PDA

View Full Version : Flight Testing and Aircraft Performance in World War II



Kettenhunde
12-11-2009, 10:45 AM
Most people here seem to value performance comparisons quite a bit.

I will post an article from the AIAA library that discusses Flight Testing technology used during World War II. The contents can be discussed and we can learn the limitations of the data that could be gathered.

I think it would add tremendously to the quality of the aircraft performance discussions taking place in these forums.

If people want to do this let me know. Moderators please watch this thread closely as it will remain a mature discussion or it will not occur.

I will post the article tonight if enough interest is out there.

Kettenhunde
12-11-2009, 10:45 AM
Most people here seem to value performance comparisons quite a bit.

I will post an article from the AIAA library that discusses Flight Testing technology used during World War II. The contents can be discussed and we can learn the limitations of the data that could be gathered.

I think it would add tremendously to the quality of the aircraft performance discussions taking place in these forums.

If people want to do this let me know. Moderators please watch this thread closely as it will remain a mature discussion or it will not occur.

I will post the article tonight if enough interest is out there.

Viper2005_
12-11-2009, 10:47 AM
Sounds interesting.

AndyJWest
12-11-2009, 10:54 AM
http://forums.ubi.com/images/smilies/25.gif

Sounds like a darned good idea. I get the impression that some people have a totally unrealistic concept of just how difficult this sort of thing was, even in peacetime.

As for keeping the discussion mature, that will mostly depend on not going off-topic into endless arguments about particular aircraft, I'd think. If we stick to general principles, we might learn a lot.

jamesblonde1979
12-11-2009, 12:15 PM
I would be very interested in seeing this.

na85
12-11-2009, 12:42 PM
I like this idea.

Kurfurst__
12-11-2009, 01:07 PM
Great idea.. many take a single spec as an absolute, but it depends on so many things... atmospheric conditions, the condition of the airframe, and individual quirks of the aircraft. The measurement methods, as well as the corrections applied to the raw data obtained were often different, and in many cases a test could be only about finding relative differences between different airframe conditions (say with or without slipper tanks), rather than seeking to give absolute values.

Manufacturers did not give +/- 3% tolerance on speed for no reason! Discussions keep revolving about wheter a plane could go say 640 or 650, but in reality acceptance tolerances were such that anything between 630 and 680 would be accepted into service.. making all the heated debates rather moot. Aircraft performing to the 'official' or nominal specs were the exception, not the rule.

stalkervision
12-11-2009, 02:22 PM
All for it. http://forums.ubi.com/groupee_common/emoticons/icon_cool.gif

Kettenhunde
12-11-2009, 02:43 PM
How about we make some rules...

I propose....

1. No discussion of specific aircraft is allowed in the thread. If you want to discuss a specific airplanes performance, start a new thread.

Reason for that is the report contains quite a bit about specific aircraft and uses their testing as examples. There is a lot of room for it to become a "see my airplane could do this" when the author makes no factual statement about any specific test. The author does not verify or examine any reports as that is beyond the scope of his subject. He uses the results for examples only.

The topic of the article as well as the thread is the instrumentation and techniques of gathering flight testing data in the Allied Flight testing programs.

We can limit ourselves to the discussion of the techniques and instruments used to gather that data.

I understand that many will not have much experience in some of this stuff too and will lurk. Your participation is welcome. I encourage everyone to participate, comment, and learn.

This particular article only covers the Allied nations. I do have some information on German flight testing equipment and techniques too.

The strengths and weaknesses of each can be discussed. It might bring out some interesting details and improve the discussions when examining a document or source.

AndyJWest
12-11-2009, 03:11 PM
Sounds good to me - just make it clear this topic is about <span class="ev_code_RED">how testing is done</span>, rather than the particular results themselves.

Maybe someone out there might have some further info too (on the subject, <span class="ev_code_RED">not</span> the results)

DrHerb
12-11-2009, 03:28 PM
I really want to see this

RegRag1977
12-11-2009, 03:35 PM
And me too!

Kettenhunde
12-11-2009, 03:54 PM
http://img264.imageshack.us/img264/9657/flighttesting1.jpg (http://img264.imageshack.us/i/flighttesting1.jpg/)

http://img264.imageshack.us/img264/5103/flighttesting2.jpg

http://img19.imageshack.us/img19/8117/flighttesting3.jpg

http://img19.imageshack.us/img19/7370/flighttesting4.jpg

Daiichidoku
12-11-2009, 04:13 PM
http://forums.ubi.com/images/smilies/blink.gif

so K, when are you going to post the informative stuff?

(although i really liked that part when he said that the 51's laminar flow wings were not affected by comrpessibility http://forums.ubi.com/images/smilies/88.gif)

(oh, and that the 38 DRF delivery was in "probably a Lockheed Hudson"...super research skillz http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif)


sorry, but this "article" seems pedestrian....this was published? please tell where sir

Daiichidoku
12-11-2009, 04:24 PM
aviation advanced techmology applications?

when the address is googlemapped, it comes up as a bungalow-looking building in the middle of what appears to be a residential neighbourhood

high school paper maybe?

Kettenhunde
12-11-2009, 04:25 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> so K, when are you going to post the informative stuff? </div></BLOCKQUOTE>

That is what I was afraid of and why I have not posted such reports before....

The really important stuff is in the report. Problem is that you have understand what your looking at to realize what is important.

For example, the airspeed calibration system of known points on the ground is used today. It is used for General Aviation aircraft and extensively in amateur built aircraft.

Why? It has some severe limitations. Even in the modern days of GPS, it can deliver results which are as much as 40% off with some common errors.

The largest one being the expression of compressibility which viper mentioned several days ago. We recognize today that it is only accurate for a narrow speed range.

http://img695.imageshack.us/img695/5901/calibrationmethods.jpg (http://img695.imageshack.us/i/calibrationmethods.jpg/)
By crumpp (http://profile.imageshack.us/user/crumpp)

AndyJWest
12-11-2009, 04:26 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">
The NACA Langley engineers noted that most of the fighters had an unbalenced static and total pressure systems which, therefore, showed an indicated airspeed which was too high (the volume of the static lines was much greater than that of the total pressure lines)
</div></BLOCKQUOTE>

Anyone have any idea how much such faults actually affect results? Are we talking a few mph, or something more significant?

There doesn't sem to be anything startlingly new here, it just reinforces my impression that expecting exact figures from tests done during the period is optimistic.

Is there much original test results material about in the form it was originally recorded? It would be nice to see cockpit instrumentation photo's for example.

Xiolablu3
12-11-2009, 04:35 PM
Good read, thanks for posting.

However one thing about the piece troubles me.

The writer talks about how the Spitfire IX handling suffered with the gain in 200hp from the mkV to the MkIX, as no more rudder area was added. But didnt most aircraft of WW2, such as the Me109, Fw190, P47, P40 etc have an increase of 200hp or more over their life and have no rudder increase?

The MkIX did actually have a rudder change, but it was a bit later.

Kettenhunde
12-11-2009, 04:40 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Anyone have any idea how much such faults actually affect results? </div></BLOCKQUOTE>


Yes they do effect results. Simply put, this is the data the engineers crunch the numbers on.

There was no universal definition of compressibility effects at this time or even today nor standards for application of instrument errors.

The compressibility is another subject for later so lets just deal with determining PEC. It is a universal environmental effect but our ability to express it is not if that makes any sense to you.

For example, trailing cone, pressure wakes, all assume a zero PEC and tower is just for calibrating the static source. It too requires a zero PEC. Tower requires some very accurate altimetry both on board the aircraft and the tower station.

The most accurate method for pressure used today is using a differential pressure gauge at the trailing static source and aircraft static source. They did not do that back then....

Now, the system they had worked very well for comparative design work. That is if they want to compare the effect of adding an external bomb and fairing to an existing design, then all the tools are present to make an accurate prediction of the results for that design. It has some severe shortcomings for comparing one design to another design especially if the data is gathered under different conditions.

Kettenhunde
12-11-2009, 04:44 PM
No specific aircraft will be discussed. Start a new thread if you wish to on that airplane.

