P-40 turn rate/Flight model check

[ACG_KaiLae]

Mostly was looking forward to getting the stuff we needed. Have now to figure out plan B.

[6./ZG26_Klaus_Mann]

Mostly was looking forward to getting the stuff we needed. Have now to figure out plan B.

They didn't ban him, they flatout deleted his account. That's pretty harsh. He may be horrible to most people most of the time, but he doesn't post Nazi Content or just outright Trolls people. I don't think he deserved that. I've done far worse and come out with temporary bans only. 

Edited January 11, 2017 by 6./ZG26_Klaus_Mann

[unreasonable]

Don't forget the Spitfire which has the same problem. It would be instructive to do more of the IL2 simulated aircrafts' empirical CLmax'es. I've already done the 109E7 and found that its calculated CLmax from the dev numbers is very similar to the measured value of CLmax that the RAE aircraft tests determined from the trailing pitot method.

 

? I am bringing the Spitfire case precisely because it has the same problem! (And was told not to....)

 

But I have found something very interesting when extending the investigation to the methodology and it's application to a wider sample: looking at the RAE report on the Bf109E here:

 

http://www.kurfurst.org/Tactical_trials/109E_UKtrials/Morgan.html

 

there are two tables that show both ASI and Vi which is the "trailing static". Vi is shown as higher than the ASI - and the size of the difference is greatest at lowest speeds.

 

See if I can attach the tables:

 

 

 

The methodology is referred to in a note, but the source unfortunately turns out to be an unpublished paper.

 

Kurfurst's insistence on white text on black makes this a hard read for me, so perhaps I have missed something, but this seems very relevant.  

This suggests that either:

 

1) The PEC charts for the P-40 are being read the wrong way round, or 

2) That they cannot be extrapolated to the stall speed region because some other phenomenon than is captured in the extrapolation affects the adjustment 

Edited January 11, 2017 by unreasonable

[6./ZG26_Klaus_Mann]

? I am bringing the Spitfire case precisely because it has the same problem! (And was told not to....)

 

But I have found something very interesting when extending the investigation to the methodology and it's application to a wider sample: looking at the RAE report on the Bf109E here:

 

http://www.kurfurst.org/Tactical_trials/109E_UKtrials/Morgan.html

 

there are two tables that show both ASI and Vi which is the "trailing static". Vi is shown as higher than the ASI - and the size of the difference is greatest at lowest speeds.

 

See if I can attach the tables:

 

109 snip 1.PNG

 

109 snip 2.PNG

 

The methodology is referred to in a note, but the source unfortunately turns out to be an unpublished paper.

 

Kurfurst's insistence on white text on black makes this a hard read for me, so perhaps I have missed something, but this seems very relevant.  

This suggests that either:

 

1) The PEC charts for the P-40 are being read the wrong way round, or 

2) That they cannot be extrapolated to the stall speed region because some other phenomenon than is captured in the extrapolation affects the adjustment 

The Ram Pressure in the Pitot Tube Changes with the positioning of the Airframe in the Air. That K-21 hasn't become stationary in the Air, it's slipping, yet the indicator goes to almost 0. 

On a K-13 it will shoot to well above 200kph or less than 0, depending on which direction you slip in. 

 

[unreasonable]

If you take the numbers from that chart for flaps up stall speed - weight and S also given, assuming corrected for SL - you can get CLmax:

 

With pilot's ASI = 2.23 

 

With trailing ASI = 1.38

 

RAE shows 1.4 - same as my result if I round.

 

IMHO this pretty conclusively demonstrates why taking stall speeds out of manuals and sticking them in the equations is an invalid method of determining CLmax empirically.

[unreasonable]

The Ram Pressure in the Pitot Tube Changes with the positioning of the Airframe in the Air. That K-21 hasn't become stationary in the Air, it's slipping, yet the indicator goes to almost 0. 

On a K-13 it will shoot to well above 200kph or less than 0, depending on which direction you slip in. 

 

 

Right, but we have to assume that slipping is not the issue, rather more likely to be the high AoA - and not just a speed effect - causing a systematic error between plane pitot measurements and trailing pitot measurements.

 

The RAE tests suggest that the plane pitot under-reads the ASI compared to a trailing pitot, by 20 mph, ie  over 20%! This is a massive error, even before the squaring effect on calculated CLmax,  and with an opposite sign to that which has been assumed in some posts.

 

The Spitfire RAE tests conclusions and manual stall speed data follow exactly the same pattern. It would be extraordinary if the same was not true for the P-40.

[6./ZG26_Klaus_Mann]

Right, but we have to assume that slipping is not the issue, rather more likely to be the high AoA - and not just a speed effect - causing a systematic error between plane pitot measurements and trailing pitot measurements.

 

The RAE tests suggest that the plane pitot under-reads the ASI compared to a trailing pitot, by 20 mph, ie  over 20%! This is a massive error, even before the squaring effect on calculated CLmax,  and with an opposite sign to that which has been assumed in some posts.

