auto-rich / auto-lean carburetor

[Guest grouby]

Hi everybody,I'm trying to simulate a carburetor for a A-20C Havoc.You know that kind of carb that has the four positions, FULL RICH (or EMERGENCY), AUTO RICH, AUTO LEAN and IDLE CUT-OFF, and that were used on many of the WWII US aircrafts and more (and that is not simulated by FS and even not CFS!!!).I think it won't be very difficult to programm a gauge that controls the mixture ratio like the carb would do.But here is the problem: what is really doing that kind of carb?I'm not talking about the idle position. That I understand... I think... Well, it's gonna be long... so if you perfectly know how such a carb is working, you may don't read what is following, which is written in poor english, and just explain it to me in detail! ;)First, when looking at the A-20 manual, it looks like the engine power is not changing whenever the mixture controls are in AUTO RICH or AUTO LEAN, or it changes a little. For example:The maximum continuous cruising settings for the Wright R-2600-23 are:in AUTO RICH, max rpm: 1900, max manifold pressure: 30 Inchesin AUTO LEAN, max rpm: 1850, max manifold pressure: 27 Inchesand the consumption per engine, for maximum cruising, are :in AUTO RICH, at 900 hp: 84 gphin AUTO LEAN, at 900 hp: 69 gphSo, it looks like in the AUTO LEAN position the same power is obtained with less rpm and less manifold pressure, and that it burns less fuel.This results have sense if we consider that in AUTO LEAN the fuel/air ratio is perfect for best power and that in AUTO RICH the ratio is much too high, providing too much fuel to the engine so that the power is decreased for the same rpm and manifold pressure.If this is true, I can understand the interest of the AUTO RICH by the fact that it's a "cooler" mixture ratio, it prevents the cylinder head temp to exceed the maximum temp of 232

[ozbeowulf 0]

Maybe I can help; maybe not.First off, the important thing about the ubiquitous WWII-era four position automatic mixture controls is that they control the fuel/air RATIO. Auto-Rich, for example, does not deliver a single metered amount of fuel regardless of power setting and/or altitude. When the throttle position is changed, more air is allowed to enter the induction system so the automatic mixture control adjusts the amount of fuel accordingly but it always maintains the same fuel/air RATIO. Likewise, during climb or descent, as the air density varies the amount of fuel delivered changes but the fuel/air ratio does not. I do not know the specific fuel/air ratios for the system installed on the A-20, but I'd bet they are the standard ratios developed for Amelia Earhart's long distance flights by Kelly Johnson (of Skunk Works fame i.e., 0.073 for economical cruise power and 0.083 for standard climb power.Last year, I developed such an automatic mixture gauge to do this job after a public query about automixture by FSAviator. This is the relevant section of xml code...(L:automix sel, enum) 0 == if{ (>K:MIXTURE1_RICH) } (A:Eng1 manifold pressure, inHg) 34 > (L:automix sel, enum) 1 == && if{ (>K:MIXTURE1_RICH) } (A:Eng1 manifold pressure, inHg) 34 < (L:automix sel, enum) 1 == && if{ (G:Var1) 0 == if{ 1 (>G:Var1) (P:Local Time, seconds) 10 + (>G:Var2) } } els{ 0 (>G:Var1) } (G:Var1) 1 == if{ (P:Local Time, seconds) (G:Var2) > if{ (>K:MIXTURE_SET_BEST) } } (L:automix sel, enum) 2 == (A:RECIP MIXTURE RATIO:1, enum) 0.073 < && if{ (>K:MIXTURE1_INCR) }(L:automix sel, enum) 2 == (A:RECIP MIXTURE RATIO:1, enum) 0.075 > && if{ (>K:MIXTURE1_DECR) }(L:automix sel, enum) 3 == if{ (>K:MIXTURE1_LEAN) } A few points: 1 - I was unable to get the code to work correctly if I used A:RECIP MIXTURE RATIO to assign values to both Auto-Rich and Auto-Lean. That's why I used K:MIXTURE_SET_BEST for Auto-Rich. That delivers a fuel/air ratio of 0.0846, which is close enough. (Accordingly, K:MIXTURE_SET_BEST can be used to, in effect, toggle FS9's Automixture function on and off.) 2 - This system worked fine during flight tests according to AFSD. However, FSAviator believes that FS9 would not behave correctly if the fuel/air ratio is set to 0.073 and would alter the flight model. I did not see any evidence of that, but my flight-testing of this system was fairly brief. I was deep in another project with manual mixture controls (but fuel/air ratio meters) at the time.FWIW,Glenn