Moderators please enforce this...

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> 1. No discussion of specific aircraft is allowed in the thread. If you want to discuss a specific airplanes performance, start a new thread.

Reason for that is the report contains quite a bit about specific aircraft and uses their testing as examples. There is a lot of room for it to become a "see my airplane could do this" when the author makes no factual statement about any specific test. The author does not verify or examine any reports as that is beyond the scope of his subject. He uses the results for examples only. </div></BLOCKQUOTE>


<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Xio says:
However one thing about the piece troubles me......Spitfire IX handling suffered with the gain in 200hp </div></BLOCKQUOTE>


<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> 51's laminar flow wings </div></BLOCKQUOTE>

Kettenhunde
12-11-2009, 04:55 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">high school paper maybe? </div></BLOCKQUOTE>

Not that easily dismissed I am afraid....


http://img52.imageshack.us/img52/4166/titlejy.jpg

AndyJWest
12-11-2009, 04:56 PM
Um, Kettenhunde, can you clarify what PEC is? Not all of us are familiar with the technical lingo involved...

Kettenhunde
12-11-2009, 05:05 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">m, Kettenhunde, can you clarify what PEC is? Not all of us are familiar with the technical lingo involved... </div></BLOCKQUOTE>

It is the position error correction of the pitot static system.

It is an error found in both the airspeed indicator and the altimeter based on the installation of the static and dynamic pressure measuring points.

The pitot static system just measures pressure. Your airspeed is the difference between the dynamic and the static pressure as the aircraft moves through the air.

If either static or dynamic pressure is not reading accurately, then your ability to measure will not be correct.

AndyJWest
12-11-2009, 05:11 PM
http://forums.ubi.com/images/smilies/25.gif

Thanks for the clarification.

Kettenhunde
12-11-2009, 05:16 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Manufacturers did not give +/- 3% tolerance on speed for no reason! Discussions keep revolving about wheter a plane could go say 640 or 650, but in reality acceptance tolerances were such that anything between 630 and 680 would be accepted into service.. making all the heated debates rather moot. Aircraft performing to the 'official' or nominal specs were the exception, not the rule.
</div></BLOCKQUOTE>

That is a good observation and very true. In fact, many manufacturers assume a fresh engine or one that is developing rated power.

As you all know, the moment we start an engine, it begins to lose power over its life cycle.

Kettenhunde
12-11-2009, 05:32 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Thanks for the clarification.
</div></BLOCKQUOTE>

You are welcome.


Does everybody understand the "tools of the trade" portion of the article, limitations of those tools, and what that means to WWII aircraft performance?

That is what this thread is about....


Is it likely we are going to get an accurate comparison for different flight test's on different designs conducted by different countries on different days by different design teams?

AndyJWest
12-11-2009, 07:04 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Is it likely we are going to get an accurate comparison for different flight test's on different designs conducted by different countries on different days by different design teams? </div></BLOCKQUOTE>

In short, no.

This is the problem that any historian faces. And like it or not, this is what we are doing here. Studying history. It is easy to take false comfort from a number written down on a piece of paper, and assume it is 'correct' when even the person writing it might well have doubts about it's validity. Taken out of context, 60-odd years later, a single number is hardly proof of anything. One can probably safely make generalisations, and assume that if (Hypothetically) Supermarine wrote that a 100-gal drop tank reduced maximum speed at 20,000 ft in a Mk IX by 35 mph TAS, the figure is about right, but if (hypothetically) Messerschmitt reckon the 109G is faster than a Spitfire mk V by 26.5 km/h at 5000m, they will probably be wrong. The data we have now can be no more reliable than that which was available at the time. Furthermore, to state the obvious, there was a war on. Absolute values were less important than measurable improvements, and even the measurement wil sometimes be just plain wrong. Sometimes decisions have to be made on less-than-perfect data. And no doubt on occasion, the figures will have been 'adjusted' to justify such decisions, on both Axis and Allied aircraft.

na85
12-11-2009, 07:16 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:

One can probably safely make generalisations, and assume that if (Hypothetically) Supermarine wrote that a 100-gal drop tank reduced maximum speed at 20,000 ft in a Mk IX by 35 mph TAS, the figure is about right, but if (hypothetically) Messerschmitt reckon the 109G is faster than a Spitfire mk V by 26.5 km/h at 5000m, they will probably be wrong. </div></BLOCKQUOTE>

Not sure I agree with that, unless you're suggesting Messerschmitt's evaluation was from combat reports.

But if Messerschmitt had a captured Spitfire MkV and in their flight testing they discovered that the 109G-6 was slower by 26 km/h then they would be correct. That particular SpitV is faster than that particular 109G-6 at that particular altitude, under those specific atmospheric conditions, by 26.5 km/h.

But the point Kettenhunde is trying to make (I think) is that if supermarine conducted the same test, with a spit V and 109G-6, there is no guarantee the results from the British test will match those of the German test. In fact it's more likely that they will not match simply because all the conditions of the tests are not the same.

Kettenhunde
12-11-2009, 07:26 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Absolute values were less important than measurable improvements, and even the measurement wil sometimes be just plain wrong. </div></BLOCKQUOTE>


Great points! You hit on the main reason firms conduct flight testing with this one too. It is a very important to the design team and as the article says, "the proof of the pudding".

The main purpose of flight testing is to measure improvements within the design. It is to take design A, establish a ruler and use it to measure the progress of improvements to design A.

It is not to compare Design A to Design B. In fact, their rulers may be very inaccurate when compared on paper.

Notice the few comparative flight test's that are out there and the formats used in that testing.

Do we commonly find charts graphing speed and climb comparison or are the tests more subjective in nature over defined parameters of flight?

Kettenhunde
12-11-2009, 07:32 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">But the point Kettenhunde is trying to make (I think) is that if supermarine conducted the same test, with a spit V and 109G-6, there is no guarantee the results from the British test will match those of the German test. In fact it's more likely that they will not match simply because all the conditions of the tests are not the same. </div></BLOCKQUOTE>

You have my point down and great summary of it so far, na85.

Andywest can probably clarify but I think you might misunderstanding each other.

I read it that Supermarine has a good idea what the improvements are with their design but Mtt using that same graph would not return an accurate measurement of the relative performance of the two airplanes.

M_Gunz
12-11-2009, 07:51 PM
There was one plane A to plane B comparison I believe you mentioned before, please correct or fill in any details
I'm going to screw up here: Hellcat and Corsair flying side by side and one was reading 20mph higher IAS.

Yah, I can say PEC too. But if you think that will stop the favorite charts boys then you've beat your head
against that wall too long, the only "FACTS!" they care about are the ones they choose based on how good the are.

AndyJWest
12-11-2009, 08:04 PM
I think that na85 and I aren't really in disagreement at all. Messerschmitt could probably get at least a ball-park figure for speed differences between two aircraft in the same conditions, if they were both available. Not to within 0.5 km/h I'm sure, but that would hardly matter. What is important is that they will be measuring TWO AIRCRAFT, not any Spit V* vs any Bf-109G6 or whatever. Even new, I'm sure that production tolerances etc will affect performance, and throw in a few hours use under combat conditions, performance from one aircraft to the next one off the production line might vary markedly.

(* THe Spit V is probably the worst example of an aircraft 'type' for this discussion anyway. Some were earlier marks rebuilt or reengined, some were new. They had many different engine sub-types fitted. Probably the only thing they had in common was that they were more 'Mk V' than anything else...)

Kettenhunde
12-11-2009, 08:04 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">There was one plane A to plane B comparison I believe you mentioned before, please correct or fill in any details </div></BLOCKQUOTE>

That is the one and a great example.

Don't get me wrong, Charts have their place and all is not doom/gloom on the subject of comparing performance. I plan on getting into that but want to make sure everyone understands some of the important technical details and the limits of our data.

It is important to know what is comparable and introduces the least amount of error.

Kettenhunde
12-11-2009, 08:15 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Even new, I'm sure that production tolerances etc will affect performance, and throw in a few hours use under combat conditions, performance from one aircraft to the next one off the production line might vary markedly. </div></BLOCKQUOTE>

Aircraft performance does vary considerably. Production tolerances are usually defined by contract as a percentage range over a mean average.