 

The Spitfire RAE tests conclusions and manual stall speed data follow exactly the same pattern. It would be extraordinary if the same was not true for the P-40.

The Axis doesn't matter, the angle does. 

[unreasonable]

The Axis doesn't matter, the angle does. 

 

If I understand what you are saying then a slip angle of say 20 degrees at zero AoA generates the same error as an AoA of 20 degrees with no slip - both at  the same speed? 

 

If that is right - and I can see why that would make sense - if your slip error (ie angle error) is as large as your example suggests, then this will be much greater than a PEC which is (if I understand correctly) only a speed adjustment, and has no implicit AoA/slip assumption.

 

edit: I have to admit some considerable uncertainty on these technical points. The RAE 109 report does contain this:

 

 

 

 

Whatever the technicalities, the RAE is calculating CLmax using a trailing pitot and a resulting speed much higher than that recorded on the pilot's gauges. 

 

 

Edited January 11, 2017 by unreasonable

[6./ZG26_Klaus_Mann]

If I understand what you are saying then a slip angle of say 20 degrees at zero AoA generates the same error as an AoA of 20 degrees with no slip - both at  the same speed? 

 

If that is right - and I can see why that would make sense - if your slip error (ie angle error) is as large as your example suggests, then this will be much greater than a PEC which is (if I understand correctly) only a speed adjustment, and has no implicit AoA/slip assumption.

Well, I may have oversimplified, the Static Ports matter as well. And often the Pitot Tube and Static Ports are positioned in a way that Sideslip interferes with it a lot more than AoA, but in general, yes. That's why there is a Special Pitot Tube extension for the K-21 as seen in the video which reduces the effect of the angle of Airflow on the reading. 

Edited January 11, 2017 by 6./ZG26_Klaus_Mann

[unreasonable]

Looking at the language of the RAE report it looks as though their "PEC" adjustment includes everything in the difference between the pilot's ASI and the trailing ASI.  20mph being one of the experimental points on the curve and also the difference in the tables at that speed region.

 

Fairly "brave" to extrapolate a curve from two points, but hard to ignore the data points that are there at low speeds, which are missing on all the P-40 PEC curves.

[BlackSix]

They didn't ban him, they flatout deleted his account. That's pretty harsh. He may be horrible to most people most of the time, but he doesn't post Nazi Content or just outright Trolls people. I don't think he deserved that. I've done far worse and come out with temporary bans only. 

6. It is forbidden to discuss the actions of moderators and administrators in any form on the forum. All questions are to be sent via "personal message" to the administrator/moderator.

 

Violations of this rule will result in the following:

 First offense - 3 days ban on entry

 

And a Second account can also help keeping your Main Account to Positive Content and if you are part of a Squad it really helps you image not being a Male Donkey all the time. 

I want to remind you the forum rules:

 

21. Registering and usage of a backup (optional) account without the consent of the forums administrator is prohibited. Violation of this rule will result in the deletion of the duplicate account and limited access rights for the main forum account for a period of 1 month.

[Holtzauge]

? I am bringing the Spitfire case precisely because it has the same problem! (And was told not to....)

 

But I have found something very interesting when extending the investigation to the methodology and it's application to a wider sample: looking at the RAE report on the Bf109E here:

 

http://www.kurfurst.org/Tactical_trials/109E_UKtrials/Morgan.html

 

there are two tables that show both ASI and Vi which is the "trailing static". Vi is shown as higher than the ASI - and the size of the difference is greatest at lowest speeds.

 

See if I can attach the tables:

 

109 snip 1.PNG

 

109 snip 2.PNG

 

The methodology is referred to in a note, but the source unfortunately turns out to be an unpublished paper.

 

Kurfurst's insistence on white text on black makes this a hard read for me, so perhaps I have missed something, but this seems very relevant.  

This suggests that either:

 

1) The PEC charts for the P-40 are being read the wrong way round, or 

2) That they cannot be extrapolated to the stall speed region because some other phenomenon than is captured in the extrapolation affects the adjustment 

 

Some good research work and analysis there unreasonable! I have that report but I did not think about that part and this shows that at least on the Me-109E, the IAS reading or ASI the pilot see's on his speed indicator is then substantially lower than the true IAS. Also agree on points 1 and 2 and I'm leaning towards 2 being true and that the PEC is not meant to be extrapolated to stall speeds.

 

Looking at the language of the RAE report it looks as though their "PEC" adjustment includes everything in the difference between the pilot's ASI and the trailing ASI.  20mph being one of the experimental points on the curve and also the difference in the tables at that speed region.

 

Fairly "brave" to extrapolate a curve from two points, but hard to ignore the data points that are there at low speeds, which are missing on all the P-40 PEC curves.

 

Well I agree and my theory is also we cannot extrapolate the P-40 PEC to stall speeds because I believe the PEC only covers higher IAS speeds for a reason and that is since this is where a pilot needs the true IAS to use to calculate a true TAS to do accurate dead reckoning navigation. For stall the pilot is not interested in this: he uses the speed indicator reading as is and consequently it makes sense to have that value in the manual, irrespective of what the true IAS really is. I spoke to Dakpilot (who AFAIK is an IRL pilot who flies for a living) about this a while back and he seems to agree anyway.

