Post by Tom/CalClassic on Aug 9, 2008 0:16:59 GMT -5
Dear Tom,dear FSAviator
Simple but probably lengthy question
If I get the FDE somewhat correct(weights,power areas etc) and have the correct wing profiles etc in the air file,what defines
A:The planes stall limit at high speed
B:How to alter
OR is there any tutorial(I found none) which describes this?
Thanks in adavance
Godspeed
Thunder100
PS:It is for a single prop fighter,not a 4 engine liner
IP Logged
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #1 on: Dec 5th, 2006, 10:41am » Quote Modify
--------------------------------------------------------------------------------
Hi,
And here's your answer from FSAviator:
Q. > Is there a web flight dynamics tutorial (I found none).
There is no web tutorial that comes close to explaining everything that needs to be understood to encode a realistic air file. The subject is too complex and requires too many diagrams for clarity.
MSFS uses real world physics throughout so the relevant tutorials are available from every bookshop. However even FD authors who normally think metric should perhaps seek textbooks published in the United States expressed in Imperial units. It may be less confusing in the long run. Unfortunately if there is a text book using Imperial units, still in print, that covers flight + engine + airscrew dynamics adequately in a single volume I have never seen it.
There is a single metric textbook that covers most of the ground. Mechanics of Flight by A.C. Kermode (ISBN 0 273 31623 0). It runs to over 500 pages with maybe 300 diagrams. It also provides sample aerofoil lift slope data as required by Table 404 of the air file, which is probably at the root of the specific question asked.
The structure of the air file, (i.e. where the concepts in textbooks are stored within FS9), is explained in the help files of freeware air file editors, about as clearly as it explained anywhere else, but be warned many of the help comments within air file editors or spread across the web explain how very old versions of MSFS and CFS worked and do not apply to FS9.
Q. > in the air file what defines the planes stall limit at high speed and how is it altered
I am not sure that I have understood the question asked, but assume you mean how is G load taken into account.
Aeroplanes stall when the wing reaches the angle of attack (AoA) at which the airflow over the wing becomes turbulent. This happens when the lift required to be in equilibrium with (equal to) the current weight is more than the wing can produce.
Current weight is current mass multiplied by current G (loading).
As weight varies the wing varies its AoA to produce current lift = current weight sustaining equilibrium. The fuselage has no choice but to pitch up and down with the wing. If AoA becomes excessive, as the wing attempts to achieve Newtonian equilibrium with current weight, the wing stalls. Stalling 'V speeds' such as Vs, Vs0, etc., denote the drag (IAS) associated with one G stall, but it is current angle of attack, not drag (IAS), or velocity (TAS) that controls the propensity to stall.
The stalling angle is defined in TBL 404 of the air file. Within TBL 404 the entire lift slope is described point by point with Co-efficient of Lift plotted versus *fuselage* AoA expressed in radians.
Fuselage AoA equals wing AoA minus mean wing incidence.
Mean wing incidence equals wing incidence plus or minus mean wing twist.
The associated comments in an aircraft.cfg are just REM statements recording the data used by the FD author to calculate the mean incidence offset to be used in TBL 404, or they can be set to zero since they are not read by FS9.
After making those calculations the flight dynamics (FD) author encodes all of the real lift slope for the specific aerofoil within TBL 404 of the specific air file. Else he or she must estimate it. It is the invariant Microsoft flight model (FM) inside fs9.exe and inside sim1.dll that calculates current weight as G varies and thus calculates the propensity to stall.
The aeroplane stalls when the current weight causes the current (fuselage) angle of attack to exceed the stalling angle (in radians) specified for CLmax within the encoded TBL 404 lift slope. The stalling angle is the fuselage AoA that the FD author associated with CLmax in TBL 404. The loss (rate) of CL between that angle (for CLmax) and the next higher encoded fuselage angle of attack in TBL 404 determines the severity of the stall.
If the aircraft has automatic slats the CLmax and fuselage AoA encoded must be for the open status with the closed status encoded as the prior matched data pair within TBL 404.
FS9 simulates post stall behaviour poorly and inaccurately by default, but altitude loss per second due to stall can be imposed at any desired rate fairly easily by the means above and stall propensity versus G is as accurate as the aerofoil specific lift slope data encoded within TBL 404.
All of which probably goes a long way to explaining why there is no web tutorial explaining everything, or even a significant fraction of the things, that an FD author must know how to calculate or estimate outside FS9 and where to subsequently store his results within FS9.
FSAviator 12/06
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
thunder100
Connie
If it looks good,it flys good(usually)
Gender:
Posts: 166
Re: Question to FSAviator on Stallvariables in FDE
« Reply #2 on: Dec 5th, 2006, 1:46pm » Quote Modify
--------------------------------------------------------------------------------
Dear Tom(if you could transfer)
Dear Fs Aviator(if you read this)
Thank you very much
the threat does not belong here,yes I know very well,but there is no other source to get the help by FS Aviator as far as I know.
Issue is:
I bought WOP Bf-109 E which stall's in deliverycondition at any speed above 130 knots at 4,5 G both positive and negative G's.
This is not correct and I failed yet to get a WOP support forum approval (5 days) although I did nothing wrong.Anyway
Why wrong?
I have a friend (90 years) which was Bf-109 pilot in WW2(and FW-190,Ju-88,He-111,Ju-52).He was also at one time instructor for Roumanian airforce during the war.He says that the real planes stall was 3 g low speed 6,5 g at mid speed and 4 G at high speed ,which he could read from a G-meter in his instructor plane.And at negative G --> 3 G was really max.
I am 48 and a graduated mecahnical engineer as well as aerodynamics(although I only used it for engine fuel/air airstreams),so I can really understand what you are saying and I will now have a deep look into the airfile to repair this by try and error till it is close to reaility(simple to learn about it).Also I had help about the real fighter behavior on the forum by a PM to Jim D and Jesse.So thank you again,you are a big knowledge source to FS community
So now I close it,too much words for an Off Calclassic topic anyway ,but THANK YOU VERY MUCH AGAIN
Godspeed
Thunder100
aka
Roland Berger(Vienna/Austria)
IP Logged
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #3 on: Dec 18th, 2006, 7:38pm » Quote Modify
--------------------------------------------------------------------------------
Hi Roland,
And here's your reply from FSAviator:
Tom has assured me that he does not object to a follow up post on this topic. I will not be drawn into reviewing individual air files released by third party developers. What follows relates to a real aeroplane or FS9 flight dynamics in general.
I will start by illustrating use of the relevant laws of dynamics, and then explain why they matter in real life, and why FS9 users need to understand the consequences of Newton’s Laws even if they do not manage to understand the equations. I will use the Bf109G as a worked example, but everything that follows also applies to propliners. Combat strategies used to illustrate the relevance of the dynamics discussed are generic and simplified.
First the equations;
According to the RAF test pilot report of a captured aircraft Bf109 CLmax was about 2.34 with Lachmann auto slats open, probably achieved at about 0.35 radians fuselage AoA. Maybe as low as 1.44 with auto slats closed. The wing area was 173.3 square feet and the following data assume (combat) mass = 6950 pounds.
To calculate gear up and zero flap 1G stall (Vs) (in Imperial units = feet per second as required by Microsoft) we proceed as follows;
1) half the density of air at sea level (measured in slugs is always 0.00119) * CLmax * wing area (measured in square feet)
Result 1 = 0.00119 * 2.34 * 173.3 = 0.483
2) Divide the (combat) mass by Result 1 above
Result 2 = 6950 divided by 0.483 = 14390
3) Determine the square root of Result 2
Vs = SQR(14390) = 120 feet per second
4) Convert to KCAS (ASI reading if no pitot tube position or other errors) using 1.68 divisor
120 ft/sec divided by 1.68 = 71.4 KCAS = auto slat open stall at 1G with mass = 6950lbs = Vs
5) Most (but not all) V speeds including Vs vary with square root of current weight and weight is just mass * G.
Vs at 4G = 71.4 * SQR(4) = 143 KCAS
6) I believe that max safe G to avoid structural failure of the Bf109G tail was + 6.0
Vs at 6G = 71.4 * SQR(6) = 175 KCAS
Of course the RAF test pilot included all those data in his test report, but we rarely have a full flight test report for an aircraft we intend to fly in FS9. The only way we can obtain those handling data is to calculate them from the data in the supplied flight dynamics, whether the supplied dynamics are realistic or not.
Now let’s explore why the data emerging from the equations above matter to FS9 users.
In FS9 the aeroplane will have exactly the dynamics specified within the FS9 FD and we must fly it accordingly. For use within FS9 it may be more useful to convert Bf109G data to metric Kph rather than imperial Knots. However in what follows I will assume zero pitot position error, zero ASI gauge error, zero compressibility error, and talk about KIAS for ease of comparison with existing propliner threads and propliner handling notes.
Contrary to popular belief aeroplanes are not kept aloft by the lift from their wings. They are kept aloft only by the power from their engines. We must never forget that any aeroplane requires cruise (horse) power just to maintain current altitude at one G. Even cruise power is a lot of horse power. Gliders do not remain aloft in still air.
When we pull 6G the Bf109G still has a mass of 6950lbs but it weighs almost 42,000lbs. It has nowhere near enough power to cruise at that weight. Pulling 6G will stall the Bf109G unless we (spiral) dive hard enough to keep the drag above 175 KIAS.
The minimum drag required to sustain the G load equal to the safe structural limit is the called 'corner' because we can 'corner' hardest under that condition. At 175 KIAS the instant we pull 6G we stall. We need to dive and keep (spiral) diving to keep the drag above 175 KIAS in order to sustain 6G. To pull 4G we only need to keep our drag above 143 KIAS, since our Bf109G 'only' weighs about 28,000lbs.
At least that is true in perfectly smooth air. Any turbulence would push the wing through the stalling angle so in real life we need to (spiral) dive hard enough to keep the drag a decent margin above the relevant stall limit (including corner) if we wish to generate a max G turn, whether to avoid an enemy fire control solution, or to obtain a fire control solution. In normal weather conditions we might target 180 KIAS, but more than that in turbulent air.
A Bf109G will stall at any G load. One part of the FS9 FD author's job is to make it stall at the correct drag (IAS) for that G load and to punish pilot error accordingly. I explained that process in detail in my previous post.
However there is another issue. FS9 users often confuse pre stall sink with post stall sink when turning hard.
On final approach we reduce power and IAS. Suppose we close the throttles completely in an aeroplane with no flaps such as the Boeing 247 and trim elevator up to target a lower drag. The drag reduces, angle of attack increases, the nose comes up and the Boeing 247 sinks towards the runway nose up, but under full control. However if we pull back on the yoke and try to maintain altitude with less than cruise power applied any aircraft can only stall.
Whenever we reduce power below cruise power an aeroplane sinks. It will sink whether nose up or nose down, whether turning or wings level. Like a glider it can only sink if it lacks cruise power in still air. It has not stalled, but it may nevertheless sink fast.
When turning hard and pulling hard on the joystick to pitch an aircraft round a steeply banked high G turn FS9 users sometimes become confused. The Bf109G cannot cruise at 42,000lbs it can only sink. It does not have enough power to cruise at 42,000lbs at any drag (IAS). By pulling 6G we have decided to make the weight more than any applied power can possibly support. We have decided to sink hard. This does not mean we have stalled. If we stall we will not be able to sustain 6G. After we stall max G is 1G. So long as we can sustain more than 1G we are not stalled, but we will sink very fast regardless. We must spiral dive in order to sustain high G turns. Even briefly available War Emergency Power (WEP) from a Daimler Benz 605 engine will not support anything like 42000lbs in level (i.e. cruising) flight.
When pulling high G, to create very high aircraft weights, to improve instantaneous turn rate, very high rates of involuntary descent are certain, but do not imply stall. They imply only the dynamic certainty of sinking hard at that high weight, just as applying less than cruise power for approach weight implies the dynamic certainty of sinking down the approach.
