FS2004 Strange False Readings

[vololiberista]

With engines off and an OAT of 35c both EGT and ITT give false readings.
(A:ENGx EXHAUST GAS TEMPERATURE, celsius) is equal to 56c
as is (A:TURB ENGx ITT, celsius). Should they not equal ambient temperature if engines not running?

With ambient temps between 0 > 15 they are the same
ambient temp of 20 they equal 5 deg more 25
ambient temp of 25 they equal 9 deg more 34


vololiberista

 

[jx_]

How long since they engines were last run? They take four to eight hours to reach ambient internally. And just like brake heating, there are times when temps will actually rise after shutdown.

This all real world of course. In the sim, those variables are based on your scaling and rate of change. So either your scaling is way off or the temps are still slowly dropping.

Most likely the scaling is too high.

 

[vololiberista]

The figures are for cold engines. The problem is that as there is no ITT gauge I'm using to scale LP cooling air (i.e. the temperatue of the LP air before it enters the core compressor. )
As far as EGT goes there was no value at all in the air file. It may be a little high but I set it for hot starts to be above ambient 25c at sea level.
For example the max. EGT is 595c at 104% n2 During start-up it peaks at 600 which is ok as there is a Top Temp limiter switch. But higher temps at sea level will almost always require a hot start.
vololiberista

 

[jx_]

okay,

Both variables are controlled by the air file. EGT is 530 and ITT is 1540. They are pseudo...or fake... meaning you have to play with them to get the output numbers right. If your not getting good outputs, play with the numbers some.


Unless you know the temps at those stations for your engine. Then of course you can find temperature from temperature ratio and thrust. But you might still need to code something to handle the cold engine.

 

[WarpD]

I was given a rather definitive statement regarding things like EGR, ITT and such in the .air file:

There are entries AIR_61S_OIL_TEMPERATURE, AIR_61S_OIL_PRESSURE, AIR_70_ITT etc... and they all have 3 fields: a 'tuning constant', a max value and a 'time constant'.

Would it be too much to ask how these values are used to calculate the results in the sim? It's been more or less a 'hit/miss' game of trying to set correct values in the 'tuning constant' and 'time constant' to get decent values returned. A better understanding of how the sim actually uses those values would g a long ways to reducing the 'guessing game'.

They have an internal normalized algorithm based on the appropriate parameters (e.g. N, combustion, damage, atmosphere, etc...). The "peak" or "max" value essentially sets the high end "normal" limit. The "tuning constant" is just a scalar on top. It essentially does the same thing as scaling the max value by the same value and leaving it at "1.0". These generally utilize 1st order lags, in which the time constant is that specified in those .air blocks.

When you state 1st order lags... would that be something like this: new_value = time_constant * raw_value + (1 - time_constant) * old_value?

Yes, that looks right.

I've had zero time trying to do anything useful with this information... but, there it is. Given that, someone who's smarter than I could probably figure out a way to always get the right values going for an engine. Or at least a lot closer than we do now.

 

[taguilo]

EGT and ITT work fine in FS for running engines (with some minor discrepancies with RW), but once you shut off those fans, all that happens is quite unrealistic in both versions.
First of all, Hot Starts are not simulated so you have to figure out a method that suits your needs; I for one use a group of polinomials directly linked to N2 rotation speed to stabilization, outside temp, fuel ON moment, etc, etc.
Second, once the engine is closed temp cooling in FS drops almost immediately, which is utterly wrong. A turbine engine needs from 40 to 60 minutes of cooling time to be manageable by tech personnel, and a lot more, like jx_ stated, to reach ambient temperature. Again, in this case I use another group of polinomials that imitate a gradual cooling depending on ambient conditions, lapse the engine had been running, starter-on forced cooling, etc.
I have an example gauge (not the final version) that I've coded for the C208 Caravan; I can attach a copy here if it might be useful to anyone.

Tom

 

[mgh]

The attached figure shows the generalised response of a first-order lag to a step input. Although a step input isn't always realistic it can be helpful in getting an understanding. The figure shows that regardless of the value of the time constant, the output reaches 63% of the input after a time equal to 1 * Time Constant: 86% after 2 * Time Constant: 95% after 3 * Time Constant: and 99% after 5 * Time Constant.

