Has anyone noticed the fuel burn in the B737-800 (default)? It sure seems high to me. I haven't been able to get beyond about 2000nm with full tanks. Doesn't match the specs listed.If wrong, how can it be corrected? Changing the fuel flow scalar in the aircraft cfg? Or does one have to use AirEd also, and if so, what needs to be changed?Thanks.

>Has anyone noticed the fuel burn in the B737-800 (default)? >It sure seems high to me. I haven't been able to get beyond>about 2000nm with full tanks. Doesn't match the specs>listed.>>If wrong, how can it be corrected? Changing the fuel flow>scalar in the aircraft cfg? Or does one have to use AirEd>also, and if so, what needs to be changed?>>Thanks. See "A321 fix or simply a change?" I commented on fuel flow in the B737-800. It appears there is no simple way (changes in aircraft.cfg) to fix the fuel flow and still have N1 reading appropriately. One would have to use Aired and modify one of the turbine tables. Also, drags. Zero Lift drag is over double a realistic value in both the A321 and the B737. Induced drag is too low at cruise altitudes. For some reason, the Lear 45 is much closer on Cdo. Ron

Sorry, I hadn't completely read your other post lately. Thanks for the info.To get things more in line with reality, I might make a copy of the 738 air file and tweak it a bit using some of the parameters from the FS9 734. This should get the performance somewhere in the ballpark. Since we don't have a default medium range airliner, I need to get as close to the spec. range of the 738 as possible for intercontinental flights.Thanks.

>>>>It appears there is no simple way (changes in aircraft.cfg) to fix the fuel flow and still have N1 reading appropriately.<<<

Ron and Douglas - Thanks for the info. I've done as you suggested and changed the two parameters. I'll fly it and see what happens.BTW, I said intercontinental flights when I mean't transcontinental. Thanks again. Perhaps someone with a lot more experience with air files than me can truly repair this.Thanks Again.

>>>>>>It appears there is no simple way (changes in>aircraft.cfg) to fix the fuel flow and still have N1 reading>appropriately.<<<<>The problem with the fuel flow being higher in the 738 is the>new

>However, TBL 401 affects Cdi (Induced Drag) as a function of Mach Hi Ron,so, if this is the case, when MSFS calculates Cdi=Cd0+K*Cl(AoA)^2, does it mean it considers not the ACTUAL Mach corrected Cl, but just the uncorrected Cl resulting just from table 404?Marco

>>However, TBL 401 affects Cdi (Induced Drag) as a function of>Mach >>Hi Ron,>so, if this is the case, when MSFS calculates>Cdi=Cd0+K*Cl(AoA)^2, does it mean it considers not the ACTUAL>Mach corrected Cl, but just the uncorrected Cl resulting just>from table 404?>Marco For a given CL, AoA decreases as the Lift Slope (TBL 401) increases. Lower AoA translates directly to lower Cdi (in FS). Actually, one has to use Effective AoA. AoA measured from zero lift AoA. That line swivels about AoA=0 as TBL 401 varies from 1.0. So, CD = Cdo + Cdm + K*AoAeff^2. K includes Oswald Efficiency, which can be considered to vary with Mach. Typically up a bit before Cruise Mach, then down rapidly as it passes through LRC Mach and beyond. The decrease in Oswald Efficiency increases Cdi, so total drag increases rapidly past LRC Mach at higher FL's and weights. The effect on Mach related Oswald efficiency is: a(M) = ao * TBL (401)^2. Where ao is Oswald Efficiency set in Aircraft.cfg. If TBL 401 peaks at 1.05, then a(M) increases from ao at low speeds to 1.05^2 ~ 1.10 * ao. Which means 10% less induced drag at the peak. Roughly 0.03 Mach past LRC Mach TBL 401 decreases to say, 0.90. So, a(M) drops to ao * 0.81; Cdi increases near 20%. If Cdi were equal to Cdp at LRC (high FL and weight), then CD would increase by 10% in 0.03 Mach. Which is why flying faster than LRC Mach rapidly increases fuel consumption and reduces range. I've noted my TBL 401's come out close to 1.00 at LRC Mach, meaning the set Oswald efficiency is correct for low Mach numbers, and at LRC Mach. For Jet Transports ao is on the order of 0.78 to 0.80. 0.80 was given for a DC-9. The trick is to figure the 'K' above. It is near the slope of TBL 403 CL vs Alpha, but with ao factored in. Herve' Sors AFSD calculates the effective slope (the one FS uses) by a geometrical process. However, that is only close when CLmax occurs at 0.27 radians. I think one can also adjust the factor to get Cdi to work out correctly so CD * q * S gives the correct drag in level flight (drag = thrust). Later, I implemented this stuff in my XML Jet Test. I put a fudge factor in the AC Constants file so I can adjust Cdi as above. Typically, the fudge factor is around 0.985. Once set it's good for all flight conditions. So, AFSD and my Jet Test calculate drags the saw way FS does. Otherwise things wouldn't 'sum out'. LG drag, spoiler Drag, and Flaps drag complicate matters, AFSD managed to account for them all correctly. Drag due to Slip was not calculated, though I think I know how. TBL 400, Ground Effect works the same as TBL 401 does. Except h/b, rather than Mach is the independent variable. This gives a reasonable GE model, AoA decreases near the ground and Cdi drops. Reasonable, but not exact. The default TBL401 is appropriate for an AR of 7:1. It would vary some with different Aspect Ratios. Finally, a negative Horiz Stab lift increases wing loading, thus Cdi. I set the Horiz Stab angle so Cl_hs is about 5% to 10% of CL_wing. Which is a typically cruise value. RonPS: I don't think Aces understands this stuff. ;)

