Aces High Bulletin Board
General Forums => Aircraft and Vehicles => Topic started by: Letalis on July 29, 2010, 04:51:01 PM
-
Came across this nice lil vid on youtube. The aircraft in the video appears to be a production F6F-3
http://www.youtube.com/watch?v=VVtVynNk6SU
Note the stall speed references starting at 2:30. Unfortunately the video does not specify test altitude, gross weight, indicated or true airspeed but the speeds seemed quite low compared to our FM.
Test flying an F6F offline at 4000 ft and 50% fuel, I found stall speeds even higher than expected after looking at the video. In fact it was impossible to get below 82 kias even after expending all ammo. (This resulted in a GW of approx 10,950lbs) Further, with the autopilot in angle mode the aircraft stagnated around 96 kts. (Which was roughly entry into slow flight.) This did not reflect the expected quickening toward stall speed as dictated by an increase in induced drag and moving the wrong way on the power curve.
The stall speeds reflected in the video seem quite different than those here: http://www.history.navy.mil/branches/hist-ac/f6f-5.pdf
On a positive note, stall characteristics appear equally docile in the video :)
Also interesting (and more concrete) is the fact flaps have no intermediate setting. I do not believe intermediate flaps were one of the mods found with the F6F-5. (Armor, cowling, spring-tab ailerons, tail structure strengthened, radio antenna etc)
Thoughts?
-
Did the video use calibrated airspeed for those stall numbers or simply the IAS?
-
Unfortunately didn't say :mad:
-
I believe we went through this a few months ago with the same video.
http://bbs.hitechcreations.com/smf/index.php/topic,283939.0.html
-
The F6F's had an infinite number of intermediate flaps. The flaps were electrically controlled and could be lowered and stopped at any position within the flap range (from 50 degrees at 93 Kts IAS to 15 degrees at 150 Kts IAS). Also they had a very nice "blow up" feature that would return them to the original position when the air speed decreased sufficiently.
And yes the current F6F-5 seems to stall "a little" (3 to 5 mph) high if you compare it to the stall chart in the POH. I suppose that could attributed to my ham-fisted flying abilities, but I think it needs to be looked at.
-
knots are not miles. FYI. You're using them as if they were.
-
I believe we went through this a few months ago with the same video.
http://bbs.hitechcreations.com/smf/index.php/topic,283939.0.html
Thanks FLS. Pitot issue on the F6F seems a good possible explanation. Given the F6F's wing loading even at low weights, I'm inclined to think the video is wrong or making some kind of omission. :salute
-
If you haven't seen it already you may find this report interesting.
http://www.wwiiaircraftperformance.org/f6f/fn322.pdf
-
Are you using this data in level flight? If you pull any G's the stall speeds will increase proportionately. So for instance, if you pull a 4G maneuver, take the square root of the force of 4g's, which is 2, multiply that by the level flight 1g stall speed. If your stall speed is 120mph and you pull a 4 g maneuver your stall speed at 4g's is 240 mph. This is why most planes auger trying to pull up at low altitudes.
Any time you turn an aircraft so that the lift force becomes more horizontal to the ground your stall speed also increases.
-
Try this and see if you can get close to the speeds on this chart. I converted knots into mph, the easiest data points are at 75 Knots and 13,000lbs (flaps up) and 15,000lbs (flaps down).
Fly it at 100 feet ASL straight and level, and once you get to the proper weight set the fuel burn to zero. Then pull back the power and decelerate to the stall speed without changing altitude, pull the nose up as needed as you approach the stall to maintain altitude.
This chart is from the F6F-5 POH 15 May 1947.
(http://332nd.org/dogs/baumer/BBS%20Stuff/F6F/F6F-5StallChart.jpg)
I have a very hard time getting close to 86mph before the Hellcat starts to drop into a good stall. Maybe it's just me, but most of the time it stalls around 90mph. Also I've tried keeping some power on (20-25 MAN) and that didn't seem to help.
