FM Speculations and Observations (FE2 and SF2)

[+VonS 1,423]

Hello gents',

 

Thought I'd post some speculations and observations on the FM aspects of some things in FE2 and SF2 that I've recently been tinkering with. This might be an interesting post for those who like to follow these technical threads.

 

I was recently installing more aircraft into my small SF2 install - something I wanted to get to last year but had no time - and I decided to install a couple of the late prop aircraft too, particularly the SeaFury and F4U-5 variant of the Corsair. Overall pleased with the handling of the types, I tweaked a few things such as desensitizing the rudder on them, and also giving them slightly heavier aileron feel - since those late model prop aircraft were fairly heavy, about 4500 kg on average.

 

Further testing involved flying and inspection of the [FlightControl] and [Engine] sections...couldn't get much beyond about 500 kph near sea level or at alt. in the types...so I commented out the MaxSpeedSL entry for the FlightControl, and tweaked prop speeds to close to historical, within the 2000 to 3000 rpm range, also pushed up the value for the SLPowerDry entry for the Engine, usually by a noticeable amount, and then tested further.

 

Some of the things I've noticed, in terms of SF2 specifically (but this post is relevant to FE2 too):

 

a) WEP power is not modelled realistically, since you can keep running it without engine damage

 

b) WEP power doesn't give much of a boost, if at all (although I've only tinkered with the SeaFury and Corsair so far)

 

c) a better way of simulating, however simplistic it is, the potential of WEP power, is to manipulate the AltitudeTableData numbers instead (for example, I plugged in values ranging from about 1.05 to 1.35 around the 6000 m alt. band, which should be where the Corsairs benefitted the most from max. throttle)

 

d) after manipulation of the AltitudeTableData, it's best to get rid of the WEP entries entirely, and set up "throttle gradients" instead under the engine sections, such as:

 

ThrottleRate=0.65
IdleThrottle=0.12
CruiseThrottle=0.45
MilThrottle=0.70
MaxThrottle=1.00

 

The numbers above I've plugged in, for example, for the Corsair. The space between Military throttle and Max throttle would be where "WEP power" would theoretically in this case be engaged, somewhere around the 0.85 mark for the band I've selected above. For this to work somewhat realistically, the player should restrict themselves to about 85% throttle max. except in extreme situations, depending of course on where military and max throttle are set. Also important for this to work is to make sure that the max SL speed under the FlightControl section is left uncapped (commented out) - otherwise artificial barriers are set up that cannot be bypassed by more than about 20-30 kph.

 

Speed caps work beautifully in SF2 for the jet engines, but, at least from my experience, don't work that well with the late prop planes since you can then never hit top max. recorded historical speeds, not even at alt. - and not at SL either. Also noticeable in the example above is that I've increased the throttle rate from the default of about 0.25. Historically, late prop aircraft would have had better acceleration than early jet planes (not better top speeds, but faster acceleration to their own max speed). By the 1960s/70s, jet engines should surpass the props even in acceleration.

 

I then re-tested by dumping the modified Corsair into my FE2 install. In SF2 it can get into the 760s kph, as historical, using the engine and throttle values I've included in my graded throttle settings above, while in FE2 the same tweaks push the top speed into the 880s kph - a difference of about 120 kph for top speeds, across FE2/SF2.

 

This is why it's not a good idea simply to transfer data inis from one of the sims to the other. I'm assuming that the discrepancy in top speeds means that TK went with calculations to give less air resistance in FE2 (maybe FE as well?) - to fit better with the more fragile and maneuverable WWI aircraft types. One thing that remains constant between FE2 and SF2 is that, at least from my observations, the WEP modelling is not very realistic - there was no high alt. band where I could push the top speed beyond, for example, a reasonable top speed that may have been possible at near sea level (such as 500 or 600 kph).

 

Some preliminary conclusions, to be tested further:

 

a) speed caps at sea level should be kept in place in FE2, as done in the data inis in my FM packs, since they fit better with the lower top speeds of the early aircraft anyway, and mach limits should also be kept fairly low (I have them set at 120% of the "top speed" of the WW1 aircraft....this allows for realistic dive speeds that I haven't been able to push beyond about the 460 kph mark even on the late war types such as the SE5a)

 

b) flying late WW2 and early Cold War prop aircraft in FE2 is not recommended, but FE2 may be good for "Golden Age" stuff, as Geezer speculated upon in another thread (can't remember the link for that now), in other words for aircraft that don't have a top speed higher than about 350 or 400 kph

 

c) the difference between top speeds at alt and at sea level are modelled just fine for the jet engines in SF2, since caps on SL speeds, and Alt speeds, work fine for those (have seen this in some Flogger tweaks that I did for my mini-SF2 install); in other words, SF2 respects the wet/dry mach values and alt table entries beautifully for jet-age stuff

 

d) I think that the MaxThrottle value in SF2, for the prop aircraft, could use some more tinkering/testing...I've set it at 1.00, but will try other values of 1.20, 1.50 etc., to see if this creates more realistic differences between regular power and WEP, or some kind of "accordion" effect

 

e) as it stands, I'm enjoying the SeaFuries and Corsairs but only with SL speed uncapped and giving historical top speeds not at alt. but near sea level (this is a compromise of sorts since, at alt, they still make about 600 kph and sometimes a bit more - so I'm not terribly irritated by not attaining 740 kph at 7000 m alt or so, for example, in the SeaFury, as recorded for that plane)

 

f) a couple of pics included below of the AI in SF2 now losing control of their F4U-5 Corsair during a scrap, and doing several back flips (too much throttle applied perhaps?); he then bails out of the otherwise undamaged plane; so, overall, I'm liking the realism of some of the haphazard maneuvers they now make, and also the increased speed at which these late prop fights now happen, even on the deck

 

g) anything that I find of further use in the SF2 data inis, and if relevant to FE2, I will tinker with more to bring even more realism into FE2 (the lockout speed trick implemented for control surfaces in the alternative Morane-Saulnier type N data ini posted under the FM thread for FE2 is an example of things discovered while combing through the SF2 files)

 

Happy flying,
Von S 

 

Edited March 20, 2019 by VonS
Fixed typos.