>For a given CL, AoA decreases as the Lift Slope (TBL 401)>increases. Lower AoA translates directly to lower Cdi (in>FS).Ok but, in all aerodynamic texts I've read (Roskam, Torenbeek, etc.), I've always found the induced drag connected to Cl^2 and not to AoA^2.In other words, the variation of the lift slope curve should have no effect on the induced drag in cruise, in real aircrafts.What do you think?Marco

>>>>>If the

Guys (especially Ron and Douglas), I took your advice and changed the two settings suggested. Today I flew from KMEM to KSFO, something over 1500NM and burned 32700 lbs. of fuel. This included a lengthy taxi on arriving at SFO, and a step climb to FL350 out of MEM. There was an average 55 kt. quartering head wind for almost all of the route. Based on this, I think your suggestions have made the fuel burn in the 738 much more realistic. As for the N1, at this time I'm not too worried about that. Thank you very much.BTW, most of your discussion is waaaaay above my head, but I would like to understand more. Are the acronyms (ie Cdo, Cdi, etc.) standard aeronautical engineering terms, or they specific to FS? I don't know what they mean, and would like to. Is there a list I can get somewhere that will explain them? I've read most of the FS9 SDK and the FSX SDK but these aren't mentioned.Thanks so much for your help.

>>For a given CL, AoA decreases as the Lift Slope (TBL 401)>>increases. Lower AoA translates directly to lower Cdi (in>>FS).>>Ok but, in all aerodynamic texts I've read (Roskam, Torenbeek,>etc.), I've always found the induced drag connected to Cl^2>and not to AoA^2. That's one neat trick in MSFS: it uses AoAe^2 (AoA past AoAo) rather than CL^2 to calculate Cdi. Cdi continues to increase past stall, in MSFS, which is realistic. CL^2 decreases past stall, so using that would give too little drag post stall. The variation with AoAe^2 is equivalent to using CL^2 in the linear part of the CL vs AoAe curve. But, allows a reasonable approximation to Ground Effect through TBL 400 (which changes Lift Slope as a function of h/:(. I found that making TBL 401 vary to control Lift Slope vs Mach allowed a pretty good adjustment for variations in Cdi vs Mach. At the expense of Pitch having to increase past LRC Mach, rather than decreasing as it would in a real jet transport. The effect is typically less than 0.5 deg low; I feel appropriate drag is much more important than less than 1.0 deg error in pitch at higher than normal Mach numbers. NACA curves from around 1946 show more of an change in Cdi wrt AoA rather than CL. At least up to typical cruise Mach. Now a better way to set the 'Drag Polar' would be to make Cd a function of Reynolds number (higher at low speeds), Mach number (Cdi and later, Cdp increase, but at higher Mach numbers they typically degrease again), and finally AoA. With CL also varying with AoA, and Mach. Only CFD can calculate these things, and it takes a supercomputer. That's why empirical (often wind tunnel) data is generally used in engineering simulators. JSBSim allows one to set drag and lift