-
We’ve had discussions about the F6F stall speeds in the past. Numbers from different sources vary. A snapshot from the Joint Fighter Conference in 1944 is interesting. At the conference flight tests for the stall speeds of the F6F-5 were (in knots):
F6f-5 (flown by11 Army, 2 British, 4 Navy, 10 Contractors)
Power off, clean: 65-81, 77 average
Power on, clean: 60-79, 69 average
Power off landing configuration: 55-75, 66 average
Power on landing configuration: 60-70, 65 average
Accelerated 3g: 105-150, 121 average
A couple of points can be drawn from this. First the variability in real life you can have in flight tests for the same plane. (I can’t remember if the F6F-5 was the exact same plane flown by every test pilot. It was at least the same configuration I believe.) Numerous real life factors play a part. In general comparing data from different flight tests can be problematic because the numbers tend to rarely line up exactly.
Second one of the causes of variability that effect stall speed is testing technique. Once upon a time we had a BBS discussion on this and how stall testing technique will result in different numbers in real life as well as in AH. I believe a key factor that could account for the lower stall speed numbers referenced in the film is due to the F6F’s nose attitude above horizon. In the film they demonstrated various stalls of the F6F. Here is a frame from the film just before the F6F enters a power-on stall in clean configuration.
(http://thetongsweb.net/images/f6fangle.jpg)
In the picture notice the nose angle (attitude) above the horizon labeled as gamma. My rough guess is that’s a ~30 degree angle for gamma in the video. This angle can greatly impact stall speed results. Why? Because it changes the amount of weight, thus lift that wings must support. Here is a free body diagram that demonstrates this.
(http://brauncomustangs.org/images/fig119a.jpg)
Notice that Lift = Weight * cosine ( gamma)
At level flight gamma=0, thus Lift=Weight since cosine(0)=1. However as nose angle (gamma) increases the amount of weight that lift must equal reduces based on the vector geometry. The greater the angle, the lower the stall speed will be because the airplane doesn’t have to produce as much lift to keep the airplane flying. In other words since lift is a function of lift coefficient and airspeed a plane can fly at Clmax at a lower velocity because the amount of lift needed to maintain flight is lower due to increased gamma.
Pyro posted some stall tests of the F4U that demonstrates this in action. The following are snapshots of power-off stalls where nose angle gamma was varied. The F4U was loaded at 12,904 lbs
Power off stall gamma = 7 degrees, weight=12,904 lbs.
(http://hitechcreations.com/pyro/7stall01.jpg)
(http://hitechcreations.com/pyro/7stall03.jpg)
Just before stall notice that the lift is 12,063 lbs, almost 850 lbs less than the 12,904lbs of weight. Stall speed is just about 103 mph.
Power-off stall at gamma=20 degrees, weight=12,904 lbs.
(http://hitechcreations.com/pyro/20stall01.jpg)
(http://hitechcreations.com/pyro/20stall03.jpg)
Just before stall lift=10,465 lbs, a full 2439 lbs less than the full weight. Stall speed is 98 mph.
The point here is that for the same airplane 1g stall speed varies depending on the amount of angle gamma above the horizon. In our example at 22 degrees stall speed is 98 mph vs. 7 degrees at 103 mph, a difference of 5 mph.
To conclude, one of the factors that could play a noticeable factor in variations of stall speeds for the F6F in question is the technique used to test stall speed, specifically the angle of the nose above the horizon (gamma).
Tango
-
Excellent writeup dtango, and (as usual) you bring up several interesting points.
Of particular point of interest to me is your discussion of the joint fighter conference report. I have a copy of the report and the testing information is what got me really looking into this. As I read the report all tests were done in the same aircraft, unless it had mechanical issues during the Conference.