[Wrench 9,429]

wonder if you didn't exceed the Structrual Limits on the Corsair? That may cause hidden damge??? 

I'd also suggest looking over the data tables in the stock SF2 P-51D and Spitfires (9c & e, 18, 22 & 24 -even though the latter are Griffon powered marks.

The S-99 (109G) and S-199 don't have WEP installed

[+Geezer 3,569]

VonS -

It's always a pleasure to read one of your analytical posts.  Keep up the good work! 

[KJakker 875]

Funny you should be talking about Corsairs and performance data. I just watched this video the other day.

Also there was an interesting comment by YouTuber Tom Everhart about Corsairs in Korea. Maybe @Wrench know some history as to whether this story is true or not.

Quote

When the U.S.  first started using Corsairs to  "High Speed Dive"   to attack caves in Korea,  they ran into a Problem.    If they were using the  "Newer F4U-5  Corsairs",   the plane could attack at 400 mph.  and  "Pull Out With No Problem".  If they were attacking with an earlier WW-2 version Corsair,   when they did a  "High Speed Pull Out",  the Corsair might roll hard to the left,  while the Corsair behind him might roll hard to the right during the same maneuver.    Not being able to figure out and solve the problem locally,  they finally go a hold of an  "Old School Chance Vought Engineer",  who told the to change out the Ailerons with  "New Ailerons".   The Mechanics protested,  saying that there was  "Nothing Wrong With The Ailerons On The Planes".    The Engineer's response was  "Just Do It And Go Try The Dives Again".  Sure enough, the problem was solved.  It turns out that the Corsairs have Wooden Ailerons and the glue was  "Old and Dried Out"  to the Point that at those  "High Speed Pull Out Stresses" would cause the Old Wooden Ailerons  to Warp,  thus causing the Plane to turn toward the  Worse Aileron during the Pull Out.  Who knew?  Regards,  Tom

 

[+VonS 1,423]

Didn't think this thread would prove so popular but happy to see some fresh posts here gents. Regarding the Corsair, I'm surprised to learn that they used wooden ailerons, fabric covered, for later models such as the F4U-4, was thinking initially that this is something only typical of early war types like the Hurricanes, early Spits, and the "birdcage" variant of the Corsair. Apparently trim tabs were also made of wood on many of the types...a legacy of WW1 maybe?...wood/fabric being easier to fix from bullet holes perhaps.

In terms of my further testing of the FMs:

It appears that MaxThrottle values manipulated beyond 1.00 (100%) to something like 1.2, 1.5, 2.0, etc., don't make any noticeable difference. I am testing this with WEP tables and values removed entirely. I do have another theory that I want to test, to see if max throttle values beyond 1.00 are somehow relevant only to WEP data, but not to regular alt table data values and regular throttle gradients.

Will look at the stock SeaFury data ini with WEP enabled and will toggle max throttle settings further to see if I can at all hit max historical top speeds at alt. I'll also look into tinkering with max throttle values of between about 1.05 and 1.10 or 1.15 - perhaps a slight increase beyond 100% will, however counter-intuitive, give better results than noticeable increments.

I'll report back if any of these tests and combinations manage to give higher top speeds at alt. than at sea level. For now, I'm still running the late prop aircraft with SL top speed uncapped and with engine power increased to give historical top speeds recorded at alt., but in this case reachable only at sea level. A compromise of sorts yes but I can tolerate it . Would be nice however to get WEP power realistically implemented (in SF2), although for FE2 there's no need for it.

Von S

[+VonS 1,423]

Been tinkering some more with the data inis, this time with the Lavochkin-11 since it flies similarly to the Extra-300 aerobatic plane from what I've noticed in SF2 and is fairly high performance. Will incorporate this into several of the late prop aircraft soon hopefully and will upload the data inis under the downloads section for SF2 if anyone's interested (SeaFury, SeaFury export version, F4U-4 and F4U-5 Corsairs, Yak-9P, LA-9 and LA-11, also "Sprint" upgrades for the Thunderbolt, and I'm tempted to look at the A-Team Superbolt what-if models and Spit Mark 14, also Paulo's Mark 19 - this is enough for an early Cold War install - the WW2 stuff I doubt I'll look at tweaking since I'm usually busy with WW1, but using the tweaks below people can cook up their own solutions to the WW2-era types).

Below follow the latest modifications for the LA-11 data ini, and I think that WEP power is now modelled correctly from what I've been able to experience. 

