Crump Posted October 27, 2016 Posted October 27, 2016 If I overinterpreted you, then I dodn^t meant to. Still I'm happy having your spreadsheet as I can use it in the way I mentioned. And it for sure it is instructive as it illustrates the relations of the variables in a more practical way than with the HP-35 pocket calculator back then. Still I am surprised to see the deviations of ~0.2 for airfoils from "real world values" on an airfoil compared to what I thought were measured values from built airfoils in a wind tunnel as well. In the article you posted these were differences seen in a front line combat aircraft of the 40's. But to say that you cannot get closer to published values...? Or can I? Why do you think one would publish data if what you get is so far off? That is actually the entire point of the article. The article does not say you will not achieve agreement from 2D data to aircraft but rather it points out what to look for if your design is NOT getting that agreement.
Crump Posted October 27, 2016 Posted October 27, 2016 I don't see anyone disputing this. However the sectional polar is not valid for finite wings. It most certainly is valid. It is the CLmax of the wing root that determines the stall in a high aspect ratio rectangular wing. At the tip of any real wing, lift goes to zero, irrespective of airfoil. If you're implying that the coefficient at the tip of the wing is 1.5 or whatever the peak of the 23009 polar is, I'm afraid you are incorrect. Has nothing to do with Clmax of a wing design in a high aspect ratio rectangular wing.
CUJO_1970 Posted October 27, 2016 Posted October 27, 2016 So, the FW-190A-3 is going to be reworked at the same time they work on the FW-190A-5? Is this correct?
Crump Posted October 27, 2016 Posted October 27, 2016 So, the FW-190A-3 is going to be reworked at the same time they work on the FW-190A-5? Is this correct? Yes
JG13_opcode Posted October 28, 2016 Posted October 28, 2016 It most certainly is valid. It is the CLmax of the wing root that determines the stall in a high aspect ratio rectangular wing. Again, that's not what I'm saying. In a simulation you need more than just the root airfoil. So, the FW-190A-3 is going to be reworked at the same time they work on the FW-190A-5? Is this correct? That is what I've been told.
Crump Posted October 28, 2016 Posted October 28, 2016 Again, that's not what I'm saying. In a simulation you need more than just the root airfoil. Certainly if you want to determine behaviors. For performance you only need the root airfoil. It defines the edge of the performance envelope. We only care that the wing is stalling for performance and cannot fly anymore. We do not care if the ailerons are effective or if the airplane inverts does a somersault to a three point landing.
CUJO_1970 Posted October 28, 2016 Posted October 28, 2016 OK guys, that is some really big news, but there is a big difference between evaluate and rework based on new information. "Rework" implies that sufficient new documentation has been provided to the developers that supersedes the documentation that resulted in the 1.17 Clmax and that they will "rework" to corrected FM. "Evaluate" simply means they can look at it and conclude that everything is just fine as is and there will be no change. I know some of you have worked behind the scenes to provided the developers documentation required....I appreciate your efforts as well as the knowledge you share here.
unreasonable Posted October 28, 2016 Author Posted October 28, 2016 Another source: the US report on RAE Fw tests. http://www.wwiiaircraftperformance.org/fw190/Fw_190_Eng-47-1658-D.pdf Two points to note: 1) The power off stall speed is given as 118 IAS - unlike the 110 in the other RAE text. Typo? In which one? Who knows. 2) Weight is recorded as "full gas and oil but no ammunition....) So illustratively, take the default Crump figures in the calculator, knock 200kg off the weight and set the speed so that the IAS Vmin = 118mph. Result - CLmax 1.29 This simply demonstrates the sensitivity of the calculated result to the data entered - the RAE result cannot be used as confirmation of anything.
ZachariasX Posted October 28, 2016 Posted October 28, 2016 we use the wing area as our reference area, we use the wing CLmax to determine performance. That is how aircraft performance math is done. I must admit, this how I was told it to back then. Other people designed R/C aircraft as well, but "just taking wing" and tacking that to an airframe. Some built after published plans, but they often enough didn't fly satisfactorily. Looking at the plans, they were drawn in a way not consistent procedure I described initially in this thread. Even though it looked in the right proportions, often enough it made me just help them collecting pieces of wood. It was before the time where you could just add a brushless motor to a brick and it would fly. The next question then becomes if you are going to lobby for a wing designed specifically for a high Clmax to have a lower Clmax then how is that going to effect your relative line up? Are you going to lobby for lower Clmax airfoils to be increased so that the relative performance picture is now skewed to their favor? If we were, this had serious implication on Russian build planes as well as early Spitfires. At Rechlin, they frowned at the building quality of those aircraft. Even with the Spitfire, they thought one should replace half of the rivets and insert them properly to make the aircraft truly airworthy. The Spitfire AFAIK also only received flush rivets when the Mk.IX came out. The Soviets had even larger tolerances. Now, if we deducted (in consequence) an even larger fraction of lift from the ClarkY of the Yak, imagine the wildfire here in the forum. It would bring CLmax below what we have correctly for the Fw-190. How that "flies", we know. Do we give all aircraft such a penalty? I mean, in the end it wouldn't it be great if we had wind tunnel data from full size aircraft and derive from there? Then we had in the game represented what that particular aircraft had as lift/drag coefficients. But it should be about consistent with what you had as service aircraft. I guess it would be at least a "politically correct figure". I know some of you have worked behind the scenes to provided the developers documentation required....I appreciate your efforts as well as the knowledge you share here. Also my thanks to anyone who passed the developers proper documentation!
ZachariasX Posted October 28, 2016 Posted October 28, 2016 This simply demonstrates the sensitivity of the calculated result to the data entered You can use the calculator like that, and I am glad you think it is interesting or handy. Just be careful..... You told me to be careful. But:
JG13_opcode Posted October 28, 2016 Posted October 28, 2016 Certainly if you want to determine behaviors. For performance you only need the root airfoil. It defines the edge of the performance envelope. We only care that the wing is stalling for performance and cannot fly anymore. We do not care if the ailerons are effective or if the airplane inverts does a somersault to a three point landing. We do in a simulation.