I had noticed from my testing that stall speed did very depending on how well (smoothly) the entry was flown. That's why I tried to be as specific when describing how I did the stall test. It's also interesting to compare some of the other stall speeds given in that report to the other planes we have in game. I was able to get lower stall speeds (7 to 10 mph) in the F4U-1C than whats listed in the POH or the Fighter Conference Report. Given how well HTC has modeled things, I have usually just chalked up this difference to my flying rather than a real issue. But it does seem odd to me that I can do the same stall test so much better in the Corsair than the Hellcat.
Just for a little added testing here are the numbers fro the F4U-1C from the Joint Fighter Conference;
F4U-1C (flown by; Army-13, British-3, Navy-4, Contractors-8) so a similar number and mix to the F6F-5 tests
Power off, clean: 65-88, with an average of 82 (5 knots above the Hellcat)
Power on, clean: 60-83, with an average of 76 (7 knots above the Hellcat)
Power off landing configuration: 63-90, with an average of 74 (8 knots above the Hellcat)
Power on landing configuration: 63-84, with an average of 70 (5 knots above the Hellcat)
Accelerated 3G: 130-190, with an average of 150 (29 knots above the Hellcat)
I know from talking with Badboy the 3G tests are not really accurate given how the tests were probably performed.
-
One of the things I've found was a copy of the US Navy Fixed Wing Performance Flight Test Manual. It has a very interesting chapter, Stall Speed Determination (Chapter 3), discussing the definition of "stall" first, then techniques for testing it. I'll include a few excerpts:
The stall speed investigation documents the low speed boundary of the steady flight envelope of an airplane. In the classic stall, the angle of attack (a) is increased until the airflow over the wing surface can no longer remain attached and separates. The resulting abrupt loss of lift causes a loss of altitude and, in extreme cases, a loss of control. The operational requirement for low takeoff and landing airspeeds places these speeds very near the stall speed. Since the stall speed represents an important envelope limitation, it is a critical design goal and performance guarantee for aircraft procurement and certification trials. Verifying the guaranteed stall speed is a high priority early in the initial testing phases of an airplane. The significance of this measurement justifies the attention paid to the factors which affect the stall speed...From a test pilot’s perspective, the task is to investigate how much lift potential can be exploited for operational use, without compromising aircraft control in the process. The definition of stall speed comes from that investigation.
(emphasis added)
While the wing is normally a predominant factor in determining minimum speed capability, the maximum lift capability frequently depends upon thrust and center of gravity (CG) location. Thrust may make significant contributions to lift through both direct and indirect effects. The location of the CG affects pitch control effectiveness, pitch stability, and corresponding tail lift (positive or negative lift) required to balance pitching moments. These effects can be significant for airplanes with high thrust to weight ratios or close coupled control configurations (short moment arm for tail lift).
The speed corresponding to CLmax may not be a reasonable limit...other potential limitations...may prescribe a minimum useable speed which is higher than the speed corresponding to CLmax. The higher speed may be appropriate due to high sink rate, undesirable motions, flying qualities, or control effectiveness limits...Any of these factors could present a practical minimum airspeed limit at a lift coefficient less than the CLmax potential of the airplane. In this case, the classic stall is not reached and a minimum useable speed is defined by another factor.
Note: This is why I prefer Badboy's Bootstrap approach to determining stall speeds in game. Since you have to maintain the turn, you're basically discovering the minimum controllable airspeed versus the classic stall speed. For our purposes of comparison, its a better number to use, and from a safety standpoint in actual aircraft, a smarter number to use. As long as the landing gear and structure can take the force of the higher landing speed, what's a better landing speed--1.2 times Vs or 1.2 times Vmincontrol? Obviously if you can maintain control all the way through the aerodynamic stall speed, a la Cessna 172, you can use them interchangeably. In my Grumman, the plane wanted to depart at Vs so using it at 1.2 times Vs wasn't quite as comfortable.
