Outerra forum

User mods, screenshots & videos => Aircraft => Flight Model Development => Topic started by: bomber on April 03, 2015, 06:38:56 am

Title: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 03, 2015, 06:38:56 am
This is how I do the engine, it's pretty straight forward at this point.

So with data from wiki and the operators manual
http://www.kalwishky.com/files/O-300%20Operators%20Manual.pdf

http://target4today.com/_posted_files_/Cesna_172/Continental_O-300.xml
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 04, 2015, 05:06:48 am
ok here's the results of the Static Engine Test on the original engine...

(http://target4today.com/_posted_files_/Cesna_172/engine-1.JPG)

Not Good, as there's insufficient power at sea level, but not to worry as this is because of the Volumetric Efficiency being only 0.756...

So we'll raise it up to 1.0 and retest.
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 04, 2015, 06:25:24 am
So I ran the engine at 2450rpm (recommended cruise RPM) at 9000ft with a wind velocity of 125mph (cruise speed) and this helped me determine the cooling factor for the engine cowling to ensure that the engine cylinder head temp doesn't go above 525F..
I adjusted the BSFC value whilst cruising to ensure that the fuel consumption was within spec 12.5gallons per hour
At sea level I adjusted the Volumetric efficiency so as the engine produced 145HP at 2700RPM as per spec.

And then performed the Static Engine Test at all alts.

(http://target4today.com/_posted_files_/Cesna_172/engine-2.JPG)
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 04, 2015, 06:42:41 am
As always here's the spreadsheet that goes with it...

http://target4today.com/_posted_files_/Cesna_172/Static_Test_Results.xlsx

3 years ago I brought the subject up over on the JSBsim dev mailing list that something might very well be wrong with the piston engine code, produced data like this for the merlins. But was met with denials and assertions that this was an FG issue.. Interestingly when I brought the subject up at FG, they said it was a JSBsim issue.

I then spent the intervening years on aerodynamics of a glider, as I couldn't trust the piston engine to give the correct values in the hope that someone would come along and help rectify the problem within the JSBsim C++ code.... that's not happened.

The bottom line here is that even with the best aerodynamic model, if the power/trust unit (engine/prop) produces the wrong figures, then your aircraft performance just wont be right...

Now I can create a JSBsim 'system' using xml to correct the error in the engine....I was just hoping that I wouldn't have too...

Regards

Simon
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: Revolver on April 04, 2015, 11:37:08 am
Hello Simon,

very well that you trouble in this direction. I have not argued yet with JSBSim, but I see graphics in your tests that there everything is not perfect. I mean with the fact, that the engine JSBSim as such permits to us not everybody original data simply purely press and with it get the same tests graphics which agree with original Work-Tests-Graphics with each other.
For me personally it becomes interesting when the test graphics are same with each other, because then it will show us to all that Motor-(X) is programmed so immediately, how in real world.
In my opinion it is also important because afterwards not only everything will correctly function, but one will have all originating questions (why we have differences etc.) with a blow away.
And another one I would like to say. For the easy user it would have been better and easier a lot if he writes the original-data purely and gets back the correct values as they should be.


BR,
Stefan

p.s. At the moment I understand so, what inscribing in the program original dates we won't receive the copy on an output which it is possible to call identical with the original? ::)
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: Ron Jensen on April 04, 2015, 02:11:29 pm
3 years ago I brought the subject up over on the JSBsim dev mailing list that something might very well be wrong with the piston engine code, produced data like this for the merlins. But was met with denials and assertions that this was an FG issue.. Interestingly when I brought the subject up at FG, they said it was a JSBsim issue.

I think your statement is an unfair assessment. I have personally spent many, many hours responding to your posts and e-mails, and tried my best to explain how and why the code works the way it does, both on the JSBSim side and the FG side. As I have stated before, I am unable to reproduce your results using valid test methods and feel your methodology of adjusting atmospheric properties while your models stay at sea-level. is the source of your error. The simulations are not designed to allow that.

For example, as I look at your spreadsheet I see you have a MAP of 15.7 (inHg?) at "24000 ft" while you have a "Barometer" of 11.6 inHg. This should not happen as MAP should always be less than Barometric pressure in a normally aspirated engine. Going back to your reference book, 15.7 inHg yields 80+ hp at 16,000 ft, so the actual result of 90 hp at 24,000ft seems reasonable since horsepower increases with altitude when MAP is held constant.

