Steam Engine Phisics

[Bill69]

Hi All,

I have been spending a lot of time checking which tags in the steam section of e-spec actually affect the performance of the loco.

Some things I found are quite the opposite to what has been posted as a fix, especially for noisy safety valves blowing off.

It has been said that if your safeties are blowing let the coal burn down and keep it at a low percentage.
This does absolutely nothing to cut down the steam generated by the boiler or reduce the blowing off.
Coal fed to the fire is totally window dressing and does nothing to control the amount of steam generated by the boiler.

Three tags that do most of the control for steam are 1. firebox-heating-surface-area. 2. max-fire-temperature. 3. min-fire-temperature.
These three tags have to be able to supply enough steam to keep the boiler up to near operating pressure but not over operating pressure.
So if your loco is continually blowing the safeties these are the tags to change first, however if reasonable changes do not produce the desired affect you should also look at the boiler-efficiency, boiler-efficiency-idle and boiler-efficiency-min tags. Depending on where your loco gains pressure. If it gains pressure doing heavy work it will be the boiler-efficiency, if it gains pressure when stationary it will be the boiler-efficiency-idle and if it gains pressure all the time it will be boiler-efficiency-min.
Reduce the relative tag gradually, only by about 0.1 between each test.

Another problem I found is the steam-chest-max-flow tag, according to the Trainz Dev Wiki this should be about 200 for a medium size loco, I find this too high.
If it is too high you will find it difficult to accelerate when you start off, it tends to bog down the loco and you will have to reduce the regulator and possibly the cutoff to get the loco going. I would suggest 150 to 180 for a medium size loco but remember the steam used here will also affect the safety valves, steam generated has to be balanced with steam used.

Have fun,
Bill69

 

[pcas1986]

These are rather difficult to test so I appreciate the effort it must have taken to reach those conclusions. Thank you.

Cheers

 

[Bill69]

Yes you are right there Paul. the testing took longer that the changes.
After every change, when things were getting close, a 30 mile test in Cab mode then a 30 mile test in DCC mode.

Cheers,
Bill

 

[Azervich]

Hi Bill,

Their are other tags that do affect the performance of a steam loco

  firebox-thermal-conductivity          5 (I set this between 4 and 5)
  firebox-plate-thickness               0.014287 (larger the plates the less heat goes to the boiler)
  fuel-energy                           29322453 (lower values means less heat energy into the fire when at max temp)
  fuel-specific-heat-capacity           1288.52002 (the fire temp for the coal to release it's set energy)

other tags that should be used are these

  starting-boiler-steam                 0
  starting-water                        9192.019531
  starting-coal                         356

I have starting water about 65 to 80% sight glass, trainz defaults about 37% and when you start a DCC/AI session it starts pumping water till it reads 70% sight glass and that's what causes safeties to go off when stationary on session startup.

Cheers.

 

[Bill69]

Hi Azervich,

I use only one of these tags you have mentioned, the firebox-thermal-conductivity, the others I do not use. They are not required in CMPs checking and not required to get a steam engine running as it should.

Bill69.

 

[Azervich]

Trainz defaults those tags when not specified in the config, like fuel-energy defaults to 32473980 (14,000 BTU Pound) and puts out large amount of heat into the fire, even at lower temps, the tag fuel-specific-heat-capacity defaults too 1200 I think it was.

Cheers.

 

[Bill69]

As I said in my first post coal in the firebox does nothing to the boiler temperature, therefore 14,000 BTUs per pound will do nothing.