But there is a third much more complex issue which few FS9 users grasp.
The force of air on the elevators must be countered by the arm muscles of the pilot; actually by just his right arm. His left arm must be operating the elevator trim wheel, the throttle, the boost override switches, the war emergency power switches, the water methanol coolant arming switches, the reflector gun sight, and the rpm lever. Pulling our plastic joystick full aft with our right arm takes little effort. Not so for the real Bf109G pilot.
In flying clothing his mass is 200lbs. At 6G his weight is 1200lbs. His right arm weighs almost 200lbs. His heart is failing to pump enough blood to his vital organs. First it shuts down blood supply to his arm muscles; then to his eyes. He is still fully conscious. His brain still has enough blood. He feels the pain. He is a fit fighter pilot. Even without a G suit he may not suffer GLOC until 6.2G, but at 6.0 he has no blood supply to his arms, or his eyes.
So you are in the gym with a tourniquet round each arm so that your arm muscles receive no blood. You are sitting down wearing a flying suit that has 700 pounds of lead weights sewn into it. You have a 200lb weight hanging from each elbow and a 100lb weight on top of your head. Now all you have to do is hold your right arm out and pull say 150 lbs continuously with your right arm on a lever whilst finely balancing the trim with your left arm with your eyes closed (blacked out) but not quite unconscious (GLOC). How long can you do this?
If FS9 users have never weighed 1,200 pounds in real life they have no idea how difficult it is to work really hard doing something really complex when they do.
What most FS9 users fail to grasp is that the FD author also has to code the consequence of all of this in the air file. The FD author models pilot surplus strength versus elevator loading. He writes the curves that limit what the pilot can do as well as what the aircraft can do. He treats the real pilot as just another aircraft system which has design limits. To obtain 6G the pilot must pull the elevators up into a drag of greater than 175 KIAS but there is always some higher drag that prevents the pilot from pulling them up far enough to generate 6G, with neutral trim, and with each other (incorrect) trim setting the IAS he cannot pull against is different.
continued in next post...
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #4 on: Dec 18th, 2006, 7:39pm » Quote Modify
--------------------------------------------------------------------------------
Most FS9 users don't grasp that. They don't trim for the target drag (IAS) and the air file curves written by the FD author calculate that the real pilot has insufficient surplus strength remaining to sustain the necessary pull in the out of trim condition. Many FS9 users mistake this for loss of control due to stall, but it is loss of control due to fatigue, due to failure to trim for the target IAS.
To sustain a high G turn it is necessary to trim into the turn until there is no back pressure required on the joystick at the correct target drag (IAS) for the high G turn. All complex aeroplanes need to be flown by the numbers and the Bf109G is no exception. For a Bf109G trimming for around 180 KIAS (just above corner) is about right for fully engaged air combat in still air. Most FS9 users fail to do this.
Before we dive to bounce the enemy, (or to prevent an enemy from maintaining a tally on us), we may have (cruise) drag in excess of 180 KIAS in a Bf109G. Even though we intend to dive (increase drag = IAS) we must trim for the drag (IAS) we intend to target in the next combat turn, whether engaging offensive or engaging defensive. This may be less than cruise drag (IAS) and is always less than dive drag (IAS) so we will often need to adjust elevator trim upward before the dive so that we are trimmed for 180 KIAS which will be our target IAS once engaged with the enemy. We will normally enter the dive by rolling nearly inverted and pulling through working with the applied target = 175 KIAS trim.
This has nothing to do with fuel flow under negative G. Daimler Benz 605 engines are fuel injected. In combat we may need to pull 6G, and we must be trimmed for about 180 KIAS to do that. If we are trimmed for 180 KIAS and we decide to pull only 4G we, (the real pilot encoded by the FD author as an aircraft system), are bound to have enough strength, but if we trim for more drag than 180 KIAS we may not have the strength to fight that out of trim condition trim to 6G, or even to 4G.
Propliners are designed so that elevator trim is neutral for design cruise drag at mid cruise weight. Fighters are designed so that elevator trim is neutral at corner drag at design combat weight.
With trim neutral the pilot of a real Bf109G could not muster the strength to apply 6G when the drag on the elevators exceeded about 300 KIAS and he could not keep his left arm in position to operate the trim wheel at 6G either. To max rate turn in a Bf109G drag must be kept between 175 KIAS and 300 KIAS with the aircraft trimmed for about 180 KIAS (normal weather = little turbulence). However the radius of turn is very much less at 180 KIAS since radius of turn varies with the *square* of the IAS (at constant altitude).
The radius of turn is three times worse at 300 KIAS compared to 175 KIAS. We must trim for a little more than corner drag and target a little more than corner drag throughout the time that we are fully engaged.
We must avoid stall when turning hard by pitching the nose to generate more than 175 KIAS and we must also pitch the nose to avoid freezing of the controls above 300 KIAS. We must go nose high or nose low to sustain a little more than corner drag whatever VSI that delivers (terrain permitting!). We change target IAS from corner only to engage offensive during acquisition of a fire control solution, but we remain trimmed for corner. When we engage defensive we turn max (structural G limit) rate just above corner drag = 175 KIAS, nose up or nose down, (terrain permitting!). We do all of this with WEP or MIL or whatever lesser power is currently available.
Real fighters are often, (but not always), designed so that their elevator authority matches corner drag (Va = Corner). Applying full up elevator above corner IAS causes the real tail to break off. However because MSFS is a fifty dollar flight simulator with a 50 cent flight model structural failure will occur with full aft stick at corner IAS if the aircraft is trimmed for less than corner drag; if elevator authority was coded as realistic. Since most MSFS users fail to grasp these essentials most FD authors code elevator authority to be much less than realistic to protect naive users who believe that full up elevator can be applied at any IAS and however badly out of trim the aircraft may be when they apply it.
When fully engaged in a Bf109G we trim neutral. We never pull hard enough to bleed drag to the 6G stall IAS = 175 = corner, let alone the 1G slat open stall drag = 72 KIAS, unless to obtain a momentary snap shot fire control solution when we are certain that no one is doing the same to us. When fully engaged we must not become obsessed by our potential target. We are always somebody else’s potential target and we must be able to fly a 6G break to foil their fire control solution, so we must be trimmed for it, and we must always have enough drag (IAS) for it.
Air combat must be a series of drive by shootings, never a close in knife fight. Knife fights let the other guy have a chance of striking a lucky blow. If we cannot organise a drive by shooting right now then we are the potential target for a drive by shooting by the enemy and we need to disengage. Disengaging does not involve going round and round in tight circles. We disengage by seeking cloud cover, or by causing the enemy to lose his tally on us (whether visual or radar) by other means.
If we cannot make the enemy lose his tally on us, provided no enemy aircraft is inside gun parameters, we must target IAS = Vy. The multiplier from Vs to Vy depends mostly on wing aspect ratio. It might be Vs * 1.5 in a typical fighter so maybe 110 KIAS = best climb rate in a Bf109G.
During phases of disengagement we must try to climb above all threats. If the engine overheats we must target a higher drag (IAS) to cool the radiator better. Engine cooling is proportional to IAS whether engines are 'air cooled' or 'liquid cooled' because the liquid coolant is air cooled in the radiators anyway. We will not reduce boost (MAP = Ata). The highest pilot in a (piston engine guns only) combat has the initiative. He can drive by again, or just drive away.
During these periods of disengagement we will be weaving and circling gently to keep all the potential threats and potential targets in sight and clear our wingman's tail. He is clearing ours. We are barely pulling any G as we weave uphill at Vy or a little more for adequate engine cooling. If the enemy is attempting to drive away we will pitch the nose to corner and follow (climbing if possible) at IAS = corner. We will exceed corner only to maintain our tally on the enemy and only if he is nearing the limits of visibility. We will cease to climb only when we reach the rated altitude for our current power setting, or when we are above all enemy aircraft on which we have a tally. Then we will pitch the nose level to target Vmax for our current power setting. We are always seeking to maximise specific energy. Once we reach the dynamic condition in which it is safe to maximise our kinetic energy (velocity = TAS) we cease to maximise our potential energy (altitude).
If at any time when we are disengaging an enemy is threatening to achieve a fire control solution we must reduce nose pitch and increase drag to corner. When the next enemy drive by starts we must max rate turn *towards* the attacker at corner, trimmed for corner, to spoil his fire control solution. When engaged defensive in a guns only fight we always turn at corner towards any threat that may have, or be about to acquire, a fire control solution.
We pitch the nose to achieve target IAS = corner before we max rate turn into the threat. We must recognise the developing threat early, but we must not abandon best climb (IAS = Vy) for IAS = corner until we need to. Climbing into thinner air and ramming fewer air molecules is the key to maximising performance and increasing specific energy (sum of potential energy (altitude) and kinetic energy (velocity = KTAS)), in any aeroplane, whether propliner, fighter, or interceptor.
When we are not yet ready to begin a drive by shooting (again), and no one is ready to drive by us, we grab as much altitude as we can, as fast as we can (target drag = IAS = Vy), whilst always weaving gently to keep all potential threats and targets in sight.
Risking stall should never be an issue, since we will never let the drag go below Vy = 110 KIAS when disengaging with minimal G, or below a little more than corner = 175 KIAS when engaging offensive, or defensive with substantial G.
continued in next post...
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #5 on: Dec 18th, 2006, 7:40pm » Quote Modify
--------------------------------------------------------------------------------
We don't really want the drag to exceed about 180 KIAS at any time during combat in normal weather, and drag exceeding 300 KIAS is very bad news in a Bf109G. We control the vertical geometry of the fight accordingly always going nose high or nose low to achieve out current drag (IAS) target and stay within our current drag (IAS) limits.
We use as much boost (Ata) as the cylinder heads will permit given the cooling drag (IAS) we are flowing through the radiator at the time. We must not boil the coolant with too much boost (Ata) or with too little (radiator) drag (IAS). We may be driven from targeting drag = IAS = Vy to targeting drag = IAS = corner by the coolant limit and best rpm will be different for the two IAS cases at constant boost (Ata).
We must target engine rpm carefully for our 110 KIAS or 180 KIAS drag targets to maximise thrust at constant boost (Ata). Perhaps 2800 rpm at Vy and perhaps 2600 rpm at corner with WEP or MIL power when engaging the enemy, even though we might cruise at 2000 rpm, even when sustaining corner for tactical cruise during ingress and egress, with only cruise boost (Ata) applied.
Fighter and interceptor pilots do not worry about stalling. They know their IAS targets and they keep current drag within the correct IAS limits by pitching the nose. They fly complex aircraft by the numbers. Their problem is maintaining adequate situational awareness, not maintaining adequate IAS.
Once an aeroplane is correctly trimmed for the current IAS target it achieves that target without needing any help from its pilot. Once an aeroplane is trimmed for target IAS pulling the stick behind neutral bleeds drag = IAS below target and pushing it ahead of neutral increases drag = IAS beyond target.
FS9 users typically fail to adopt the correct (IAS) target and consequently they fail to trim for that target. By failing to adopt the correct target, and by failing to trim for that target, they leave themselves without tactile feedback from their joystick concerning (extent and direction of) deviation from IAS target.
Propliners often have a VSI target, or a power target, rather than an IAS target, but many FS9 users fail to trim for any of the targets (in the supplied handling notes). Consequently dynamics wander off target and the aeroplane fails to achieve target (book) performance. Fighters and interceptors are no different, and in the presence of the enemy they always have an IAS target. They need to be trimmed to that IAS target. Failure to calculate Vy and Corner, followed by failure to trim for Vy or Corner as the fight shifts between engaged and disengaged phases leaves the pilot without tactile feedback of dynamic target compliance.
Learning to trim for the current handling target is a major part of learning to fly any aeroplane. It is a lesson that must not be skipped. We must learn how to trim for any IAS. The Bf109G is not the place to start. Start with a nice stable propliner.