In other words the output has stabilised after a time equal to 5 * Time Constants has elapsed for practical purposes. So, if some parameter needs 30 mins to stabilise the time constant should be 30 * 60 / 5 = 360 sec.

Of course this is in principle – FS may be different.

 

[vololiberista]

EGT and ITT work fine in FS for running engines (with some minor discrepancies with RW), but once you shut off those fans, all that happens is quite unrealistic in both versions.
First of all, Hot Starts are not simulated so you have to figure out a method that suits your needs; I for one use a group of polinomials directly linked to N2 rotation speed to stabilization, outside temp, fuel ON moment, etc, etc.
Second, once the engine is closed temp cooling in FS drops almost immediately, which is utterly wrong. A turbine engine needs from 40 to 60 minutes of cooling time to be manageable by tech personnel, and a lot more, like jx_ stated, to reach ambient temperature. Again, in this case I use another group of polinomials that imitate a gradual cooling depending on ambient conditions, lapse the engine had been running, starter-on forced cooling, etc.
I have an example gauge (not the final version) that I've coded for the C208 Caravan; I can attach a copy here if it might be useful to anyone.

Tom

I've played With different numbers including the "set" numbers and still get the same differences between ambient and actual if the ambient temp is above 15c. The numbers for EGT are the maximum operating temps for the engine. That all works well except fo the above. The manuel states that no start is allowed if the needle is "in" the numbers. Which means if I'm at a hot airfield I'm stuck!!! I'm happy to leave as is to simulate stopovers with the sim still live. Ambient temps would be better with a cold start so I can shave of any excess quite easily as L:vars always start at zero?? What happens in a "saved flight" do the L:vars retain their values?
Another option would be to somehow time the engine from shut down to start-up and create an artificial scale to add back into EGT and ITT.
vololiberista

 

[vololiberista]

okay,

Both variables are controlled by the air file. EGT is 530 and ITT is 1540. They are pseudo...or fake... meaning you have to play with them to get the output numbers right. If your not getting good outputs, play with the numbers some.


Unless you know the temps at those stations for your engine. Then of course you can find temperature from temperature ratio and thrust. But you might still need to code something to handle the cold engine.

Mine are set for a max EGT of 595c and engine overheat of 430c (i.e. value of LP cooling air).
vololiberista

 

[jx_]

Well they shouldn't be set to max EGT, that's the first mistake. All engines will exceed max temps if run too hard for too long.

What you need is the highest temp the engine can reach under the roughest operating conditions. In a real engine: with anti-ice on at emergency power at high density. This should be your simulated max.

That number is, while giving consideration for the vast number of different engines configurations, "usually" between 1200C-1800C EGT, and 1800C-2500C ITT.

Then use the rate of change variable to adjust things so that normal takeoff limited thrust (5-minutes at takeoff power) does not exceed the max rated temp. This will usually give a good fit to the engines overall performance, but will not help you with start temperatures. MSFS doesn't add FUEL until combustion has started at a point that is guaranteed to be successful. If we were to engage the starter and open the fuel valve at the same time in the real world, we would simply have a nice gas oven back there that would warm up the turbines. As you know, this added temp is what causes hot starts.


I believe you said you are using the ITT variable for HP intake temperature. That would never reach 430C. If you have the pressure ratio or temperature ratio of the LP fan you can simply multiply ambient total temp * temperature ratio or find temp from pressure through density.

 

[jx_]

and be sure your airfile entry is in Rankine.

If max reachable temp is 1500C that is 2700R

 

[vololiberista]

Thanks for all that I'll fiddle around a bit more. For the moment I have a work around for the ambient temperature compensation. Doing the calc 1% lower allows the needle to rise smoothly.
As for the 430c. That is correct. It's the max allowed temperature for LP cooling air at the back of the LP compressor. These are RR Conways!!!
BTW although I set 595c as the maximum the engines do exceed that value in certain atmospheric conditions and hot starts waiting for n1 to catch-up! The max ambient starting temp for these engines is 50c.
vololiberista

 

[jx_]

Thanks for all that I'll fiddle around a bit more. For the moment I have a work around for the ambient temperature compensation. Doing the calc 1% lower allows the needle to rise smoothly.
As for the 430c. That is correct. It's the max allowed temperature for LP cooling air at the back of the LP compressor. These are RR Conways!!!
BTW although I set 595c as the maximum the engines do exceed that value in certain atmospheric conditions and hot starts waiting for n1 to catch-up! The max ambient starting temp for these engines is 50c.
vololiberista

Yes but it should be set to max possible not max allowable. A real engine won't stop rising just because max allowable has been reached.