Edited January 11, 2017 by Holtzauge

[ZachariasX]

Some good research work and analysis there unreasonable! I have that report but I did not think about that part and this shows that at least on the Me-109E, the IAS reading or ASI the pilot see's on his speed indicator is then substantially lower than the true IAS. Also agree on points 1 and 2 and I'm leaning towards 2 being true and that the PEC is not meant to be extrapolated to stall speeds.

 

 

Well I agree and my theory is also we cannot extrapolate the P-40 PEC to stall speeds because I believe the PEC only covers higher IAS speeds for a reason and that is since this is where a pilot needs the true IAS to use to calculate a true TAS to do accurate dead reckoning navigation. For stall the pilot is not interested in this: he uses the speed indicator reading as is and consequently it makes sense to have that value in the manual, irrespective of what the true IAS really is. I spoke to Dakpilot (who AFAIK is an IRL pilot who flies for a living) about this a while back and he seems to agree anyway.

 

I agree, unreasonable got an important point there. In the case of the ASK-21 glider (that I know well from presonal experience), the the angle of your airframe chord to flight direction has a big impact on the readings on the gauge.

 

For instance, making the speedometer drop to zero is a nice indicator when slipping "enough". You can do that easily. You even sense how the airflow starts to bend the canopy a bit. Same holds true for high AoA flying. PEC is not meant to corrections in that department. On one side you have the problem of the pitot to be not pointing into the wind, on the other hand you have the static pressure sensor on the side. If that one is turned into an unfavorable angle (either directly into the wind or covered behind something) you lose calibration.

 

Because of that, if there's an "onboard" measurement to be trusted in some way, I'd go with the trailing device that is completely out of the turbulence of the aircraft.

[=362nd_FS=Hiromachi]

Mostly was looking forward to getting the stuff we needed. Have now to figure out plan B.

Just contact them ? I'm in continuous relation with them since 2014 and have ordered numerous documents. https://airandspace.si.edu/archives-research-request

Fill the above research inquiry, specify what you are looking for (but remember, sometimes it might need more general description, those who will do the research are volunteers and they sometimes dont have knowledge in specific area) and wait. In 3 weeks to 1 month you should receive a letter from them containing Copy Form and list of documents they have in archive, then you have to mark documents you are interested in, calculate total cost, shipping cost and payment method (those are limited, I believe only check, money order or visa/mastercard credit cards) and you wait another weeks (up to 2 months actually) for them to process it. 

This has however limits - per order you cannot request more than 50 pages of document in paper or 30 rolls of microfilm. This is biggest flaw of NASM in my humble opinion, they should make an effort to digitize their archives since rarely anyone this days has stuff to preview those reels.

I have requested this very Monday some data from them and in a week or so letter should get to them. I will try to come back when order reaches me so you will know for sure how long it may take. 

[unreasonable]

Thinking about the implications of this a step further, if this is a general phenomenon then presumably it applies to other aircraft as well, (perhaps to a greater or lesser degree).

 

Given that in BoX as of now there is no PEC at all (as far as we can determine) it makes comparing manual stall speeds with game speeds problematic, whatever the reason. For instance, stall speed range for the 109 E-7 in BoX Tech Specs is  152-159kph = 94-99 mph.  

 

RAE report (I know slightly different mark) is 75mph pilot AIS, 95.5 trailing AIS.  So BoX corresponds aerodynamically very close to what RAE thinks is right, but us sitting in the cockpit are seeing something some 20mph different from what a 109 pilot would have seen.

 

Possible rule of thumb - single engine plane stall speeds in BoX will be about 20 mph higher than those shown in manuals? Works for the P-40E at least when taking the more fully loaded of my range of estimates. 109 mph vs about 90 for the manuals.

[Holtzauge]

I agree, unreasonable got an important point there. In the case of the ASK-21 glider (that I know well from presonal experience), the the angle of your airframe chord to flight direction has a big impact on the readings on the gauge.

 

For instance, making the speedometer drop to zero is a nice indicator when slipping "enough". You can do that easily. You even sense how the airflow starts to bend the canopy a bit. Same holds true for high AoA flying. PEC is not meant to corrections in that department. On one side you have the problem of the pitot to be not pointing into the wind, on the other hand you have the static pressure sensor on the side. If that one is turned into an unfavorable angle (either directly into the wind or covered behind something) you lose calibration.

 

Because of that, if there's an "onboard" measurement to be trusted in some way, I'd go with the trailing device that is completely out of the turbulence of the aircraft.

 

Was some time ago but I also used to fly the ASK-21. But I learned to fly on the Bergfalke III though! Most epic flight: In a Spatz being towed by a Tiger Moth over Ålleberg in Sweden: I remember the pilot complete with white scarf, brown leather helmet and goggles turning around and giving me the sign to release the tow line.......