Crump Posted October 28, 2016 Posted October 28, 2016 We do in a simulation. Not to determine the stall point and the performance of the aircraft. You need behavior math after that in a sim but that is not performance math and will not tell you the correct stall point.
Crump Posted October 28, 2016 Posted October 28, 2016 I must admit, this how I was told it to back then. It is pretty much standard. It is standard practice in performance math to use just the wing root is what does the heavy lifting for our aircraft and when it stalls, the airplane is not flying anymore. If we were, this had serious implication on Russian build planes as well as early Spitfires. At Rechlin, they frowned at the building quality of those aircraft. Even with the Spitfire, they thought one should replace half of the rivets and insert them properly to make the aircraft truly airworthy. The Spitfire AFAIK also only received flush rivets when the Mk.IX came out. The Soviets had even larger tolerances. Now, if we deducted (in consequence) an even larger fraction of lift from the ClarkY of the Yak, imagine the wildfire here in the forum. It would bring CLmax below what we have correctly for the Fw-190. How that "flies", we know. Do we give all aircraft such a penalty? I mean, in the end it wouldn't it be great if we had wind tunnel data from full size aircraft and derive from there? Then we had in the game represented what that particular aircraft had as lift/drag coefficients. But it should be about consistent with what you had as service aircraft. I guess it would be at least a "politically correct figure". Everybody had quality control issues and everybody conducted acceptance checks as well as other programs to fix any quality control issues. Aircraft in the field require maintenance but for the most part, the ground crews did an outstanding job. They are kind of the unsung heroes in this... It is a very dangerous road to go down. It will completely upset any relative performance line up and leave a skewed performance picture if weathering effects are ever added. You are talking about altering the design characteristics of the aircraft and morphing it into something else beside the intended model. Modeling factory a properly constructed and finished aircraft also gives some easy options for further development. A weathering slider for example should be connected to Clmax. The rougher the paint job the lower the Clmax. The longer the aircraft has been in the field the engine power begins to suffer. It allows them to have a measured starting point to begin to simulate wear and tear on the airframe as part of a campaign.
Crump Posted October 29, 2016 Posted October 29, 2016 Another source: the US report on RAE Fw tests. http://www.wwiiaircraftperformance.org/fw190/Fw_190_Eng-47-1658-D.pdf Two points to note: 1) The power off stall speed is given as 118 IAS - unlike the 110 in the other RAE text. Typo? In which one? Who knows. 2) Weight is recorded as "full gas and oil but no ammunition....) So illustratively, take the default Crump figures in the calculator, knock 200kg off the weight and set the speed so that the IAS Vmin = "ABOUT" 118mph. Result - CLmax 1.29 This simply demonstrates the sensitivity of the calculated result to the data entered - the RAE result cannot be used as confirmation of anything. Fixed it for you.... Where is the PEC correction and at what altitude did they perform the stall testing? Oh yeah, it is not there in the report because there was none and the airspeed indicator was not calibrated. You do realize the report actually evens says "about" before the listed speed? You think the error lies in the "about" speed measurement from one hasty flight or in the years of design team testing and evaluating the wing design?
NZTyphoon Posted October 29, 2016 Posted October 29, 2016 Another NACA report that may prove useful: Summary of Measurements in Langley Full-Scale Tunnel of Maximum Lift Coefficients and Stalling Characteristics of Airplanes. The abstract states: The results of measurements in the Langley full-scale tunnel of the maximum lift coefficients and stalling characteristics of airplanes have been collected. The data have been analyzed to show the nature of the effects on maximum lift and stall of wing geometry, fuselages and nacelles, propeller slipstream, surface roughness, and wing leading-edge appendages such as ducts, armaments, tip slats, and airspeed heads. Comparisons of full-scale-tunnel and flight measurements of maximum lift and stall are included in some cases, and the effects of the different testing techniques on the maximum-lift measurements are also given. Some conclusions that were reached (page 37): 5. Propeller operation will generally increase the severity of the stall, especially on single-engine airplanes, by producing an asymmetrical stall pattern and by cleaning up the inboard sections of the wings. 6. The maximum lift coefficient of an airplane may be appreciably increased by the elimination of wing surface roughness and air leakage through the wing. 7. The detrimental effect of placing machine gun and cannon at the leading edge of a wing may be reduced considerably by properly locating the guns in the wings. Highest maximum lift coefficients were measured for machine-gun installations in which the ends of the barrels were flush with the wing surface at the leading edge and slightly above the wing chord line and for cannon installations that were submerged in the wings.