Each performance parameter will have its own range, too.

Throw in maintenance/wear, instrument errors, piloting skill, and our range of error from that paper flight tested performance just got MUCH larger.

That is one reason why airplanes that are within 10% or more in some cases on paper are considered to have the same performance tactically.

There just isn't any real difference in the air.

na85
12-11-2009, 09:49 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:

That is one reason why airplanes that are within 10% or more in some cases on paper are considered to have the same performance tactically.

There just isn't any real difference in the air. </div></BLOCKQUOTE>

Very true. In Shaw's excellent book that's the definition of "similar aircraft" - aircraft whose performances are within 10% of each other.

M_Gunz
12-12-2009, 03:03 AM
The big difference in contemporary WWII fighters is who has more than 10% advantage in energy since that does
give a like advantage in maneuver while it lasts. Your energy state may be worth much more than your charts.

Kettenhunde
12-12-2009, 06:37 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Your energy state may be worth much more than your charts.
</div></BLOCKQUOTE>

Definately, this is why real Fighter pilot are concerned with positional tactics more so than aircraft performance.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Very true. In Shaw's excellent book that's the definition of "similar aircraft" - aircraft whose performances are within 10% of each other. </div></BLOCKQUOTE>

Exactly! This is why the standards are set so high today to be a test pilot. Firms are looking for pilots who can eliminate as much of the pilot error as possible to get things closer to just the manufacturing, environment, and equipment errors.

Like Andywest points out...

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> This is the problem that any historian faces. And like it or not, this is what we are doing here. Studying history. It is easy to take false comfort from a number written down on a piece of paper, and assume it is 'correct' when even the person writing it might well have doubts about it's validity. </div></BLOCKQUOTE>

What is the best way to go back 60 years and try to determine the performance line up?

Calculations, flight testing, anecdotes,.....??

Kettenhunde
12-12-2009, 10:29 AM
Here are some stated goals of test flying...

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> The objects of test flying were: —
(1) To ensure that the aircraft's flying characteristics met the requirements of the operational pilot as far as safety and being
pleasant to fly were concerned.
(2) To ensure that the performance was up to the standard required by the design specification as regards speed, ceiling,
fuel consumption, etc.
(3) To ensure that the aircraft was a good engineering product and that maintenance was simple and straightforward. </div></BLOCKQUOTE>

http://www.flightglobal.com/pd...1955%20-%201449.html (http://www.flightglobal.com/pdfarchive/view/1955/1955%20-%201449.html)

What is the best way to go back 60 years and try to determine the performance line up?

Calculations, flight testing, anecdotes,.....??

psykopatsak
12-12-2009, 12:11 PM
just a mention, "today" in the article was 21-22 years ago.

Kettenhunde
12-12-2009, 12:17 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">just a mention, "today" in the article was 21-22 years ago. </div></BLOCKQUOTE>


The instruments and methods it discusses are over 60 years old.....

na85
12-12-2009, 01:29 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:

What is the best way to go back 60 years and try to determine the performance line up?

Calculations, flight testing, anecdotes,.....?? </div></BLOCKQUOTE>

I personally don't put much weight in pilot anecdotes from WW2. A lot of people here treat them as the word of god.

I would wager that a combination of theoretical calculations and subsequent validation by flight testing would be the best way.

Kettenhunde
12-12-2009, 02:07 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> I personally don't put much weight in pilot anecdotes from WW2. </div></BLOCKQUOTE>

I agree anecdotes are very rocky territory especially for absolute performance declarations. Just because a story is told of an airplane out performing another has little to contribute.

Anecdotes can work for stability and control and other technical observations for operating the aircraft, IMHO. One thing that has to be considered is the level of training of the pilot. What is easy and comfortable to a familiar pilot is not so easy to one not so familiar.

A good technical background is a must to sort the wheat from the chaff regarding anecdotes.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> I would wager that a combination of theoretical calculations and subsequent validation by flight testing would be the best way. </div></BLOCKQUOTE>

I agree that flight testing and calculations have the most useful information to tell us.

Which is the most accurate in your opinion and why?

na85
12-12-2009, 02:15 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:

I agree that flight testing and calculations have the most useful information to tell us.

Which is the most accurate in your opinion and why? </div></BLOCKQUOTE>

Accurate in terms of what?

If you have six specific aircraft and want to rank those 6 specific craft in terms of performance I would suggest simply taking each one for a test flight.

For an entire class of aircraft, or to make general statements like "Bf109Fs can get to X altitude faster than Spitfire MkVs" then... I don't really know.

You can infer a good amount from your test aircraft, and pit them against each other in a race to 30000 feet, but I'd imagine it would be necessary to do some statistics work to extend that data to all aircraft of the same type.

Kettenhunde
12-12-2009, 04:02 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Accurate in terms of what? </div></BLOCKQUOTE>


If we have a calculation and a flight test, which has the least amount of error?

I am not asking anyone to make an absolute statement. Just in terms of margin of error which do you think generally has less error.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> If you have six specific aircraft and want to rank those 6 specific craft in terms of performance I would suggest simply taking each one for a test flight. </div></BLOCKQUOTE>

Yes! I agree side by side flight testing is the most accurate method.

However to clarify my question, we only have to choose between a calculation and a flight test.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> For an entire class of aircraft, or to make general statements like "Bf109Fs can get to X altitude faster than Spitfire MkVs" then... I don't really know. </div></BLOCKQUOTE>

That is extremely difficult to do in reality and just not very accurate.

psykopatsak
12-12-2009, 04:37 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">quote:
just a mention, "today" in the article was 21-22 years ago.



The instruments and methods it discusses are over 60 years old..... </div></BLOCKQUOTE>

my point is that the argument "is still being used today" might not be fuly valid. or it might, i dont know about that, however i do know to take sources with an ounce of salt.

Kettenhunde
12-12-2009, 04:39 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">my point is that the argument "is still being used today" might not be fuly valid. or it might, i dont know about that, however i do know to take sources with an ounce of salt. </div></BLOCKQUOTE>

What argument? You mean a reciprocal ground course for determining airspeed indicator error?

It is still used today but for specific applications. As I said, mostly General Aviation and Amateur built aircraft.

In fact I did it for a friend about a month ago to get his home built Vari-eze a PEC.

You can use it for just about any aircraft but it loses accuracy if outside of the narrow parameters listed. You are not going to get as accurate a return from a 350 mph airplane as you will a 100mph airplane with it.

psykopatsak
12-12-2009, 05:15 PM
no need to defend your argument here mate, im just noting that its an old text, so that some facts might be out of date.

Kettenhunde
12-12-2009, 05:34 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">no need to defend your argument here mate, </div></BLOCKQUOTE>

I think something is getting lost in the translation.

There is no argument at all and I appreciate your question as well as participation.

I just was letting you know that reciprocal course is used today as the author points out. That is all.

http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

It is just not the latest technique or even considered a very accurate one outside of its narrow application.

M_Gunz
12-12-2009, 05:38 PM
You only get one test flight? Or a full series to average out?

Here's a question back: how well does IL2 compare as calculations method?
The aerobatics guys felt it's got 2nd order motions which no others they knew had.
They also said it's not perfect.

AndyJWest
12-12-2009, 05:42 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">...2nd order motions... </div></BLOCKQUOTE>

Um, more tech-speak there. M_Gunz, could you oblige with a definition?

Kettenhunde
12-12-2009, 05:46 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">You only get one test flight? Or a full series to average out? </div></BLOCKQUOTE>

It does not matter although the more test flights you do, the lower the margin of error for test flying.

M_Gunz
12-12-2009, 07:13 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">...2nd order motions... </div></BLOCKQUOTE>

Um, more tech-speak there. M_Gunz, could you oblige with a definition? </div></BLOCKQUOTE>

Well first of all, those guys used that term and mentioned some examples -- IIRC last was after 4.08
in Oleg's Ready Room so IIRC only not completely sure I could fill in all of it but, slip-roll couple,
gyro and propwash effects? The problem being is that I thought at least one other sim does those
though maybe in a way that disagrees with the aerobatics fliers like EcoDragon and Iceman, like as
in "oh look, they got this right for a change!".