 

Edit: BTW I'm not THAT old: It was part of instructor training and you were supposed to become familiar with a lot of different types..........

Edited January 11, 2017 by Holtzauge

[Holtzauge]

Thinking about the implications of this a step further, if this is a general phenomenon then presumably it applies to other aircraft as well, (perhaps to a greater or lesser degree).

 

Given that in BoX as of now there is no PEC at all (as far as we can determine) it makes comparing manual stall speeds with game speeds problematic, whatever the reason. For instance, stall speed range for the 109 E-7 in BoX Tech Specs is  152-159kph = 94-99 mph.  

 

RAE report (I know slightly different mark) is 75mph pilot AIS, 95.5 trailing AIS.  So BoX corresponds aerodynamically very close to what RAE thinks is right, but us sitting in the cockpit are seeing something some 20mph different from what a 109 pilot would have seen.

 

Possible rule of thumb - single engine plane stall speeds in BoX will be about 20 mph higher than those shown in manuals? Works for the P-40E at least when taking the more fully loaded of my range of estimates. 109 mph vs about 90 for the manuals.

 

That is my take as well: And there is probably no PEC in any other flight sim either. Makes sense if you think about it: When you calculate dynamic pressure you enter the square of the TAS not IAS. I'm pretty sure they calculate the dynamic pressure continuously in this way in BoX to calculate forces which are then integrated into motion and then use the density from the atmospheric model they have to calculate an IAS to be displayed in the cockpit. This IAS will of course be "perfect" CAS in that there are no calibration or instrumentation errors.....

 

One could of course argue that it would be nice to implement a PEC in each plane so you get the same reading as IRL but I would most certainly put that quite low on a scale of what to do. Much lower than introducing new planes and certainly way lower than addressing the turn rate issue that was discussed in the "another look a turn times" thread.

Edited January 11, 2017 by Holtzauge

[unreasonable]

In terms of programming in a PEC (or whatever this specific error should be called)  I agree - not a priority. What would be useful, however, is a developer statement, perhaps as a sticky attached to the Tech Specs, that since there is no PEC the speeds given in BoX, while correct, will not match those in manuals, especially at low speeds, and that this difference could be up to 20mph at the stall. 

 

That is if we have analysed this correctly.

 

It might help reduce the number of threads about FM errors or at least resolve them quickly.

Edited January 11, 2017 by unreasonable

[Holtzauge]

Now that sounds like a very good idea!

[ZachariasX]

 

 

What would be useful, however, is a developer statement, perhaps as a sticky attached to the Tech Specs, that since there is no PEC the speeds given in BoX, while correct, will not match those in manuals, especially at low speeds, and that this difference could be up to 20mph at the stall.

 

Not sure if this would be needed as it probably would cause more confusion to the "gamer". I mean, those subtleties on the speedometer are only relevant if you are performing longer flights and old style navigation. For dead reckoning, you have to know exactly what you are reading from the dials.

 

I feel stuff like that has more of a place in FSX and the likes. In a combat sim, all they do is taking away ressources. As long as you have a plausible IAS, you can do all computations necessary for simulating flight and it still leaves it up to the pilot to figure out TAS. If you go futher, there are many effects like AoA, slip, icing, etc... All for something that you not really want. You can simulate icing rather simple. There's a pitot heater switch, so simulating icing gives that one purpose, same as carb heater switch. But in a combat sim, I'd set the criteria for that up in a simple way... Save ressources for important stuff, like damage modelling.

 

As for flaring out, I mean, when you're right before touchdown, you don't really look at the speedometer, what is relevant is the attitude and the "feel" of the aircraft when you settle it.

 

What we're doing/discussing here is not the way the average customer should use those plane. If every customer was to reverse engeneer the flight dynamics, the devs had a bad life...

[JG13_opcode]

Agree re: developer statement. I posted a question in the sticky thread but Han didn't answer.

[ZachariasX]

Agree re: developer statement. I posted a question in the sticky thread but Han didn't answer.

In the end, why would he really want to disclose what they are doing in deatil? If he does, he gives the competition valuable info on how far they themselves should go with their efforts.

 

All 777 needs to do is deliver a good product that is of use to us. The "debuging info" is more something for the beta crew.

 

As much as I sympathise with the disclosure of the inner mechanics of the sim as you and unreasonable mentioned, I'm not sure I would do so if I was in Han's position. For 99% of the users this info is of no use and would just represent the sim, and not the actual aircraft. But if you are beta crew, you might have to sign something and then you can see what is concerning the sim.

[ACG_KaiLae]

Just contact them ? I'm in continuous relation with them since 2014 and have ordered numerous documents. https://airandspace.si.edu/archives-research-request

Fill the above research inquiry, specify what you are looking for (but remember, sometimes it might need more general description, those who will do the research are volunteers and they sometimes dont have knowledge in specific area) and wait. In 3 weeks to 1 month you should receive a letter from them containing Copy Form and list of documents they have in archive, then you have to mark documents you are interested in, calculate total cost, shipping cost and payment method (those are limited, I believe only check, money order or visa/mastercard credit cards) and you wait another weeks (up to 2 months actually) for them to process it. 