We must always remember that there is no such thing as pitch trim, only elevator trim. The pitch that emerges from a given elevator trim can be nose up or nose down depending on the applied thrust. With autopilot off, when in the cruise in a propliner with 'realistic' flight dynamics, alter MAP at constant trim. Watch what happens for 30 seconds. Now, again from the cruise, alter rpm at constant trim. Watch what happens for 30 seconds. Elevator trim demands a specific IAS and (re)captures the demanded IAS by pitching the nose versus the horizon. The applied trim is resisted by (the encoded) inertia, stability and potentially by aerodynamic damping too. These are all high in a propliner and low in a fighter.
When the pilot alters any dynamic variable the applied elevator trim will pitch the aircraft to recapture the demanded (trimmed) IAS (drag). WITH THE JOYSTICK NEUTRAL. The joystick is used to overshoot or undershoot trimmed IAS. The dynamic target that is trimmed must be the correct one for each phase of flight, in a propliner, or an interceptor, or a fighter. Correct elevator trim provides our tactile frame of reference versus the current handling target.
If we pull hard on a joystick to deliberately and continuously undershoot trimmed IAS we will eventually stall, whether in a B247 on approach at 1G with power off, or in a Bf109 pulling high G in a dogfight with War Emergency Power applied.
We must always know what our current dynamic handling target is and we must always be trimmed for it, otherwise we will spend the entire flight fighting against the aeroplane, instead of occasionally intervening to change one dynamic target to another at the end of a specific phase of the flight. All aircraft missions have phases. In combat they may alternate rapidly, but they are still phases and they have dynamic handling targets to be calculated, noted, adopted, and trimmed.
When the mission profile was sweep/intruder/light strike, once over enemy territory, many combat formation leaders chose to tactical cruise at drag = corner if fuel state made that choice compatible with the mission objective. Most such missions were conducted using vintage era navigation by visual reference to the surface techniques described in a recent post and so although penetration was optimally at or just above rated altitude actual ingress and egress levels were restricted by cloud base and visibility.
A key skill for the Bf109G pilot was shifting from trim = 175 KIAS = corner (neutral trim at design combat weight) to trim = 110 KIAS = Vy, but learning to trim quickly for any and every dynamic handling target is a key skill in any aeroplane.
This text assumes zero flap in all cases. Extending flap to modify wing camber alters trim and being able to retrim for any dynamic target with any flap setting (wing camber) is also a key skill. Bf109G CLmax(augmented) was about 2.84, but deploying max flap massively reduces max safe G load and the Bf109G flaps were not rated for combat use.
Learning how to engage defensive in FS9 implies on line participation with an adversary, but learning how to engage offensive, disengage, and re-engage offensive, only requires AI traffic to be present. Stalling should never be an issue.
Whether or not the flight dynamics in a given freeware or payware product are realistic they have an encoded structural limit, wing area and CLmax. Max safe G is encoded in REC 1101 of the air file. Wing area is encoded in the [airplane_geometry] section of the aircraft.cfg and CLmax is in TBL 404 of the air file (see post 2 of this thread). Always ignore any data in the [Reference Speeds] section of the aircraft.cfg and make your own calculations.
Combat mass may be taken as max *clean* gross minus 20% max internal fuel for an interceptor or minus 40% max internal fuel for a fighter. Strike aircraft may be treated as fighters, but add relevant payload when calculating combat mass.
Weight and fuel data are in the aircraft.cfg. Microsoft assume 6.0lbs per USG of modern unleaded AVGAS but leaded fuel from WW2 was around 6.5 so realistic flight dynamics for aircraft with leaded fuel require an aircraft.cfg within which the real tankage is overstated by 8% else they are underweight for all purposes, including calorific value of the fuel mass in relation to both endurance and range.
Consequently IAS = Vs, IAS = Vy and IAS = Corner can be calculated, noted and adopted during handling even if they are unrealistic. Elevator trim for Vy and corner at combat mass can be user tested once Vy and Corner are noted. Indeed they must be known to the Fs9 user of combat aircraft. My two posts in this thread provide all the necessary formulae for calculating them and the basic handling strategies associated with them.
Without handling notes to operate them by all flight dynamics are useless, whether realistic or otherwise.
FSAviator 12/06
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
thunder100
Connie
If it looks good,it flys good(usually)
Gender:
Posts: 166
Re: Question to FSAviator on Stallvariables in FDE
« Reply #6 on: Dec 19th, 2006, 2:26am » Quote Modify
--------------------------------------------------------------------------------
Dear Tom,dear Fs Aviator
Thank you so much for the time you invest in my question.It will give me a lot of trying over christmas.
Part(little) I found by myself.Also now I know(by error) that due to the high consumption of the V12s in front of a pilot the weight of fuel pays a critical role in stall behavior(4 engine props weight % age of fuel is much less) .Also now I understand why the P51 pilots had to drop the external tanks once enganged in dogfight
Anyway,again thank you so much,Merry Christmas to all of you and a healthy next year
Godspeed
Thunder100
IP Logged
thunder100
Connie
If it looks good,it flys good(usually)
Gender:
Posts: 166
Re: Question to FSAviator on Stallvariables in FDE
« Reply #7 on: Dec 20th, 2006, 1:35pm » Quote Modify
--------------------------------------------------------------------------------
Dear Fs Aviator
Again thank you
Some comments:
1.)I was wrong on WOP 109 flight Dynamic by astupid mistake.I thought I had reduced fuel to 60 % and stored the plane to go flight testing,but by typing error it was still 80 %.So I could never pull more then 4,8 G in a dive down corner.With 60 % became 6 g with 20 % 6,5 G.my friend says that the G-Limit on E to G model was 6,5 G for the last K arround 7 G.The WOP FDE/air File combination does this correctly
2.) The 109 has a fuselage tank behind and under the seat what makes it very tail heavy at the start and also during flying with full tank.As you correctly say it needs trim all the time.And this makes the stall even more entertaining with full tanks as also the position of CG vs Center of lift plays a role as well in increasing the AOA
3.)I cannot say anything about the stick moving forces but my friend says it was not that heavy probably due to the "horn equalizer(correct english?)" to offset too high forces.
The highest G I ever flew in real life was probably about 3 to 5 G in a Pilatus PC-9.
4.)I was beta test for the late John ,our Connie Guru(with your final FDE) he learned me that the highest target in his FS world is accuracy.I am still temped by this and sometimes await too much from FS and try to get too much
5.)I have ordered the book,should arrive in 2 weeks and I gathered some Bf 109 profile and other informations.I bought the Flight replicas Bf-109 K which has the usual FS never stall FDE(while exterior model is a real ringer).There I will try first to come close to the WOP and then with my friend to try to come close to reality
6.)I am still annoyed by the stall behaviour modelled.My friend says,when the 109 has accelerated stall it usually drops the nose sharply into the corner you turn when beeing low fuel,when heavy it drops the rear out of the corner and increases AOA to totally lose control(Because of the heavy tank).the WOP goes into a vicious shake left and right stall with little altitude lost.But to make this more accurate will be a second step.
7.)and Last
My friend has flown with H.J Marseile in the training(which crashed a lot of planes at that time/Vienna Schwechat) and in the very end a short bit with Hartmann.Both of them were to his knowledge the non respecters of the "Iam higher I win rule".Hartmann in particular mainly attacked from rear and below and after firing sharply break away(high G).Therefore he never switched to FW-190,because the Bf-109 Slats made this sharp break aways possible.
Again thank you for this big help ,which cannot be obtained anywhere else.
godspeed
Thunder100
And also thanks to TOM to accept this in a propliner forum as off off topic and for the communication
Merry Christmas
IP Logged
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #8 on: Dec 30th, 2006, 1:07pm » Quote Modify
--------------------------------------------------------------------------------
<<I was beta test for the late John ,our Connie Guru(with your final FDE) he learned me that the highest target in his FS world is accuracy.I am still temped by this and sometimes await too much from FS and try to get too much>>
I remember and am grateful for your effective contribution to the L-049A flight dynamics beta testing process. You are however correct. You still expect too much from a desktop flight simulator. The following post has general application to FD authoring of 'stall' in particular and 'performance envelopes' in general, but again the Bf109G is used as the example.
<< I thought I had reduced fuel to 60 % and stored the plane to go flight testing, but by typing error it was still 80 %.So I could never pull more then 4,8 G in a dive down corner.>>
Corner IAS does vary (slightly) with CoG displacement from CoL in real life, but probably not in FS9. If you are experiencing significantly variable corner IAS due to displacement of CoG from CoL it is probably due to CoL variation with Mach rather than CoG variation with fuel status.
However I doubt your difficulty in achieving 6G relates to the aircraft at all.
Real pilots must be aware of their personal G tolerance limit. If it is less than the tolerance of the airframe for practical purposes maximum instantaneous turn rate is defined by the pilot limit and 'corner' targets are defined by that lower personal limit. If the FS9 FD author decides that the real pilot limit is less than the airframe limit he should restrict the available G to the implied pilot limit. Even 6G may be optimistic for (prolonged) fighter pilot tolerance without a G suit.
<<I cannot say anything about the stick moving forces but my friend says it was not that heavy>>
As explained the stick force required to pull 6G during a 'break turn' from wings level when trimmed for the 6G stall was 150lbs or more in the real aircraft. This was considered to be 'high' in the relevant timeframe. The FD author may decide that he will not allow this whilst the real pilot weighs 1200lbs. Realistic FS9 flight dynamics incorporate human dynamic (physiological) limits as well as airframe dynamic limits.
<<I have ordered the book,should arrive in 2 weeks>>
Remember aircrew fatigue limits in the air file cannot be researched in books about aeroplane variants. One of the themes that I try to explain here frequently is that the data in any book are just an example for a specific circumstance and may lack relevance to the writing of air files unless they are intended to replicate that specific circumstance. The average Bf109K pilot of 1945 was less well trained, less experienced and had a lower G tolerance than a 1942 Bf109G pilot, even if the aircraft had higher limits.
I would not be surprised if the official G limit of the Bf109 airframe was increased in the later variants as a result of combat experience, but any increase beyond 6.0 would be irrelevant. Each FD author of each air file is free to impose an 'historically realistic' physiological limit instead. I believe you are probably assuming that you are being allowed to achieve full elevator deflection just because you can pull your desktop joystick full aft. Pilot fatigue is a much more powerful influence than fuel status.
<<6.)I am still annoyed by the stall behaviour modelled. My friend says,when the 109 has accelerated stall it usually drops the nose sharply into the corner you turn when beeing low fuel,when heavy it drops the rear out of the corner and increases AOA to totally lose control(Because of the heavy tank).the WOP goes into a vicious shake left and right stall with little altitude lost.But to make this more accurate will be a second step. >>
MSFS calculates a single vector value for lift. It does not calculate for the left and right wing vectors independently. Consequently it cannot model realistic yaw or roll behaviour during stall or post stall. Certain behaviours that 'indicate stall' can be encoded and certain post stall behaviours can also be encoded. These can only be generic in character even though their severity may be varied. Pitching behaviour during the 1G stall can be coded more or less realistically, but will then induce unrealistic pitching behaviour during the high G stall since the real yaw and roll characteristics which would usually swamp the real pitch characteristic cannot be present. Many values encoded in MSFS flight dynamics must be appropriate compromise values for all cases. It is usually an error to encode perfect behaviour for one case even if that possibility exists.
Stall is simply not an important event during air combat handling since no one should ever pull to the stall. Those who do induce chaos can be punished with loss of pitch, roll and yaw, stability, damping and authority.
