Hurricane engine torque

[TomcatViP]

Home computating power is not the real limiter for flight Sim.
Think that your FM does not hve to compute how your plane will behave to any kind of solicitation. Most of the Maths are done once for all B4 you release your FM.

For example if I want to simulate flight buffeting on low level flight at high speed I won't hve to actually calculate for each image projected on the screen how my ctrl surface will behave and than what wld be the amplitude of the plane then pilot head shaking !! An average non harmonic solicitation will do the trick

Good Phy and Math are essential to hve a nice FM. But an overall comprehension of what you hev to look too or generally speaking a comprehension of the flight phenomenas (and I didn't say "sciences") is of utmost importance.

Years ago Engineers where computing real FM with computers that had the power of a today pocket calculator. Nowadays a millions more power is used at every steps of the process by any one in the dev chain (thx to CAO requirements)... And we don't do better planes (F35/Typhoon/A400M etc...)

[Sternjaeger II]

Originally Posted by TomcatViP

Home computating power is not the real limiter for flight Sim.

erm... it is actually. There is virtually no limit to the intricateness of a flight model, but it will need more and more computing power. You could build a CAD model of a Spitfire and simulate fatigue on each single screw if you wanted to, but it will need a lot of computing power.

Think that your FM does not hve to compute how your plane will behave to any kind of solicitation. Most of the Maths are done once for all B4 you release your FM.

For example if I want to simulate flight buffeting on low level flight at high speed I won't hve to actually calculate for each image projected on the screen how my ctrl surface will behave and than what wld be the amplitude of the plane then pilot head shaking !! An average non harmonic solicitation will do the trick

Good Phy and Math are essential to hve a nice FM. But an overall comprehension of what you hev to look too or generally speaking a comprehension of the flight phenomenas (and I didn't say "sciences") is of utmost importance.

Years ago Engineers where computing real FM with computers that had the power of a today pocket calculator. Nowadays a millions more power is used at every steps of the process by any one in the dev chain (thx to CAO requirements)... And we don't do better planes (F35/Typhoon/A400M etc...)

I'm sorry but I don't think you have a full understanding of the topic man, or if you do, your explanation is ridden with generalisations.

Engineers don't calculate flight models, a flight model, even when accurate, is an approximation used for simulators. To design a plane you can render in CAD the surface and structure and calculate loads or aerodynamic coefficients, but you will still need to build and test prototypes.

[jf1981]

Originally Posted by Viper2000

Most of the aeroplanes I've flown in the sim will actually sort themselves out in cruise. However, you have to be flying at the correct speed, boost, rpm and altitude, otherwise it's not going to work.

Filling in the numbers I get roughly 4808 lb-ft. The maximum weight of the aeroplane is rather less than 7000 lb. Go figure...

Hi,

I double checked and found roughly your figures. I further estimate the plane would fall (roll) on the side by 10 to 15 degree in a second if you would'nt compensate.

How comes the counter torque effect is lowered with hight speed (when not compensated, it would roll fast at low sped, but roll less at high speed) ?

Regards
JF

[Viper2000]

Originally Posted by Peril

Your a funny man Viper, you think like an opposing team not wanting improvements.

Only if your definition of "improvement" is "change for the sake of change".

Originally Posted by Peril

For info it was a Hurricane II with a Merlin XX, Rotol constant Speed prop, document is 'Measurements of the Flying Qualities of a Hawker Hurricane, by Neilsen and phillips. You could have this test data on your HD already I suppose??

Nope. If I had the data I wouldn't have been asking about it.

Originally Posted by Peril

But if you don't and want to improve the sim, 'please supply more data' would be a good start. I tested the Hurricane in game marked as a Rotol prop Hurricane, not sure if that is a definitive match, it's hard to tell based on that description.

They are different aeroplanes with very different engines (Merlin XX is single stage 2 speed with a redesigned supercharger), a different wings (Hurricane II has a 12 gun/4 cannon metal wing vs the 8 gun metal or wooden wing of the Hurricane I) and a different fuselage length (Merlin XX is longer so the Hurricane II has a longer fuselage).

Obviously these are quite substantial design changes, and for this reason it isn't obvious that the Hurricane II would have the same trim behaviour as the Hurricane I.