I've seen computer flight sims since 81-82 on a terminal in 2 tones of plasma orange on black clear
to youtube shows of sims I don't have enough computer to run. The fidelity of the sim to the reality
is not just chart numbers but how it handles. Flight sims I experienced flew pretty much "on-rails"
(you were easily able to fly right up to and hit the speed in the tables) until I got Red Baron 2.
RB2 was not on rails nor was it cracked. The cracks are the places where the model breaks down or
up or whatever. And then there are the dreaded canned effects and possibly semi-canned effects like
spins in EAW.
Departures are where I like to explore in a new sim. How much my control inputs do, what if anything
and what are the conditions that start and end the departure? Some early sims pretty much just
ignored departure from flight altogether. Now with IL2 I usually have some effects as told by the
aerobatics as well as aero guys in response to my control inputs. Is it all done in the same code as
ordinary flight or is there some form of "canning" as I've read claimed on this forum? I don't know.
How do you run flight and not flight with the same code anyway? I would think that some branching to
include completely different algorithms would be necessary no matter how you slice it.

IL2 was a real eye-opener even in just the demo. The IL2 demo was way better than what was sold for
$40 or $50 even with so few planes. IL2 set new records, more than one, in combat flight sims. Since
then -some- have been beat while at the same time IL2 was being constantly upgraded and expanded and
set new high marks here and there. They just aren't your color-by-numbers-coders that turn things
like Microsoft Help out.

M_Gunz
12-12-2009, 07:34 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">You only get one test flight? Or a full series to average out? </div></BLOCKQUOTE>

It does not matter although the more test flights you do, the lower the margin of error for test flying. </div></BLOCKQUOTE>

I guess that answers the test flight vs calculation as an approach question then, more test flights accumulate
accuracy while calculations only improve to a certain point by including more or less what is known *caveat*
not including the test flight data, it's one -or- the other! But maybe 1,000,000 test flights per test would
be taking it too severe, is there some number of tests under 100 flights per test that would beat sliderules
and a chalkboard or a good PC and a compiler? I would guess so, the equations usually don't include a lot of
minute differences the real plane has. You could take the real plane up and tune the sim to that and then take
a newer serial production of the same plane and find a very different plane like the mid-42 Spit VB as opposed
to those produced in 1941... big difference. The one we have labeled as 41 is really the 42, check the engine
listed in the Airplane Guide pdf.

Kettenhunde
12-12-2009, 08:55 PM
There has to be some more opinions or folks to offer their answer to the question!

Come on, it is a discussion!

Good Job, M_Gunz.

AndyJWest
12-12-2009, 08:56 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">...propwash effects.. </div></BLOCKQUOTE>

There is definite evidence for these, or someone has tweaked the FM to mimic them for a particular case. Try starting just the front engine of a Do-335 while on the ground. It turns easily with rudder while taxiing at low speeds. Then try the rear engine only (prop behind the rudder) - much harder to turn.

I've also seen effects from damage being quite convincing - losing the fin & rudder on a Me 109 (collision with target - still do it occasionally http://forums.ubi.com/images/smilies/blush.gif ) results in 'dutch roll' or something similar - getting worse as speed decreases.

The effects of slip-roll coupling, or something closely related, became apparent with my autopilot project - I had real problems with multi-engined aircraft (high polar inertia in roll) at low speeds.

Gyroscopic effects are immediately obvious on the takeoff run - as the nose lifts, the plane yaws, which can then force the nose down further - gyroscopic precession.

The extent to which all these effects are due to a realistic basic physical model, as opposed to special code for particular situations, is of course unknown to me. Maybe one day the source code will be released and we will find out, but I'd like to think that most of this is inbuilt - coding for exceptions is never as satisfactory as getting the basic algorithm right.

What I can say is that in developing my autopilot project, I assumed the physics was 'real', and rather than trying to control 'the sim', I treated the data in the same way I would expect to do in a real autopilot. O.k, I know very little about how they actually work (should probably have done some research first..), but nothing I did seemed counter-intuitive to me.

This project also showed me some of the difficulties of measuring performance, even under the controlled conditions of a sim. Supposing you want to measure maximum sustained speed in level flight. You need to consider weight (more so at high altitude) prop pitch, mixture, radiator position... Results vary from one map to another, probably due to pressure differences, but temperature may also be a factor. I generally got results more or less as IL-2 Compare suggests, but there are just to many variables to state anything definitive about performance without more work than would seem sensible in a 'real world' situation in the middle of a war. In most cases, I suspect that testing was only done to the minimum extent that was necessary, rather than an exhaustive analysis of the limits. Any data we have now should be looked at in that light.

na85
12-12-2009, 09:03 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">...2nd order motions... </div></BLOCKQUOTE>

Um, more tech-speak there. M_Gunz, could you oblige with a definition? </div></BLOCKQUOTE>

Pretty sure "2nd order motions" refers to the coupled behavior exhibited by control surfaces, such as induced yaw/roll etc. The equations governing the resulting changes in roll/pitch/yaw moments are 2nd order differential equations, and are computationally very expensive. Most consumer-level sims don't include this, or use linear approximations at best.

JuHa-
12-13-2009, 03:25 AM
Interesting. They do mention that a captured plane with damaged propellor was fitted with Hamilton standard one. (Or maybe just the blades were replaced) This would be pretty serious modification, which should be mentioned in the results. But given the difficulties of obtaining OEM spare parts...

This gives a bit more credit to the old tale of an FW190 that had been fixed with a propellor from Stuka for test flights.

jamesblonde1979
12-13-2009, 03:46 AM
I found the bit about the Germans putting a DB engine in a Spitfire for testing and the US fitting a Hamilton Standard prop on the Zero quite interesting. It speaks volumes about just how little each side really knew about the other.

I wonder if there are any pics or stories relating to that DB Spitfire?

Kettenhunde
12-13-2009, 04:23 AM
Start another thread if you want to discuss specific aircraft.

Between a flight test and calculation, which is the most accurate in your opinion and why?

Kettenhunde
12-13-2009, 08:40 AM
I think it would be easier for people to see things if they understood the margins of errors a little better.

We know about flight testing and can give it it about a 10% margin.

What about the margin of error for calculations?

Are they greater than or less than 10%?

What are our major sources of error in the math?


Let's assume our math formula is correctly applied and we have no a process errors.

One of the major sources is the expression of forces. Our coefficient of drag for example is nothing more than the ratio of drag pressure to dynamic pressures. We know the weight, wing design, and reference area, our drag to due to lift is pretty accurate on paper. There is some room for error but the power of induced drag is first power and makes up ~15% or less of our total drag at Vmax. Also by this time in history, wing design was pretty well advanced and it would be very surprising to find a firm that could not at a minimum hit the average numbers.

Not much influence on our margin of error outside of ability to express compressible dynamic pressure.

The other major component of drag, parasitic does have a large influence on Vmax forces. The power of parasitic drag is cubed with velocity.

However, it is easier to express than the drag due to lift! The reference areas are not unknown. We can even force gauge models in the wind tunnel with scaling effects and almost eliminate all the unknowns except one major one, our ability to accurately express compressible dynamic pressure.

Now of course we can and do approximate the effects. It works very well this way. There is not any rules though on exactly when to apply corrections from system to system. It is a universal environmental effect so whichever system is in use, application is universally applied to all things in that system.

What does that mean....

The basic source of error for both flight testing and calculations stem from the same place, compressibility.

AndyJWest
12-13-2009, 09:02 AM
I'm not sure it is particularly helpful to try to determine whether calculation or flight testing gives more accurate results. Presumably calculation is done at the design stage, and then flight testing is done to confirm the calculated performance values. At least, in an ideal world, that is how it should be done. In a war-time situation, I'd expect a more pragmatic approach on the lines of 'we've got a new engine upgrade that gives us 200 hp more at 20,000ft, what effect is that going to have on max speed?' Out with the slide-rule and an answer is obtained, 'humm, 30 knots or thereabouts, not bad...'. The new engine is installed, the test pilot gives it a few runs, confirms the plane is faster, though given circumstances and individual variations, an exact figure is neither obtainable nor particularly useful.

