxBoilerOutput ( 20070lb/h ) as you increase cut-off beyond 75%. Is this borne out in running?