I'm all in favour of making the sim as good as it can be, but I don't think that there's much to be said for deliberately asking for trim behaviour to be based upon that of a substantially different, albeit related, aeroplane.

Given a comprehensive "changelog" between the Mark I and Mark II aeroplanes (which I don't have, before you ask; I know the broad brush strokes, but I can't immediately bring to mind the sort of specific details required to get trim behaviour right) we could potentially try to reverse engineer Mark I data from the Mark II data that you've found.

Originally Posted by Peril

So you are aware, I spent the last 6 years building and researching FM for another sim focused on realism, so I have some experience and data to call upon and know what is required. I too didn't modify my FMs without good data but I'm not one to ignore what is logical either. If it is of interest we had to apply on average 0.2 deg of wing twist to obtain an outcome that match reality re roll trims. Sims are great representations but unfortunately they are rarely perfect in result because of using the simplified code they have to use due to the limits of home computing. Before you go off all negative, think about what we have proved here and work backwards from there, I always found the answer could be somewhere in the middle.

Feel free to use any alt, power and the glide test data to confirm, I think you'll find it likely has a problem

That would confirm some changes are needed, some of us are here to 'help' where we can...the only agenda I have is FM realism.

Building sim models is all about matching.

The problem is that there are lots of variables to consider, and depending upon the underlying flight model of the simulator itself it may or may not be possible to accurately match all of the available test data.

For example, you can't make an accurate WWII fighter in X-Plane without recourse to plugins because X-Plane just can't model a WWII engine properly; you can't get realistic supercharger behaviour because it only allows for turbonormalising, and you can't get a satisfactory model of exhaust thrust either. Which means that most people end up with inaccurate airframe drag and/or an inaccurate propeller model in an attempt to match the headline top speed of the aircraft. This then results in incorrect cruise and glide behaviour. Essentially, the tighter you squeeze any one parameter, the more likely the others are to slip through your fingers.

The most dramatic instance of this I ever encountered was a Corsair which flew nicely and looked pretty but was almost impossible to land. What happened was that the guy who'd made the model didn't realise that the R-2800 has a 0.5:1 reduction gear. So he had a very supersonic prop with awful performance, and this had forced him to dramatically reduce the airframe drag in order to match the stated top speed; it was therefore basically impossible to get the thing to descend for landing!

Until such time as the SDK is released and documented, and the major engine bugs are squashed, it's pretty difficult to work out the specifics of the model used in CoD, and therefore it's hard to form a sensible opinion as to how good the models are given whatever the limitations of the overall FM are.

[Sternjaeger II]

Originally Posted by jf1981

Hi,

I double checked and found roughly your figures. I further estimate the plane would fall (roll) on the side by 10 to 15 degree in a second if you would'nt compensate.

How comes the counter torque effect is lowered with hight speed (when not compensated, it would roll fast at low sped, but roll less at high speed) ?

Regards
JF

because at higher speeds, when the wing lift is more pronounced (efficient), torque is partially compensated by your wing lift.

[Sternjaeger II]

Originally Posted by Viper2000

The most dramatic instance of this I ever encountered was a Corsair which flew nicely and looked pretty but was almost impossible to land. What happened was that the guy who'd made the model didn't realise that the R-2800 has a 0.5:1 reduction gear. So he had a very supersonic prop with awful performance, and this had forced him to dramatically reduce the airframe drag in order to match the stated top speed; it was therefore basically impossible to get the thing to descend for landing!

wow, I'm surprised the thing gets airborne in the first stance! Does X-Plane model propeller cavitation too?

[TomcatViP]

Originally Posted by Sternjaeger II

erm... it is actually. There is virtually no limit to the intricateness of a flight model, but it will need more and more computing power. You could build a CAD model of a Spitfire and simulate fatigue on each single screw if you wanted to, but it will need a lot of computing power.



I'm sorry but I don't think you have a full understanding of the topic man, or if you do, your explanation is ridden with generalisations.

Engineers don't calculate flight models, a flight model, even when accurate, is an approximation used for simulators. To design a plane you can render in CAD the surface and structure and calculate loads or aerodynamic coefficients, but you will still need to build and test prototypes.

Dear SternJ,

I am sry but you didn't took the train on time. See bellow post of Vip to understand that I was talking abt tweaking to simulate complicated probs.