somewhat as above. I suggested Reynold's number be added, though it's 'polymorphism' allows one to set most anything as a function of anything else in the XML data files. Orbiter doesn't account for variation of Cdi wrt Mach. It is known to have a limited drag model. It does allow for R in the Drag Polar, while 'Mach Drag' is limited to setting a couple of constants to fit a predetermined shape. A shape that may fit some AC, but none that I know of. ;) I've found engineering books often go into a lot of detail on some effects, but completely miss others. Probably based on the knowledge of the author. It's only by skimming over a lot of information and actually using my techniques to set drags that I've gotten pretty good PPH fit over a wide range of flight conditions. +/- 2% in the typically cruise condition, increasing to maybe +/- 10% at conditions well off the norm. Rather tedious, and often I don't have enough information to fit PPH over the full flight envelope. -------------- Note the above only sets Drag, thus fuel flow. Getting N1, N2, EPR, and EGT to come out is another problem. Guys like Ian Kerr seem to understand the details and I've had him work out SD's, drags, and turbine tables. Then, hack them some if I think appropriate. There is also the fact that TSFC varies with operating conditions, including Mach. I've made use of one turbine table that gives Corrected Fuel Flow to get a JT8D to give a static TSFC of 0.58, increasing to about 0.84 at Cruise conditions. I have to 'uncorrect' the CFF variable, but unfortunately can't use the 'more accurate' fuel flow to drain the tanks. However, my Jet Test gauge shows all sorts of parameters, I even forget what some of them mean. ;) Gives me something to watch on long test flights. >In other words, the variation of the lift slope curve should>have no effect on the induced drag in cruise, in real>aircrafts.>Marco True, at least not obviously. But, it does allow me to contour Cdi(M) quite well in MSFS. This was known in February of 2000, a while after FS2K came out. I've adjusted TBL 401 in every jet I've done since that time. Ron

> That's one neat trick in MSFS: it uses AoAe^2 (AoA past>AoAo) rather than CL^2 to calculate Cdi. Cdi continues to>increase past stall, in MSFS, which is realistic. CL^2>decreases past stall, so using that would give too little drag>post stall.Mmm, pretty neat actually! Though I think it underestimates of one order of magnitude the drag at very high AoA. From my knowledge the total Cd of a wing at 90degrees tops around 1.3-1.8, MSFS formula gives around 0.12-0.15 for common GA Aspect Ratios.Neverthless MSFS seems to be able to model spins quite well.>I found that making TBL>401 vary to control Lift Slope vs Mach allowed a pretty good>adjustment for variations in Cdi vs Mach.My doubt is: shouldn't Cdi always _increase_ with increasing Mach, in RL?> NACA curves from around 1946 show more of an change in Cdi>wrt AoA rather than CL. At least up to typical cruise Mach.I would be really interested in those! Do you have any references (authors, etc.)?Marco