Crump Posted October 29, 2016 Posted October 29, 2016 Thanks for sharing that! In fact the main conclusion of that report is: A comparison of full-scale-tunnel and flight measurement of the maximum lift coefficient of an airplane showed that satisfactory agreement may be obtained if the comparison is made under similar test conditions, such as Reynold number, slipstream, and time rate of change of angle of attack. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19930091906.pdf Perhaps all this resistance to established engineering practices is because you do not think design teams are capable of achieving a properly constructed and finished aircraft. Let's see exactly what finish we have achieve before surface roughness begins to degrade our airfoil.... It is known, at one extreme, that the surfaces do not have to be polished or optically smooth. Such polishing or waxing has shown no improvement in tests in the Langley two-dimensional low-turbulence tunnels when applied to satisfactorily sanded surfaces. So polishing has NO effect on the if the finish is properly prepared.... Large models having chord lengths of 5 to 8 feet tested in the Langley two dImensional low-turbulence tunnels are usually finished by sanding in the chordwise direction with No. 320 carborundum paper when an aerodynamically smooth surface is desired. Experience has shown the resulting finish to be satisfactory at flight values of the Reynolds number. Any' rougher surface texture should be considered as a possible source of transition, although slightly rougher surfaces have appeared to produce satisfactory results in some cases. https://www.amazon.com/Load-Silicon-Carbide-Sandpaper-Sheets/dp/B0006M2SI4 That is a pretty low standard actually. If you do not believe me, go rub your car down with 320 grit paper and get back to us with the result. Let us know what you think is smoother, the paint or the sanded finish... This data was collected with a surface roughness equal or slightly worse to being rubbed down with 320 grit sandpaper as the finish. No paint, no primer, no other surface preparation....just rubbed down with 320 grit paper is all. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19930090976.pdf
Crump Posted October 29, 2016 Posted October 29, 2016 (edited) From the FW-190 Operating Instructions using the devs translations: "Upon failure of the flap adjustment has to be considered that the ailerons are more sensitive in the rash." Outside of "the rash" being mistranslated and should probably mean reversal or stall, it makes sense, the ailerons adjustments were problematic even in the Luftwaffe. If improperly adjusted, the ailerons will reverse before the wing is stalled. That is not good and leads to early stalling of the wing as our pilot will not be able to match the angle of attack rate of change the 1.58 Clmax is based on. That is also why your good flight instructor will smite you if you touch the ailerons in a stall. The touchdown speed increases by about 35 km / h. Standard instructions for most aircraft when performing a no flaps landing, the speed increase is slightly higher but considering the aileron characteristics, understandable. This would be the increase on the airspeed indicator as read by the pilot over the normal landing configuration stall speed. If 165kph is our average 1G Indicated Stall Speed then this gives us a normal buffer above the stall for approach. Vref is usually found between 1.2 to 1.3 stall speed in the landing configuration. 165kph + 35kph = 200kph 200kph/165kph = 1.2 Vs1 Since it is very different in the individual aircraft stalling speed is appropriate at high altitude detected with idle speed (is around 160-170 km / h) and this speed of 20 km / h increases considered landing speed." Great! we have a 1G power off stall speed at high altitude! It does not say anything about landing condition stall speed except to point out your landing speed must increase and it will be a 20kph buffer over landing speed. I like it that you picked the middle of the road as the range is given for analog airspeed indicator error and NOT Position Error Correction. Focke Wulf used IAS in the Operating Instructions as is standard practice. In the stall, pilots do not care about TAS, Calibrated, Equivilent of True Airspeed. All of that performance listed in the Flugzeug Handbuch is indicated airspeed Geschwindigkeit Angezeigt (Va) unless otherwise noted as Geschwindigkeit Wahre (Vw). Late in the war, like North American and a few other manufacturers Focke Wulf begins using Position Error Correction (Geschwindigkeit Kalibrierte, Vwc) and Compressibility Error Corrections (Geschwindigkeit Kompressibilität, Vwk) to arrive at True Airspeed as we know it...or Vwck in some of their reports but never in the Flugzeug Handbuch. That is raw indicated airspeed and has nothing to do with what the airplane is actually doing. So lets see if our German Data and Our Allied data give good agreement in this case. Power off stall speed in the FW-190A3. It is pretty well documented the stalling speed of the aircraft. Faber's WNr 313 was one of the best documented examples and we will use as well as some other FW-190A3/A4 aircraft that are also equally well documented as well the leading expert on the FW-190, Focke Wulf, GmbH data. Let's see how the British, American, and German data aligns within instrument margin of error with the fixed and defined relationship of Coefficient of lift and angle of attack to get a good ballpark of what performance is plausible and how far off that mark we can be for our FM to be a good representation of the aircraft. The problem is that airspeed changes due to weight, instrument error, and density altitude. It is NOT an absolute. The only absolute relationship is a wing stalls at the same Angle of Attack. This is one of the defining relationships in aircraft performance calculations. Due to propeller effects, power on and power off stall angle of attack will not be the same with power on being the higher of the two, generally speaking. Faber's WNr. 313 stalled at 110mph IAS clean configuration and 105 IAS mph in landing configuration, gear down and flaps 60 degrees. Well our gauge in the cockpit is a poor indicator of the speed the wing is actually feeling as it moves thru the air. An airplane is the only vehicle whose speedometer does not tell you how fast you are going and the altimeter does not tell you the altitude. It only sort of tells us what the airplane feels. In order to know what the airplane is feeling, we have to use Equivalent Airspeed which also called "perfect indicated airspeed" or "engineers speed". To do that we have to convert Indicated Airspeed to Equivalent Airspeed. Indicated airspeed + Position Error Correction = Calibrated Airspeed Calibrated Airspeed + Compressibility Correction = Equivalent Airspeed To get our calibrated airspeed we need a Position Error Curve. This is nothing more than a curve that shows us the instrument error. This is due to a number of factors mainly manufacturing differences and integrity of the pitot static system. It generally measured by flying a course over the ground between know distances, accounting for wind speed, and using the rate formula to figure out the difference between the rate on the airspeed indicator and what the rate formula tells us it should be... Position