The definition of stall airspeed is linked to the practical concept of minimum useable airspeed. Useable means controllable in the context of a mission task. The stall speed might be defined by the aerodynamic stall, or it might be defined by a qualitative controllability threshold. The particular controllability issue may be defined precisely, as in an abrupt gbreak, or loosely, as in a gradual increase in wing rock to an unacceptable level. Regardless of the particular controllability characteristic in question, the stall definition must be as precise as possible so the stall speed measurement is consistent and repeatable. Throughout the aerospace industry the definition of stall embraces the same concept of minimum
useable speed
(emphasis added).
Now, this is a current document. I've got a feeling that this hadn't been widely accepted at the time of that conference, at least not industry wide since they were still developing then what has become standardized aerodynamic theory/practice now. Obviously from the portion I emphasized, they didn't have a definition precise enough to keep from having a +/- 10 knot threshold among the pilots. Also, what was the manner in which the stall airspeed was recorded? Did the pilot just write a note after he recovered of what he saw on the analogue ASI? No room for error there... :)
Anyway, just a few observations...
-
All very salient points Stony, and I completely agree that we should have a clear definition of what it means to "Stall" in Aces High.
My main concerns with using Badboy's approach are, repeatability, and accuracy of timing. This is more of my own hesitation to use data I've generated with his approach than anything else. I know that the math is completely correct, but my sloppy flying can generate a wide range of data points.
Also, having been more involved with modifying game controllers lately, I feel that different controls (x52/CH/etc.) on the same PC can generate different flight performance. This is not a fault with Aces High, but merely a point that various equipment setups can generate different outcomes.
-
Check out dtangos data. The climb rates are all over the board. If there is any climb or decent rate the aircraft is not in a 1g state. I would try the test again and note the speed the aircraft is unable to maintain a 1g neutral state (ie the speed in which the aircraft starts to decend) and use that as your stall definition. Noting the point at which you lose lift is more repeatable than observing control surface performance, buffeting ect.
I would immagine as far as the f6f goes, knowing the speed the aircraft is unable to maintain 1g would be most valuabe when landing on an aircraft carrier. There would be little room to correct for any errors to prevent hitting the stern of the ship or pancaking on deck.
-
Check out dtangos data. The climb rates are all over the board. If there is any climb or decent rate the aircraft is not in a 1g state.
You can be in a climb or dive at 1g. An unaccelerated steady climb is at 1g.
-
I see your missunderstanding, but no, 1g is one gravity or a neutral state, where the lift component has the same force as gravity. Im at 1g just sitting in my chair. If you change any portion of the lift component your not at 1g anymore. A 0g state is anything in free fall. The same thing also applies to the thrust and drag components. Look at it this way. If your climbing at 1500 ft. per minute you are actually accelerating away from the force of gravity at 1500 ft. per minute.
So in your test, if you put the plane in a climb your actually increasing the stall speed do to the fact that your changing the airplanes 1g state. If your stall speed is 75mph and your plane is climbing at 4g's(for ease of this example lets say), your stall speed would increase to the square root of 4, which is 2 times your 1g stall speed of 75mph. The plane will stall at 150 mph.
In the case of a 1g stall in level flight, whats really going on is the slower the airspeed the more the wing has to change its angle of attack to create the lift nessesary to counter the gravitation force its opposing. Eventually the flow of air will separate from the airfoil and that is the stall. Your not stalling the plane, your stalling the wing. Dont confuse the angle of attack with the planes attitude (compared to the horizon). The planes attitude doesnt even factor. You can have a combination of the four vectors at various speeds and create the same effect. For example. Fly strait then turn the plane on its side. It can remain at that altitude but now your lift component is being maintained by the rudder and fusalage not the wings. The planes attitude is now perpendicular to the horizon, no effect on the 1g state. The plane is currently on its side flying level, now pull the stick back, now your changing the thrust and drag components but not lift and gravity. The airfoil doesnt know the difference and will stall the same as if it where horizontal.
Hope this helps.
-
Not to sound overly contrite, but 1G is 1G no matter what your orientation is. If your accelerometer is aligned to measure in the direction of motion and it registers 1G and is not changing then you are not accelerating.