>> That's one neat trick in MSFS: it uses AoAe^2 (AoA past>>AoAo) rather than CL^2 to calculate Cdi. Cdi continues to>>increase past stall, in MSFS, which is realistic. CL^2>>decreases past stall, so using that would give too little>drag>>post stall.>>Mmm, pretty neat actually! Though I think it underestimates of>one order of magnitude the drag at very high AoA. From my>knowledge the total Cd of a wing at 90degrees tops around>1.3-1.8, MSFS formula gives around 0.12-0.15 for common GA>Aspect Ratios. "Flat Plate" drag Cd may reach 2.0, so 1.3 to 1.8 appears reasonable. The induced drag constant for the C172 is about 0.050. Cdi = 0.050 at CL=1.0. Which occurs around 10 deg AoA. With Cdi varying as AoA^2, Cdi would be (90/10)^2 * 0.050 at 90 deg. 81*0.05 is 4.05, which is actually more than twice the max drag expected. I saw some curves by 'Horner'. They show CL may peak again around 35 deg at low Reynolds Numbers. Seems CL eventually varies with cos (AoA), going through 0 at AoA = 90 deg. While Cdi may eventually vary with sin (AoA), peaking at 90 deg. >Neverthless MSFS seems to be able to model spins quite well. Not in my experience. Unless one takes large liberties on some of the SD tables. I don't know if MSFS accounts for 'gyroscopic' coupling between Ixx, Iyy, and Izz. But, such coupling is one important factor in spins. Missing in MFSF is the ability to vary Yaw derivatives as a function of Alpha and Beta. They can only be modified wrt Alpha. My feeling is that both need to be adjustable to get realistic spins. JSBSim allows that, but it looks like no one has figured out how to get spins in JSBSim so far.>>I found that making TBL>>401 vary to control Lift Slope vs Mach allowed a pretty good>>adjustment for variations in Cdi vs Mach.>>My doubt is: shouldn't Cdi always _increase_ with increasing>Mach, in RL? In some cases it may, but remember, q (dynamic pressure) also increases with Mach. The Concorde TBL 401 appears to be about right, it drops slowly past Mach 1.0 or so. However, the higher q more than makes up for the decreasing Oswald Efficiency. At least at appropriate FL's for the weight. While Oswald Efficiency in the FA-18E drops rapidly near Mach 1.0, but increases again past 1.05. Flying near Mach 1 is very inefficient! Cdp tends to increase as M increases, but in many AC, Cdp peaks near Mach 1.0, then drops more or less as a hyperbola. The result in the Concorde is that range is higher at Mach 2.0 than at 1.90. Cdp has dropped and so has Cdi. Typically, drags and SD's get simple and smooth past Mach 1.4 or so, things are also simple in the hypersonic realm. It's the transonic region that is almost impossible to calculate, though various approaches exist to estimate it. Note virtually all jet transports are similar, only small changes in wing profile and fuselage fineness ratio were made for new generations, it was too risky to try anything significantly different from what worked in the past. It turned out that the B707 didn't have a large enough vertical stab. A small fin was added below the fuselage in some versions. Eventually a larger vertical fin became standard. I've also noted that 'tail volume', Aspect Ratio, and some other basics are very similar in most all jet transports. Don't mess with what worked before; changes after the first prototype are very expensive to incorporate. Further, Airlines are conservative and don't care to gamble on radical new designs (even if they were superior). The "Boeing Flight Engineering Performance Manual" explains a lot on drag and turbines. Though, the stuff on drags isn't directly applicable to MSFS one can get some ideas from it. Some of the turbine formulas apply exactly to the MSFS turbine tables. >> NACA curves from around 1946 show more of an change in Cdi>>wrt AoA rather than CL. At least up to typical cruise Mach.>>I would be really interested in those! Do you have any>references (authors, etc.)?>Marco "Theory of Wing Sections". I think some transonic data confiscated from the Italians at the end of WWII. Regardless, a lot of NACA R&D papers up to 1958 are available on line. Maybe from Langley. More than one wants. A 1946 paper described turbines pretty well, about the only differences in modern turbofans is that they are more efficient. In the end, the trick is to get through all the detail and figure out the relatively simple things one can set in MSFS AIR files. That's true for most things in life. ;) And, there is a criteria of success in MSFS. One can compare FM tables with the FS model and see how close they correspond. I got a 727-200 so I could set EPR to the FM value for a specific weight and FL at Mach 0.800 and the AC would slowly settle down within about 0.005 Mach of the FM value. Probably as close as a real 727 would. However, EPR was off some at Mach 0.82, and also at low thrust levels. "Tom", in the AC/Panel Forum claims he has gotten EPR worked out for all conditions by putting two of the turbine tables in his XML gauge and calculating EPR from them (and flight conditions). Actually, his EPR changes with throttle setting; it shows the pilot where to set the throttle without waiting for the turbine to spool up to the steady state value. Eventually, the calculated EPR reaches the pre-calculated EPR set by the throttle. Some MSFS jets try to do that, but are way off. ;)Herve' Sors AFSD will predict all the way back to throttle position. Based on calculations of drags and turbine tables. The fact actually throttle position eventually stabilizes within 0.25% of predicted throttle position indicates that 'we' understand exactly how FS uses the tables (and constants). That includes TBL 401, 403, 430, and all but one of the turbine tables. Also, HStab Lift, and how it varies with M. Ron