Error Correction is not universal but with a range of error, it is close enough. The more data points one gathers, the better. So let's see what we have for Position Error Curves on the same pitot static system design for the FW-190A3. Remember, the pitot static system was changed in later variants and moved to the outboard wing tip. Let's compare two PEC curves and get an idea of the Position Error Correction. Here is the provisional investigation PEC curve done by the RAE on WNr 313. Here is an FW-190A4 EB-104 PEC curve: The data points are plotted in each curve. The RAE curve is derived from 5 data points while the USAAF curve was constructed from 9 measured points. Atmospheric conditions such as gusty winds effect these datum points. The fact that the USAAF curve has more points and the points are more consistent without the large spread that appears in the RAE PEC curve leads me to believe the USAAF datum was collected under more ideal conditions. At 110mph the PEC curves both show the FW-190 airspeed indicator is showing the airplane is moving faster than it actually is going. Neither curve goes to 110 mph IAS so we must ballpark it and our error shows the airspeed indicator anywhere from about 4 mph to 10 mph faster than the airplane is going. So our calibrated airspeed is going to be a range from 106 mph CAS to 100 mph CAS. Now to get our Equivalent Airspeed we need to account for compressibility effects. In modern terms, below 10,000 feet there is no reason to apply a correction unless we are going over 200 knots CAS. It is negligible. In World War II, the expression is not universal and ranges from no compressibility at all to somewhat large compressibility corrections by today's standards. I find it highly unlikely that any test pilot performed stall testing on a unknown enemy design below 10,000 feet. Why? Stalling runs the risk of spinning. In, many World War II fighters the minimum safe altitude for spin recovery was 10,000 feet. About .5 mph is a reasonable compressibility error correction. Our Focke Wulf Flugzeug Handbuch even acknowledges this fact: Since it is very different in the individual aircraft stalling speed is appropriate at high altitude detected with idle speed (is around 160-170 km / h) and this speed of 20 km / h increases considered landing speed." So, Our Calibrated Airspeed brings our Equivalent airspeed range to 105.5mph to 99.5mph. Our FW-190 should stall somewhere in that EAS mile per hour airspeed range. Let's see if our absolute relationship of angle of attack and coefficient of lift, falls within that speed range as measured by the RAE and USAAF. Focke Wulf used a CLmax of 1.58 for the FW-190A series as the airplanes design CL max. That is found here, in "Widerstandsdaten von Flugzeugen". The FW-190A design maximum coefficient of lift is 1.58. At 8580 lbs weight for WNr 313 at measured at a stall speed of 110mph IAS let's see what our Equivilent Airspeed has to be in order to achieve a Clmax of 1.58 and does that Equivalent Airspeed fall without our range of possible speeds. WNr. 313 is a Type II fighter variant with 4 wing cannon. The weight difference is due to the fact the RAE used natural Petroleum AVGAS which weighs ~7.48lbs per imperial gallon. C-3 fuel is closer to jet fuel and is a light oil with a distinctive burnt coal smell and weights ~7.86 lbs per imperial gallon. Using the BGS system and a conversion factor for knots... EAS means sigma (density ratio) = 1 Velocity in KEAS = SQRT{295 (8580lbs) / (1.58 * sigma * 197 ft^2) Velocity equals = 90.2 Knots Equivilent Airspeed 90.2 KEAS * 1.15 = 104 mph EAS. 110 mph IAS - 104 mph EAS = Correction error of 6 mph. 104 mph EAS + .5 mph CEC = 104.5 mph CAS = 104.5 mph CAS + 5.5 mph PEC = 110 mph IAS as recorded by the RAE on WNr 313. So at the weight that the RAE corrected WNr 313 data to of 8580lbs the pilot reading 110mph IAS would be traveling at ~90.2Knots EAS. That fits well within our range of airspeed measurement error and agrees with Focke Wulf operating instructions and the published Clmax of their design. The only absolute relationship of stall vs angle of attack easily returns a stall speed that is well within the margin of error found in airspeed measurement. Given the fact the Allied data and German data both give good agreement regarding stall data we can gain some insight referencing both sets of data. The 1G level power off stall point for a Type II fighter is the same airspeed as the Landing speed. Both equal 110 mph IAS. The most damning evidence outside of Focke Wulf's published data for the Clean Wing CLmax is the fact our Touchdown Speed and 1G clean stall speed are the exact same in the FW-190. Both are 110mph Indicated Airspeed measured by the RAE and including a PEC curve. A Clmax of 1.58 represents the clean configuration (gear up, flaps up) design CLmax. This can be used to find the correct stall speed for any weight or configuration of the aircraft. Touchdown Speed also requires a CL of 1.58 as the lift required does not care about configuration. At the same weight and dynamic pressure, we require the same coefficient of lift. That can easily be confirmed with the Landing data chart found in the Flugzeug Handbuch for the FW-190A8; same wing design as the FW-190A3: Focke Wulf used a CLmax of 1.58 for their design. If touchdown speed = 1G stall speed then the Coefficient of lift is the same for both conditions of flight. The clean wing Clmax MUST be 1.58. That is backed up by Allied testing of the design and independent measurement of wing design of the same airfoil selection by Grumman Aerospace. Edited October 29, 2016 by Crump
JG13_opcode Posted October 29, 2016 Posted October 29, 2016 Not to determine the stall point and the performance of the aircraft. You need behavior math after that in a sim but that is not performance math and will not tell you the correct stall point. Agree to disagree, then. As I have heard back from Han that they will not be touching this issue until the A-5 work is underway, any further discussion is moot at this point.
Crump Posted October 29, 2016 Posted October 29, 2016 Agree to disagree, then. As I have heard back from Han that they will not be touching this issue until the A-5 work is underway, any further discussion is moot at this point. Ok. In the meantime, please share with me your 2D airfoil theory that supersedes measured data or even gives good agreement at anything above low to moderate angles of attack! The foundation of Lifting Line theory and what makes it give good agreement with measured data is the fact the 2D section data MUST agree with the measured results or be within the confines and limitations of airfoil theory. One of those limitations is that none of the current airfoil theories do well at large angles of attack so I am curious as to how we are going to solve this without CFD or measured data. It is a very flawed foundation to do an analysis of CLmax based on airfoil theory math!
CUJO_1970 Posted October 29, 2016 Posted October 29, 2016 As I have heard back from Han that they will not be touching this issue until the A-5 work is underway, any further discussion is moot at this point. We only have to live with that 1.17 CLmax screw up for about one year then when (or if) it gets fixed. Nice to know we are always one well-meaning chart away from disaster here.