There is of course technology available now that could obtain better theoretical performance figures that could have been done at the time, but the value of this depends entirely on the accuracy of the input data. Even with a complete set of blueprints for airframe and engine, one would still be looking at an 'idealised' aircraft rather than a typical one just off the production line, with all the variations in tolerences etc that this implies.

Kettenhunde
12-13-2009, 09:37 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> I'm not sure it is particularly helpful to try to determine whether calculation or flight testing gives more accurate results. </div></BLOCKQUOTE>

Yes and no. One has to realize the limitations and what each approach can tell us.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Out with the slide-rule and an answer is obtained, 'humm, 30 knots or thereabouts, not bad...'. The new engine is installed, the test pilot gives it a few runs, confirms the plane is faster, though given circumstances and individual variations, an exact figure is neither obtainable nor particularly useful.
</div></BLOCKQUOTE>

Very astute observations.....We call this performance trends and your correct, all of it whether flight or calculated is only accurate "within the realm of significant digits".

Neither approach can determine precision results, both make predictions on trends.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Presumably calculation is done at the design stage, and then flight testing is done to confirm the calculated performance values. </div></BLOCKQUOTE>

You bring up a good point in this part.

Yes and no, calculations are done throughout the entire process.

In fact, the data from a flight test are calculated. They are just calculated from the measured points. Those measured points are gathered as discussed in the article. One major source of error is the same one for both calculations and flight testing.

What happens if our flight test does not meet the expected values determined by calculations?

Do you think the design team goes:

1. We were wrong at the drawing board and chucks the calculations.

2. We were wrong at the flight test line or when we built it, in piloting it, or accounting for the environment, and chucks the flight test results.

3. We checked the calculations and they are correct, fix the prototype and lets find out what went wrong in the flight test.

4. We checked the calculations and found an error, let's recalculate and see where we are at...

You could say that while calculations and flight testing share common source of error, flight testing has significant other sources to contend with as well.

JtD
12-13-2009, 09:41 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I'm not sure it is particularly helpful to try to determine whether calculation or flight testing gives more accurate results. Presumably calculation is done at the design stage, and then flight testing is done to confirm the calculated performance values. At least, in an ideal world, that is how it should be done. </div></BLOCKQUOTE>

Todays practice is to validate calculations by tests.
You first do your calculations, then you build, then you test, and if you find a good agreement between test and calculation, you'll use your calculation for further studies.
For instance, despite all the calculations the engineers do around car crashes, the safety of the vehicle is eventually rated by a crash test. Now if that looked very much like the calculated results (with Western cars it usually does), it's much easier (and cheaper) to analyze modifications and improvements through the now validated calculations then through repeated tests.

Kettenhunde
12-13-2009, 09:54 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Todays practice is to validate calculations by tests. </div></BLOCKQUOTE>


It was the same then too.

psykopatsak
12-13-2009, 10:02 AM
only that today you can calculate every nanometer of the aircraft, then you couldn't.

Kettenhunde
12-13-2009, 10:12 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">only that today you can calculate every nanometer of the aircraft, then you couldn't. </div></BLOCKQUOTE>

What do you mean by "calculate every nanometer"? Are you trying to say they were not capable of determining wetted area?

We are not anymore accurate today than they were especially when it comes to high power piston engine designs. The theory is the same and in fact, many of the "advances" we use today are the result of their work. There just has not been any real advancement in subsonic piston aircraft design since World War II.

We can do things much faster with fewer resources and know more about the transonic as well as supersonic realms of flight. There was not a whole lot of transonic flight going on during WWII. Encountering it stumped a few folks back then which goes right back the point I made about being able to describe compressibility.

In fact, we were not able to solve some of the equations in those regions until advancements in computer technology made it possible in the 1970's.

R_Target
12-13-2009, 10:33 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by JuHa-:
Interesting. They do mention that a captured plane with damaged propellor was fitted with Hamilton standard one. (Or maybe just the blades were replaced) This would be pretty serious modification, which should be mentioned in the results. </div></BLOCKQUOTE>

Not at all, as the Sumitomo on the A6M was a license-built copy of a Hamilton prop.


Sorry Crumpp! http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

Kettenhunde
12-13-2009, 10:51 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Sorry Crumpp! </div></BLOCKQUOTE>


I was dying to point that out and almost did....

Don't do it again!

http://forums.ubi.com/images/smilies/16x16_smiley-wink.gif

JuHa-
12-13-2009, 03:21 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">the Sumitomo on the A6M was a license-built copy of a Hamilton prop.
</div></BLOCKQUOTE>

Ah! So now I know more http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif

Waldo.Pepper
12-13-2009, 08:00 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I'm not sure it is particularly helpful to try to determine whether calculation or flight testing gives more accurate results. </div></BLOCKQUOTE>

Surely the answer is dependent upon the qualifications, temperament and level of professionalism of the people involved in the flight testing, and the determining/calculating the flight performance.

After all if the structural engineer makes an error in calculation, the bridge falls down.

ElAurens
12-13-2009, 09:44 PM
If I may be so bold to ask, where are you going with this discussion Crumpp?

How does it pertain to our little simulation?

I understand that if you test aircraft X and end up with performance number Y, that it is valid for the conditions at the time of that test, and if say, a month later you test the same aircraft again, with a different pilot, and in different conditions the numbers generated will in all likely-hood be different.

However, what we are doing here is not testing real aircraft, we are simulating them within the constraints of the available technology. So we need reliable, accurate numbers, in order to do that.

I'm not sure what you are up to here.

Yes, we all know the testing numbers from WW2, or any era fro that matter, have a margin of error. I'm totally in agreement with that.

But for the purpose of simulation you need some fixed set of performance figures to go from. You just can't throw up your hands and say it's all up to tactical advantage so the numbers don't matter. If that were the case you could just give all the planes in the sim the same flight model.

I don't think that would go over very well.

Kettenhunde
12-14-2009, 03:43 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Surely the answer is dependent upon the qualifications, temperament and level of professionalism of the people involved in the flight testing, and the determining/calculating the flight performance.

After all if the structural engineer makes an error in calculation, the bridge falls down. </div></BLOCKQUOTE>

Yes! And when a design does not conform to its calculated specifications, we have the design team equivalent of the bridge falling down.

Which brings us back to the question...

Between a flight test and calculation, which is the most accurate in your opinion and why?

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> ElAurens
</div></BLOCKQUOTE>

The first post answers your questions.

Please participate but stick to the topic of course.

Kettenhunde
12-14-2009, 04:47 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Anyone have any idea how much such faults actually affect results? Are we talking a few mph, or something more significant? </div></BLOCKQUOTE>

Did this question ever get answered as far as putting a number on it for you, AndyJWest?

I went back to review the topic before answering another participant and realized that numbers where never attached to the explanation when your question was addressed.

The error can be significant, as high as + or - ~40 mph in the air at altitude.

AndyJWest
12-14-2009, 08:57 AM
If ASI readings could possibly be out by as much as + or - 40mph, it makes me wonder what other methods might have been used to measure speeds at altitude, and whether they were any more accurate. The 'obvious' solution of timing over a known distance is problematic because determining when the 'start line' is crossed would need some fairly sophisticated equipment on the ground, duplicated at the finish. One would also have to consider wind, making runs both ways essential, though even then a crosswind will introduce an error.

Were there any other ways used to determine groundspeed during high-altitude flight? The GEE (http://en.wikipedia.org/wiki/GEE_(navigation)) navigation system developed by the British during WWII might have been useful, at least for bombers, I suppose but nothing else springs to mind immediately.

Kettenhunde
12-14-2009, 03:15 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">If ASI readings could possibly be out by as much as + or - 40mph, </div></BLOCKQUOTE>

That is not the Position error although that is a part of it.