The realistic consequences of turbulent flow are not present in FS9. Nor should they be. Turbulent = chaotic behaviour is very difficult to document, let alone model and replicate. It is sufficient to substitute generic chaotic behaviour that signals significant pilot error and imposes the key negative consequence which is degradation of behaviour and control all kinds in all three axes. The location of the many relevant variables is explained in air file editor help files. However FS9 users (and FD authors) must come to terms with the reality that the chances of any of the values for chaotic flow being known and documented for a given aircraft, at every state of trim, in every state of load, is very slight. It is an error to suppose that flight simulators designed to replicate the behaviour of hundreds of different aircraft need a mathematical model of chaos good enough to replicate things that have never been documented for 99.9% of those aircraft during chaotic flow. Realistic models of chaotic flow belong in flight simulators dedicated to a single aircraft type for which relevant and extensive documentation exists.
FS9 remains the most realistic multi aircraft desk top flight simulator available and that is unlikely to change for a very long time. Behaviour during chaotic flow is not worth worrying about. All the available desk top flight simulators, released before or after FS9, have more important realism issues that need to be addressed more urgently.
<<I bought the Flight replicas Bf-109 K which has the usual FS never stall FDE>>
They all stall. The problem is that the accelerated stall is not recognised by the user and not remedied by the user.
If a tiny shotgun pellet and a very large heavy plank of wood are dropped from a tower at the same moment then the lightweight rounded pellet hits the ground first because it has the lower co-efficient of drag. Both objects are subject to an acceleration of 1G but the plank hits more air molecules on the way down and they retard its chaotic fall.
When the pilot of a Bf109G decides to target corner drag and pull 6G his aeroplane is beautifully aerodynamic and his rate of descent is very high as he spiral power dives at 180 KIAS (drag) and maybe 300 KTAS (velocity). If he allows the wing to stall chaos ensues. He is suddenly sitting on a plank. The co-efficient of drag becomes huge. Falling planks and stalled falling wings are not aerodynamic.
The consequence of stalling from level flight at 1G is a sudden increase in rate of descent. Conversely the consequence of stalling from a spiral power dive at 6G is a sudden reduction in rate of descent. Some FS9 users find this confusing, but it is correct.
As I explained in my first post the lift slope in TBL 404 of the air file can be as accurate as the data you may obtain for CL above and below the stalling angle. I have supplied the real CL for each stall case. The lift post stall is nowhere near zero even though it is far below CLmax. The post stall lift must deliver the rate of descent of a plank in chaotic flow, not the rate of descent of a falling shot gun pellet, and certainly not the rate of descent of a power diving aerodynamic fighter. In the high G case stall arrests descent.
Any of the air files you are describing could have false values for lift in chaotic flow within TBL 404. If you have better values substitute them, but they are not absent, nor is the current value in any of them necessarily false. They may also have false values for degradation of control authority, stability and damping in each axis, for each trim state, for each AoA, post stall. Again if you find chaotic flow data that is more accurate you can substitute it, but representative generic data is unlikely to be absent and the data encoded may be a reasonable generic approximation of pilot induced chaos.
Continued in next post...
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #9 on: Dec 30th, 2006, 1:08pm » Quote Modify
--------------------------------------------------------------------------------
Realistic handling characteristics at the moment of stall are irrelevant. User induced chaos only needs to be signalled and then punished if uncorrected. No one has ever suggested that desk top flight simulators can be used to teach stalling behaviour or incipient spin behaviour at any G, any control surface status, any trim status, any Mach status, etc combined in any way with one another. Users can however teach themselves to avoid chaotic flow conditions by noting, targeting and achieving the relevant IAS for the current phase of flight.
The equations that govern aeroplanes do not differentiate between propliners and fighters and so the issues illuminated above are at least marginally on topic here. I am not sure that we should stray into discussing wider aspects of air combat at all, but anyway there are two cases.
Doctrine is one case and the laws of dynamics are the other. Only doctrine is debatable. It may or may not prove useful, but I will resort to posting one of Newton's equations in the hope of illustrating the dividing line.
In air combat using fixed guns, if two pilots are aware of one another's presence, and both have the same personal or aircraft structure G limit, the pilot with sufficient specific energy (Es) advantage can usually obtain a (fixed gun) fire control solution and can always avoid being the subject of one. He can just drive away.
Specific energy (Es) is sometimes pronounced Energy state. The equations that govern the equilibrium exchange of potential and kinetic energy above planet Earth, published by Newton, are not subject to debate.
Es = (H + v^2) / (2 * G)
Or this can be broken up into two parts;
Es = (H/G) + (v^2/G)
or in Imperial aviation terms as required by Microsoft
Add altitude in feet to TAS squared in feet per second and then divide the total by 64.4
This formula tells us exactly how many feet of altitude one knot of velocity = KTAS is worth in air combat.
10 KTAS = 16.8 ft/sec which when squared is about 280
280 feet of altitude are worth 10 KTAS. They make the same contribution to Es.
Until a fire control solution is imminent we will not trade altitude for velocity at more than 280 feet per 10 KTAS gained. Now remember that diving increases drag = IAS and therefore reduces velocity = TAS.
As Part 1 of the Propliner Tutorial explains;
<<<<It takes MSFS users, (and many real pilots), a long time to understand that if a fighter pilot power dives his fighter from 250 KIAS at 40,000 feet to 400 KIAS at low level he has decelerated from about 500 KTAS to about 400 KTAS. As the fighter pilot dives hard and watches the ASI needle proceed from 250 to 400 he is watching the drag rise, he hears the wind noise screaming ever louder as he decelerates a hundred knots in no time at all. A drag of 400 KIAS at low level ensures that the fighter is much slower than it is with a drag of 250 KIAS at high level. It's just a lot more drag, so we hear much more wind noise. Wind noise isn't an indicator of velocity; it's just an indicator of drag. IAS isn't an indicator of velocity, it's just an indicator of drag.>>>>
Consequently it is always an error to reduce altitude in fixed gun combat until the moment comes to seek a fixed gun fire control solution against a target with a lower (Es) energy state (or to avoid an enemy fire control solution). Only when that moment comes do shooter specific and target specific combat doctrine take over and control the sequence of events.
This thread is about flight dynamics. It is about the size of the theoretical aircraft performance envelope, how much of the theoretical performance envelope is actually usable in real life, how to restrict the FS9 user to those real limits, and how to punish attempts to exceed those real limits. However in any wider context we must not forget that weapon platform performance envelope (battlefield mobility) is a tiny component of combat lethality. The primary components of combat lethality are firepower and durability.
Combat doctrine is devised by intelligence analysts after comparing the firepower and durability of the weapon platform to the equivalent values for the target. The Bf109G had exceptionally low firepower and fairly low durability by contemporary standards. Consequently it needed to acquire an enduring fire control solution from a position where it was safe from target firepower retaliation. Luftwaffe doctrine for attacking a target of a given type using a Bf109G had to take that into account.
Combat begins with detection of the enemy. A 'contact' with the enemy may endure for many minutes. In real air combat the time spent acquiring and then prosecuting a fire control solution, or avoiding enemy fire control solutions is a tiny percentage of that contact. Doctrine applies only to the target acquisition and prosecution phases of the contact. The rest of the contact is driven only by the laws of dynamics and the relentless pursuit of specific energy (Es) advantage.
Having perfected one doctrine relating specifically to the advantages and disadvantages of one particular weapon platform real weapon platform operators are often reluctant to begin the climb along a new learning curve for a different weapon platform which requires different operating doctrine, even if that new weapon platform has a superior performance envelope, superior firepower, and superior durability. To understand the history of the Bf109 it is necessary to understand that it was very popular long after it was outclassed, because the Lachmann auto slats conferred a very low corner drag that gave it a very wide usable performance envelope.
Replicating the dynamics of 'corner' and Vy, both of which are multiples of slat open and flaps retracted stall (Vs), is the key to delivering 'realistic' combat flight dynamics. However replicating the stall characteristics, at the stalling angle associated with Vs, is almost entirely irrelevant. Post stall lift, drag and control authority do not need to be 'accurate'. They only need to be degraded into representative chaos. It is lift, drag and control authority within the usable envelope that need to be carefully researched and replicated and it is only the usable envelope that needs to be carefully limited for both airframe fatigue and aircrew fatigue.
The laws of dynamics that this thread is all about apply equally and without exception to every aeroplane whether or not it is a weapon platform and every pilot whether or not he or she is a combat pilot. However combat doctrine, including fire control doctrine, must also take into account everything else resulting from intelligence analysis that is not equal for all weapon platforms and all pilots.
Continued in next post...
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #10 on: Dec 30th, 2006, 1:10pm » Quote Modify
--------------------------------------------------------------------------------
If Tom agrees that it is at all appropriate to this forum I offer the following thoughts on WW2 air combat doctrine and WW2 fire control doctrine.
[Sure, no problem. TG]
The Luftwaffe did not have access to gyroscopic, lead computing optical sights (LCOS technology) in the relevant timeframe. Consequently hardly any Luftwaffe fighter pilots were able to calculate a fire control solution with shooter G applied. Luftwaffe fire control doctrine therefore called for all shots to be taken at 1G (i.e. wings level with pitch roll and yaw increment all equal to zero).
Furthermore without LCOS technology the vast majority of fighter pilots could not demonstrate the ability to obtain and prosecute a 1 G fire control solution from the beam or the quarter, even against a non evading sleeve target drone. They failed to apply appropriate deflection, because they consistently failed to estimate range correctly. Having failed to estimate range they could not resolve the Pythagorean fire control solution and could not calculate the angular lead required.
Consequently standard Luftwaffe fire control doctrine called for (sequential) stern attack with zero deflection and with decaying overtake drag (IAS) from six o'clock low after diving from higher altitude with superior Es. The closure began at more than corner IAS at the bottom of the dive and the pull into six o'clock low was placed and timed in 4D so that the subsequent climbing 1 G wings level attack terminated at corner well inside effective range, after obtaining and prosecuting a 1G fire control solution with zero deflection.
Contrary to popular opinion most fighter pilots are not especially gifted and the best pilots may not be posted to fighter units. Consequently many other air forces imposed identical fire control doctrine in the relevant timeframe. The air forces that failed to implement a close replica of this fire control doctrine scored significantly fewer victories per hundred contacts until they were equipped with LCOS.
Marseille is famous for publicly refuting this doctrine and refusing to adopt it. Marseille, and a few other aces, could calculate and prosecute fire control solutions involving both deflection and shooter G, without LCOS, because they alone could estimate both range and lead without them. German propaganda referred to Marseille as 'the flying computer'. However what Marseille failed to understand, but the relevant intelligence analysts understood all too well, was that the Luftwaffe and every other air force had very, very few pilots who could emulate his skill in computing a fire control solution. Marseille is famous for shooting down 8 aircraft in a single sortie and 17 in a single day.
Hartmann by contrast is famous for scoring 352 victories steadily over several years. Unlike Marseille he survived the war and never lost a wingman. Hartmann's combat doctrine was simply Luftwaffe Bf109 combat doctrine, which was to seek only one (successful) fire control solution per sortie and then drive away. Hartmann's personal fire control doctrine was, "Get close .. when he fills the entire windscreen ... then you can't possibly miss.". This doctrine is a clear acknowledgement of the lack of capability of the Revi reflector sight as a fire control computer, but is also a reflection of the increasingly inadequate firepower of the Bf109.
Luftwaffe fire control doctrine for the Bf109 was limited by the fire control technology available, and the often low levels of installed firepower, but was clearly effective, both in terms of achieving kills and in terms of avoiding enemy fire control solutions. Of course many fighter pilots, of many air forces, disregarded fire control doctrine and attempted to prosecute fire control solutions whilst pulling G, but more than 99% of them failed to score enough hits to have a material effect, more than 99% of the time. The durability of the target was too high compared to the firepower brought to bear by the shooter.
Continued in next post...
Simple but probably lengthy question
If I get the FDE somewhat correct(weights,power areas etc) and have the correct wing profiles etc in the air file,what defines
A:The planes stall limit at high speed
B:How to alter
OR is there any tutorial(I found none) which describes this?