NZTyphoon Posted October 29, 2016 Posted October 29, 2016 (edited) In fact the main conclusion of that report is: A comparison of full-scale-tunnel and flight measurement of the maximum lift coefficient of an airplane showed that satisfactory agreement may be obtained if the comparison is made under similar test conditions, such as Reynold number, slipstream, and time rate of change of angle of attack. In other words, unless flight test conditions match wind tunnel tests almost exactly, there is no guarantee that the two results will agree. This is why the words may and if have been used in the report, not will and when... The actual, concluding paragraph of this report states: 10. In a single instance where great care was taken to reproduce the test conditions of Reynolds number, propeller operation and the time rate of change of angle of attack, satisfactory agreement of the maximum lift coefficients determined from full-scale tunnel and flight tests was obtained. It is believed that equally satisfactory agreement may be obtained with other airplanes provided that sufficient care is taken to reproduce the test conditions. For the rest of it, this NACA report again shows that other factors, not just the airfoil shape, come into play when determining CLmax numbers on aircraft, reinforcing Holtzauge's comments In addition to it being difficult to predict how small deviations in the wing profile on a micro scale like joints, bulges, gun ports, hatches, rivets and surface finish etc. the IRL wing on a macro level will also deviate from a wind tunnel model in that there are gaps between ailerons, flaps and access hatches etc. that are not airtight and since air will flow from high to low pressure you will get small disturbing “fountains” of air seeping out from under and inside the wing to the upper suction side disturbing the flow. In a solid smooth wind tunnel model and a CFD model this is of course not an issue. You also have effects from propeller wash, wing root fillets etc. the effects all of which may be difficult to predict. So even if you can run at full scale Reynolds numbers (which was the exception not the rule) like in NACA’s variable density wind tunnel or get CFD results you still have to do some guesswork. Edited October 30, 2016 by NZTyphoon
Crump Posted October 29, 2016 Posted October 29, 2016 In other words, unless flight test conditions match wind tunnel tests almost exactly, there is no guarantee that the two results will agree. This is why the words may and if have been used in the report, not will and when... Really?? You know we go to school for this in aeronautical science. If you car does not work like it is supposed to you troubleshoot and fix it right? You think engineers just throw up their arms and say, "Shucks, Jim...that mystery wingding thing did not measure up....dang it all we have to start all over again!" In fact, these reports you quote contain the troubleshooting instructions! You just keep wanting to read them as impasses and impossible odds that cannot be overcome. I will admit that sanding even a small airplane with 320 grit sandpaper is not a task I would relish but there are laborers for that stuff and they can be supervised. 10. In a single instance where great care was taken to reproduce the test conditions of Reynolds number, propeller operation and the time rate of change of angle of attack, satisfactory agreement of the maximum lift coefficients determined from full-scale tunnel and flight tests was obtained. It is believed that equally satisfactory agreement may be obtained with other airplanes provided that sufficient care is taken to reproduce the test conditions. That is really taking this in a misleading way, LOL. The summary is quite clear in what they mean as it is a foundation of aeronautical science that 2D measured data agrees with the airplane. You're taking to be that they could only get a single airplane to agree with the 2D data. That is baloney and patently false. One airplane was tested and great care was taken to ensure the conditions of Reynolds number, rate of change in angle of attack, and dynamic pressure values all were under the same conditions!! Wha-Lah!! They got the expected good agreement and aeronautical science is safe for another day!! You see this getting 2D data to agree with an airplane as some frontier to be explored. I see it just as my professor told me, "This is what you do and where you get the information..." Those guys at the NACA were taught the same thing, LOL. They simply confirmed that the engineering norm was correct. You think they needed to do that to every airplane? No, one is enough. Here is yet another study on the foundation of aeronautical science! It says we can keep going as a science and not religion too!! If you do not get good agreement, then the first thing you need to do is check the conditions of the test to ensure they match. Kind of funny that is exactly what Focke Wulf did at Chalise Muerdon. http://forum.il2sturmovik.com/topic/25504-moderators-please-lock-thread/?p=395596 It kind of worked for them didn't it? They got that wing-dingy thing to work as expected LOL.
JG13_opcode Posted October 30, 2016 Posted October 30, 2016 In the meantime, please share with me your 2D airfoil theory that supersedes measured data or even gives good agreement at anything above low to moderate angles of attack! Nothing I wrote conflicts with thin airfoil or Prandtl's theory. You're just not getting what I'm saying and that's OK.
unreasonable Posted October 30, 2016 Author Posted October 30, 2016 Agree to disagree, then. As I have heard back from Han that they will not be touching this issue until the A-5 work is underway, any further discussion is moot at this point. Indeed. At least they now seem to accept that there is a possible issue with the 190 and it is not just about "emotions" - but we will have to see what they decide is appropriate. Perhaps if anyone wishes to continue to lobby/propagandize for a particular number they should now start their own thread. If, on the other hand, anyone has anything to add on how the CLmax of other aircraft can or cannot be determined empirically using the equations in the calculator, please go ahead - this is after all the topic.
NZTyphoon Posted October 30, 2016 Posted October 30, 2016 (edited) Condescending blather... In fact, these reports you quote contain the troubleshooting instructions! You just keep wanting to read them as impasses and impossible odds that cannot be overcome. I will admit that sanding even a small airplane with 320 grit sandpaper is not a task I would relish but there are laborers for that stuff and they can be supervised. ...more condescending, irrelevant remarks... Had Crump actually read the report, which he obviously hasn't, he would know that the data on WING SURFACE ROUGHNESS AND LEAKAGE (starting on page 22) has little to do with the paint finish; in fact it deals with ...access doors, inspection plates, gun ports ammunition ejection slots, and many other items that tend to make the wings extremely rough and to allow air leakage through the wings....The data include lift measurements with the wings in service and with the wings faired and sealed in attempts to increase the maximum lift coefficients of these airplanes. Instead, Crump has resorted to his usual smart-alec, condescending style, that isn't worth honoring with a reply. Edited October 30, 2016 by NZTyphoon
Crump Posted October 30, 2016 Posted October 30, 2016 he would know that the data on WING SURFACE ROUGHNESS AND LEAKAGE Really... Leakage is fixed by proper construction NzTyhpoon. If an assembly line worker builds a access panel and it does not fit properly... then we have leakage. The data include lift measurements with the wings in service and with the wings faired and sealed in attempts to increase the maximum lift coefficients of these airplanes. Do you know what faired and sealed means? It is not that hard to do, guy. You are kind of reminding me of the defense you put up denying the early marque Spitfires were longitudinally neutral to unstable at normal to rearward CG limits!!
NZTyphoon Posted October 30, 2016 Posted October 30, 2016 A couple more interesting NACA reports: The experimental and calculated characteristics of 22 tapered wings. Comparative Flight and Full-Scale Wind-Tunnel Measurements of the Maximum Lift of an Airplane.
JtD Posted October 30, 2016 Posted October 30, 2016 Indeed. At least they now seem to accept that there is a possible issue with the 190 and it is not just about "emotions" - but we will have to see what they decide is appropriate. Perhaps if anyone wishes to continue to lobby/propagandize for a particular number they should now start their own thread. If, on the other hand, anyone has anything to add on how the CLmax of other aircraft can or cannot be determined empirically using the equations in the calculator, please go ahead - this is after all the topic. It would help the developers if one complete set of data could be tracked down, meaning a stall speed corrected for position error together with aircraft weight and test conditions. Even better if the test conditions represented what we want to see in game, for instance it should be a fighter version in mint condition.