 

SLPowerDry=2037317.5 ----------> this value had to be increased noticeably from stock, sometimes by values of half a million, sometimes more, to give the engines the necessary brute force
ThrustAngles=0.0,0.0,0.0
ThrustPosition=0.0,1.2,0.0 ----------> I've slightly reduced the middle thrust value on several of the types
ThrottleRate=0.50 ----------> higher value for more acceleration, needed on these late prop, early Cold War aircraft
IdleThrottle=0.12
CruiseThrottle=0.55
MilThrottle=0.80
WEPThrottle=0.95 ----------> important discovery, the WEP value should follow somewhere between military power and max throttle, not after the max throttle value (higher than max throttle)
MaxThrottle=1.25 ----------> the "accordion" effect works when the wep value is lower than max throttle, then once we expand max throttle beyond a value of 1.00, the extra power kicks in (this I may tamper with further but wep values of 0.95 and max throttle of about 1.25 work well, giving a boost of about 15 kph once we go into the wep range at altitude, but not at sea level where it's disabled now)
AltitudeTableNumData=61 ----------> this was incorrectly recorded as 64 initially, now fixed
AltitudeTableDeltaX=350 ----------> smaller intervals in meters for more subtle changes
AltitudeTableStartX=0.0 ----------> start at the zero interval for alt table data (sea level)
AltitudeTableData=1.00,0.9999183,0.999698,0.999339,0.9988413,0.9982049,0.9974298,0.996516,0.9954635,0.9942724,0.9929425,0.991474,1.0297504,1.107311,1.224733,1.3020163,1.1291609,1.0261668,0.987311,0.984733,0.982016,0.980609,0.976166,0.963034,0.9597625,0.9563524,0.9528035,0.949116,0.9452898,0.9413249,0.9372213,0.932979,0.9285981,0.9240784,0.9194201,0.914623,0.9096873,0.9046129,0.8993998,0.894048,0.8885576,0.8829284,0.8771606,0.8712541,0.8652088,0.8590249,0.8527023,0.8462411,0.8396411,0.8329024,0.8260251,0.8190091,0.8118543,0.8045609,0.7971289,0.7895581,0.7818486,0.7740005,0.7660136,0.7578881,0.7496239

 

// ----------> wep values below are basically a reduplication of the non-wep alt. values included above (I've copied these from the Corsair files and then tweaked further as per aircraft type)

WEPAltitudeTableNumData=61 ----------> fixed to 61, from 64
WEPAltitudeTableDeltaX=350
WEPAltitudeTableStartX=12.0 ----------> this took a while to figure out but is important since otherwise WEP does not work, the value corresponds to the value in the wep alt table data below, counting from zero, which is the sea level value (with a value of 12.0 for the startx value, the wep boost that gives me 125% throttle kicks in somewhere around 4700 meters alt. for the Lavochkin in the case of my tweaks, close to the "1.107311" value below, the power boost being most noticeable around "1.3020163" below, which places us around 5200 meters altitude where I can get to about 680 kph on full wep power)

 

WEPAltitudeTableData=1.00,0.9999183,0.999698,0.999339,0.9988413,0.9982049,0.9974298,0.996516,0.9954635,0.9942724,0.9929425,0.991474,1.0297504,1.107311,1.224733,1.3020163,1.1291609,1.0261668,0.987311,0.984733,0.982016,0.980609,0.976166,0.963034,0.9597625,0.9563524,0.9528035,0.949116,0.9452898,0.9413249,0.9372213,0.932979,0.9285981,0.9240784,0.9194201,0.914623,0.9096873,0.9046129,0.8993998,0.894048,0.8885576,0.8829284,0.8771606,0.8712541,0.8652088,0.8590249,0.8527023,0.8462411,0.8396411,0.8329024,0.8260251,0.8190091,0.8118543,0.8045609,0.7971289,0.7895581,0.7818486,0.7740005,0.7660136,0.7578881,0.7496239
PropDiameter=3.375 ----------> this I increased slightly since it looked too small but such details are easy to modify
PropPFactorCoefficient=0.0
PropTorqueFactor=0.02
PropRotationDirection=1
ReversePropOrientation=TRUE
SpinnerNodeName=Spinner
StaticPropNodeName=SlowProp
SlowPropNodeName=SlowProp
FastPropNodeName=FastProp

 

// ----------> prop rpm values below are estimates but fairly close, most of these aircraft wouldn't/couldn't exceed about 3000-3100 rpm

SlowPropRPM=1700
FastPropRPM=2500
MaxPropRPM=2900
IdlePropRPM=500
BSFC=0.01
ExhaustEmitterName=CleanExhaustEmitter
ExhaustPosition=0.0,1.5,-0.3
FireSuppression=TRUE
MinExtentPosition=-0.5,2.2,-0.5
MaxExtentPosition= 0.5,0.8, 0.5
MaxPowerAnimationTime=0.1
MinPowerAnimationTime=0.3
ConstantSpeed=TRUE

 

// ----------> borrowed the more subtle prop efficiency numbers from the Corsair

PropEfficiencyAdvanceRatioTableNumData=27 ----------> this was incorrectly recorded as 26, now fixed to 27
PropEfficiencyAdvanceRatioTableDeltaX=0.10
PropEfficiencyAdvanceRatioTableStartX=0.0
PropEfficiencyAdvanceRatioTableData=0.1039061,0.2017555,0.3307556,0.4630645,0.5773853,0.6674284,0.7350242,0.7854134,0.8242808,0.8561408,0.8837222,0.9080403,0.9388862,0.9455014,0.9572477,0.9641239,0.997018, 0.9976275,0.968019, 0.9698392,0.9732296,0.9955389,0.9699651,0.9443028,0.8800089,0.7018427,0.4483748
MomentOfInertia=4.0 ----------> this number was too powerful on some of the types, or too weak on others (lower number is more powerful)....the WWI types are usually in the 7.0 to 12.0 range, values of about 3.0 to 6.0 work better for these post-WW2 prop types

 

Three other important things to be aware of:

1. leave the max sea level top speed commented out in the FlightControl section of the data inis since that entry is no good for late prop types (it will artificially cap top speeds at all altitudes, not just sea level)

2. for max realism, use only full military throttle near sea level...top speeds near sea level should be around 600-650 kph then....seems reasonable, less/more depending on what the top speed at alt. is anyway on the various aircraft types...the good thing now is that even if you go full throttle at sea level, you are not engaging wep boost that pushes beyond 100% throttle since you are not at the necessary alt. (determined by the startx value in the wep chart above)

3. I'm still thinking about whether or not it's a good idea to include the values for max time of wep power...since, technically, once a set time runs out, you can again lower throttle, go back up again, and you have wep power, unless there's some kind of timeout command that can be used to force an "interval" between wep on/off...I might not bother therefore with wep max time limits since wep already guzzles gas once you punch it up to max, so that's a good enough limiter itself 

Pics below of the LA-11 during tests and another Corsair that the AI lost control over during excessive maneuvers at high speeds - I'm thoroughly enjoying the increased realism now with these "brute force" machines.

Von S

 

Edited June 2, 2019 by VonS
Fixed errors in the post.

[+Coupi 4,279]

Hello VonS, 

Very interested in your SF2 FM stuff which is high level experience ! 

After my job time... I go to test your WEP data.

Flying your FM work is the opportunity to learn more about the aerodynamics possibilities of FE/SF2 games. 

Thanks a lot, 

Best regards,

Coupi.

Edited March 25, 2019 by Coupi

[+mue 707]

Does anyone have a documentation (e.g. formulas) how the resulting thrust is computed from the FM ini values at given input values (throttle, altitude, velocity)?

[+VonS 1,423]

10 hours ago, mue said:

Does anyone have a documentation (e.g. formulas) how the resulting thrust is computed from the FM ini values at given input values (throttle, altitude, velocity)?

A good question...I think that only TK really knows how those formulas add up and work in FE2/SF2...my impression is that much of the number crunching that we tweak into the data inis are manipulations and changes to "already established" formula averages that were directly programmed into the TW series of sims. The trick is to figure out what units TK was using...most of the data inis are in metric (m/s, etc.), but some of the ratio numbers took a bit of guesswork to figure out, and actual testing in the sims.

Perhaps Baffmeister knows more about thrust formulas, etc. - might be a good idea to ask him. Most of the mods I've done for FE2 are all hand-edited and then tested, over and over again, within FE2 - an approach that, while slow, works well for WW1, but I haven't really touched any of the jet-age stuff where thrust is handled in more complicated ways. Here are a few links I've managed to find on CombatAce that might be of some help:

https://combatace.com/forums/topic/61789-calculating-a-boosters-acceleration/

https://combatace.com/forums/topic/93066-used-fm-units-of-measure/

https://combatace.com/forums/topic/89652-aircraft-acceleration-values/

Von S

[+Coupi 4,279]

Hello,

I'm testing the WEP with the La-11... 

In the [Engine] section is it necessary to add the following line : HasWEP=TRUE ?

Best regards,

Coupi.

[+VonS 1,423]

1 hour ago, Coupi said:

Hello,

I'm testing the WEP with the La-11... 

In the [Engine] section is it necessary to add the following line : HasWEP=TRUE ?

Best regards,

Coupi.

While I've noticed that I get about 12-15 extra kph with the La-11 at around 5200 m alt. already, it would be a good test to include the HasWEP=TRUE value as well, under the Engine section. I am also comparing some of the types further with modifications to the Engine section based on my long post above...currently working on the Thunderbolt "M" upgrade tweaks and the what-if Superbolts from the A-Team...will post if I find any important differences between performance on the different types. Also, I recommend testing with the SLPowerWEP value also incorporated into the Engine section of the data ini, and then without that value included. I am currently testing with SLPowerWEP included for the Superbolt, and I am noticing some extra, subtle acceleration - although no change to the top speed. Fortunately I'm able to run it in the 780s kph which is close to the historical top speed for the first prototype of the Superbolt. It took some time to get everything running with the Superbolt since I had to incorporate many alpha and data table values for the wing sections, borrowed from the Thunderbolt data inis hosted on CombatAce. Otherwise, in stock form, the A-Team Superbolt would start descending, above about 5000 m alt., if running beyond about 650 kph, with no ability to correct the descent - now it works fine with the borrowed wing values in place.

My engine dry/wep values for the P-72A-1 prototype are:

SLPowerDry=3072665.0
SLPowerWEP=3572665.0

I will test further by reducing wep to about a value of 3272665 - I think a difference of a value of 200,000 units, approximately, between the non-wep and wep entries, should be adequate for subtle performance differences.

Also, I have noticed some other oddities regarding the late prop aircraft that are worthy of some info. here, regarding the wep manifold pressures.

From what I've been able to read over the past few days...manifold pressures, if wep power was available, for early WW2 types were around 52 inches, about 64 inches for late WW2 types, with a possible 70 inches being used on these post-war variants I'm currently tinkering with for my Cold War SF2 install. Manifold pressures of 72, 74 inches, etc., are only used on modern performance and racing planes. What is confusing in the data inis is that some files record these values in inches, others in "ata" (absolute atmosphere) units...I think TK was most likely using the ata units. The general rule is 30 inches of manifold pressure equalling about 1.0 ata units, so:

early WW2 types, WEPManifoldPressure=1.73

late WW2 types, WEPManifoldPressure=2.13

post-WW2 types, WEPManifoldPressure=2.33

I've as well increased the WEPTempDelta value from 10.0 to 20.0, hopefully for more overheating chance although I haven't seen any differences so far.