That is the margin of error of the flight test results between what is recorded and the actual performance of the aircraft.

Kettenhunde
12-14-2009, 03:23 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">t makes me wonder what other methods might have been used to measure speeds at altitude, and whether they were any more accurate. </div></BLOCKQUOTE>


The primary method used by both the United States and the United Kingdom was ground course.

Now some of the other nations used other techniques which also have their own set of issues. Some are more accurate than others.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> The 'obvious' solution of timing over a known distance is problematic because determining when the 'start line' is crossed would need some fairly sophisticated equipment on the ground, duplicated at the finish. One would also have to consider wind, making runs both ways essential, though even then a crosswind will introduce an error. </div></BLOCKQUOTE>

Yes it is hard to time but it is usually done at sea level with ground observers marking the time.

The wind speed and direction is noted and the course is flown in four cardinal directions or a triangle. At least one is as close to into and with the wind as possible.

Times are recorded and wind direction/speed is accounted for....

That becomes your TAS once wind is factored and they we can use that to convert to EAS and finally by removing our compressibility correction, CAS. It is done as close to sea level as possible because TAS = EAS at sea level.

The difference between that CAS and your cockpit ASI forms the PEC curve.

That is the process in a nutshell.

psykopatsak
12-14-2009, 03:59 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">quote:
only that today you can calculate every nanometer of the aircraft, then you couldn't.



What do you mean by "calculate every nanometer"? Are you trying to say they were not capable of determining wetted area?

[...]

We can do things much faster with fewer resources and know more about the transonic as well as supersonic realms of flight. There was not a whole lot of transonic flight going on during WWII. Encountering it stumped a few folks back then which goes right back the point I made about being able to describe compressibility.

In fact, we were not able to solve some of the equations in those regions until advancements in computer technology made it possible in the 1970's. </div></BLOCKQUOTE>

what you said yourself http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif

Kettenhunde
12-14-2009, 08:08 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">what you said yourself </div></BLOCKQUOTE>

Did you understand what I wrote? They were more accurate back then on a prototype than we are today. You start into the supersonic and hypersonic realms, things get hard to predict.

The math is the same then as today but subsonic designs are much easier as well as more accurate to predict performance.

To be clear this...

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> In fact, we were not able to solve some of the equations in those regions until advancements in computer technology made it possible in the 1970's. </div></BLOCKQUOTE>

Is the equations of motions that describe this....

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> That becomes your TAS once wind is factored and they we can use that to convert to EAS and finally by removing our compressibility correction, CAS. </div></BLOCKQUOTE>


Our compressibility and therefore....

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> calculations ..........eliminate all the unknowns except one major one, our ability to accurately express compressible dynamic pressure.

Now of course we can and do approximate the effects. It works very well this way. There is not any rules though on exactly when to apply corrections from system to system. It is a universal environmental effect so whichever system is in use, application is universally applied to all things in that system.

What does that mean....

The basic source of error for both flight testing and calculations stem from the same place, compressibility.
</div></BLOCKQUOTE>

So which has less unknowns to deal with, Calculations or a Flight Test....

What I am getting at is today we have seen a reversal from WWII. It was easier to predict subsonic designs than today's super/hypersonic designs.

Back then, if a first run prototype was not within 7% of the calculations, it was generally not because of the calculations.

It was an issue with the prototype which would be fixed.

Once established, the calculations could be expected to be with 1% margin of error for what the design was capable of achieving.

Stability and control was much more of an issue for a WWII designer than any performance predictions.

All the best,

Crumpp

Waldo.Pepper
12-14-2009, 08:29 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Between a flight test and calculation, which is the most accurate in your opinion and why? </div></BLOCKQUOTE>

It depends upon the quality of the individual conducting the flight testing or the calculations and how high their standards are, and how rigidly they are adhered to.

Kettenhunde
12-14-2009, 08:32 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">It depends upon the quality of the individual conducting the flight testing or the calculations and how high their standards are, and how rigidly they are adhered to. </div></BLOCKQUOTE>


Good point. Yes it does and mistakes do happen.

Waldo.Pepper
12-14-2009, 09:41 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">It depends upon the quality of the individual conducting the flight testing or the calculations and how high their standards are, and how rigidly they are adhered to. </div></BLOCKQUOTE>


Good point. Yes it does and mistakes do happen. </div></BLOCKQUOTE>

And because of that the debate is unsolvable.

Kettenhunde
12-15-2009, 03:55 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">And because of that the debate is unsolvable. </div></BLOCKQUOTE>


I will agree that it is not solvable to a precision numbers like many seem to expect. Aircraft performance is about trends and the resources are not available to get things within 1%.

I disagree that the debate is unsolvable. In fact that is the point of the thread is to recognize when a solution exist's.

The performance trends are definitely solvable as are the correct behaviors for each design.

We have to assume no gross errors in the calculations due to incompetence just as we assume there are no gross errors due from incompetence in the flight testing.

There are many more opportunities for error in the flight testing portion.

We are just defining the limits of our accuracy from each approach at the moment, flight testing and calculations.

psykopatsak
12-15-2009, 05:27 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">what you said yourself </div></BLOCKQUOTE>

Did you understand what I wrote? They were more accurate back then on a prototype than we are today. You start into the supersonic and hypersonic realms, things get hard to predict.

The math is the same then as today but subsonic designs are much easier as well as more accurate to predict performance.

To be clear this...

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> In fact, we were not able to solve some of the equations in those regions until advancements in computer technology made it possible in the 1970's. </div></BLOCKQUOTE>

Is the equations of motions that describe this....

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> That becomes your TAS once wind is factored and they we can use that to convert to EAS and finally by removing our compressibility correction, CAS. </div></BLOCKQUOTE>


Our compressibility and therefore....

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> calculations ..........eliminate all the unknowns except one major one, our ability to accurately express compressible dynamic pressure.

Now of course we can and do approximate the effects. It works very well this way. There is not any rules though on exactly when to apply corrections from system to system. It is a universal environmental effect so whichever system is in use, application is universally applied to all things in that system.

What does that mean....

The basic source of error for both flight testing and calculations stem from the same place, compressibility.
</div></BLOCKQUOTE>

So which has less unknowns to deal with, Calculations or a Flight Test....

What I am getting at is today we have seen a reversal from WWII. It was easier to predict subsonic designs than today's super/hypersonic designs.

Back then, if a first run prototype was not within 7% of the calculations, it was generally not because of the calculations.

It was an issue with the prototype which would be fixed.

Once established, the calculations could be expected to be with 1% margin of error for what the design was capable of achieving.

Stability and control was much more of an issue for a WWII designer than any performance predictions.

All the best,

Crumpp </div></BLOCKQUOTE>

you make it sound as if we've gotten worse att calculating subsonic pops, just because todays fighter designs are supersonic jets. no logig there at all.

Kettenhunde
12-15-2009, 05:38 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">you make it sound as if we've gotten worse att calculating subsonic pops, </div></BLOCKQUOTE>

In some ways, we have gotten worse at it. We are not any more accurate today for subsonic propeller aircraft.

They also had much more experience at cutting edge subsonic propeller aircraft design than we do today. There are only a handful of companies that even work with subsonic propeller aircraft designs in today's market.

There just is not a large market for 2000hp piston engined aircraft.

Can you think of a new 2000hp piston engine aircraft design in the last 40 years?

M_Gunz
12-15-2009, 09:17 AM
If framerate and time to write was not a problem, how good of a flight simulator (even in slow motion only) can be made?
Are there virtual wind tunnels that are very accurate into the transonic range? There have been decades of study since
computers became part of the process, there must be some very good models.. doesn't every PhD have to do a major project?
How many wouldn't add something to the ability to predict flight through calculation to some close degree?

M_Gunz
12-15-2009, 09:25 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
Can you think of a new 2000hp piston engine aircraft design in the last 40 years? </div></BLOCKQUOTE>

IIRC the last of the big flying boats _may_ have been a postwar design and _might_ have used big P&W's.
Have to remember that some things were salvaged from scrapped airplanes or at least were going to.
There was one scheme to use Packard Merlins on big farm tractors and don't laugh, I have a friend with
a dairy farm and his tractor is in that power range, he raises a lot of cow feed for 40 cows though his
boy is probably big enough now to help. IIRC the Merlins also got put into boats and racing planes.