Hutzlipuh Posted October 30, 2016 Posted October 30, 2016 Agree to disagree, then. As I have heard back from Han that they will not be touching this issue until the A-5 work is underway, any further discussion is moot at this point. We only have to live with that 1.17 CLmax screw up for about one year then when (or if) it gets fixed. Nice to know we are always one well-meaning chart away from disaster here. I bet you if the FW 190 was overperforming it would have been fixed within 2 weeks... I just cant understand the devs work ethics... if i screwed up the first priority i had to do is fix it .... Also i am missing a documentation on all planes WHAT data they used..i know that requires some work , but since they already have to gather documentation and filter it , there must be a list. Sorry to say this, but for me there seems to be a lack of interest from the dev side. 2
unreasonable Posted October 30, 2016 Author Posted October 30, 2016 It would help the developers if one complete set of data could be tracked down, meaning a stall speed corrected for position error together with aircraft weight and test conditions. Even better if the test conditions represented what we want to see in game, for instance it should be a fighter version in mint condition. Right - but short of us going and buying our own plane, it seems unlikely that we are going to find that now, but who knows. Perhaps the truth is out there.... I bet you if the FW 190 was overperforming it would have been fixed within 2 weeks... I just cant understand the devs work ethics... if i screwed up the first priority i had to do is fix it .... Also i am missing a documentation on all planes WHAT data they used..i know that requires some work , but since they already have to gather documentation and filter it , there must be a list. Sorry to say this, but for me there seems to be a lack of interest from the dev side. This kind of speculation is only going to get you banned and/or the thread locked, if the moderators are still alive. 1
ZachariasX Posted October 30, 2016 Posted October 30, 2016 I bet you if the FW 190 was overperforming it would have been fixed within 2 weeks... I just cant understand the devs work ethics... if i screwed up the first priority i had to do is fix it .... Also i am missing a documentation on all planes WHAT data they used..i know that requires some work , but since they already have to gather documentation and filter it , there must be a list. Sorry to say this, but for me there seems to be a lack of interest from the dev side. Given the the info we have at hands, I find this an utterly unfair statement. The devs made a decision based on the info they had at hand. It has been discussed that there is better info at hands and there are dedicated members of this forum passing this on to the devs. As the in-game Fw would rather need a basic overhaul than a tweak (it matches published speeds and climb afer all!), the descision to do so when creating the A5 does make a lot of sense. People need to get paid to make this overhaul. As said, especially on second thought I am happy unreasonable provided his his spreadsheet as it illustrates some fundamentals of aerodynamics, especially to the layman like myself. Reading through the tread, I find a lot of arguments are made using science in a negative way. With this I mean to disprove something rather than providing means for a correct predictions within defined constraints. This seems to be ok for some as: Z, I am not trying to design an aircraft! therefore you do not need any specific number, one is just happy that it may be somewhere that consits with one's current understanding of the material. Given what we have, unreasonables spreadsheet however does not rule out the "infamous 1.58" that at least Focke Wulf GmbH was comfortable to put in their documentation. There is good reason to assume that final "true real world values" might differ slightly from technical data as Holzauge and JtD explained. Now how much that ultimately is, most obviously difficult to say. Just deducing .2 based on one report I find drastic. After all, it reflects a lift equivalent of missing 20% airfoil between "theory" (even though that "theory" might have been derived in a wind tunnel) and real world. Now, the fraction of people that doesn't want to build aircraft might walk away happy and satisfied. Negative science (in this case disproving the 1.58 works well and is even very plausible) is good enough. The devs in turn they have to design an aircraft. THEY NEED A NUMBER, even though the aircraft is just algorithms. As they cannot insert "~1.3 to ~1.4" in their formula. We need a "correct" value. And in science, you simply don't vote on the correctness of a value. It is either correct or it isn't. Like it or not. In this sense the statement (although probably attributed to the high probability of attracting bad conversation when discussing such a topic) Perhaps if anyone wishes to continue to lobby/propagandize for a particular number they should now start their own thread. is very appropriate, because the correct number doesn't need lobbying. It is just correct. It however implies that one is happy not knowing CLmax. In process, one is happy having understood that basic science cannot determine some of the most fundamental properties of aerodynamics. So only fancy computers doing computations we don’t understand and don’t care to understand can do it, producing rainbow colored models of your plane? But even that doesn’t reflect your aircraft, because it has a “perfect wing finish”? What is all that good for then? In short, I am left with a great explanation why basic applied aerodynamics is nothing is nothing but chance in the end. In the light of your background unreasonable, yes, applied economics is nothing but religion. So you may be happy with that concept. Coming from more “exact” sciences, this makes me spit my coffee. My suggestion of using wind tunnel data for political reasons was just to use a number that bears some justification that gives even the most illiterate with an opinion something that makes sense to him. (I do like the idea of the CLmax slider for weathered skins. Imagine how polished aircraft would look on WoL.) But where would that leave other aircraft? The NACA 230x profiles have their thickest part rather far forward. After that point, there is usually a turbulent airflow. This means that I would expect this kind of profile to be rather resistant to surface defects, as most of those will be located in an area of turbulent airflow. This in contrast to a laminar flowing wing, like on the P-51. AFAIK the very least of all P-51 in the field ever had laminar flow around the wing, as it is enough for a rigger and fitter to walk on the wing producing dents and scratches big enough to distort the airflow. If we had an aircraft that supposedly should deviate most from theoretical values, it should be the P-51. Yet we know it was still one of the most efficient airframes (in the real world) of the time. Messerschmitt and Kurt Tank on purpose didn’t use laminar flow wings as they knew they were most affected by surface imperfections and considered unsuitable for field use. How much the profile affects the aerodynamic robustness to imperfect surface can be illustrated here: That one has the thickest part of the wing profile very much forward. You hardly notice a difference when it’s wet, when there are bugs on it, etc. and you do need substantial airbrake for it to have an effect. And even though the surface is cloth, it is (or should be) much finer than 320 grid sandpaper. Take this one: and you notice every mosquito sticking on your wing. And just touch the brakes and watch the variometer. Surface finish is basically similar as wind tunnel airfoils, else you have to return the plane to the shop. You may argue, the Falke, engine off, is dropping like a piano anyway compared to the Nimbus. But so is the Fw-190. Morien Morgan didn’t have super computers, but he still could design something that was a Concorde later on. Was he just lucky? 1
ZachariasX Posted October 30, 2016 Posted October 30, 2016 It would help the developers if one complete set of data could be tracked down, meaning a stall speed corrected for position error together with aircraft weight and test conditions. Even better if the test conditions represented what we want to see in game, for instance it should be a fighter version in mint condition. That would be great, but it probably would redirect a lot of resources for communication to "get the best data together with the community". AFAIK the devs are asking already for a certain set of info on the respective plane and they take from that what they can use. Adding to the spec sheet what we are discussing here would be irrelevant info to probably most of their clients at the cost of additional exposure.