I've currently gotten rid of the entry for max alt. at which WEP operates, so that it operates no higher than the max alt included at the top of the data ini, and I've removed the max. time for WEP duration entry since it seems largely a nuisance at this point, although I will keep testing further.

Von S 

Edited March 26, 2019 by VonS
Added info.

[+mue 707]

3 hours ago, VonS said:

My engine dry/wep values for the P-72A-1 prototype are:

SLPowerDry=3072665.0
SLPowerWEP=3572665.0

I will test further by reducing wep to about a value of 3272665 - I think a difference of a value of 200,000 units, approximately, between the non-wep and wep entries, should be adequate for subtle performance differences.

The unit is Watt (W).

 

3 hours ago, VonS said:

From what I've been able to read over the past few days...manifold pressures, if wep power was available, for early WW2 types were around 52 inches, about 64 inches for late WW2 types, with a possible 70 inches being used on these post-war variants I'm currently tinkering with for my Cold War SF2 install. Manifold pressures of 72, 74 inches, etc., are only used on modern performance and racing planes. What is confusing in the data inis is that some files record these values in inches, others in "ata" (absolute atmosphere) units...I think TK was most likely using the ata units. The general rule is 30 inches of manifold pressure equalling about 1.0 ata units, so:

early WW2 types, WEPManifoldPressure=1.73

late WW2 types, WEPManifoldPressure=2.13

post-WW2 types, WEPManifoldPressure=2.33

MinManifoldPressure, MaxManifoldPressure and WEPManifoldPressure are not used for the flight model. Those values are used only for the display of the cockpit manifold pressure gauges.

[+VonS 1,423]

47 minutes ago, mue said:

The unit is Watt (W).

 

MinManifoldPressure, MaxManifoldPressure and WEPManifoldPressure are not used for the flight model. Those values are used only for the display of the cockpit manifold pressure gauges.

Thank you for the info. - this will save me a lot of time with my SF2 install, not having to tweak such entries if they are only cosmetic. (Makes me wonder about what other entries are only cosmetic and that I've maybe been approaching too seriously.) 

In terms of watts used for powerdry/powerwep values, etc., one word of advice for those tweaking FMs is to be aware that historical watt numbers for the engines don't necessarily and always translate to historical results when flying in the TW sims. At least for my WW1 tweaks package, endless testing was needed to get the models behaving historically within FE2 - although sometimes this would require unhistorical engine power numbers, or higher/lower rpm values for the props, usually a balancing act between the two, etc., and of many other factors.

Theoretically, historical numbers might work well in the data inis all the time, but this would probably require entire revamps of several data inis, a very time-consuming process....touching on drag values, lift/data tables for wings and stabs, etc. My approach with the WW1 models has been to "hammer" the FMs into a shape that works well historically, based on the wonderful FM work initially done by Ojcar, TexMurphy, and others - and even then it takes at least a couple of years to comb through the data inis. The A-team data inis usually require more work...I have limited myself to touching only 5 or 6 of their WW1 aircraft.

Sure wish we had some kind of program for FE2/SF2 that would "visualize" (as in a wind tunnel) what's happening when we change and crunch numbers into the data inis...would be more efficient than doing endless cycles of testing in the sim itself...but it is what it is. 

Happy flying,

Von S

Edited March 26, 2019 by VonS
Added info.

[+mue 707]

After sifting through the archive of the thirdwire forum: Third Wire.7z I found the formula for thrust computation:

Thrust = Power / Velocity * Efficiency
Power is dependent on throttle setting and is multiplied with the coefficient from the altitude table
Efficiency is taken from the prop efficiency table

With the FM set to Normal the thrust (from debug output) matches this formula exactly. But with the FM set to Hard, the thrust in game is smaller.
In the forum archive there is a thread "TK, ran into a constant speed prop thrust issue" that covers this issue:
From TK:

Quote

Hm, actually, looking at that graph, although the equation of thrust is the same, the V probably isn't, between NORMAL and HARD FM with 1.0 Advanced Ratio Table, that is... I'm probably adding induced velocity (correctly or incorrectly), causing the V to be slightly higher (and thrust to be slightly slower). In NORMAL FM, velocity is the same all over the aircraft, but in HARD, each component has slightly different local velocity.

and

Quote

The code snippets for induced velocity (at prop position, its different at further downstream):

  InducedVelocity = 0.5 * (-vel + sqrt(vel*vel + (2.0*thrust) / (rho*disc_area) ));

If I understand this correctly then the InducedVelocity must be added to the Velocity in the thrust formula. But in the above formula the InducedVelocity is dependent on thrust (but for the thrust calculation I need the InducedVelocity). Maybe the thrust variable in the InducedVelocity formula is different to thrust calculated by the thrust formula? Does anybody have an idea and can shed some light on this?

[+VonS 1,423]

Hello gents,

I wouldn't be surprised at all if TK's formula for thrust in InducedVelocity is different from thrust calculations in the thrust formula - I haven't been able to solve that riddle but thanks for letting us know about it Mue. Maybe some SF2 jet specialists will be able to pitch in with their thoughts on it. 

On a different note, I'm happy with my latest tests for these post-WW2 prop aircraft so will most likely continue installing some of the tweaked types into my SF2 install as is, after the modifications. I've tried to go with an "accurate enough" approach for this stuff, while I tend to be nit-picky with the WW1 FMs.