-
Not to sound overly contrite, but 1G is 1G no matter what your orientation is. If your accelerometer is aligned to measure in the direction of motion and it registers 1G and is not changing then you are not accelerating.
Yep, you are absolutely correct. Im making more of this that I should be. Im trying to make the point that his climb rate is not constant in any of the tests that where done. Try the test while maintaining a 0 rate of climb for accuracy. If im flying level getting close to a stall, if I start to climb (start jerking the stick) I am changing the dynamic. I'm going to pull +1g's and that is going to affect the stall speed. The same is true if I let the plane decend. Im creating and accelerated stall not a normal stall.
Normal stall speeds are always determined from strait and level flight.
Just trying to help out.
-
Yep, you are absolutely correct. Im making more of this that I should be. Im trying to make the point that his climb rate is not constant in any of the tests that where done. Try the test while maintaining a 0 rate of climb for accuracy. If im flying level getting close to a stall, if I start to climb (start jerking the stick) I am changing the dynamic. I'm going to pull +1g's and that is going to affect the stall speed. The same is true if I let the plane decend. Im creating and accelerated stall not a normal stall.
Normal stall speeds are always determined from strait and level flight.
Just trying to help out.
#1 They weren't my tests. They were tests done by Pyro.
#2 I'm not the one that's misunderstanding it. It's you :). You can absolutely be at 1g in a climb or a dive. Infact for normal steady climbs that's assumed. Maybe someone else has the energy to explain it to help you work through the kink in your physics. I can't think of a pithy way of helping you work through your thinking except to say that if you were at a load factor greater or less than 1g then your flight path would be curved and not straight.
#3 "Normal" stall speeds are determined a various number of ways, not just level flight.
Cheers!
Tango
-
#2 I'm not the one that's misunderstanding it. It's you Smiley. You can absolutely be at 1g in a climb or a dive. Infact for normal steady climbs that's assumed. Maybe someone else has the energy to explain it to help you work through the kink in your physics. I can't think of a pithy way of helping you work through your thinking except to say that if you were at a load factor greater or less than 1g then your flight path would be curved and not straight.
Not sure what you thinking is today dtango, I'm surprised you seem to be missing this one. I may be wrong but I believe load factor is lift / weight of plane. IF this is the case then in any non accelerated flight, I.E. same speed and not turning load factor is 1 only in level flight. Flying straight up or down it is 0, and more specifically it is the cosine of the climb or decent angle not accounting for any engine incidence or AOA's.
HiTech
-
May be semantics, but load factor is a dimensionless "g" that isn't supposed to be used interchangeably with the gravitational "g", right?
-
Not sure what you thinking is today dtango, I'm surprised you seem to be missing this one. I may be wrong but I believe load factor is lift / weight of plane. IF this is the case then in any non accelerated flight, I.E. same speed and not turning load factor is 1 only in level flight. Flying straight up or down it is 0, and more specifically it is the cosine of the climb or decent angle not accounting for any engine incidence or AOA's.
HiTech
Well, I'm full of surprises I suppose :D, sometimes wrong ones! Yes you're right HT. I mixed up load factor with the fact that in an unaccelerated climb or dive forces normal (perpendicular) to the flight path cancel each other out just like they do in level flight. So my original response about 1g climbs & dives to jamdive was incorrect.
What I meant to say is that in an unaccelerated climb or dive there is no net acceleration normal to the aircrafts flight path which means lift = weight * cosine (climb_angle). In other words the amount of lift equals the amount of weight normal to the flight path just like it is in level flight. The stall tests performed in the F6F video in this thread appear to be in this climbing flight mode and thus because of the effect of climb angle gamma could explain why the stall speeds quoted are as low as they are.
Cheers,
Tango
-
May be semantics, but load factor is a dimensionless "g" that isn't supposed to be used interchangeably with the gravitational "g", right?
The dimensionless "G" (lift/weight) is usually used with a capital. Lower case "g" is usually used for the gravity acceleration.
p.s.