These are the tags I use in all my steam section e-specs.

number-cylinders
number-power-strokes
firebox-heating-surface-area
boiler-volume
steam-chest-volume
steam-chest-max-flow
max-fire-temperature
min-fire-temperature
initial-boiler-temperature
max-coal-mass
ideal-coal-mass
shovel-coal-mass
safety-valve-low-pressure
safety-valve-high-pressure
safety-valve-low-flow
safety-valve-high-flow
water-injector-rate
piston-volume-min
piston-volume-max
piston-area
burn-rate
burn-rate-idle
speed
piston-angular-offsets
firebox-thermal-conductivity
super-heating-constant
firebox-efficiency
blower-effect
blower-max-flow
boiler-efficency
boiler-efficency-idle
boiler-efficency-min
valve-lap-percent
cutoff
hand-brake-max-force


Bill69

 

[JamesMoody]

It has been said that if your safeties are blowing let the coal burn down and keep it at a low percentage.
This does absolutely nothing to cut down the steam generated by the boiler or reduce the blowing off.
Coal fed to the fire is totally window dressing and does nothing to control the amount of steam generated by the boiler.

That is incorrect.

Heat to boiler comes from the combination of the firebed temperature and the coal mass.

Less coal mass == fire goes colder faster == less heat transfer. No coal mass == no heat transfer at all.

The firebox surface area, wall thickness and conductivity changes only *how fast* the heat gets to the boiler, not *how much* heat gets there.

However, note that this is an over-time effect, not an instant effect. The boiler retains temperature and pressure, and a fire that is in the process of burning down will also be /hotter/, as it isn't having to heat new coal to start that new coal burning. So you won't see the low mass effect come in immediately, only after the fire has been low for some time and has been subsequently allowed to cool.

Nevertheless, if you are having to keep the coal at a low level to avoid safety valves popping, it sounds like there's a problem with the enginespec. It might not be where you think it is though.


Most changes in a steam loco enginespec have two effects, and because of this you generally spend a while chasing a problem around in a steam loco espec until it comes out at one end or the other - either the coal settings or the cylinders.

The biggest thing I've found that needs to be correct is the piston minimum volume. This is the volume of not only the clearance at the end of the piston, but of the steam passages between the cylinder and the valve. Get this too small by even a tiny margin, and you get a loco that is significantly overpowered, and operates way too efficiently. Then you naturally compensate for this by driving it too gently, which results in more efficiency, and in even less steam usage. Result is you never use even remotely close to enough steam, and that means you never use even remotely close to enough water, either. This config would also tend to build pressure and blow safeties.

Check to see if you are using enough water. If not, your front end is too efficient, and increasing the cylinder minimum volume is how you fix it.

Once you've got the front end set up right, then you can fine-tune the boiler and firebox to match the front end.

Three tags that do most of the control for steam are 1. firebox-heating-surface-area. 2. max-fire-temperature. 3. min-fire-temperature.

With regard to the heating surface area - set this as correctly as you can. Do not tune this value.

Instead tune the maximum firebed to boiler heat transfer by one or other of the the various transmission tags (e.g. heat conductivity and wall thickness). These are all used in the same calculation. Doubling the transfer area and halving the thickness will have the same effect on the code, but halving the thickness is way less likely to get someone to declare "that's bull****" when reading your config.
Note that tuning these values is a zero sum game - it's not about how much heat gets through, but about how quickly it gets through. If you can't keep the fire hot, look to turn this down. If you can't raise enough steam with a hot fire (and the fire then stays hot for ages no matter what you do), then then look to turn this up.

Max and min fire temperature are intended to be 'edge conditions' to stop crazy situations arising. If you are running into those, and they are affecting you in significant ways, you have bigger problems elsewhere.

If you have a problem with safeties blowing at idle, look at increasing the boiler-heat-loss tag. This creates a constant reduction on the boiler temperature, tending to bring it down. It should be set to a level that will balance out the idle fire input to stop the pressure building forever at idle with the loco not moving. The efficiencies you mention are also worth looking at, but not sufficient on their own without the boiler-heat-loss tag.

Another problem I found is the steam-chest-max-flow tag, according to the Trainz Dev Wiki this should be about 200 for a medium size loco, I find this too high.
If it is too high you will find it difficult to accelerate when you start off, it tends to bog down the loco and you will have to reduce the regulator and possibly the cutoff to get the loco going.