Thanks in adavance
Godspeed
Thunder100
PS:It is for a single prop fighter,not a 4 engine liner
IP Logged
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #1 on: Dec 5th, 2006, 10:41am » Quote Modify
--------------------------------------------------------------------------------
Hi,
And here's your answer from FSAviator:
Q. > Is there a web flight dynamics tutorial (I found none).
There is no web tutorial that comes close to explaining everything that needs to be understood to encode a realistic air file. The subject is too complex and requires too many diagrams for clarity.
MSFS uses real world physics throughout so the relevant tutorials are available from every bookshop. However even FD authors who normally think metric should perhaps seek textbooks published in the United States expressed in Imperial units. It may be less confusing in the long run. Unfortunately if there is a text book using Imperial units, still in print, that covers flight + engine + airscrew dynamics adequately in a single volume I have never seen it.
There is a single metric textbook that covers most of the ground. Mechanics of Flight by A.C. Kermode (ISBN 0 273 31623 0). It runs to over 500 pages with maybe 300 diagrams. It also provides sample aerofoil lift slope data as required by Table 404 of the air file, which is probably at the root of the specific question asked.
The structure of the air file, (i.e. where the concepts in textbooks are stored within FS9), is explained in the help files of freeware air file editors, about as clearly as it explained anywhere else, but be warned many of the help comments within air file editors or spread across the web explain how very old versions of MSFS and CFS worked and do not apply to FS9.
Q. > in the air file what defines the planes stall limit at high speed and how is it altered
I am not sure that I have understood the question asked, but assume you mean how is G load taken into account.
Aeroplanes stall when the wing reaches the angle of attack (AoA) at which the airflow over the wing becomes turbulent. This happens when the lift required to be in equilibrium with (equal to) the current weight is more than the wing can produce.
Current weight is current mass multiplied by current G (loading).
As weight varies the wing varies its AoA to produce current lift = current weight sustaining equilibrium. The fuselage has no choice but to pitch up and down with the wing. If AoA becomes excessive, as the wing attempts to achieve Newtonian equilibrium with current weight, the wing stalls. Stalling 'V speeds' such as Vs, Vs0, etc., denote the drag (IAS) associated with one G stall, but it is current angle of attack, not drag (IAS), or velocity (TAS) that controls the propensity to stall.
The stalling angle is defined in TBL 404 of the air file. Within TBL 404 the entire lift slope is described point by point with Co-efficient of Lift plotted versus *fuselage* AoA expressed in radians.
Fuselage AoA equals wing AoA minus mean wing incidence.
Mean wing incidence equals wing incidence plus or minus mean wing twist.
The associated comments in an aircraft.cfg are just REM statements recording the data used by the FD author to calculate the mean incidence offset to be used in TBL 404, or they can be set to zero since they are not read by FS9.
After making those calculations the flight dynamics (FD) author encodes all of the real lift slope for the specific aerofoil within TBL 404 of the specific air file. Else he or she must estimate it. It is the invariant Microsoft flight model (FM) inside fs9.exe and inside sim1.dll that calculates current weight as G varies and thus calculates the propensity to stall.
The aeroplane stalls when the current weight causes the current (fuselage) angle of attack to exceed the stalling angle (in radians) specified for CLmax within the encoded TBL 404 lift slope. The stalling angle is the fuselage AoA that the FD author associated with CLmax in TBL 404. The loss (rate) of CL between that angle (for CLmax) and the next higher encoded fuselage angle of attack in TBL 404 determines the severity of the stall.
If the aircraft has automatic slats the CLmax and fuselage AoA encoded must be for the open status with the closed status encoded as the prior matched data pair within TBL 404.
FS9 simulates post stall behaviour poorly and inaccurately by default, but altitude loss per second due to stall can be imposed at any desired rate fairly easily by the means above and stall propensity versus G is as accurate as the aerofoil specific lift slope data encoded within TBL 404.
All of which probably goes a long way to explaining why there is no web tutorial explaining everything, or even a significant fraction of the things, that an FD author must know how to calculate or estimate outside FS9 and where to subsequently store his results within FS9.
FSAviator 12/06
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
thunder100
Connie
If it looks good,it flys good(usually)
Gender:
Posts: 166
Re: Question to FSAviator on Stallvariables in FDE
« Reply #2 on: Dec 5th, 2006, 1:46pm » Quote Modify
--------------------------------------------------------------------------------
Dear Tom(if you could transfer)
Dear Fs Aviator(if you read this)
Thank you very much
the threat does not belong here,yes I know very well,but there is no other source to get the help by FS Aviator as far as I know.
Issue is:
I bought WOP Bf-109 E which stall's in deliverycondition at any speed above 130 knots at 4,5 G both positive and negative G's.
This is not correct and I failed yet to get a WOP support forum approval (5 days) although I did nothing wrong.Anyway
Why wrong?
I have a friend (90 years) which was Bf-109 pilot in WW2(and FW-190,Ju-88,He-111,Ju-52).He was also at one time instructor for Roumanian airforce during the war.He says that the real planes stall was 3 g low speed 6,5 g at mid speed and 4 G at high speed ,which he could read from a G-meter in his instructor plane.And at negative G --> 3 G was really max.
I am 48 and a graduated mecahnical engineer as well as aerodynamics(although I only used it for engine fuel/air airstreams),so I can really understand what you are saying and I will now have a deep look into the airfile to repair this by try and error till it is close to reaility(simple to learn about it).Also I had help about the real fighter behavior on the forum by a PM to Jim D and Jesse.So thank you again,you are a big knowledge source to FS community
So now I close it,too much words for an Off Calclassic topic anyway ,but THANK YOU VERY MUCH AGAIN
Godspeed
Thunder100
aka
Roland Berger(Vienna/Austria)
IP Logged
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #3 on: Dec 18th, 2006, 7:38pm » Quote Modify
--------------------------------------------------------------------------------
Hi Roland,
And here's your reply from FSAviator:
Tom has assured me that he does not object to a follow up post on this topic. I will not be drawn into reviewing individual air files released by third party developers. What follows relates to a real aeroplane or FS9 flight dynamics in general.
I will start by illustrating use of the relevant laws of dynamics, and then explain why they matter in real life, and why FS9 users need to understand the consequences of Newton’s Laws even if they do not manage to understand the equations. I will use the Bf109G as a worked example, but everything that follows also applies to propliners. Combat strategies used to illustrate the relevance of the dynamics discussed are generic and simplified.
First the equations;
According to the RAF test pilot report of a captured aircraft Bf109 CLmax was about 2.34 with Lachmann auto slats open, probably achieved at about 0.35 radians fuselage AoA. Maybe as low as 1.44 with auto slats closed. The wing area was 173.3 square feet and the following data assume (combat) mass = 6950 pounds.
To calculate gear up and zero flap 1G stall (Vs) (in Imperial units = feet per second as required by Microsoft) we proceed as follows;
1) half the density of air at sea level (measured in slugs is always 0.00119) * CLmax * wing area (measured in square feet)
Result 1 = 0.00119 * 2.34 * 173.3 = 0.483
2) Divide the (combat) mass by Result 1 above
Result 2 = 6950 divided by 0.483 = 14390
3) Determine the square root of Result 2
Vs = SQR(14390) = 120 feet per second
4) Convert to KCAS (ASI reading if no pitot tube position or other errors) using 1.68 divisor
120 ft/sec divided by 1.68 = 71.4 KCAS = auto slat open stall at 1G with mass = 6950lbs = Vs
5) Most (but not all) V speeds including Vs vary with square root of current weight and weight is just mass * G.
Vs at 4G = 71.4 * SQR(4) = 143 KCAS
6) I believe that max safe G to avoid structural failure of the Bf109G tail was + 6.0
Vs at 6G = 71.4 * SQR(6) = 175 KCAS
Of course the RAF test pilot included all those data in his test report, but we rarely have a full flight test report for an aircraft we intend to fly in FS9. The only way we can obtain those handling data is to calculate them from the data in the supplied flight dynamics, whether the supplied dynamics are realistic or not.
Now let’s explore why the data emerging from the equations above matter to FS9 users.
In FS9 the aeroplane will have exactly the dynamics specified within the FS9 FD and we must fly it accordingly. For use within FS9 it may be more useful to convert Bf109G data to metric Kph rather than imperial Knots. However in what follows I will assume zero pitot position error, zero ASI gauge error, zero compressibility error, and talk about KIAS for ease of comparison with existing propliner threads and propliner handling notes.
Contrary to popular belief aeroplanes are not kept aloft by the lift from their wings. They are kept aloft only by the power from their engines. We must never forget that any aeroplane requires cruise (horse) power just to maintain current altitude at one G. Even cruise power is a lot of horse power. Gliders do not remain aloft in still air.
When we pull 6G the Bf109G still has a mass of 6950lbs but it weighs almost 42,000lbs. It has nowhere near enough power to cruise at that weight. Pulling 6G will stall the Bf109G unless we (spiral) dive hard enough to keep the drag above 175 KIAS.
The minimum drag required to sustain the G load equal to the safe structural limit is the called 'corner' because we can 'corner' hardest under that condition. At 175 KIAS the instant we pull 6G we stall. We need to dive and keep (spiral) diving to keep the drag above 175 KIAS in order to sustain 6G. To pull 4G we only need to keep our drag above 143 KIAS, since our Bf109G 'only' weighs about 28,000lbs.
At least that is true in perfectly smooth air. Any turbulence would push the wing through the stalling angle so in real life we need to (spiral) dive hard enough to keep the drag a decent margin above the relevant stall limit (including corner) if we wish to generate a max G turn, whether to avoid an enemy fire control solution, or to obtain a fire control solution. In normal weather conditions we might target 180 KIAS, but more than that in turbulent air.
A Bf109G will stall at any G load. One part of the FS9 FD author's job is to make it stall at the correct drag (IAS) for that G load and to punish pilot error accordingly. I explained that process in detail in my previous post.
However there is another issue. FS9 users often confuse pre stall sink with post stall sink when turning hard.
On final approach we reduce power and IAS. Suppose we close the throttles completely in an aeroplane with no flaps such as the Boeing 247 and trim elevator up to target a lower drag. The drag reduces, angle of attack increases, the nose comes up and the Boeing 247 sinks towards the runway nose up, but under full control. However if we pull back on the yoke and try to maintain altitude with less than cruise power applied any aircraft can only stall.
Whenever we reduce power below cruise power an aeroplane sinks. It will sink whether nose up or nose down, whether turning or wings level. Like a glider it can only sink if it lacks cruise power in still air. It has not stalled, but it may nevertheless sink fast.
When turning hard and pulling hard on the joystick to pitch an aircraft round a steeply banked high G turn FS9 users sometimes become confused. The Bf109G cannot cruise at 42,000lbs it can only sink. It does not have enough power to cruise at 42,000lbs at any drag (IAS). By pulling 6G we have decided to make the weight more than any applied power can possibly support. We have decided to sink hard. This does not mean we have stalled. If we stall we will not be able to sustain 6G. After we stall max G is 1G. So long as we can sustain more than 1G we are not stalled, but we will sink very fast regardless. We must spiral dive in order to sustain high G turns. Even briefly available War Emergency Power (WEP) from a Daimler Benz 605 engine will not support anything like 42000lbs in level (i.e. cruising) flight.
When pulling high G, to create very high aircraft weights, to improve instantaneous turn rate, very high rates of involuntary descent are certain, but do not imply stall. They imply only the dynamic certainty of sinking hard at that high weight, just as applying less than cruise power for approach weight implies the dynamic certainty of sinking down the approach.
But there is a third much more complex issue which few FS9 users grasp.