This has however limits - per order you cannot request more than 50 pages of document in paper or 30 rolls of microfilm. This is biggest flaw of NASM in my humble opinion, they should make an effort to digitize their archives since rarely anyone this days has stuff to preview those reels.

I have requested this very Monday some data from them and in a week or so letter should get to them. I will try to come back when order reaches me so you will know for sure how long it may take.

 

Generally, how much does that cost you?

[unreasonable]

@Zacharias I sort of agree - clearly most people are not bothered about this nor should the developers waste a lot of time modeling it. But if there are going to be threads saying "Plane X has a porked FM because the stall speed is massively wrong" as a result of this phenomenon, which inevitably get repeated in General Discussion, then there is an issue of damage control. Easy way to deal with it IMHO would be to refer people to an official note that explains the issue briefly but clearly.

 

Perhaps this debate over the P-40 is a one off: interesting that there were no complaints about the stall speed discrepancy for the 109E. We shall see.

 

@JG13_opcode  I saw that: you got the reply initially IIRC. Lost in translation? 

[Holtzauge]

Basically agree with all of the above BUT it would like unreasonable points out help also the developers if they confirmed there is no IAS error on the the IL-2 speed indicators in that they would not have to deal with a lot of complaints about porked FM's. This is not the first thread in which PEC's have been discussed and how/if they should be applied.......

[ZachariasX]

But if there are going to be threads saying "Plane X has a porked FM because the stall speed is massively wrong" as a result of this phenomenon, which inevitably get repeated in General Discussion, then there is an issue of damage control. Easy way to deal with it IMHO would be to refer people to an official note that explains the issue briefly but clearly.

 

 

True. But it should teach us to be careful with what we are "measuring" in the sim and we have to be careful with original reports. So the bar is high for our hobby here. The thing is, whatever implementation we have on speed readings, it is most likely systemic to all aircraft.

 

This means, whatever "error" we find, it should be an error as well on other aircraft.

 

It is a bit bold to say that aircraft X stalls at that speed in the sim, when in fact you cannot know exactly how this "speed" figure was devised. On the other hand, if something is off with one plane, it could be assumed that it should also be off on other planes. In other words, all aircraft should be discounted from theoretical values in a consistent manner. We can only find that kind of consistency and compare that one to original source data.

 

In sum, we know very little, but we learn a lot. Certainly I do.

Basically agree with all of the above BUT it would like unreasonable points out help also the developers if they confirmed there is no IAS error on the the IL-2 speed indicators in that they would not have to deal with a lot of complaints about porked FM's. This is not the first thread in which PEC's have been discussed and how/if they should be applied.......

You're speaking from my heart. I was just trying to see the other side ofthe coin.

[Holtzauge]

True. But it should teach us to be careful with what we are "measuring" in the sim and we have to be careful with original reports. So the bar is high for our hobby here. The thing is, whatever implementation we have on speed readings, it is most likely systemic to all aircraft.

 

This means, whatever "error" we find, it should be an error as well on other aircraft.

 

Exactly! In fact I found some interesting info in NACA report 420 which is an entire report dedicated to aircraft speed measurements that seems to indicate just that: An aircraft’s fixed pitot system does not give a good indication of the true IAS at stall speeds due to a too high crosswise flow over the static ports combined with a loss of total pressure in the pitot itself.

 

To be honest, I have not read through it completely but just browsed parts so I may have missed something but there is some rather interesting info on pages 391 and 392 (attached) that I have tried to summarize below:

 

As I understand it, it looks like there is a general issue with a pitot/static setup in that especially the static ports are sensitive to a change in the angle of attack: As long as the flow is parallel to the pitot tube things are hunky dory but if you get a component of pitch or yaw the static ports start trolling the measurements: Apparently the pitot is rather unsensitive up to an angle of attack or yaw of around 15 degrees but the static ports act up much earlier which as the author points out is not so strange since instead of a flow parallel to the ports, some will have a component of flow in and some out. Apparently the pitot is somewhat less sensitive and it looks like it is only at over 15 degrees aoa that you start to see noticeable errors. Figure 5 shows this graphically and indicates that the pressure differential gives a rather large error when the flow is offset at an angle of 15 to 20 degrees like when close to stall. Also note that the author mentions that for flight testing it is recommended to use a pitot static arrangement suspended from the plane and there is also a figure 7 showing that once you get the system trailing about 30 feet from the plane you start getting reliable IAS readings.

 

So another thing that would be good to look into is how the IAS was measured on the P-40: If it used a pitot static system similar to what is described in NACA 420 then it is not strange that we see the low stall IAS reading that we do in the P-40 manual, i.e. just as unreasonable’s earlier posts highlighted for Me-109, the P-40 IAS reading in the cockpit speed indicator will also be off from the true IAS value and by the looks of it this is simply a consequence of using a pitot/static system that does not align with the flow direction, which of course will be a universal effect affecting all aircraft equipped with a pitot/static system and not just limited to the aircraft we have been talking about here.