Crump Posted October 30, 2016 Posted October 30, 2016 CONCLUSIONS 1.Satisfactory agreement exist between the maximum lift coefficients measured in flight and those measured in the full- scale wind tunnel. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19930091693.pdf The calculated and experimental values of CLmax are not always in good agreement. the predicted CLmax would be 1.74. The CLmax actually measured was 1.81 There is no airfoil theory that predicts the CLmax of an airfoil accurately. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19930091703.pdf
Crump Posted October 30, 2016 Posted October 30, 2016 (edited) n the light of your background unreasonable, yes, applied economics is nothing but religion. So you may be happy with that concept. Coming from more “exact” sciences, this makes me spit my coffee. Me too. A degree in Philosophy is a degree in the art of argument that everything is uncertain, LOL. That works well for contemplating questions of existence of consciousness and our place in the universe. Yes, the arrow has to fly 1/2 of the distance to the target, 1/4 distance, 1/8 distance, 1/16 distance...etc etc etc..mathematically it could be divided until infinity and never the reach the target. We know that is not true. Shoot the arrow and it hits the target in a relation to D=RT formula. My suggestion of using wind tunnel data for political reasons was just to use a number that bears some justification that gives even the most illiterate with an opinion something that makes sense to him. (I do like the idea of the CLmax slider for weathered skins. Imagine how polished aircraft would look on WoL.) But where would that leave other aircraft? I was disappointed in CloD when that slider had no effect at all on performance. Not using a Clmax gathered under specific and known conditions would leave this and other designs in a quagmire. The relative performance would be skewed at the mercy of the inaccuracies of airspeed measurement in the early 1940's. The foundation of aeronautical science is the ability to relate measured 2D data to a full sized aircraft and return the same results. It has taken several pages but people seem to realize it is much more complicated than just pulling back the stick and seeing what the airspeed indicator reads when the stall happens. The aerodynamic conditions must be the same. The root airfoil in a high aspect ratio rectangular wing or one with moderate taper is like the first domino in a chain. When it falls, it starts a chain reaction across the wing pushing the other sections to higher coefficients of lift until they reach Clmax. Once that root section stalls, the aircraft is no longer flying. Whatever degree of control and the stall behaviors depends on specifics of the wing design but it is totally irrelevant to the performance or when the wing stalled. Knowing the established conditions and measured value, in this case CINA atmosphere at sea level, gives a solid foundation and known starting point to adjust to other conditions of flight. Edited October 30, 2016 by Crump
Crump Posted October 30, 2016 Posted October 30, 2016 Yet we know it was still one of the most efficient airframes (in the real world) of the time. Messerschmitt and Kurt Tank on purpose didn’t use laminar flow wings as they knew they were most affected by surface imperfections and considered unsuitable for field use. Kurt Tank did a pretty good job of making the FW-190 somewhat impervious to surface conditions. One the design specifications was that it be able to operate for extended periods of time from unimproved airstrips at or near the front lines. This result shows that the normal painting should be regarded as adequate in terms of surface quality and that a further effort for surface improvement is hardly worth it. http://www.wwiiaircraftperformance.org/fw190/fw190-A-2_216_220.html Not only was the airfoil selected for its resistance to surface imperfections....You can see why looking from the tail of the design as much as possible is hidden behind that engine.
ZachariasX Posted October 30, 2016 Posted October 30, 2016 (edited) A degree in Philosophy is a degree in the art of argument that everything is uncertain Well the last thing i want to be is being dismissive of unreasonables argument per se. After all, he provided with his spread sheed a nice "diagnostic" tool that I was never thinking of, as I always shopped CLmax values just of the shelf for the wing root section and was happy with that. Now I can get an idea of the plausibility of real world data. So in terms of philosophy, I find this rather paractical. Besides I apreciate people literate in conversation. I was hyperventliating above much more because I treat this arument here under the premisis that everyone, and certainly Holzauge and JtD as well, know more about aerodynamics that I know. I'm here to listen to you guys. I have my opinion anyway as much as I reserve the right to change it, so no purpose in airing it here. What made it difficult for me is that some lines of argument are just leaving me with nothing, if you follow them through all the way. A mathematical proof that I'm wrong doesn't help at all; I need to know what is the sensible and productive way to do something without being fully at the mery of chance. Ah, Faust: Habe nun, ach! Philosophie, Juristerei und Medizin, Und leider auch Theologie Durchaus studiert, mit heißem Bemühn. Da steh ich nun, ich armer Tor! Und bin so klug als wie zuvor; (Ah! Now I’ve done Philosophy, I’ve finished Law and Medicine, And sadly even Theology: Taken fierce pains, from end to end. Now here I am, a fool for sure! No wiser than I was before:) Edited October 30, 2016 by ZachariasX
JtD Posted October 30, 2016 Posted October 30, 2016 In the light of your background unreasonable, yes, applied economics is nothing but religion. So you may be happy with that concept. Coming from more “exact” sciences, this makes me spit my coffee.Well, engineering isn't an exact science either. It's taking shortcuts based on other peoples empircally collected experiences. If it was science, all aircraft build for the same purpose would be the same, as would be the result of every test made on the same topic. They aren't and therefore there is in fact a correct range for a clmax, and it's a task for the developers to chose which one they think is most suitable to portrait the aircraft in game. What they like is a source, giving them a specific number (like the 1.17 they are using now), but if it is not available, they, at some point, will have to simply decide what number will give them the best result. Based on accumulated experience and knowledge, not as a result of a single process giving them the ultimate truth.