>Guys (especially Ron and Douglas), I took your advice and>changed the two settings suggested. Today I flew from KMEM to>KSFO, something over 1500NM and burned 32700 lbs. of fuel.>This included a lengthy taxi on arriving at SFO, and a step>climb to FL350 out of MEM. There was an average 55 kt.>quartering head wind for almost all of the route. Based on>this, I think your suggestions have made the fuel burn in the>738 much more realistic. As for the N1, at this time I'm not>too worried about that. Thank you very much. Good.>BTW, most of your discussion is waaaaay above my head, but I>would like to understand more. Are the acronyms (ie Cdo, Cdi,>etc.) standard aeronautical engineering terms, or they>specific to FS? I don't know what they mean, and would like>to. Is there a list I can get somewhere that will explain>them? There are lots of sites Google will find. The idea is to find one that isn't filled with math, but doesn't oversimplify things either. NASA has some explanations of aerodynamics for 'K-12'. The stuff is actually higher level. One can adjust some parameters and see the effect on flight. Cdo, Cdi, etc. are standard nomenclature for "Coefficient of Drag, Zero Lift", and "Coefficient of Induced Drag" (a function of CL or AoA). "Cdp" is "Coefficient of Drag, parasitic". Similar to Cdo, but I use it to mean the sum of Cdo + Cdm (Mach Drag). When applicable Cdp might include LG, Flaps, etc. drag. http://perso.orange.fr/hsors/FS_Soft/index.html is Herve' Sors FS site. It has links to aeronautical and powerplant information. Much is on the more advanced side, but one link often leads to another. I skim over such links and learn a little at a time. Herve's site also includes some of my short MSFS documents, one is a commented aircraft.cfg file that many have said clears up lots of things. I also UL versions of Aired.ini to Herve's site. That file has concise comments and formulas on may things (Info), but one has to have some understanding to get much out of them. If nothing else, I have a place to find Prop and other formulas that I'd lose otherwise. Finally, "Aerodynamics for Naval Aviators" is an inexpensive reprint of a Navy manual that explains a lot of prop and turbine AC. It has some formulas, but not too many. While it won't tell one just how to set up MSFS AC, it does cover many basics pretty well. For both pilots and those who want to understand a bit of how AC perform. One can skim over a bit at a time, if he gets more interested he can find more detail in the paperback. >I've read most of the FS9 SDK and the FSX SDK but these aren't>mentioned. The MS SKD's are useful, but the terms are often 'vulgar' (common) rather than precise engineering definitions. One doesn't know what some mean without some background not given in the SDK's. I suspect the MS guys who write them up don't understand some of the terms either. Finally, the SDK info on aircraft.cfg is anemic, comments are often useless, sometimes backwards from reality. One reason I wrote a general aircraft.cfg that includes comments on what really counts, and how to calculate it. Many adjustments are for 'scalars'. Which multiply things like Cdo+Cdm in the specific AIR by some factor typically near 1.0. Since many of the default AC are way off, one might have to set such scalars to 0.6, 1.5, etc. It would be much better to know the actual values of the variables, especially since most all jet transports run about the same. While Single Engine Land AC usually run higher, depending on retractable gear, etc. one can estimate a reasonable value for a start. Some URL's give actual values for Cdo: I've used a NASA table at their "History of Aviation" site. The pros set Cdo, etc. in the AIR file and generally have the aircraft.cfg 'scalars' set to 1.00. OTOH, adjusting lines in aicraft.cfg is easier, and OK for simple adjustments to an AC. Assuming one can get a reasonable combination of the things I've brought up. It doesn't take an engineer to set up reasonable flight models. If one grasps the basics and uses them in the appropriate order. To get really good flight and powerplant models is more difficult, but heck, if there weren't a challenge I'd be doing something else. Note the Panel and AC Forum contains some information on AIR files, etc. Though, some is back a year or more. Avhistory.org covers AIR files, but has been rather dead for some time. Ron