[Holtzauge]

OK, so I found a series of pictures of the P-40 pitot installation and it is just the type of installation that NACA report described as problematic: A pitot with static ports set perpendicular to the flow and we now know from NACA report 420 that this type of arrangement gives a too low reading at angles of attack of 15 degrees and higher, i.e. close to stall. So based on this I would say that the stall speeds given in the P-40 manual need to be adjusted upwards not downwards.

[Kurfurst]

In addition, remember that the RAE also measured the Clmax of a captured Me-109 using the very same method to 1.4 and that the Germans themselves estimated the Me-109 Clmax to.........you guessed it.....1.4.

 

I thought they did not measure it on the 109 but guesstimated from that measured on the Spitfire.

 

"The " stall boundary " depends on an estimate of CL max at full throttle. In the case of the Spitfire this has been measured in flight, while the Me.109 figures were based on the Spitfire results; tables of the assumed values of CL max are given in Fig. 17. CL max falls off as g is increased, because the stalling speed increases as g gets larger, thus lessening the slipstream effect."

 

Its always worth to check the primary sources first: http://kurfurst.org/Tactical_trials/109E_UKtrials/Morgan.html

[=362nd_FS=Hiromachi]

Generally, how much does that cost you?

Well, I'm not very generous and try not to throw money left and right so I never bought a single roll, though I know eventually I will have to since some documents like D4Y1 manual exceed 270 pages. Thus I stick to paper copies only - 0.3 $ per page with 50 pages limit. You have to keep in mind that for every document you have to add 2 pages for front and back cover but I always cheat on that one This time I packed order for 4 documents with 51 pages for all of them + 8 "mandatory pages" = 59 pages total. That cost me 17.7 $, to that you gotta add shipping costs according with the following table - https://airandspace.si.edu/research/resources/archives/pricing-fees so 8 $. So overall it's not extremely expensive, particularly if you take into account that I ordered captured Japanese documents from 1935-1945 period. For rare documents prices are quite ok, I could maybe complain that Australian War Memorial sends all this stuff on CD but its still so much better then London National Archives at Kew (Brits out there are damn greedy). 

[Holtzauge]

I thought they did not measure it on the 109 but guesstimated from that measured on the Spitfire.

 

"The " stall boundary " depends on an estimate of CL max at full throttle. In the case of the Spitfire this has been measured in flight, while the Me.109 figures were based on the Spitfire results; tables of the assumed values of CL max are given in Fig. 17. CL max falls off as g is increased, because the stalling speed increases as g gets larger, thus lessening the slipstream effect."

 

Its always worth to check the primary sources first: http://kurfurst.org/Tactical_trials/109E_UKtrials/Morgan.html

 

No, you are confusing the power on "stall boundary" with the Clmax on the glide. The Clmax of 1.4 refers to the power off or glide Clmax and has nothing to do with the Clmax at full throttle.

 

And yes, I agree that it is always best to check with the the primary sources, i.e. original documents instead of relying on transcripts like on your site. That is why I checked the original pdf I have of RM 2361 from which the attached page comes from. In this it is made quite clear that the Clmax was determined in flight trials with a swiveling pitot system and not as a "guesstimate" as you suggest.

[unreasonable]

Exactly! In fact I found some interesting info in NACA report 420 which is an entire report dedicated to aircraft speed measurements that seems to indicate just that: An aircraft’s fixed pitot system does not give a good indication of the true IAS at stall speeds due to a too high crosswise flow over the static ports combined with a loss of total pressure in the pitot itself.

 

To be honest, I have not read through it completely but just browsed parts so I may have missed something but there is some rather interesting info on pages 391 and 392 (attached) that I have tried to summarize below:

 

As I understand it, it looks like there is a general issue with a pitot/static setup in that especially the static ports are sensitive to a change in the angle of attack: As long as the flow is parallel to the pitot tube things are hunky dory but if you get a component of pitch or yaw the static ports start trolling the measurements: Apparently the pitot is rather unsensitive up to an angle of attack or yaw of around 15 degrees but the static ports act up much earlier which as the author points out is not so strange since instead of a flow parallel to the ports, some will have a component of flow in and some out. Apparently the pitot is somewhat less sensitive and it looks like it is only at over 15 degrees aoa that you start to see noticeable errors. Figure 5 shows this graphically and indicates that the pressure differential gives a rather large error when the flow is offset at an angle of 15 to 20 degrees like when close to stall. Also note that the author mentions that for flight testing it is recommended to use a pitot static arrangement suspended from the plane and there is also a figure 7 showing that once you get the system trailing about 30 feet from the plane you start getting reliable IAS readings.