Using my methods give me the attached graphs. The red dashed line is the expected horsepower
(http://www.jentronics.com/fgfs/temp/b01.png)
(Edit: Add the image as an inline link instead of an attachment)
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 05, 2015, 02:29:44 am

I think your statement is an unfair assessment. I have personally spent many, many hours responding to your posts and e-mails, and tried my best to explain how and why the code works the way it does, both on the JSBSim side and the FG side.

As I have stated before, I am unable to reproduce your results using valid test methods and feel your methodology of adjusting atmospheric properties while your models stay at sea-level. is the source of your error. The simulations are not designed to allow that.

I'm sorry you feel that way Ron, I've always appreciated the time and effort you've put in helping me understand how the engine and thruster file contents work...

but that's a separate issue to this topic and I think it's unfair to attempt to mix the two together...

Your only response to the original post I made 3 years ago, was that and to para-phrase "you can't do it that way, I can't reproduce it, and anyway FG uses a different atmospheric model which makes it their fault"..
There was never hours spent on it, you've posted more on this topic yesterday than ever you did before and this is the first time I've ever seen any supporting data...

It wasn't my intention to be unfair but to resolve this issue... If it turns out that testing the piston engine the way I've been doing it is wrong and the piston engine is correct, then I'll be over-joyed and the issue will be resolved...

Now I will go have a look at your comments, and post my findings as soon as possible....

Quote
For example, as I look at your spreadsheet I see you have a MAP of 15.7 (inHg?) at "24000 ft" while you have a "Barometer" of 11.6 inHg. This should not happen as MAP should always be less than Barometric pressure in a normally aspirated engine. Going back to your reference book, 15.7 inHg yields 80+ hp at 16,000 ft, so the actual result of 90 hp at 24,000ft seems reasonable since horsepower increases with altitude when MAP is held constant.
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 05, 2015, 07:46:38 am
This is the piston.cp file...

https://gitorious.org/jsbsim/jsbsim/source/eef1ac615c3d804ceaa947e839e921950e93b486:src/models/propulsion/FGPiston.cpp#LNaN-NaN

line 448
Quote
//%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 
void FGPiston::ResetToIC(void)
{
  FGEngine::ResetToIC();
 
  ManifoldPressure_inHg = in.Pressure * psftoinhg; // psf to in Hg
  MAP = in.Pressure * psftopa;
  TMAP = MAP;
  double airTemperature_degK = RankineToKelvin(in.Temperature);
  OilTemp_degK = airTemperature_degK;
  CylinderHeadTemp_degK = airTemperature_degK;
  ExhaustGasTemp_degK = airTemperature_degK;
  EGT_degC = ExhaustGasTemp_degK - 273;
  Thruster->SetRPM(0.0);
  RPM = 0.0;
  OilPressure_psi = 0.0;
  BoostLossHP = 0.;
}
 
//%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

line 358
Quote
  property_name = base_property_name + "/map-inhg";
  PropertyManager->Tie(property_name, &ManifoldPressure_inHg);


So the next search is to find in.Pressure to see how this is derived...

Any idea Ron where it might be ?
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: Ron Jensen on April 05, 2015, 01:59:49 pm
So the next search is to find in.Pressure to see how this is derived...

Any idea Ron where it might be ?

This is set by a call to Atmosphere->GetPressure(); in FGFDMExec.cpp (http://jsbsim.cvs.sourceforge.net/viewvc/jsbsim/JSBSim/src/FGFDMExec.cpp?revision=1.171&view=markup) line 454.
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 05, 2015, 02:53:14 pm
forgot to post this earlier, had family round..

(http://target4today.com/_posted_images_/pistonengine_entity_relationships/doMap.png)
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 05, 2015, 07:11:49 pm
Ok as it would now seem that testing the engine the 'Rolls Royce' way of adjusting the pressure and temp within a testbed doesn't work I've created a floating plane that hovers to the required altitude.... in effect allowing me to do a static engine test.

So I'm now looking forward to some more engine testing and hopefully getting values that agree with those of Ron.

(http://target4today.com/_posted_files_/Cesna_172/engine-test.png)

There is one little niggle I have however, and that's this....... Will JSBsim be able to cope with differing planetary values for pressure and temp with alt ?
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: Uriah on April 05, 2015, 11:00:14 pm
Nice work Simon!!!
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 12, 2015, 05:18:47 pm
OK this is how the engine tests out using the new method....