Have flown the first prototype of the Superbolt further, with the debug mode set to true in the huddata.ini file in the Flight folder, and was observing what happens. At alt., for example, it's misleading to look at the kph speed listed in the bottom left corner of the SF2 screen. Looking at the KIAS and KTAS numbers in the debug screen is more important. Flying the Bolt at an alt. of about 8120 meters, and at full WEP throttle, I get numbers like KIAS 334.3 and KTAS 514.0, while 93% throttle, just under WEP power, gives KIAS 324.2 and KTAS 498.6. I don't know how accurate the KIAS/KTAS numbers are in SF2...KTAS seems a little too high compared to the KIAS numbers, but an ave. of the low and high values at 93 and 100% (WEP) throttle gives me numbers of 411.4 knots and 424.15 knots. Such numbers translate to about 762 kph at alt. at non-wep, and 786 kph at full wep power. Seems like a reasonable difference to me, of about 25 kph with wep engaged, and I'm then hitting the 780s kph which is close to historical top speeds at alt. for the Superbolt (first prototype). [EDIT: have decided that it's better to set a mach limit value under the FlightControl section of the data ini for these post-WW2 prop aircraft than to use averages of KIAS and KTAS - setting the mach limit value about 100 kph above the historical top speed recorded at alt. for these types usually pushes them to within 20-30 kph of top historical top speed at alt. - avoid mach limit numbers identical to top speeds since top speeds are then too low by about 100 kph - this info is not relevant to the FE2 tweaks but SF2 is a learning process and anything of value, if possible, will be re-integrated into the FE2 data inis.]

I'm not going to bother bringing the KTAS values down into the historical range since then KIAS will drop too much also - this will end up requiring some fairly low watt numbers for the engine power value under the Engine section of the data ini - so, if anyone wants to tinker further with this, I also recommend checking my lengthier post above where I've indicated some important entries, in bold, in the data ini - in that case for the La-11.

Three things that haven't changed since I began testing these tweaks for my SF2 install:

a) leave SL top speed uncapped under the Flight control section of the data inis for WW2 and post-WW2 prop aircraft, since it creates an artificial barrier for these powerful late-prop aircraft, and run at no more than "full military power" at near sea level (whatever the throttle setting might be in your data inis)....I've set this to about 80 or 85% depending on the aircraft...as indicated in one of my posts above, even if you push to full wep power at SL, you will get a slight boost in acceleration possibly, but it won't push the top speed that much if at all, until the relevant alt. table values for wep kick in (determined by the number indicated for the wep altitude startx value)

b) a good difference between max engine power non-wep and with wep, under the Engine section, is about 200,000 watts...this gives a noticeable but subtle difference...no rocket-powered, artificial boosts are felt this way

c) the third point: set the entry under the Engine section, for constantspeed for the prop, to FALSE (that's right) - most of those data inis have it set to TRUE - but testing in debug mode shows that the prop speed is then constantly at max. rpm value, no matter the throttle position, and the prop efficiency is constantly in the 95-100% range - setting that entry to False solves the problem (and then there are nice changes to prop rpm numbers, and efficiency, across the throttle bands) - hopefully that entry is not simply a cosmetic one ...but here are some numbers for the tweaked Bolt data ini, with constantpropspeed set to False:

3200 rpm, full wep

3067 rpm, minimum wep

3007 rpm, just under wep (93% throttle)

2407 rpm, 70% throttle (bit less than full military power)

1900 rpm, half throttle (cruise)

847 rpm, 10% throttle

 

I might eventually post some of these tweaked data inis for the post-WW2 prop machines once I finish tinkering with my SF2 Cold War install....in the meantime the more relaxing skies of FE2 are waiting for me. 

Happy flying,

Von S

Edited April 2, 2019 by VonS
Added some info. and threw in an important edit.

[+Geezer 3,569]

7 hours ago, VonS said:

....in the meantime the more relaxing skies of FE2 are waiting for me. 

Warm up your white scarf and enjoy yourself.  

[Wrench 9,429]

I would like to pin this thread, it that's ok with you,VonS.

There is a WEALTH of data that shouldn't be lost. I'll also create a link to it in the SF2 Knowledge Base.

thanks to all! 

[+VonS 1,423]

4 hours ago, Wrench said:

I would like to pin this thread, it that's ok with you,VonS.

There is a WEALTH of data that shouldn't be lost. I'll also create a link to it in the SF2 Knowledge Base.

thanks to all! 

Sure thing Wrench, it's fine with me if you want to sticky this thread.

Once I tweak these data inis further I might post them under the downloads section for SF2 (I still need to comb over the A-team Spit 14, Paulo's Spit 19, and I found three more I want to check over for my install...Paulo's La-7, F-6 (French) version of the Mustang, plus the F4U-7 variant of the Corsair used by French naval services. That should be enough variety for an early Cold War prop install in SF2.

Von S 

Edited March 30, 2019 by VonS
Added info.

[+LloydNB 212]

[+mue 707]

On 30.3.2019 at 2:21 AM, VonS said:

At alt., for example, it's misleading to look at the kph speed listed in the bottom left corner of the SF2 screen. Looking at the KIAS and KTAS numbers in the debug screen is more important.

The speed in the bottom left corner and the KIAS number in the debug screen show the same airspeed: indicated airspeed (IAS). Only the unit in the bottom left can be different (km/h or mph instead of knots).