unreasonable Posted October 30, 2016 Author Posted October 30, 2016 Given what we have, unreasonables spreadsheet however does not rule out the "infamous 1.58" that at least Focke Wulf GmbH was comfortable to put in their documentation. There is good reason to assume that final "true real world values" might differ slightly from technical data as Holzauge and JtD explained. Now how much that ultimately is, most obviously difficult to say. Just deducing .2 based on one report I find drastic. After all, it reflects a lift equivalent of missing 20% airfoil between "theory" (even though that "theory" might have been derived in a wind tunnel) and real world. Now, the fraction of people that doesn't want to build aircraft might walk away happy and satisfied. Negative science (in this case disproving the 1.58 works well and is even very plausible) is good enough. The devs in turn they have to design an aircraft. THEY NEED A NUMBER, even though the aircraft is just algorithms. As they cannot insert "~1.3 to ~1.4" in their formula. We need a "correct" value. And in science, you simply don't vote on the correctness of a value. It is either correct or it isn't. Like it or not. In this sense the statement (although probably attributed to the high probability of attracting bad conversation when discussing such a topic) is very appropriate, because the correct number doesn't need lobbying. It is just correct. It however implies that one is happy not knowing CLmax. In process, one is happy having understood that basic science cannot determine some of the most fundamental properties of aerodynamics. Just want to address some of these points - The calculator does not rule out the 1.58: but it requires one particular set of assumptions to generate this figure. These assumed values are logically possible, but certainly not in the middle of the pack of what was likely, given what else we know about the test - for instance that there was no ammunition carried and that we have two numbers for the stall speed. Use these assumptions and get 1.29 as I pointed out before. Remember that when Crump used the same calculations to show "good agreement" with 1.58, all he did was assume 1.58 (weight etc) and then calculate a Vmin. As the calculated Vmin was considerable lower than than the Vmin required to make the equation balance, he assumed that the difference between the figures was the correction for PEC and compressibility. While the difference number could have been just that, or close to that, we just do not know. Had the difference number been enormous, or had a different sign, then everyone (perhaps) would agree that the empirical test was incompatible with 1.58 As it is, it is not an absolute disproof, but nor is it evidence that 1.58 is likely, as the sensitivities show. I understand how science works - as it happens my original scholarship to a certain fenland UK university was in Natural Sciences - essentially Biology. (I switched because while I am interested in the project of science, I decided I had no desire to be a scientist). Biologists are much less worried about messiness and uncertainty than most scientists: try defining a species, for instance. There is no clear rule. Interestingly, what we are doing here is in some ways more akin to biological science than physics, in that we are trying to describe a feature of a population in a general way, and may end up with a number/ratio that actually does not apply exactly to any single member of the population at all! Every single aircraft ever made has its own CLmax at each point in time. You are right that the developers have to chose just one per type to use in their model, but that does not imply that there is a "right answer". To some extent then it is a matter of chance: we might get a set of data that matches the required accuracy for the inputs, only to find later that the particular aeroplane tested was unusually rough, for instance. So we are just looking for a representative number that gives the best fit. (Preferably derived on a consistent basis across the plane set - but that is another can of worms). Finally, I still think that you are making too much of the idea that the difference between wing and plane CLmax represents any kind of problem. Aeroplane CLmax is determined empirically by measuring the plane, not by measuring the airfoil, as the NACA documents make clear (and every thing else I have read while doing homework on this subject agrees). The airfoil just sets the absolute upper limit. So I can see no reason whatever for an assumption that there should be any particular difference, or that 0.2 is somehow more of an issue than 0.02
ZachariasX Posted October 30, 2016 Posted October 30, 2016 Well, engineering isn't an exact science either. It's taking shortcuts based on other peoples empircally collected experiences.If it was science, all aircraft build for the same purpose would be the same, as would be the result of every test made on the same topic. They aren't and therefore there is in fact a correct range for a clmax, and it's a task for the developers to chose which one they think is most suitable to portrait the aircraft in game. What they like is a source, giving them a specific number (like the 1.17 they are using now), but if it is not available, they, at some point, will have to simply decide what number will give them the best result. Based on accumulated experience and knowledge, not as a result of a single process giving them the ultimate truth. But that just it. For instance airliners are built to similar specs and they differ in in the end just cosmetically. Even supersonic airliners of the 1970's produced an identical design independent from each other. Building to similar specs produces similar (successful) designs. Apart from that, I'm still wondering how I best would proceed to design a simple aircraft to given simple specs without giving myself away to chance excessively. Or is this the nature of things anyway?
Crump Posted October 30, 2016 Posted October 30, 2016 Remember that when Crump used the same calculations to show "good agreement" with 1.58, all he did was assume 1.58 (weight etc) and then calculate a Vmin. LOL... I did not assume 1.58....see this kind of wordsmithing games does not do your argument any good, unreasonable. That is what Focke Wulf measured as the CLmax after comparative study of the wing design. I simply proved that the value Focke Wulf measured is correct and fits known data on the aircraft as well as the fact landing speed = 1G stall speed in the FW-190A. Frankly, from the science POV, Focke Wulf measurement of the wing Clmax is much more "correct" than the amatuer speculation based upon inaccurate airspeed information!! Aeroplane CLmax is determined empirically by measuring the plane, not by measuring the airfoil, as the NACA documents make clear (and every thing else I have read while doing homework on this subject agrees). The airfoil just sets the absolute upper limit. So I can see no reason whatever for an assumption that there should be any particular difference, or that 0.2 is somehow more of an issue than 0.02 Not true or correct.
Recommended Posts