Why bother with a 2000hp piston-banger when you could be running a turbo-prop?

Kettenhunde
12-15-2009, 10:08 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Are there virtual wind tunnels that are very accurate into the transonic range? </div></BLOCKQUOTE>

Yes, we have made significant strides in the transonic, supersonic, and hypersonic realms of flight.

We are talking about subsonic designs though.

In fact, the difficulties in describing the transonic realm are a major source of our errors for both flight testing and calculations in the WWII design era.

psykopatsak
12-15-2009, 12:27 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">you make it sound as if we've gotten worse att calculating subsonic pops, </div></BLOCKQUOTE>

In some ways, we have gotten worse at it. We are not any more accurate today for subsonic propeller aircraft.

They also had much more experience at cutting edge subsonic propeller aircraft design than we do today. There are only a handful of companies that even work with subsonic propeller aircraft designs in today's market.

There just is not a large market for 2000hp piston engined aircraft.

Can you think of a new 2000hp piston engine aircraft design in the last 40 years? </div></BLOCKQUOTE>

no, but quite a few turbo-props which isn't far away. both subsonic, as far as the argument concerns.

Waldo.Pepper
12-15-2009, 02:26 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
I disagree that the debate is unsolvable. In fact that is the point of the thread is to recognize when a solution exist's.
</div></BLOCKQUOTE>

If I make the childish assumption that every test I see is done by perfectly trained individuals who are operating under the highest morale character using only the best testing equipment available, then if I make such assumptions I have achieved nothing.

Instead of solving the dilemma about whether ...

"Between a flight test and calculation, which is the most accurate in your opinion and why?"

All that has been achieved it to perform a regression into the realm of pondering the character of those individuals involved in said testing. Which solves nothing.

Or do you not see this?

psykopatsak
12-15-2009, 02:36 PM
lets just say they are german prefectionists, all professors and doctors who are handcrafting the ultimate plane, and want to calculate a "Kalkül" or something http://forums.ubi.com/groupee_common/emoticons/icon_wink.gif

Kettenhunde
12-15-2009, 03:17 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">f I make the childish assumption </div></BLOCKQUOTE>

Why are you being insulting?

So, I should just assume your input is a childish assumption? Come on man....

You think we should assume that every test we see was done by people who where stupid and could not do the job.

IMHO, It seems very illogical to assume that every test we see is the result of design team incompetence. Does that seem logical to you?

The design team making a mistake is very much an outlier and not the rule.

Making the assumption the team could do their job seems logical.

Maybe you think the folks conducting the flight test are not part of that design team?

How can you take that statement to be perfect instruments as well when most of the thread has been on the topic of margin of error for techniques and our ability to measure?

We just assume that the guy making the measurements knows how to read the scale, count, and do the math formulas.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> no, but quite a few turbo-props which isn't far away. </div></BLOCKQUOTE>

Far enough away to be completely different. Turboprops have their own set of formulation and principles in predicting performance. They take some of the characteristics of power producers and some of thrust producers. They even have their own special units of power called equivalent shaft horsepower.

I get the feeling you are not very familiar with the details of the science. Nothing wrong with that at all.

Our prediction accuracy is the same today with subsonic designs as it was in WWII.

Kettenhunde
12-15-2009, 06:52 PM
The margins of error are defined for each, btw. It is not really an opinion.

The margins for a calculation are defined as ~7% for a first flown prototype and 1% after that prototype is developed.

The margins for a single flight test data point are ~10%.

So which is more accurate, a flight test or a calculation?

na85
12-15-2009, 08:57 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
The margins of error are defined for each, btw. It is not really an opinion.

The margins for a calculation are defined as ~7% for a first flown prototype and 1% after that prototype is developed.

The margins for a single flight test data point are ~10%.

So which is more accurate, a flight test or a calculation? </div></BLOCKQUOTE>

Well the answer is obvious... but let me ask you this: How can there be a universal margin of error for test flights? That seems counter-intuitive to me.

na85
12-15-2009, 09:11 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:

Far enough away to be completely different. </div></BLOCKQUOTE>

That would be because a turboprop is essentially a super-high bypass turbofan jet engine, correct?

Viper2005_
12-15-2009, 10:03 PM
If you want to design a fast piston engined aeroplane then EHP is much more useful than BHP. In fact, the concept of EHP essentially arrived as a concept of the work of ****** et al on the Merlin (The Performance of a Supercharged Aero Engine) in about 1941.

As for PhDs advancing the state of the art, what we actually do is add to the information overload. Educational advancement means that it simply isn't possible to read all the literature anymore, just as a modern fighterpilot probably can't process all of the information that his aeroplane can technically make available to him. I believe the phrase is "drinking from a firehose"?

My thesis is well over the 100 equation mark already, and you probably don't even want to think about the computer code that those equations produce.

Na85 is correct that as BPR tends towards a large number, your engine basically tends towards a turboprop. You even start needing variable geometry of some sort to make the beast match.

As for whether flight test or calculation is more accurate, beyond the philosophical details, it depends upon which you've done more of, and which team has the best resources available. Generally it is foolish to value one dramatically over the other.

Some people really do put Merlins into tractors. There is such a sport as tractor pulling, and IMO it is one of the most tragic waste of engine life imaginable. But each to their own I guess...

As for computation, you will generally find that computations are simpler in supersonic and hypersonic flow than in subsonic flow, because the information only flows downstream, and the "approximations" generally work better than their subsonic equivalents; WWII aircraft designers were quite capable of producing supersonic aeroplanes; they had good data because supersonic windtunnels were old hat. What they didn't have the ability to do was deal with the transonic regime, provide sufficient thrust to go supersonic without a gun or a rocket, or indeed produce a supersonic aeroplane which would also fly reasonably in the subsonic regime inevitably required for takeoff or landing. NB - since Miles were working on the M.52 during WWII this is something of a grey area; equally the work conducted in Germany regarding a winged V2 might arguably constitute a supersonic aeroplane. In general I have found that the more deeply you research any given subject, the greyer it becomes...

Transonic flow can be a nightmare, as of course can the boundary layer of supersonic and hypersonic flow simulation (if zero-slip holds then any supersonic flow must be transonic somewhere, provided that you zoom in far enough).

The WWII guys had real problems with this; if you look at the achieved performance compared with the predicted performance of most WWII fighters, you'll see that a gap starts to open up once the predictions exceed roughly 400 mph TAS. This was largely due to transonic effects, and increased with the margin over 400 mph TAS. Towards the end of the war, it could be in the region of 50 mph TAS, which is scary considering the v^3 characteristic of "power producers" (though you should see my note on EHP above).

Without wishing to get specific, I would note that at least one reference in the paper supports my position regarding the Mach number achieved by the Spitfire in diving flight, which would at least tend to support the general assertion that there is uncertainty in all data, and that uncertainty (like entropy) tends to increase with time.

One of the features of the report which I find most interesting is the fact that in both the case of the Spitfire and that of the P-47, the Mach number is reported to 2 decimal places. This carries with it implications as to the accuracy and certainty of the measurements.

Whilst I am personally inclined to agree with the accuracy of the measurements obtained from tests with individual aircraft, it strikes me that the variation within the operational fleet would probably be such as to render such precision meaningless.

If you look at the wing of a cruising airliner or bizjet, you will see that the shocks aren't straight due to imperfections in surface finish which are hardly perceptible when you look at the (often natural metal & highly polished) aircraft on the ground. The visible imperfections on many WWII aircraft strongly suggest much larger variations in shock geometry, and hence one assumes not inconsiderable performance variations.

Indeed, since the maximum speed of these machines was often determined by transonic drag rise (of the prop if nothing else), it doesn't seem unreasonable to postulate that the aircraft-to-aircraft variation in tactical Mach number might be of the same order as the variation in top speed.