The force of air on the elevators must be countered by the arm muscles of the pilot; actually by just his right arm. His left arm must be operating the elevator trim wheel, the throttle, the boost override switches, the war emergency power switches, the water methanol coolant arming switches, the reflector gun sight, and the rpm lever. Pulling our plastic joystick full aft with our right arm takes little effort. Not so for the real Bf109G pilot.
In flying clothing his mass is 200lbs. At 6G his weight is 1200lbs. His right arm weighs almost 200lbs. His heart is failing to pump enough blood to his vital organs. First it shuts down blood supply to his arm muscles; then to his eyes. He is still fully conscious. His brain still has enough blood. He feels the pain. He is a fit fighter pilot. Even without a G suit he may not suffer GLOC until 6.2G, but at 6.0 he has no blood supply to his arms, or his eyes.
So you are in the gym with a tourniquet round each arm so that your arm muscles receive no blood. You are sitting down wearing a flying suit that has 700 pounds of lead weights sewn into it. You have a 200lb weight hanging from each elbow and a 100lb weight on top of your head. Now all you have to do is hold your right arm out and pull say 150 lbs continuously with your right arm on a lever whilst finely balancing the trim with your left arm with your eyes closed (blacked out) but not quite unconscious (GLOC). How long can you do this?
If FS9 users have never weighed 1,200 pounds in real life they have no idea how difficult it is to work really hard doing something really complex when they do.
What most FS9 users fail to grasp is that the FD author also has to code the consequence of all of this in the air file. The FD author models pilot surplus strength versus elevator loading. He writes the curves that limit what the pilot can do as well as what the aircraft can do. He treats the real pilot as just another aircraft system which has design limits. To obtain 6G the pilot must pull the elevators up into a drag of greater than 175 KIAS but there is always some higher drag that prevents the pilot from pulling them up far enough to generate 6G, with neutral trim, and with each other (incorrect) trim setting the IAS he cannot pull against is different.
continued in next post...
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #4 on: Dec 18th, 2006, 7:39pm » Quote Modify
--------------------------------------------------------------------------------
Most FS9 users don't grasp that. They don't trim for the target drag (IAS) and the air file curves written by the FD author calculate that the real pilot has insufficient surplus strength remaining to sustain the necessary pull in the out of trim condition. Many FS9 users mistake this for loss of control due to stall, but it is loss of control due to fatigue, due to failure to trim for the target IAS.
To sustain a high G turn it is necessary to trim into the turn until there is no back pressure required on the joystick at the correct target drag (IAS) for the high G turn. All complex aeroplanes need to be flown by the numbers and the Bf109G is no exception. For a Bf109G trimming for around 180 KIAS (just above corner) is about right for fully engaged air combat in still air. Most FS9 users fail to do this.
Before we dive to bounce the enemy, (or to prevent an enemy from maintaining a tally on us), we may have (cruise) drag in excess of 180 KIAS in a Bf109G. Even though we intend to dive (increase drag = IAS) we must trim for the drag (IAS) we intend to target in the next combat turn, whether engaging offensive or engaging defensive. This may be less than cruise drag (IAS) and is always less than dive drag (IAS) so we will often need to adjust elevator trim upward before the dive so that we are trimmed for 180 KIAS which will be our target IAS once engaged with the enemy. We will normally enter the dive by rolling nearly inverted and pulling through working with the applied target = 175 KIAS trim.
This has nothing to do with fuel flow under negative G. Daimler Benz 605 engines are fuel injected. In combat we may need to pull 6G, and we must be trimmed for about 180 KIAS to do that. If we are trimmed for 180 KIAS and we decide to pull only 4G we, (the real pilot encoded by the FD author as an aircraft system), are bound to have enough strength, but if we trim for more drag than 180 KIAS we may not have the strength to fight that out of trim condition trim to 6G, or even to 4G.
Propliners are designed so that elevator trim is neutral for design cruise drag at mid cruise weight. Fighters are designed so that elevator trim is neutral at corner drag at design combat weight.
With trim neutral the pilot of a real Bf109G could not muster the strength to apply 6G when the drag on the elevators exceeded about 300 KIAS and he could not keep his left arm in position to operate the trim wheel at 6G either. To max rate turn in a Bf109G drag must be kept between 175 KIAS and 300 KIAS with the aircraft trimmed for about 180 KIAS (normal weather = little turbulence). However the radius of turn is very much less at 180 KIAS since radius of turn varies with the *square* of the IAS (at constant altitude).
The radius of turn is three times worse at 300 KIAS compared to 175 KIAS. We must trim for a little more than corner drag and target a little more than corner drag throughout the time that we are fully engaged.
We must avoid stall when turning hard by pitching the nose to generate more than 175 KIAS and we must also pitch the nose to avoid freezing of the controls above 300 KIAS. We must go nose high or nose low to sustain a little more than corner drag whatever VSI that delivers (terrain permitting!). We change target IAS from corner only to engage offensive during acquisition of a fire control solution, but we remain trimmed for corner. When we engage defensive we turn max (structural G limit) rate just above corner drag = 175 KIAS, nose up or nose down, (terrain permitting!). We do all of this with WEP or MIL or whatever lesser power is currently available.
Real fighters are often, (but not always), designed so that their elevator authority matches corner drag (Va = Corner). Applying full up elevator above corner IAS causes the real tail to break off. However because MSFS is a fifty dollar flight simulator with a 50 cent flight model structural failure will occur with full aft stick at corner IAS if the aircraft is trimmed for less than corner drag; if elevator authority was coded as realistic. Since most MSFS users fail to grasp these essentials most FD authors code elevator authority to be much less than realistic to protect naive users who believe that full up elevator can be applied at any IAS and however badly out of trim the aircraft may be when they apply it.
When fully engaged in a Bf109G we trim neutral. We never pull hard enough to bleed drag to the 6G stall IAS = 175 = corner, let alone the 1G slat open stall drag = 72 KIAS, unless to obtain a momentary snap shot fire control solution when we are certain that no one is doing the same to us. When fully engaged we must not become obsessed by our potential target. We are always somebody else’s potential target and we must be able to fly a 6G break to foil their fire control solution, so we must be trimmed for it, and we must always have enough drag (IAS) for it.
Air combat must be a series of drive by shootings, never a close in knife fight. Knife fights let the other guy have a chance of striking a lucky blow. If we cannot organise a drive by shooting right now then we are the potential target for a drive by shooting by the enemy and we need to disengage. Disengaging does not involve going round and round in tight circles. We disengage by seeking cloud cover, or by causing the enemy to lose his tally on us (whether visual or radar) by other means.
If we cannot make the enemy lose his tally on us, provided no enemy aircraft is inside gun parameters, we must target IAS = Vy. The multiplier from Vs to Vy depends mostly on wing aspect ratio. It might be Vs * 1.5 in a typical fighter so maybe 110 KIAS = best climb rate in a Bf109G.
During phases of disengagement we must try to climb above all threats. If the engine overheats we must target a higher drag (IAS) to cool the radiator better. Engine cooling is proportional to IAS whether engines are 'air cooled' or 'liquid cooled' because the liquid coolant is air cooled in the radiators anyway. We will not reduce boost (MAP = Ata). The highest pilot in a (piston engine guns only) combat has the initiative. He can drive by again, or just drive away.
During these periods of disengagement we will be weaving and circling gently to keep all the potential threats and potential targets in sight and clear our wingman's tail. He is clearing ours. We are barely pulling any G as we weave uphill at Vy or a little more for adequate engine cooling. If the enemy is attempting to drive away we will pitch the nose to corner and follow (climbing if possible) at IAS = corner. We will exceed corner only to maintain our tally on the enemy and only if he is nearing the limits of visibility. We will cease to climb only when we reach the rated altitude for our current power setting, or when we are above all enemy aircraft on which we have a tally. Then we will pitch the nose level to target Vmax for our current power setting. We are always seeking to maximise specific energy. Once we reach the dynamic condition in which it is safe to maximise our kinetic energy (velocity = TAS) we cease to maximise our potential energy (altitude).
If at any time when we are disengaging an enemy is threatening to achieve a fire control solution we must reduce nose pitch and increase drag to corner. When the next enemy drive by starts we must max rate turn *towards* the attacker at corner, trimmed for corner, to spoil his fire control solution. When engaged defensive in a guns only fight we always turn at corner towards any threat that may have, or be about to acquire, a fire control solution.
We pitch the nose to achieve target IAS = corner before we max rate turn into the threat. We must recognise the developing threat early, but we must not abandon best climb (IAS = Vy) for IAS = corner until we need to. Climbing into thinner air and ramming fewer air molecules is the key to maximising performance and increasing specific energy (sum of potential energy (altitude) and kinetic energy (velocity = KTAS)), in any aeroplane, whether propliner, fighter, or interceptor.
When we are not yet ready to begin a drive by shooting (again), and no one is ready to drive by us, we grab as much altitude as we can, as fast as we can (target drag = IAS = Vy), whilst always weaving gently to keep all potential threats and targets in sight.
Risking stall should never be an issue, since we will never let the drag go below Vy = 110 KIAS when disengaging with minimal G, or below a little more than corner = 175 KIAS when engaging offensive, or defensive with substantial G.
continued in next post...
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #5 on: Dec 18th, 2006, 7:40pm » Quote Modify
--------------------------------------------------------------------------------
We don't really want the drag to exceed about 180 KIAS at any time during combat in normal weather, and drag exceeding 300 KIAS is very bad news in a Bf109G. We control the vertical geometry of the fight accordingly always going nose high or nose low to achieve out current drag (IAS) target and stay within our current drag (IAS) limits.
We use as much boost (Ata) as the cylinder heads will permit given the cooling drag (IAS) we are flowing through the radiator at the time. We must not boil the coolant with too much boost (Ata) or with too little (radiator) drag (IAS). We may be driven from targeting drag = IAS = Vy to targeting drag = IAS = corner by the coolant limit and best rpm will be different for the two IAS cases at constant boost (Ata).
We must target engine rpm carefully for our 110 KIAS or 180 KIAS drag targets to maximise thrust at constant boost (Ata). Perhaps 2800 rpm at Vy and perhaps 2600 rpm at corner with WEP or MIL power when engaging the enemy, even though we might cruise at 2000 rpm, even when sustaining corner for tactical cruise during ingress and egress, with only cruise boost (Ata) applied.
Fighter and interceptor pilots do not worry about stalling. They know their IAS targets and they keep current drag within the correct IAS limits by pitching the nose. They fly complex aircraft by the numbers. Their problem is maintaining adequate situational awareness, not maintaining adequate IAS.
Once an aeroplane is correctly trimmed for the current IAS target it achieves that target without needing any help from its pilot. Once an aeroplane is trimmed for target IAS pulling the stick behind neutral bleeds drag = IAS below target and pushing it ahead of neutral increases drag = IAS beyond target.
FS9 users typically fail to adopt the correct (IAS) target and consequently they fail to trim for that target. By failing to adopt the correct target, and by failing to trim for that target, they leave themselves without tactile feedback from their joystick concerning (extent and direction of) deviation from IAS target.
Propliners often have a VSI target, or a power target, rather than an IAS target, but many FS9 users fail to trim for any of the targets (in the supplied handling notes). Consequently dynamics wander off target and the aeroplane fails to achieve target (book) performance. Fighters and interceptors are no different, and in the presence of the enemy they always have an IAS target. They need to be trimmed to that IAS target. Failure to calculate Vy and Corner, followed by failure to trim for Vy or Corner as the fight shifts between engaged and disengaged phases leaves the pilot without tactile feedback of dynamic target compliance.
Learning to trim for the current handling target is a major part of learning to fly any aeroplane. It is a lesson that must not be skipped. We must learn how to trim for any IAS. The Bf109G is not the place to start. Start with a nice stable propliner.