That's called wheelslip. (The non-animated version, sadly). Yes, you have to be careful with the regulator when starting a steam loco. They do have a tendency to spin.

I would suggest 150 to 180 for a medium size loco but remember the steam used here will also affect the safety valves, steam generated has to be balanced with steam used.

150 to 180 still sounds pretty similar to the suggested "about 200" - the advice was "probably not 20 unless your loco is truly tiny, and not 2000 unless you are really modelling an ocean liner or something."

You should tune this value so the steam chest pressure does what you'd expect across a range of cutoff/regulator settings. Too much flow and you'll find you can't reduce power with the regulator very well and you get problems controlling the steam chest pressure at start. Too little flow and the steam chest will go flat even with the regulator wide open, and you'll get no power. (OTOH, that's a legit issue with some real steam locos - so if that's the effect you want to produce, that's how you do it.)

Also check what your local meaning of "medium size" is versus the rest of the world. What a British person thinks of as a medium sized loco is unlikely to be "medium" to American eyes.

 

[JamesMoody]

Their are other tags that do affect the performance of a steam loco

  firebox-thermal-conductivity          5 (I set this between 4 and 5)
  firebox-plate-thickness               0.014287 (larger the plates the less heat goes to the boiler)

These values affect the speed at which the heat gets into the boiler, not so much the amount of heat. But you are certainly better off tuning these two than adjusting the firebox heating surface area. Tuning these has the same effect.

  fuel-energy                           29322453 (lower values means less heat energy into the fire when at max temp)

That's the amount of energy that is released by burning a set amount of coal. I think the numbers are in joules per kilogram?
It does vary between different types of coal. Our default is for bituminous coal, which is one of the most energy dense types. Anthracite would be a bit lower, and poor quality coal types are a lot lower.

  fuel-specific-heat-capacity           1288.52002 (the fire temp for the coal to release it's set energy)

That's the amount of energy you need to raise the temperature of the coal by a certain amount. I think it's joules per kilogram per kelvin. It's not an absolute temperature. It affects how fast the firebed will change temperature in response to stimuli - how much 'stored heat' there is in the coal. Higher values will mean more 'stored heat', so a slower response to change.

Again, coal varies per type. Our default is for bituminous coal, and is one of the highest. Anthracite would be lower.


If you are modelling a loco that didn't typically run on excellent quality coal, you certainly should be tuning both of these values, probably down from the defaults by quite some margin

other tags that should be used are these

  starting-boiler-steam                 0
  starting-water                        9192.019531
  starting-coal                         356

Starting boiler steam at zero is a good move - but I think is also the default, so is probably unnecessary. If you put a value in here, you can get an "artificially high" pressure by bottling steam, and make the safeties pop instantly. (If you want to check your safeties are working, set this to a big number.)
Starting water you should set to a good value based on your boiler volume and sight glass levels. Our default is a fairly low guess, but reasonable - it is, at least, somewhere on the sight glass. Don't copy/paste one from someone else without thinking about it - if you get outside this range, you will have big problems.
Again with starting-coal, base this from your coal levels rather than cut'n'paste.

I have starting water about 65 to 80% sight glass, trainz defaults about 37% and when you start a DCC/AI session it starts pumping water till it reads 70% sight glass and that's what causes safeties to go off when stationary on session startup.

This ... is actually an issue. Trainz doesn't currently model steam re-condensing to water if the temperature of the boiler is below the pressure on the pressure/temperature curve. When a loco is being driven, this is mostly irrelevant. Assuming the regulator is open even 1%, there will always be a drain of steam from the boiler to some degree, and this will generally fix it within a few frames. But if a loco is stationary with the regulator totally closed, no steam gets out, and it will 'bottle' the steam it has rather than allowing it to re-condense. (The opposite effect, where steam is taken out, resulting in a lower pressure and water spontaneously boiling until equilibrium is met again *is* modelled. That's important to keep a loco operating smoothly).
Anyway, if you add water, this reduces the space for steam, and if the steam isn't going out, it results in higher pressure. (It shouldn't - it should recondense. But as I said, we don't currently model that effect). If it gets too high, the safeties will pop and release it. It should only be a short burst though, not a continuing blow, as the actual steam reserves in the boiler (not counting the water...) turns out to be tiny. If you've got continuing blow, you have another problem somewhere else. As this is an idle condition, it could well be insufficient boiler-heat-loss.