One of the biggest general problems with experimental (as opposed to production) flight test work is that inherently involves the questionable statistics of small samples. It's hardly surprising that the hand-built prototypes often out-performed the mass produced examples seen in service...

Waldo.Pepper
12-15-2009, 10:25 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
Why are you being insulting?
</div></BLOCKQUOTE>

Not my intent and I apologize. But frankly what I think I am getting at is ...

"There are no facts, only interpretations." -
-- Friedrich Nietzsche

I would have thought that you would have heard of that quote.

M_Gunz
12-15-2009, 11:52 PM
@Viper, Thanks and Salute!

Niche specialization in professional fields... yup, wherever you have a lot of specialists toiling deep in the mines
you know it's hard enough to get to one of those spots let alone think of covering them all! My HS physics teacher
had fled the narrowing EE field to teach the basics in public school and he says he's much happier for it, but Mr.
McKee was a pretty crazy in an entertaining way individual.

As to "how much code"... I wrote a lot more code for money in 19 years than you'd like to see, it wouldn't have
bothered me code an engine that runs equations as data and applies the output where needed. Then you plug equations
in as a text file, it compiles that and runs. I wonder how few lines of Perl might do it, hehehe?

The equations made are always in terms of this and that with conditions stated, and as general-case as possible?
I would have thought that with all the data available that the explanations for variances would have led to better
models. More this and that terms able to describe the airframe or parts of it and the air flow better than before.

What I see written seems to say that nothing has changed much about understanding of subsonic flight since the late
40's if not before. Does that include modeling and prediction of performance including how many predictions you can
accurately make?

I think that most of the posters in this thread have different POVs on the same thing from nearly the same side and
the rest don't. Writing about it well enough to be understood, the differences get bigger.

Kettenhunde
12-16-2009, 02:27 AM
The transonic, supersonic, and hypersonic realm are all about the formation and behavior of normal shock. If you cannot accurately describe one realm, you cannot describe any of them.

You may be able to describe components of normal shock and the behavior, but that is the easy part of the picture and yes they could do that during the war.

It is the ability to accurately describe the onset and formation of normal shock that stumped them. Even today, there is not very good agreement as to the onset.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">"There are no facts, only interpretations." -
-- Friedrich Nietzsche </div></BLOCKQUOTE>

Nietzsche..... http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif

There are plenty of facts available. This is not some mystery or unsolvable puzzle. It is just applied physics and reasonable solutions predicting general performance trends are not hard to do.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Whilst I am personally inclined to agree with the accuracy of the measurements obtained from tests with individual aircraft, it strikes me that the variation within the operational fleet would probably be such as to render such precision meaningless. </div></BLOCKQUOTE>

You already stated your preference for individual flight tested data so I was looking forward to your participation.

I was not even including the variation in performance of the fleet in service. The margin of error in this case is from the instrument, pilot, environmental, and production errors.

I definitely agree that if you factor the maintenance, wear, generally speaking, lower average pilot skill of the fleet in that margin, the results are meaningless.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Without wishing to get specific, I would note that at least one reference in the paper supports my position regarding the Mach number achieved by the Spitfire in diving flight, which would at least tend to support the general assertion that there is uncertainty in all data, and that uncertainty (like entropy) tends to increase with time.

One of the features of the report which I find most interesting is the fact that in both the case of the Spitfire and that of the P-47, the Mach number is reported to 2 decimal places. This carries with it implications as to the accuracy and certainty of the measurements. </div></BLOCKQUOTE>

We are not discussing any specific aircraft, only test measurements.

The errors in the static system design are well defined and not unknown. Simply put, that destroys any claim of accuracy no matter how many decimals places the result is recorded too.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> it doesn't seem unreasonable to postulate that the aircraft-to-aircraft variation in tactical Mach number might be of the same order as the variation in top speed. </div></BLOCKQUOTE>

No that does not seem unreasonable that it would vary in the same order but I don't think sharing a common feature of a propeller disc lends itself to a wide variation in magnitude. We would simply need to see some drastic changes in propeller design.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> One of the biggest general problems with experimental (as opposed to production) flight test work is that inherently involves the questionable statistics of small samples. It's hardly surprising that the hand-built prototypes often out-performed the mass produced examples seen in service... </div></BLOCKQUOTE>

Well yes and no.....

In my experience, it takes time to work out the teething problems and there is always an issue in the first flights.

I have seen issues such as cowl pressure and uneven cooling cause significant performance issues in addition to being very difficult to track down at times. In an installation lacking EGT/CHT for each cylinder, this can take some time to figure out.

You fix that kind of stuff and most likely your prototype will be performing to within 1% of performance predictions.

Kettenhunde
12-16-2009, 03:22 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Does that include modeling and prediction of performance including how many predictions you can accurately make? </div></BLOCKQUOTE>

The short answer is yes when it comes to subsonic design.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">How can there be a universal margin of error for test flights? </div></BLOCKQUOTE>

It does not mean every test flight will be within 7%. If the error falls outside of that 7%, you probably have a mistake in the math or the data and not a minor technical issue with the prototype.

It is a guideline and not a hard fast law.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> If you want to design a fast piston engined aeroplane then EHP is much more useful than BHP. In fact, the concept of EHP essentially arrived as a concept of the work of ****** et al on the Merlin (The Performance of a Supercharged Aero Engine) in about 1941. </div></BLOCKQUOTE>

Yes and if piston engine aircraft had continued to develop it would have been very useful. Engineering simply limited them and the application has seen little use today for the power ranges of today's piston engine aircraft.

Additionally, exhaust thrust, like propeller thrust, has some unique characteristics.

It is not the same as turbine thrust in a turboprop and the analogy that we can equate turboprop design to 2000hp piston engine aircraft design does not work.

Turboprops are very easy in comparison.

Kettenhunde
12-17-2009, 03:42 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">One of the biggest general problems with experimental (as opposed to production) flight test </div></BLOCKQUOTE>


Aircraft intended for a production contract generally go through prototype, category testing, and the operational testing.

It is during the category testing that envelope expansion and testing the performance of the design would occur.

They are not conducting experimental testing, which is completely different.

Kettenhunde
12-21-2009, 07:55 AM
Just to increase the readers understanding of the processes involved in flight testing here is a paper written in 1939 on correcting altimetry data to standard conditions. This was done for each data point.

Altmetry is much easier to deal with than airspeed.

These data points then are plugged into the exact same formula's used in the initial performance predictions.

http://img684.imageshack.us/img684/6818/climbdatatostandardcond.jpg (http://img684.imageshack.us/i/climbdatatostandardcond.jpg/)


More to follow on the history of compressibility research if people are interested.

M_Gunz
12-21-2009, 09:33 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kettenhunde:
I definitely agree that if you factor the maintenance, wear, generally speaking, lower average pilot skill of the fleet in that margin, the results are meaningless. </div></BLOCKQUOTE>

Which is about the opposite of grabbing an anecdote and waving it around as an absolute truth about all planes with
the same name.

Kettenhunde
12-21-2009, 10:15 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Which is about the opposite of grabbing an anecdote and waving it around as an absolute truth about all planes with the same name.
</div></BLOCKQUOTE>

I guess it is about as far as you can get from that...

Eventually I want to get into some of the performance planning. The point of the thread is to help folks recognize good data and provide some of the tools to analyze it for comparison.

TS_Sancho
12-21-2009, 01:22 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">More to follow on the history of compressibility research if people are interested. </div></BLOCKQUOTE>

Yes, please continue. This thread is a fantastic opportunity for some of us to advance our understanding of the science and context behind all the performance numbers we see quoted in aircraft discussions.

Thank you again for taking the time to share your knowledge on the subject.

Kettenhunde
12-21-2009, 04:12 PM
Thanks TS_Sancho.

Kettenhunde
12-27-2009, 10:34 AM
I have not forgotten. Just been too busy with the Holidays for a long post.

M_Gunz
12-27-2009, 11:36 AM
Hopefully not too busy for a tall drink! Merry Christmas! Happy New Year!

Kettenhunde
12-27-2009, 01:10 PM
Thanks!

To all of you on Ubi forums, Happy Holidays and I wish you the best!