Although you are correct that under various atmospherics you could exceed, can you exceed at takeoff thrust with the brakes locked for 1 hour??? Most likely not in a realistic manner.

 

[vololiberista]

Yes but it should be set to max possible not max allowable. A real engine won't stop rising just because max allowable has been reached.

Although you are correct that under various atmospherics you could exceed, can you exceed at takeoff thrust with the brakes locked for 1 hour??? Most likely not in a realistic manner.

The problem with setting max possible is that hot starts will occur normally even at temperatures below ISA. Whereas in real life they would be rare.
For example the RR conway is allowed to operate at 595c for 5 minutes even up to 690c for 2 seconds!!! But a typical start EGT at ISA would be 510-530c. So I'm going for a happy medium. It's a pity that engine wear and tear isn't modelled. If it can that would be a bonus.
vololiberista

 

[taguilo]

So, why don't you don't model yourself a decent EGT readout?
FS's own EGT values are totally fake; I mean you can put any value in .air or .cfg files and this won't affect engine's behavior at all. Using custom polinomials in a gauge, or even in the .MLD itself you can obtain proper readings at any moment of the flight, as I stated before.

Tom

 

[vololiberista]

So, why don't you don't model yourself a decent EGT readout?
FS's own EGT values are totally fake; I mean you can put any value in .air or .cfg files and this won't affect engine's behavior at all. Using custom polinomials in a gauge, or even in the .MLD itself you can obtain proper readings at any moment of the flight, as I stated before.

Tom

That's what I'm trying to do. What I'm aiming for is about 530c start up EGT in ISA conditions but peaking above 595c in hot weather. I'm fiddling with the numbers to get a broad brush readout. I'm no mathematician so complex equations are out!
vololiberista

 

[EduHir]

I have an example gauge (not the final version) that I've coded for the C208 Caravan; I can attach a copy here if it might be useful to anyone.

Tom, could i have a look at it??

Edi

 

[taguilo]

Tom, could i have a look at it??

Edi

Indeed. Here you have the gauge, in its former state (has been improved by now, but I can't post the final version)

All those polinomials are calculated using Excel's trend lines in graphics. You design the 2D table, build the proper graphic and the spreadsheet does the rest!

Hope this helps.

Tom

 

[jx_]

The problem with setting max possible is that hot starts will occur normally even at temperatures below ISA. Whereas in real life they would be rare.
For example the RR conway is allowed to operate at 595c for 5 minutes even up to 690c for 2 seconds!!! But a typical start EGT at ISA would be 510-530c. So I'm going for a happy medium. It's a pity that engine wear and tear isn't modelled. If it can that would be a bonus.
vololiberista

I think you misunderstood me.

The values in the airfile should be set to max possible. If you are getting spikes that are causing hot starts every time, then your rate of change is set too high. Reduce that until the spike is correct. You now have the base temp for a good engine start in place. You then need to add the all the scenarios you want to model on top of that, and any custom temperature character based on the Conway specifically.

You won't need complex math. And while Tom's approach is equally good, half the work is already done for you by the hard coded variable...

So assuming the max possible temp for your engine is about 1000c, you would need to set temp to 1800 in the airfile, and reduce the rate of change until you get the peak around 520c during start up. Remember, max possible is what you would see in a severe failure where temps go far beyond the limits. This is "usually" around 1.5 times the allowed emergency temp.

Also, don't be shy with these variables...they are just numbers. You could set it to any number with any rate of change that gives you the proper ratio between start, idle, and takeoff power temps, then scale those variables in the gauge display.

 

[EduHir]

Tom,
thanks a lot!

Interesting thread at all!

Edi