How does a G meter on a plane actually work? I guess it measures acceleration in a given axis set to be aligned with the lift.
In that case, when pointing at an angle "a" above the horizon, the acceleration in the measured axis is:
L/m-g*cos(a)
and the G load it will show is:
G=L/(m*g)+(1-cos(a))
Since in a stead climb L=mg*cos(a):
G=cos(a)+(1-cos(a)) = 1.
So the G meter in that case will show 1.
-
This is another excerpt that may or may not have any bearing:
Deceleration rate has a pronounced affect on lift coefficient. Changes to the flow pattern within 25 chord lengths of an airfoil have been shown to produce significant nonsteady flow effects. The lift producing flow around the airfoil (vorticity) does not change instantaneously. During rapid decelerations the wing continues to produce lift for some finite time after the airspeed has decreased below the steady state stall speed. The measured stall speed for these conditions is lower than the steady state stall speed. For this reason, a deceleration rate not to exceed 1/2 kn/s normally is specified when determining steady state stall speed for performance guarantees.
Could be that during the stall speed testing, this wasn't being accounted for?
-
The dimensionless "G" (lift/weight) is usually used with a capital. Lower case "g" is usually used for the gravity acceleration.
p.s.
How does a G meter on a plane actually work? I guess it measures acceleration in a given axis set to be aligned with the lift.
In that case, when pointing at an angle "a" above the horizon, the acceleration in the measured axis is:
L/m-g*cos(a)
and the G load it will show is:
G=L/(m*g)+(1-cos(a))
Since in a stead climb L=mg*cos(a):
G=cos(a)+(1-cos(a)) = 1.
So the G meter in that case will show 1.
Didn't run threw your math (sorry lazy today), but you must be missing something, if the meter is positioned to measure force in the lift axis, then straight up the meter would show zero G regardless of acceleration. If you wish I can test it some time this week in the RV.
With a quick scan it appears you are measuring earth relative instead of object relative? I.E. you are saying the G meter measures mass/gravity?
HiTech
-
Didn't run threw your math (sorry lazy today), but you must be missing something, if the meter is positioned to measure force in the lift axis, then straight up the meter would show zero G regardless of acceleration. If you wish I can test it some time this week in the RV.
With a quick scan it appears you are measuring earth relative instead of object relative? I.E. you are saying the G meter measures mass/gravity?
HiTech
Ignore my previous post, that is an error... (sorry, I cannot remove it now)
I forgot to add the earth's gravity along the tube which is not 1g but g*cos(a). This cancel the g*cos(a) in the net acceleration of the plane frame of reference L/m-g*cos(a) and all that is left is G=L/(m*g) aways. For no net accelerations (L/m=g*cos(a)) it means G=cos(a) of the climb angle.
The simplest device I can think of is just a spring measuring the weight of a ball that can slide along a tube. Zero G would be the 0 point of the spring. 1G would be the point of spring load equals m*g of the ball. 2G is 2*m*g, etc. In a level flight, the tube is vertical to the ground and the ball weighs m*g against the spring. In a free fall, or a vertical climb (horizontal tube) the ball will apply no force on the spring and the device will read 0G.
I have no idea if this is how it works in practice.
-
What ww2 aircraft had an actual g-meter? I don't ever remember actually seeing one in any cockpit photos or museum exhibits.
-
No G meter in ww2 planes.
However you feel G forces in a plane flying it , you don't in a sim without motion
-
p.s.
How does a G meter on a plane actually work? I guess it measures acceleration in a given axis set to be aligned with the lift.
Although everything is digital today and MEMS accelerometers are commonly used, back in the 40s and 50s, they used strain gauge accelerometers. Essentially, a beam with a mass on the end of it. If the accelerometer (or aircraft) is dropping at the rate of gravity (32.2 ft/sec/sec) the relative g loading goes to zero and is indicated as such on the g meter (often referred to as "zero g, absolute"). Pilots will "unload" their aircraft to zero g to increase acceleration.