 

So another thing that would be good to look into is how the IAS was measured on the P-40: If it used a pitot static system similar to what is described in NACA 420 then it is not strange that we see the low stall IAS reading that we do in the P-40 manual, i.e. just as unreasonable’s earlier posts highlighted for Me-109, the P-40 IAS reading in the cockpit speed indicator will also be off from the true IAS value and by the looks of it this is simply a consequence of using a pitot/static system that does not align with the flow direction, which of course will be a universal effect affecting all aircraft equipped with a pitot/static system and not just limited to the aircraft we have been talking about here.

 

This is close to my limit, so I might have this wrong... as I understand it:

 

1) The speedo will be measuring the difference between the total pressure reading from the pitot tube and the pressure reading from the static port. Ie P-S, not the pitot reading P.

 

2) If the difference is = impact pressure (our old friend 1/2rhoV^2) , then the speedo reads accurately, ie has a ratio of 1.0 on the graph

 

3) If  the ratio is higher than 1.0, then the speedo will be overestimating the speed, and visa versa.

 

4) I take line P and S to be independent and represent measured quantities in this graph, while the difference is the line PS: which goes above 1.0 up to about 20 degrees yaw, because although P goes below 1.0 with progressive yaw, S goes down even faster at least for the first 20-25 degrees.

 

5)  This would mean that at these angles (and we can safely ignore anything much beyond 20 degrees) the Pitot-static system used in the graphic would over estimate speed as it inclines 20 degrees or so: not underestimate it as in the RAE example.

 

6) In the RAE test we are looking for about a 20% error in speed, which I think would require a 44% error in P-S. 

 

First off it it is entirely possible that I have the wrong end of the stick in the above. But if not, it suggests that we are looking for a much bigger effect that dramatically underestimates P-S, which the graph only starts to do at angles of almost 40 degrees.

 

Or that it is the same yaw/inclination effect as in the graph, but with markedly different quantities? I am still confused....   

Edited January 11, 2017 by unreasonable

[Kurfurst]

No, you are confusing the power on "stall boundary" with the Clmax on the glide. The Clmax of 1.4 refers to the power off or glide Clmax and has nothing to do with the Clmax at full throttle.

 

And yes, I agree that it is always best to check with the the primary sources, i.e. original documents instead of relying on transcripts like on your site. That is why I checked the original pdf I have of RM 2361 from which the attached page comes from. In this it is made quite clear that the Clmax was determined in flight trials with a swiveling pitot system and not as a "guesstimate" as you suggest.

 

There is no confusion, the CLmax at full throttle - the engineers like Morgan would use for estimating turning capacity of a plane - wasn't measured as this is clearly stated in the Morgan report.

 

Unless of course you're are on the opinion that you need to use the power off Clmax, i.e. the one for gliding conditions, for power on full throttle conditions like turning, which may be very much the reason to the anomalies of your calculations for the P-40. You are perhaps assuming and using the wrong kind of Clmax for it, hence they do not match the measured turn data the type. My advice to you - always stick to and try to correctly interpret the source data, lest we find yourself in the pitfalls of misinterpreting and subsequently using wrong base data and wrong results, see for example interpretation pitfalls of the RAE TN 1106, or more famously, the MT 215 trial report.

 

Long story short - if you are using gliding CLmax for your 109 or P-40 turning calculations, you are doing it wrong.

[-WILD-AlbinoHA5E]

So, does that mean that most Stall Speeds of the ingame Aircraft should be reviewed, since the numbers in the Manuals are most likely wrong for many of them?

I imagine the 109s and Fw would also have to be reviewed.

[Holtzauge]

This is close to my limit, so I might have this wrong... as I understand it:

 

1) The speedo will be measuring the difference between the total pressure reading from the pitot tube and the pressure reading from the static port. Ie P-S, not the pitot reading P.

 

2) If the difference is = impact pressure (our old friend 1/2rhoV^2) , then the speedo reads accurately, ie has a ratio of 1.0 on the graph

 

3) If  the ratio is higher than 1.0, then the speedo will be overestimating the speed, and visa versa.

 

4) I take line P and S to be independent and represent measured quantities in this graph, while the difference is the line PS: which goes above 1.0 up to about 20 degrees yaw, because although P goes below 1.0 with progressive yaw, S goes down even faster at least for the first 20-25 degrees.

 

5)  This would mean that at these angles (and we can safely ignore anything much beyond 20 degrees) the Pitot-static system used in the graphic would over estimate speed as it inclines 20 degrees or so: not underestimate it as in the RAE example.

 

6) In the RAE test we are looking for about a 20% error in speed, which I think would require a 44% error in P-S. 

 

First off it it is entirely possible that I have the wrong end of the stick in the above. But if not, it suggests that we are looking for a much bigger effect that dramatically underestimates P-S, which the graph only starts to do at angles of almost 40 degrees.

 

Or that it is the same yaw/inclination effect as in the graph, but with markedly different quantities? I am still confused....   