 

[Bill69]

I'm sorry but I don't agree, Your description may be how it is meant to work but it does NOT work. To test it pick a loco that is continually blowing the safeties and run it till the coal in the firebox gets way down, even down to two percent, you will find the loco is still using water and still blowing the safeties. As far as 'medium' is concerned that's the word used in the wiki. The piston min volume I calculate by subtracting the swept volume from the cylinder volume, if it is required to be any more accurate than that you would have to take the piston rod size into consideration. Heating surface area is often given in loco data on the Wiki. I usually keep it as given if it is not given I use PEV's tools to measure the firebox in the mesh. After much testing I can make an e-spec which will allow the boiler pressure to deviate by only 30psi. in DCC mode and can be controlled to the same level in CAB mode.


Bill69

 

[nimec]

Some of you have gone to a lot of trouble to explain what should happen. I say SHOULD but the reality is not always like that. I totally support Bill69 and his comments about adding coal etc as one of my 'trips' from the top of the bank on Connellsville to Cumberland, the loco kept howling off continuously all the way from the top until the end of the game for probably nearly an hour. I went over the top with the steam well back and the boiler low in water to try to combat it. For those of you who have never really worked on a steam engine, 'howling off' was the term used in New Zealand for the safeties going off. I let the coal get down to 2%, yes James Moody 2% and still the safeties kept howling off. There are many steam locos in this game that do this and do not reflect reality and howl off repeatedly and unnecessarily which cannot be controlled with the coal or water as suggested above. This is not an isolated case but is regular. And in case you are wondering, I do know how to drive main line steam engines. In this game, I do not use DCC only cab driving mode so have control over the loco performance and most do not respond as you say. In fact I think there is something seriously wrong with the physics of steam locos in this game. Many may respond appropriately while steaming or on the up hill but when drifting or on the downhill are uncontrollable even with the coal at 0%.

In another thread I saw a few weeks ago, they were also talking about loco's howling off and most were calling for an option in the new T:ANE to turn off that noise. Even experienced and respected trainzers like Phil Skene supported it.
It's not everyone who can adjust an e-spec or even wants to try and while the support for such an option was very strong in that thread, that guy Murphy of N3V was going against the tide and saying that it shouldn't be done. Hopefully the designers of T:ANE have a more sensible view and include it as an option. It should be all about what the user, the customer, the client wants.
Make a provision in T:ANE for people to turn off the safeties of and you will win friends but a head in the sand approach and telling people how they should drive a loco is just arrogant.
Many of the people on this forum have never driven a steam engine so make it a bit easier on them. Get the physics sorted.

 

[Azervich]

If you are modelling a loco that didn't typically run on excellent quality coal, you certainly should be tuning both of these values, probably down from the defaults by quite some margin

Hi James,

It's been said that Victorian Black Coal from the Wonthaggi Mines was very poor over the Qld and NSW Coal, I believe it's between 12,000 BTU Pound to 13,000 BTU Pound, but their are no books reporting just how much it was, only that it was cheaper then getting coal from NSW, as I like Steam Locomotives between 1857 through to 1890's they used Coke Coal and NSW Coal so my e-specs are set to use those depending on the specific fuel the loco had, from 1910 they started using Victorian Black coal so I model this into the e-specs for the era, then in 1920's we experimented with Pulverized Brown Coal (the experiments failed due to lack of technology and was poor burning) and in 1940's to Oil Burning, P.B.C started again in 1948-49 on the X Class 2-8-2 and was a Success

So I think it's important that we use this tag and turn efficiencies to ballence the boiler steaming, I try to use real firetemps too, so in trainz I use a max between 1640 Kalvin (2492 Fah) to 1760 (2708 Fah) Kalvin.