On 30.3.2019 at 2:21 AM, VonS said:

Flying the Bolt at an alt. of about 8120 meters, and at full WEP throttle, I get numbers like KIAS 334.3 and KTAS 514.0, while 93% throttle, just under WEP power, gives KIAS 324.2 and KTAS 498.6. I don't know how accurate the KIAS/KTAS numbers are in SF2...KTAS seems a little too high compared to the KIAS numbers, but an ave. of the low and high values at 93 and 100% (WEP) throttle gives me numbers of 411.4 knots and 424.15 knots. Such numbers translate to about 762 kph at alt. at non-wep, and 786 kph at full wep power. Seems like a reasonable difference to me, of about 25 kph with wep engaged, and I'm then hitting the 780s kph which is close to historical top speeds at alt. for the Superbolt (first prototype).

True airspeed (KTAS) at a given altitude translates to an indicated airspeed (KIAS) and vice versa. (In the game TK uses equivalent airspeed (EAS) for indicate airspeed) An online converter for airspeeds you can find here: http://www.hochwarth.com/misc/AviationCalculator.html
Therefore IMHO it's not correct to average both types of velocities and compare it with real life data. Depending on what real life velocity is measured (indicated airspeed or true airspeed) the according numbers (KIAS or TIAS) should be compared.

On 30.3.2019 at 2:21 AM, VonS said:

c) the third point: set the entry under the Engine section, for constantspeed for the prop, to FALSE (that's right) - most of those data inis have it set to TRUE - but testing in debug mode shows that the prop speed is then constantly at max. rpm value, no matter the throttle position, and the prop efficiency is constantly in the 95-100% range - setting that entry to False solves the problem (and then there are nice changes to prop rpm numbers, and efficiency, across the throttle bands) - hopefully that entry is not simply a cosmetic one ...but here are some numbers for the tweaked Bolt data ini, with constantpropspeed set to False:

I think with ConstantSpeed=TRUE a constant speed propeller is simulated in the game. (AFAIK most WW-2 and post WW-2 prop aircraft have constant speed propellers.) The corresponding automatic variable pitch is simulated with the high prop efficiency over a broad range of the advance ratio.

From TK:

Quote

PropEfficiencyAdvanceRatio is the table of Prop Efficiency vs Prop Advance Ratio. For fixed pitch propeller (ConstantSpeed=FALSE), it will have a peak at designed Advance Ratio (ratio of air speed vs propeller tip speed). For variable pitch prop, you can enter the envelope of the tables for all the pitch angles (to simulate the prop always being at the best pitch).

and

Quote

In real life, propeller efficiency is a function of advance ratio and prop pitch, and you'd have different efficiency curves for different prop pitch angles. Low prop pitch angle results in a peak at lower speed (and lower advance ratio), and higher prop pitch angle has a peak at higher speed (and higher advance ratio). Usually, this curve is fairly narrow for a specific prop pitch angle. But for automatic prop pitch control, you can just give it a wide curve where it stays at 1.0 over a range of advance ratio to model that fact that prop pitch is being adjusted constantly to give it max peak.

HTH

[+VonS 1,423]

On 4/1/2019 at 5:06 AM, mue said:

(Therefore IMHO it's not correct to average both types of velocities and compare it with real life data. Depending on what real life velocity is measured (indicated airspeed or true airspeed) the according numbers (KIAS or TIAS) should be compared.

I think with ConstantSpeed=TRUE a constant speed propeller is simulated in the game. (AFAIK most WW-2 and post WW-2 prop aircraft have constant speed propellers.) The corresponding automatic variable pitch is simulated with the high prop efficiency over a broad range of the advance ratio.

Thank you for the info. - have been testing further and I've discarded my average of kias and ktas - found a better way of capping the ktas to within reasonable limits, using the MachLimit value under the FlightControl section of the data ini...was testing with the Spit 14 and now I can't push beyond about the 720s or 730s kph (when converted from the ktas value) at an alt. of about 7100 meters - the key is to "peg" the machlimit value appropriately. Limiting it to the historical top speed at alt. seems to "buffer" top speeds into somewhat low, unrealistic values...but capping the machlimit at a value that converts to about 100 kph higher than top speed at alt. works fine to keep the aircraft to within about 20-30 kph of historical top speed at alt. - so I'm happy with the results. I will still however leave the SL top speed uncommented since I'm getting better performance that way.

For those who are following this thread, I recommend capping the machlimit value at about 100kph higher than historical top speed at alt. - seems to work well (this advice is for late prop aircraft in SF2...not related to FE2). The other good thing is that even with this cap in place, it doesn't limit dive speeds...some of the types like the Spits can approach Mach 0.9 in dives at full throttle anyway.

In terms of the constantspeed "true" value, I agree that it's realistic in terms of post-WW2 and WW2 aircraft props. This I recommend tweaking as per taste or leaving it on for those who don't like fiddling with the values and want to have true auto prop behavior.

Technically, on those constant speed props, you had a separate lever that you could use to chop down the rpms to taste - I don't know if this is possible in SF2 so I've resorted to turning the constantspeed off since otherwise I get constant, max rpms at all throttle settings, even at idle. Turning it off is a crude approximation of being able to control the rpms/prop pitch manually I suppose  - at any rate it gives me a nice rpm of about 2300-2500 at full military throttle and rpms below 1000 at idle. I'm assuming that the automatic variable pitch then becomes "manual" variable pitch since the prop efficiency percentage changes across throttle settings...I usually get best prop efficiency (around 90 to 95%) somewhere in the 40 to 80% throttle range with constantspeed set to off - this seems historical...jet engines from what I've read are more efficient in the 80+ % throttle range. Once I start pushing towards full throttle I drop to about 70% efficiency at lower alts., but the efficiency improves slightly at max throttle (wep) at higher alts., etc. I suppose it's a tradeoff of sorts but seems to work well enough in SF2.