 

Well the way I read it is like this: With a perfect pitot the curve P should be a straight line at 1 but it is not: it looses total pressure as the angle of yaw or attack goes up. Now with the static ports it is the opposite: they should also be a straight line indicating they are insensitive to a change in angle of attack or yaw but they instead get a slight increase in pressure. Now since the ASI relies on measuring the difference, the indicated speed should go down relative the true as the pressure in the pitot head is slightly reduced while the pressure in the static part is increased. As to how much this pressure error would change the actual reading on the speed indicator in the cockpit of a P-40 I would not care to guess because for that we would need that particular calibration curve but I believe the figure is valid in indicating the trend though.

[unreasonable]

 

 

Long story short - if you are using gliding CLmax for your 109 or P-40 turning calculations, you are doing it wrong.

 

No-one in this thread is doing any of that. The CLmax that is being discussed is at the power off stall speed, and the bone of contention is the difference between the CLmax implied by stall speeds given in manuals, those calculated from special tests like the RAE examples, and those calculated from the developers' in-game data. That is all - no-one has made any extrapolation of values to the case of turning or high speeds in this thread IIRC. There is merely the assumption that if the power off value is "wrong", then other aerodynamic properties are likely to be wrong too.

 

 

So, does that mean that most Stall Speeds of the ingame Aircraft should be reviewed, since the numbers in the Manuals are most likely wrong for many of them?

I imagine the 109s and Fw would also have to be reviewed.

 

I very much doubt it. AFAIK the developers have taken the relevant data from controlled tests - like the RAE reports, or wind tunnel tests - rather than from manuals. The Fw190 case is peculiar because of the interpretation of one particular test. The manual numbers are not wrong in that they are what an actual pilot in RL would observe on his instruments: they just do not take into account various measurement errors. The numbers you see in game do not have the measurement errors so they reflect the actual behaviour of the aircraft (as best as the developers can model it) but are different from the manual numbers.

Edited January 11, 2017 by unreasonable

[unreasonable]

Well the way I read it is like this: With a perfect pitot the curve P should be a straight line at 1 but it is not: it looses total pressure as the angle of yaw or attack goes up. Now with the static ports it is the opposite: they should also be a straight line indicating they are insensitive to a change in angle of attack or yaw but they instead get a slight increase in pressure. Now since the ASI relies on measuring the difference, the indicated speed should go down relative the true as the pressure in the pitot head is slightly reduced while the pressure in the static part is increased. As to how much this pressure error would change the actual reading on the speed indicator in the cockpit of a P-40 I would not care to guess because for that we would need that particular calibration curve but I believe the figure is valid in indicating the trend though.

 

But the graph and the text both say that the pressure at the static port falls with yaw, not increases. Text in section (d) just under Fig. 6

 

Hence the line PS in Fig. 5 goes above 1.0 in the 0-20 degree zone.

 

So you might be right that that is how a particular installation could behave, but it is not what the graph shows.

[Holtzauge]

There is no confusion, the CLmax at full throttle - the engineers like Morgan would use for estimating turning capacity of a plane - wasn't measured as this is clearly stated in the Morgan report.

 

Unless of course you're are on the opinion that you need to use the power off Clmax, i.e. the one for gliding conditions, for power on full throttle conditions like turning, which may be very much the reason to the anomalies of your calculations for the P-40. You are perhaps assuming and using the wrong kind of Clmax for it, hence they do not match the measured turn data the type. My advice to you - always stick to and try to correctly interpret the source data, lest we find yourself in the pitfalls of misinterpreting and subsequently using wrong base data and wrong results, see for example interpretation pitfalls of the RAE TN 1106, or more famously, the MT 215 trial report.

 

Long story short - if you are using gliding CLmax for your 109 or P-40 turning calculations, you are doing it wrong.

 

Well I'm sorry to hear you think I'm doing it wrong but there are different ways of doing turn analysis and the one I'm using uses the Clmax at the glide as a base. For me the convention has always been to use the power off Clmax as a base so I would love to hear how you do it since you say I'm doing it wrong. What is your method Kurfurst?

 

That being said I too get a much higher Clmax with power on just like M. B. Morgan and in my C++ simulation I get a power off Clmax of 1.36 and a full power on Clmax of around 1.8 much like M. B. Morgan does.

 

As to the "anomalies" of my P-40 calculation you talk about what are those? My estimate (24.5 s) is currently within 1% of the current IL-2 estimate of 24.3 s so we both have it wrong then? Can you please outline where the Il-2 developers and I got this wrong and what the "anomalies" are?

 

And what are the "pitfalls" of RAE TN 1106? Did we all miss something there?

[unreasonable]

@Holtzauge - should leave you to reply to posts directed to you, apologies. Thought you had gone off to do something else. 

[Holtzauge]

But the graph and the text both say that the pressure at the static port falls with yaw, not increases. Text in section (d) just under Fig. 6

 

Hence the line PS in Fig. 5 goes above 1.0 in the 0-20 degree zone.

 

So you might be right that that is how a particular installation could behave, but it is not what the graph shows.

Yup I see what you mean and I think I misinterpreted figure 5: The text under figure 6 is quite clear on the point so as you say so there seems to be more to this after all......