We must always remember that there is no such thing as pitch trim, only elevator trim. The pitch that emerges from a given elevator trim can be nose up or nose down depending on the applied thrust. With autopilot off, when in the cruise in a propliner with 'realistic' flight dynamics, alter MAP at constant trim. Watch what happens for 30 seconds. Now, again from the cruise, alter rpm at constant trim. Watch what happens for 30 seconds. Elevator trim demands a specific IAS and (re)captures the demanded IAS by pitching the nose versus the horizon. The applied trim is resisted by (the encoded) inertia, stability and potentially by aerodynamic damping too. These are all high in a propliner and low in a fighter.
When the pilot alters any dynamic variable the applied elevator trim will pitch the aircraft to recapture the demanded (trimmed) IAS (drag). WITH THE JOYSTICK NEUTRAL. The joystick is used to overshoot or undershoot trimmed IAS. The dynamic target that is trimmed must be the correct one for each phase of flight, in a propliner, or an interceptor, or a fighter. Correct elevator trim provides our tactile frame of reference versus the current handling target.
If we pull hard on a joystick to deliberately and continuously undershoot trimmed IAS we will eventually stall, whether in a B247 on approach at 1G with power off, or in a Bf109 pulling high G in a dogfight with War Emergency Power applied.
We must always know what our current dynamic handling target is and we must always be trimmed for it, otherwise we will spend the entire flight fighting against the aeroplane, instead of occasionally intervening to change one dynamic target to another at the end of a specific phase of the flight. All aircraft missions have phases. In combat they may alternate rapidly, but they are still phases and they have dynamic handling targets to be calculated, noted, adopted, and trimmed.
When the mission profile was sweep/intruder/light strike, once over enemy territory, many combat formation leaders chose to tactical cruise at drag = corner if fuel state made that choice compatible with the mission objective. Most such missions were conducted using vintage era navigation by visual reference to the surface techniques described in a recent post and so although penetration was optimally at or just above rated altitude actual ingress and egress levels were restricted by cloud base and visibility.
A key skill for the Bf109G pilot was shifting from trim = 175 KIAS = corner (neutral trim at design combat weight) to trim = 110 KIAS = Vy, but learning to trim quickly for any and every dynamic handling target is a key skill in any aeroplane.
This text assumes zero flap in all cases. Extending flap to modify wing camber alters trim and being able to retrim for any dynamic target with any flap setting (wing camber) is also a key skill. Bf109G CLmax(augmented) was about 2.84, but deploying max flap massively reduces max safe G load and the Bf109G flaps were not rated for combat use.
Learning how to engage defensive in FS9 implies on line participation with an adversary, but learning how to engage offensive, disengage, and re-engage offensive, only requires AI traffic to be present. Stalling should never be an issue.
Whether or not the flight dynamics in a given freeware or payware product are realistic they have an encoded structural limit, wing area and CLmax. Max safe G is encoded in REC 1101 of the air file. Wing area is encoded in the [airplane_geometry] section of the aircraft.cfg and CLmax is in TBL 404 of the air file (see post 2 of this thread). Always ignore any data in the [Reference Speeds] section of the aircraft.cfg and make your own calculations.
Combat mass may be taken as max *clean* gross minus 20% max internal fuel for an interceptor or minus 40% max internal fuel for a fighter. Strike aircraft may be treated as fighters, but add relevant payload when calculating combat mass.
Weight and fuel data are in the aircraft.cfg. Microsoft assume 6.0lbs per USG of modern unleaded AVGAS but leaded fuel from WW2 was around 6.5 so realistic flight dynamics for aircraft with leaded fuel require an aircraft.cfg within which the real tankage is overstated by 8% else they are underweight for all purposes, including calorific value of the fuel mass in relation to both endurance and range.
Consequently IAS = Vs, IAS = Vy and IAS = Corner can be calculated, noted and adopted during handling even if they are unrealistic. Elevator trim for Vy and corner at combat mass can be user tested once Vy and Corner are noted. Indeed they must be known to the Fs9 user of combat aircraft. My two posts in this thread provide all the necessary formulae for calculating them and the basic handling strategies associated with them.
Without handling notes to operate them by all flight dynamics are useless, whether realistic or otherwise.
FSAviator 12/06
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
thunder100
Connie
If it looks good,it flys good(usually)
Gender:
Posts: 166
Re: Question to FSAviator on Stallvariables in FDE
« Reply #6 on: Dec 19th, 2006, 2:26am » Quote Modify
--------------------------------------------------------------------------------
Dear Tom,dear Fs Aviator
Thank you so much for the time you invest in my question.It will give me a lot of trying over christmas.
Part(little) I found by myself.Also now I know(by error) that due to the high consumption of the V12s in front of a pilot the weight of fuel pays a critical role in stall behavior(4 engine props weight % age of fuel is much less) .Also now I understand why the P51 pilots had to drop the external tanks once enganged in dogfight
Anyway,again thank you so much,Merry Christmas to all of you and a healthy next year
Godspeed
Thunder100
IP Logged
thunder100
Connie
If it looks good,it flys good(usually)
Gender:
Posts: 166
Re: Question to FSAviator on Stallvariables in FDE
« Reply #7 on: Dec 20th, 2006, 1:35pm » Quote Modify
--------------------------------------------------------------------------------
Dear Fs Aviator
Again thank you
Some comments:
1.)I was wrong on WOP 109 flight Dynamic by astupid mistake.I thought I had reduced fuel to 60 % and stored the plane to go flight testing,but by typing error it was still 80 %.So I could never pull more then 4,8 G in a dive down corner.With 60 % became 6 g with 20 % 6,5 G.my friend says that the G-Limit on E to G model was 6,5 G for the last K arround 7 G.The WOP FDE/air File combination does this correctly
2.) The 109 has a fuselage tank behind and under the seat what makes it very tail heavy at the start and also during flying with full tank.As you correctly say it needs trim all the time.And this makes the stall even more entertaining with full tanks as also the position of CG vs Center of lift plays a role as well in increasing the AOA
3.)I cannot say anything about the stick moving forces but my friend says it was not that heavy probably due to the "horn equalizer(correct english?)" to offset too high forces.
The highest G I ever flew in real life was probably about 3 to 5 G in a Pilatus PC-9.
4.)I was beta test for the late John ,our Connie Guru(with your final FDE) he learned me that the highest target in his FS world is accuracy.I am still temped by this and sometimes await too much from FS and try to get too much
5.)I have ordered the book,should arrive in 2 weeks and I gathered some Bf 109 profile and other informations.I bought the Flight replicas Bf-109 K which has the usual FS never stall FDE(while exterior model is a real ringer).There I will try first to come close to the WOP and then with my friend to try to come close to reality
6.)I am still annoyed by the stall behaviour modelled.My friend says,when the 109 has accelerated stall it usually drops the nose sharply into the corner you turn when beeing low fuel,when heavy it drops the rear out of the corner and increases AOA to totally lose control(Because of the heavy tank).the WOP goes into a vicious shake left and right stall with little altitude lost.But to make this more accurate will be a second step.
7.)and Last
My friend has flown with H.J Marseile in the training(which crashed a lot of planes at that time/Vienna Schwechat) and in the very end a short bit with Hartmann.Both of them were to his knowledge the non respecters of the "Iam higher I win rule".Hartmann in particular mainly attacked from rear and below and after firing sharply break away(high G).Therefore he never switched to FW-190,because the Bf-109 Slats made this sharp break aways possible.
Again thank you for this big help ,which cannot be obtained anywhere else.
godspeed
Thunder100
And also thanks to TOM to accept this in a propliner forum as off off topic and for the communication
Merry Christmas
IP Logged
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #8 on: Dec 30th, 2006, 1:07pm » Quote Modify
--------------------------------------------------------------------------------
<<I was beta test for the late John ,our Connie Guru(with your final FDE) he learned me that the highest target in his FS world is accuracy.I am still temped by this and sometimes await too much from FS and try to get too much>>
I remember and am grateful for your effective contribution to the L-049A flight dynamics beta testing process. You are however correct. You still expect too much from a desktop flight simulator. The following post has general application to FD authoring of 'stall' in particular and 'performance envelopes' in general, but again the Bf109G is used as the example.
<< I thought I had reduced fuel to 60 % and stored the plane to go flight testing, but by typing error it was still 80 %.So I could never pull more then 4,8 G in a dive down corner.>>
Corner IAS does vary (slightly) with CoG displacement from CoL in real life, but probably not in FS9. If you are experiencing significantly variable corner IAS due to displacement of CoG from CoL it is probably due to CoL variation with Mach rather than CoG variation with fuel status.
However I doubt your difficulty in achieving 6G relates to the aircraft at all.
Real pilots must be aware of their personal G tolerance limit. If it is less than the tolerance of the airframe for practical purposes maximum instantaneous turn rate is defined by the pilot limit and 'corner' targets are defined by that lower personal limit. If the FS9 FD author decides that the real pilot limit is less than the airframe limit he should restrict the available G to the implied pilot limit. Even 6G may be optimistic for (prolonged) fighter pilot tolerance without a G suit.
<<I cannot say anything about the stick moving forces but my friend says it was not that heavy>>
As explained the stick force required to pull 6G during a 'break turn' from wings level when trimmed for the 6G stall was 150lbs or more in the real aircraft. This was considered to be 'high' in the relevant timeframe. The FD author may decide that he will not allow this whilst the real pilot weighs 1200lbs. Realistic FS9 flight dynamics incorporate human dynamic (physiological) limits as well as airframe dynamic limits.
<<I have ordered the book,should arrive in 2 weeks>>
Remember aircrew fatigue limits in the air file cannot be researched in books about aeroplane variants. One of the themes that I try to explain here frequently is that the data in any book are just an example for a specific circumstance and may lack relevance to the writing of air files unless they are intended to replicate that specific circumstance. The average Bf109K pilot of 1945 was less well trained, less experienced and had a lower G tolerance than a 1942 Bf109G pilot, even if the aircraft had higher limits.
I would not be surprised if the official G limit of the Bf109 airframe was increased in the later variants as a result of combat experience, but any increase beyond 6.0 would be irrelevant. Each FD author of each air file is free to impose an 'historically realistic' physiological limit instead. I believe you are probably assuming that you are being allowed to achieve full elevator deflection just because you can pull your desktop joystick full aft. Pilot fatigue is a much more powerful influence than fuel status.
<<6.)I am still annoyed by the stall behaviour modelled. My friend says,when the 109 has accelerated stall it usually drops the nose sharply into the corner you turn when beeing low fuel,when heavy it drops the rear out of the corner and increases AOA to totally lose control(Because of the heavy tank).the WOP goes into a vicious shake left and right stall with little altitude lost.But to make this more accurate will be a second step. >>
MSFS calculates a single vector value for lift. It does not calculate for the left and right wing vectors independently. Consequently it cannot model realistic yaw or roll behaviour during stall or post stall. Certain behaviours that 'indicate stall' can be encoded and certain post stall behaviours can also be encoded. These can only be generic in character even though their severity may be varied. Pitching behaviour during the 1G stall can be coded more or less realistically, but will then induce unrealistic pitching behaviour during the high G stall since the real yaw and roll characteristics which would usually swamp the real pitch characteristic cannot be present. Many values encoded in MSFS flight dynamics must be appropriate compromise values for all cases. It is usually an error to encode perfect behaviour for one case even if that possibility exists.
Stall is simply not an important event during air combat handling since no one should ever pull to the stall. Those who do induce chaos can be punished with loss of pitch, roll and yaw, stability, damping and authority.
The realistic consequences of turbulent flow are not present in FS9. Nor should they be. Turbulent = chaotic behaviour is very difficult to document, let alone model and replicate. It is sufficient to substitute generic chaotic behaviour that signals significant pilot error and imposes the key negative consequence which is degradation of behaviour and control all kinds in all three axes. The location of the many relevant variables is explained in air file editor help files. However FS9 users (and FD authors) must come to terms with the reality that the chances of any of the values for chaotic flow being known and documented for a given aircraft, at every state of trim, in every state of load, is very slight. It is an error to suppose that flight simulators designed to replicate the behaviour of hundreds of different aircraft need a mathematical model of chaos good enough to replicate things that have never been documented for 99.9% of those aircraft during chaotic flow. Realistic models of chaotic flow belong in flight simulators dedicated to a single aircraft type for which relevant and extensive documentation exists.