(http://target4today.com/_posted_files_/Cesna_172/engine-4.JPG)
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 13, 2015, 07:34:59 am
Ok, it's time for some good reading...

http://www.epi-eng.com/piston_engine_technology/engine_technology_contents.htm
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: HiFlyer on April 13, 2015, 07:39:26 am
Ok, it's time for some good reading...

http://www.epi-eng.com/piston_engine_technology/engine_technology_contents.htm

Little technical. I'm waiting for the action comic addition........  =D
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 14, 2015, 10:34:52 am
Read this....

http://www.pilotfriend.com/training/flight_training/fxd_wing/props.htm

If you take nothing from this document other than there being 3 types of propellor.... no not fixed pitch, constant speed, variable pitch....

but...... Climb, Cruise and Speed props....

So when you're aiming for a plane that 'hits all the numbers'..... ask yourself how's that even possible if I can only use a single prop at a time ?
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 14, 2015, 01:47:32 pm
Now have a little look around the next tool....javaprop

http://www.mh-aerotools.de/airfoils/jp_applet.htm
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 14, 2015, 03:31:37 pm
Ok right mouse button, save as

http://target4today.com/_posted_files_/Cesna_172/Prop-Cesna_172.xlsx

Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 15, 2015, 03:09:48 pm
If you're unlucky to have very little information about the prop.... which depending upon your era could be all of your planes, you're going to need to do design your prop to the known values you could have of the plane... ie it's engine, top speed etc...

First question is "how big is my prop"

So I've added the next tab on the spreadsheet which determines the maximum prop size you can have without running into helical tip issues...

http://target4today.com/_posted_files_/Cesna_172/Prop-Cesna_172.xlsx
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 17, 2015, 03:05:15 pm
http://forum.outerra.com/index.php?topic=2851.0

turning

(http://target4today.com/_posted_files_/Cesna_172/javaprop-1.JPG)

to this

(http://target4today.com/_posted_files_/Cesna_172/javaprop-2.JPG)
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 17, 2015, 03:20:29 pm
More stuff to read

http://web.mit.edu/16.unified/www/FALL/thermodynamics/notes/node86.html

hahahah don't worry, just go right to the bottom and look at the graphs..... the shapes of these graphs are important, if your results don't look like these then how can they be right.
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 18, 2015, 06:08:46 am
http://www.mccauley.textron.com/tech_guide.pdf
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 18, 2015, 07:17:51 am
ok so let use the prop

McCauley 1A200DFA 82/42. 82” 2-blade aluminum prop. (Fixed pitch)
McCauley 2-blade w/spinner 48 pounds
McCauley 2-blade (2450 RPM) 750 pounds
McCauley 2-blade 95 Kts TAS

so we're looking for a static test (0m/s) of 750lbs (3.34kN) of thrust at 2450rpm, using all the grunt the engines got, ie 145hp (108kW)

Adding these values into the first tab and using the magic of javaprop we get this....
(http://target4today.com/_posted_files_/Cesna_172/javaprop-3.JPG)

producing a set of tables like this
(http://target4today.com/_posted_files_/Cesna_172/javaprop-4.JPG)

and a nice set of graphs for Ct and Cp that look like this
(http://target4today.com/_posted_files_/Cesna_172/javaprop-5.JPG)

However this is the current prop graphs for the jsbsim Cessna
(http://target4today.com/_posted_files_/Cesna_172/javaprop-6.JPG)


I guess we need Ron to explain the differences.

Regards

Simon
 p.s. updated
http://target4today.com/_posted_files_/Cesna_172/Prop-Cesna_172.xlsx
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 18, 2015, 01:38:28 pm
I've been thinking about it and it seems that it could be linked to a windmilling prop.
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 18, 2015, 04:09:26 pm
creating the same spread on the V/(nD) of 0.05 and using xcel to mirror the values for windmilling we get the following...

I need to understand why the Cp is negative on the previous chart

(http://target4today.com/_posted_files_/Cesna_172/javaprop-7.JPG)
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 19, 2015, 07:58:04 am
more reading on stationary vs windmilling props

http://lancair.net/lists/lml/Message/44132.html?Language=
Title: Re: Piston Engine Modelling..... including it's prop.
Post by: bomber on April 21, 2015, 02:45:53 pm
MT MTV-15-B/210-58. 83” 2-blade composite prop. (Constant speed)


(http://target4today.com/_posted_files_/Cesna_172/prop-1.JPG)