Cheers.

 

[Bill69]

quote>If you've got continuing blow, you have another problem somewhere else. As this is an idle condition, it could well be insufficient boiler-heat-loss.<


Or fire temperatures set too high.

Bill69

 

[Bill69]

Here is a very short video of wheelslip that James Moody says is not supported.
I had to keep it very short otherwise it would take hours to upload.




I am getting very sick of the crap that comes out of N3V about what happens and what is and is not supported, and I don't give toss if I get banned for this.


Bill69

 

[Azervich]

Hi Bill,

The thing about the wheelspin you see is a forced wheelspin, I've run some tests and found that trainz reduces the piston output power to the wheels, think of it this way, if I have 20,000lbs of force and I start slipping the piston should still be putting out 20,000lbs of force, this is not the case in trainz, it reduces the piston power during wheelspin, I've found that output power can drop down to 7,000lbs which causes very slow wheelpin or the wheels just sliding and not turning at all.

Cheers.

 

[Bill69]

Yes I know it's forced wheelspin I banged the regulator open as soon as the brakes released. The thing is N3V said it was not supported.
I can also get quite fast wheelspin on a more powerfull loco but can't be bothered uploading another video.

Bill69

 

[ZecMurphy]

That's a slightly different effect Bill. We had, in TC3, a fully animated wheelslip action (as in, full steam locomotive wheelslip). This is the older diesel style wheel slip, which is really just 'wheel slide'.

If you read James' post, you'll note that he did NOT say it wasn't occurring in physics. Just noting it wasn't the fully animated version (again, what you are seeing is not the same wheelslip animation that TC3 had, which broke in TS2009 and wasn't noticed until around mid-late last year...).

Regards

 

[ParFive]

New around here.
Have been playing around with TS10 and some steam.
Back to safety valves.
I volunteer as a conductor on a steam train and my understanding of the safety valves is that they are set higher than the wanted boiler pressure.
A boiler designed for a running pressure of 200 psi would have the pops set at say 205 psi. I have no idea if these figures are correct.
However when I run the steam in TS10 the maximum pressure seems to be 201 psi.
The strange thing is that they pops do not go off then. They are continually going off ant 197 or 198 psi.
I would think they shouldn't go off at a lower pressure but at the high pressure to protect against explosion.

 

[ZecMurphy]

Good Morning ParFive
This depends on the location (e.g. different countries have different rules, or had different rules, on the pressures to use for the safety valves as against the boiler pressure), and how well the safety valves have been set. Either way, the pressure at which the safety valves go off is set in the locomotive's enginespec.

The pressure at which the safety valves will re-seat themselves is actually lower than the stated maximum boiler pressure in some cases. It is below the pressure at which they are set to go off, so as to actually expel enough steam to control the boiler pressure. So if you set them to open (blow off) at 205PSI, then they should re-seat at a lower pressure (say 195-200PSI, possibly lower in some cases, possibly higher in others).

One thing we don't currently model is the feathering of safety valves. This is where, when within about 5-10PSI of the blow-off point, the safety valves start to sizzle and release steam to some degree. Pop type safety valves tend to do this less than things like the ramsbottom safety valves, where you really were just pulling against a simple spring setup. How much steam is expelled, and when it occurs, will depend on how well the safety valves are maintained, and what pressure you are at.

Regards

 

[Falcus]

Personally, I'm waiting for a more robust Boiler Firing Simulation before I get really excited about the steam valves.... The game barely takes thermodynamics (Or the "momentum" there of) into account, and that played a huge role in whether a fireman was going to pop his valves or not.... But I realize to do this is not going to be an easy task, so whatever. I'll just keep using Philskenes, lilbs, and Klene&Lerro's Engine Specs till then, as these have tended to be the most forgiving for not running down the tracks hearing "PFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFT" the whole friggin way, lol.

Falcus