FS9 remains the most realistic multi aircraft desk top flight simulator available and that is unlikely to change for a very long time. Behaviour during chaotic flow is not worth worrying about. All the available desk top flight simulators, released before or after FS9, have more important realism issues that need to be addressed more urgently.
<<I bought the Flight replicas Bf-109 K which has the usual FS never stall FDE>>
They all stall. The problem is that the accelerated stall is not recognised by the user and not remedied by the user.
If a tiny shotgun pellet and a very large heavy plank of wood are dropped from a tower at the same moment then the lightweight rounded pellet hits the ground first because it has the lower co-efficient of drag. Both objects are subject to an acceleration of 1G but the plank hits more air molecules on the way down and they retard its chaotic fall.
When the pilot of a Bf109G decides to target corner drag and pull 6G his aeroplane is beautifully aerodynamic and his rate of descent is very high as he spiral power dives at 180 KIAS (drag) and maybe 300 KTAS (velocity). If he allows the wing to stall chaos ensues. He is suddenly sitting on a plank. The co-efficient of drag becomes huge. Falling planks and stalled falling wings are not aerodynamic.
The consequence of stalling from level flight at 1G is a sudden increase in rate of descent. Conversely the consequence of stalling from a spiral power dive at 6G is a sudden reduction in rate of descent. Some FS9 users find this confusing, but it is correct.
As I explained in my first post the lift slope in TBL 404 of the air file can be as accurate as the data you may obtain for CL above and below the stalling angle. I have supplied the real CL for each stall case. The lift post stall is nowhere near zero even though it is far below CLmax. The post stall lift must deliver the rate of descent of a plank in chaotic flow, not the rate of descent of a falling shot gun pellet, and certainly not the rate of descent of a power diving aerodynamic fighter. In the high G case stall arrests descent.
Any of the air files you are describing could have false values for lift in chaotic flow within TBL 404. If you have better values substitute them, but they are not absent, nor is the current value in any of them necessarily false. They may also have false values for degradation of control authority, stability and damping in each axis, for each trim state, for each AoA, post stall. Again if you find chaotic flow data that is more accurate you can substitute it, but representative generic data is unlikely to be absent and the data encoded may be a reasonable generic approximation of pilot induced chaos.
Continued in next post...
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #9 on: Dec 30th, 2006, 1:08pm » Quote Modify
--------------------------------------------------------------------------------
Realistic handling characteristics at the moment of stall are irrelevant. User induced chaos only needs to be signalled and then punished if uncorrected. No one has ever suggested that desk top flight simulators can be used to teach stalling behaviour or incipient spin behaviour at any G, any control surface status, any trim status, any Mach status, etc combined in any way with one another. Users can however teach themselves to avoid chaotic flow conditions by noting, targeting and achieving the relevant IAS for the current phase of flight.
The equations that govern aeroplanes do not differentiate between propliners and fighters and so the issues illuminated above are at least marginally on topic here. I am not sure that we should stray into discussing wider aspects of air combat at all, but anyway there are two cases.
Doctrine is one case and the laws of dynamics are the other. Only doctrine is debatable. It may or may not prove useful, but I will resort to posting one of Newton's equations in the hope of illustrating the dividing line.
In air combat using fixed guns, if two pilots are aware of one another's presence, and both have the same personal or aircraft structure G limit, the pilot with sufficient specific energy (Es) advantage can usually obtain a (fixed gun) fire control solution and can always avoid being the subject of one. He can just drive away.
Specific energy (Es) is sometimes pronounced Energy state. The equations that govern the equilibrium exchange of potential and kinetic energy above planet Earth, published by Newton, are not subject to debate.
Es = (H + v^2) / (2 * G)
Or this can be broken up into two parts;
Es = (H/G) + (v^2/G)
or in Imperial aviation terms as required by Microsoft
Add altitude in feet to TAS squared in feet per second and then divide the total by 64.4
This formula tells us exactly how many feet of altitude one knot of velocity = KTAS is worth in air combat.
10 KTAS = 16.8 ft/sec which when squared is about 280
280 feet of altitude are worth 10 KTAS. They make the same contribution to Es.
Until a fire control solution is imminent we will not trade altitude for velocity at more than 280 feet per 10 KTAS gained. Now remember that diving increases drag = IAS and therefore reduces velocity = TAS.
As Part 1 of the Propliner Tutorial explains;
<<<<It takes MSFS users, (and many real pilots), a long time to understand that if a fighter pilot power dives his fighter from 250 KIAS at 40,000 feet to 400 KIAS at low level he has decelerated from about 500 KTAS to about 400 KTAS. As the fighter pilot dives hard and watches the ASI needle proceed from 250 to 400 he is watching the drag rise, he hears the wind noise screaming ever louder as he decelerates a hundred knots in no time at all. A drag of 400 KIAS at low level ensures that the fighter is much slower than it is with a drag of 250 KIAS at high level. It's just a lot more drag, so we hear much more wind noise. Wind noise isn't an indicator of velocity; it's just an indicator of drag. IAS isn't an indicator of velocity, it's just an indicator of drag.>>>>
Consequently it is always an error to reduce altitude in fixed gun combat until the moment comes to seek a fixed gun fire control solution against a target with a lower (Es) energy state (or to avoid an enemy fire control solution). Only when that moment comes do shooter specific and target specific combat doctrine take over and control the sequence of events.
This thread is about flight dynamics. It is about the size of the theoretical aircraft performance envelope, how much of the theoretical performance envelope is actually usable in real life, how to restrict the FS9 user to those real limits, and how to punish attempts to exceed those real limits. However in any wider context we must not forget that weapon platform performance envelope (battlefield mobility) is a tiny component of combat lethality. The primary components of combat lethality are firepower and durability.
Combat doctrine is devised by intelligence analysts after comparing the firepower and durability of the weapon platform to the equivalent values for the target. The Bf109G had exceptionally low firepower and fairly low durability by contemporary standards. Consequently it needed to acquire an enduring fire control solution from a position where it was safe from target firepower retaliation. Luftwaffe doctrine for attacking a target of a given type using a Bf109G had to take that into account.
Combat begins with detection of the enemy. A 'contact' with the enemy may endure for many minutes. In real air combat the time spent acquiring and then prosecuting a fire control solution, or avoiding enemy fire control solutions is a tiny percentage of that contact. Doctrine applies only to the target acquisition and prosecution phases of the contact. The rest of the contact is driven only by the laws of dynamics and the relentless pursuit of specific energy (Es) advantage.
Having perfected one doctrine relating specifically to the advantages and disadvantages of one particular weapon platform real weapon platform operators are often reluctant to begin the climb along a new learning curve for a different weapon platform which requires different operating doctrine, even if that new weapon platform has a superior performance envelope, superior firepower, and superior durability. To understand the history of the Bf109 it is necessary to understand that it was very popular long after it was outclassed, because the Lachmann auto slats conferred a very low corner drag that gave it a very wide usable performance envelope.
Replicating the dynamics of 'corner' and Vy, both of which are multiples of slat open and flaps retracted stall (Vs), is the key to delivering 'realistic' combat flight dynamics. However replicating the stall characteristics, at the stalling angle associated with Vs, is almost entirely irrelevant. Post stall lift, drag and control authority do not need to be 'accurate'. They only need to be degraded into representative chaos. It is lift, drag and control authority within the usable envelope that need to be carefully researched and replicated and it is only the usable envelope that needs to be carefully limited for both airframe fatigue and aircrew fatigue.
The laws of dynamics that this thread is all about apply equally and without exception to every aeroplane whether or not it is a weapon platform and every pilot whether or not he or she is a combat pilot. However combat doctrine, including fire control doctrine, must also take into account everything else resulting from intelligence analysis that is not equal for all weapon platforms and all pilots.
Continued in next post...
IP Logged
--------------------------------------------------------------------------------
Tom Gibson
California Classic Propliners: www.calclassic.com/
AlcoHauler Locomotive Page: www.calclassic.com/alco/
Freeflight Design Shop: www.freeflightdesign.com/
Tom Gibson
Forum Administrator
Jets are for Kids!
Posts: 2892
Re: Question to FSAviator on Stallvariables in FDE
« Reply #10 on: Dec 30th, 2006, 1:10pm » Quote Modify
--------------------------------------------------------------------------------
If Tom agrees that it is at all appropriate to this forum I offer the following thoughts on WW2 air combat doctrine and WW2 fire control doctrine.
[Sure, no problem. TG]
The Luftwaffe did not have access to gyroscopic, lead computing optical sights (LCOS technology) in the relevant timeframe. Consequently hardly any Luftwaffe fighter pilots were able to calculate a fire control solution with shooter G applied. Luftwaffe fire control doctrine therefore called for all shots to be taken at 1G (i.e. wings level with pitch roll and yaw increment all equal to zero).
Furthermore without LCOS technology the vast majority of fighter pilots could not demonstrate the ability to obtain and prosecute a 1 G fire control solution from the beam or the quarter, even against a non evading sleeve target drone. They failed to apply appropriate deflection, because they consistently failed to estimate range correctly. Having failed to estimate range they could not resolve the Pythagorean fire control solution and could not calculate the angular lead required.
Consequently standard Luftwaffe fire control doctrine called for (sequential) stern attack with zero deflection and with decaying overtake drag (IAS) from six o'clock low after diving from higher altitude with superior Es. The closure began at more than corner IAS at the bottom of the dive and the pull into six o'clock low was placed and timed in 4D so that the subsequent climbing 1 G wings level attack terminated at corner well inside effective range, after obtaining and prosecuting a 1G fire control solution with zero deflection.
Contrary to popular opinion most fighter pilots are not especially gifted and the best pilots may not be posted to fighter units. Consequently many other air forces imposed identical fire control doctrine in the relevant timeframe. The air forces that failed to implement a close replica of this fire control doctrine scored significantly fewer victories per hundred contacts until they were equipped with LCOS.
Marseille is famous for publicly refuting this doctrine and refusing to adopt it. Marseille, and a few other aces, could calculate and prosecute fire control solutions involving both deflection and shooter G, without LCOS, because they alone could estimate both range and lead without them. German propaganda referred to Marseille as 'the flying computer'. However what Marseille failed to understand, but the relevant intelligence analysts understood all too well, was that the Luftwaffe and every other air force had very, very few pilots who could emulate his skill in computing a fire control solution. Marseille is famous for shooting down 8 aircraft in a single sortie and 17 in a single day.
Hartmann by contrast is famous for scoring 352 victories steadily over several years. Unlike Marseille he survived the war and never lost a wingman. Hartmann's combat doctrine was simply Luftwaffe Bf109 combat doctrine, which was to seek only one (successful) fire control solution per sortie and then drive away. Hartmann's personal fire control doctrine was, "Get close .. when he fills the entire windscreen ... then you can't possibly miss.". This doctrine is a clear acknowledgement of the lack of capability of the Revi reflector sight as a fire control computer, but is also a reflection of the increasingly inadequate firepower of the Bf109.
Luftwaffe fire control doctrine for the Bf109 was limited by the fire control technology available, and the often low levels of installed firepower, but was clearly effective, both in terms of achieving kills and in terms of avoiding enemy fire control solutions. Of course many fighter pilots, of many air forces, disregarded fire control doctrine and attempted to prosecute fire control solutions whilst pulling G, but more than 99% of them failed to score enough hits to have a material effect, more than 99% of the time. The durability of the target was too high compared to the firepower brought to bear by the shooter.
Continued in next post...