Happy flying,

Von S 

Edited April 2, 2019 by VonS
Added info. and some edits.

[+VonS 1,423]

Pleased to report that, after a couple of weeks of further tinkering, I've managed to put a cap on the SL speeds for the post-WW2 types being tested while managing to exceed those SL speeds noticeably at higher alts. Further discovery - best is to cap the MachLimit speed at about 30 to 60 or so kph higher than historical max speeds at alt. - similar to the "120%" rule I've been using for mach limits in FE2. This gives more accurate results at alts. in SF2 than the more liberal mach cap of 100 kph higher than historical (which leaves too much room for inaccuracies in speeds).

Also been tinkering with the climb rates on some of the types, as indicated in the pic below I found on one of the WW2 forums. More thorough info. regarding the tweaks will be included in the Read Me file for the package itself - now to double-check the files and upload everything under the SF2 downloads section in the near future. All in all an interesting experiment and I'm happy with the results - hope others enjoy flying the tweaked types too.

If I ever decide to do a WW2 install in SF2, I might use some of the discoveries from these post-WW2 prop. tweaks and then "roll them back" into the WW2 types - but this is a project for a later date. Once I upload this package I want to return to modding some of the types not yet touched in FE2...the Snipe comes to mind, as do some of the two-seaters.

Happy flying,

Von S

 

Edited April 16, 2019 by VonS
Fixed typos.

[+Coupi 4,279]

I like your stuff ! Many many thanks !!! 

[+VonS 1,423]

Stephen's beautiful model for the Da Vinci flyer got me thinking today about butterflies, dragonflies, mechanical birds, early flight in general, also horse power and ox power and other (obsolete) power measurements and ratios (oxes are like torque since they have more "pull power" than a draft horse, but the horse always has more hp...I think the old rule is that an ox can do about 65% of the work of a draft horse). In other words, it only has about 2/3s of the hp of a horse, but that makes sense since it's not a real horse.

It seems that the major problem with early powered flight (man or engine powered) was the power to weight ratio (inefficient) of early engines, especially steam engines. Clement Ader's "Eole" flyers of the 1890s (the ones with bat-like wings) were about 150 to 250 kg in weight (empty), and had cumbersome steam engines to power them...of about 20-30 hp. It's possible that they got off the ground briefly but had no pilot control.

The Wright brothers also encountered a similar power-to-weight ratio problem but overcame this with a somewhat lightweight combustion engine of about 15 hp, and a fairly long wingspan, plus the biplane construction that gives further wing area for better lift. The Flyer B and C would be further improvements...with engines of about 25 to 35 hp. The Wright flyers had similar empty weights to the Eole variants, of about 200 to 300 kg, but a better wing design and lighter (combustion) engine - and also the capability for controlled flight.

Man-powered "ornithopters" are another problem entirely, since, to speak in very rough averages, a man has about 1/3 the horsepower of a single horse (race horse), even less if you measure with a draft horse (which may be how the Germans were measuring horse power prior to the 1920s...perhaps this is why German hp ratings are lower on the Mercedes D.III engine when compared to British tests on the same engine...an interesting side-topic).

Something like the Da Vinci flyer might actually work but you would have to be able to pump out, consistently, about 40 to 60 horses given the construction materials of the time (linen, wood, steel tubing perhaps, etc.). I've gone with estimated empty weights for the Da Vinci and Der Flieger variant of between about 150 and 190 kg - which seems reasonable given historical construction materials - about half the weight of the Eole or Wright flyer - but with an "invisible" combustion engine included to simulate extreme pedal power. This gives the Da Vinci flyers enough power to take off.

There were some experiments in the 1980s from MIT with pedal-powered flight, but those required extremely long wingspans for lift, also cutting-edge construction materials, and saving on weight wherever possible - and they flew at something like 45 kph and maybe 20-30 meters alt. (haven't looked up those statistics in a while).

To get something like the Da Vinci off the ground would require about 50 hp, from what I've been able to test in FE2, similar to what's needed for the early MS types and Bleriot XI to get off the ground too (those have longer wings for better lift when compared to something like the Da Vinci so the Anzani 25 hp engine works but barely works on those, as was the case historically when Bleriot flew across the English Channel in 1909). For some of the types in FE2, I've gone with a lower-than-historical hp rating since the climb rate and lift was already too good on some of the wing types, and they still fly well enough for early aircraft (often you have to juggle between the engine rating, empty weight, and other values, to get the overall result right, without overhauling an entire data ini).

I did some more tinkering today with the Da Vinci and provide one more variant in the package below...for this one, I've shaved off the empty weight to 70 kg (average weight of one man maybe), and was able to reduce the power on the invisible combustion engine to only 10 hp - about the power of 10 to 20 men pedaling away  - anything less and it can't take off...and I think a lower empty weight also wouldn't be realistic (even this weight is pushing it) - considering typical construction materials for the type.

So, this one is as close as I can get it to how birds fly and how a real pedal-powered contraption might behave (one made from materials in the garage or a WWI factory, not with fancy ultralight components or cutting-edge technology) - something to enjoy in FE2. The ai flies it just fine, but requires some time to get to alt. The ceiling on this one seems to be around 500 m, top speed is only about 65 kph. This one is even more delicate than the other two variants posted under the Da Vinci thread, so watch dives and hard maneuvers.

To install, simply rename Stephen's DaVinci aircraft folder to TouringEdition and copy over the skin folder and ini files.

Von S 

EDIT: File removed, will be included in mini ver. 9.9 update of the FM pack under the relevant thread.

Edited November 27, 2019 by VonS
Deleted file and link.