Moreover it's over insupportable to hve to read on each of your post basically the "You don"t know as your are not this or that and blablabla"

Firsty I have reasonable confidence of my knowledge and I am always pls to learn/re-learn new things
Secondly I am debating here with all the people that are enjoying on this forum a new sim that is a step further toward our expectation regarding realism. I am not talking to you abt myself, you or my own lovely stern.

Pls open up and be alert and positive

[Viper2000]

Originally Posted by Sternjaeger II

wow, I'm surprised the thing gets airborne in the first stance! Does X-Plane model propeller cavitation too?

You can't get cavitation in air; cavitation is a phase-change phenomenon.

You can certainly stall the propeller in X-Plane, and it will suffer shock losses. Actually the propeller model is one of its nicer features.

However, in general the underlying assumptions are undocumented and subject to change without notice. So you build a nice model, and then find that Austin has decided to tweak something in the next version and suddenly the predicted performance changes dramatically.

For this reason, it can't be used for serious work where fidelity is important.

Austin is a law unto himself and his system does not necessarily follow the conventions one would expect (eg although the underlying model seems to use SI, airframe dimensions are input in decimals of feet. Meanwhile, propeller root and tip chords are input in decimals of inches...). The engine model is a bit crazy, kinetic heating data is very questionable for M>>2, transonic behaviour is obviously lacking in fidelity, etc etc.

It's very good at replicating Austin's personal flying experience, but that's basically limited to GA piston singles. So although the simulator will allow you to build a rocketship and fly into LEO, the results have problems.

Of course, the average user doesn't understand the limitations of the code, not least because the underlying assumptions are undocumented, and therefore tends to believe that because the computer told them that their design would do this then it must be true.

So I fear that it's probably only a matter of time before some fool conducts virtual spin testing of their homebuilt masterpiece in X-Plane and then wins a Darwin award by assuming that X-Plane has validity in this regime simply because it produces output...

[Sternjaeger II]

Originally Posted by Viper2000

You can't get cavitation in air; cavitation is a phase-change phenomenon.

You can certainly stall the propeller in X-Plane, and it will suffer shock losses. Actually the propeller model is one of its nicer features.

However, in general the underlying assumptions are undocumented and subject to change without notice. So you build a nice model, and then find that Austin has decided to tweak something in the next version and suddenly the predicted performance changes dramatically.

For this reason, it can't be used for serious work where fidelity is important.

Austin is a law unto himself and his system does not necessarily follow the conventions one would expect (eg although the underlying model seems to use SI, airframe dimensions are input in decimals of feet. Meanwhile, propeller root and tip chords are input in decimals of inches...). The engine model is a bit crazy, kinetic heating data is very questionable for M>>2, transonic behaviour is obviously lacking in fidelity, etc etc.

It's very good at replicating Austin's personal flying experience, but that's basically limited to GA piston singles. So although the simulator will allow you to build a rocketship and fly into LEO, the results have problems.

Of course, the average user doesn't understand the limitations of the code, not least because the underlying assumptions are undocumented, and therefore tends to believe that because the computer told them that their design would do this then it must be true.

So I fear that it's probably only a matter of time before some fool conducts virtual spin testing of their homebuilt masterpiece in X-Plane and then wins a Darwin award by assuming that X-Plane has validity in this regime simply because it produces output...

yeah sorry, we erroneously refer to stalling as cavitation, but in fact it's propeller stall. Sounds interesting, does it also make the typical noise?

As for your last statement, there's nothing better than natural selection man

[Sternjaeger II]

Originally Posted by TomcatViP

Dear SternJ,

I am sry but you didn't took the train on time. See bellow post of Vip to understand that I was talking abt tweaking to simulate complicated probs.

Moreover it's over insupportable to hve to read on each of your post basically the "You don"t know as your are not this or that and blablabla"

Firsty I have reasonable confidence of my knowledge and I am always pls to learn/re-learn new things
Secondly I am debating here with all the people that are enjoying on this forum a new sim that is a step further toward our expectation regarding realism. I am not talking to you abt myself, you or my own lovely stern.

Pls open up and be alert and positive

no man, I'm really trying to be nice, but you were talking baloney.. and fair enough, it's not like we can know everything about everything, but if there's a topic you don't have a complete understanding of, just don't give statements like that, because people like me, who are driven by a particularly pedantic nature, won't like it.. it's for the sake of science really..