[hervesors 1]

Hi all,I believe FS9/FSX models may behave differently (at least for best power mixture setting) according to the presence or not of Table 507 in the air file. Table 507 is a torque factor (ratio) versus F/A ratio. Maximum value of this table (usually 1.0) will determine the best power F/A ratio. When this table is missing, best power (according to BHP value) is obtained for F/A ratio = 0.083 (FS9 and FSX). My tests also show that "best economy" (according to peak EGT value definition) is obtained for a F/A ratio of 0.075.Herv

[ozbeowulf 0]

Very interesting, Herv

[Guest grouby]

Hello!thanks for answering.I'm afraind fs9 doesn't read at all the table 507. I've tried to change it, remove it, and nothing changed. The best power is always for a ratio of 0.083.With a ratio of 0.09 you'll have the worst consumption per HP.Here is some curves from the AF 52-12 manual (Powerplant maintenance):http://g.rouby.free.fr/Images/Carbcurves.jpgThe A-20 carb is a pressure injection.The typical F/A ratio curve shows that the F/A ratio is richer in idle and at high power than both RICH and LEAN position at cruise.According to the AF 52-12 manual, the F/A ratio for idling is set just a little richer than the best power ratio (because it can be adjusted to the best power ratio... don't ask me why).So we might think that the best power ratio is only reached when idling or near the take-off power.In the cruise range, the F/A ratio is at or leaner than the best power ratio in AUTO-RICH, and even leaner in AUTO-LEAN.The main problem for me is to find the good settings so that the performance of the A-20 are the same than those in the manual's charts.First, the consumption should fall from 100% in auto-rich to 70 to 80% in auto-lean. According to this, if the RICH position is set for a F/A ratio of 0.083, then the LEAN position F/A ratio should be around 0.065 !!! Which is far leaner than Amelia's setting !For example: My engine is set to follow the brake power chart, with the best power F/A ratio (table 508 and 509 torque vs rpm).With 30.3 Mpsi and 1900 rpm, my engine give 869 hp at sea level with a F/A ratio of 0.083. If the fluel flow scalar (aircraft.cfg) is set to 1.16, the consuption is 82.5 gph.Then, in LEAN position, with a F/A ratio of 0.067, the fuel consumption falls to 67.5 gph, which is correct. But the brake hp also falls to 771 hp.With those settings, the more economical cruise (25.2 Mpsi, 1400 rpm, 31.5 gph per engine) is obtained with a F/A ratio of 0.062 !!! with a difference in consumption of 70% from AUTO-RICH to AUTO-LEAN.(All this above is calculated this way: for a F/A ratio of 0.083, you can find the consumption for a given HP, then you can search for which F/A ratio the consumption will be the same than the one given in the manual, and finally you can calculate the power loss for this F/A ratio)I think the brake power chart is for auto-rich, and that in auto-lean the power is decreased because if not, then the aircraft should have an incredebly high induced drag, with an oswald coefficient around 0.4 where it should be around 0.7.The fact is that if I follow the consumption and power differences given above, the aircraft should have very few induced and parasite drag to reach the given cruise speed.And I'm not talking about propeller efficiency !!!Well, well...Maybe all this is picturing very well the problems a flight model designer have to deal with: lack of information about how the chart were made, back in the mid 40s, difficulties to have the fs9 physical model match with the real world one, and the worst is the manner fs9 calculates some parameters, just like prop efficiency: If you know from table 511 that the prop peak efficiency is 0.85 for a blade angle of 35

[hervesors 1]

>>I'm afraind fs9 doesn't read at all the table 507. I've tried to change it, remove it, and nothing changed. The best power is always for a ratio of 0.083<

[Guest grouby]

If it is still interesting someone...I've found some answers about carb settings on the NTRS NASA site.It looks like the F/A ratio settings for WWII carbs was a lot studied (is that english i'm writting??) by the NACA.For a 2600 cubic inch engine, as well as many others, the "normal" F/A ratio settings were:around 0.065 for AUTO-LEAN in the cruise range,near or less than 0.083 for AUTO-RICH at max cruise power (around 900 hp)up to 0.11 / 0.13 at rated and take-off power in FULL RICHthis gives a BSFC (fuel consumption per hp) of 0.42 lb/hp/hr in AUTO-LEAN, 0.52 in AUTO-RICH and around 0.75 for max power!That looks like what i've found in the A-20C manual, giving 200 gph per engine during take-off and 64 gph at max cruising.Now there is (almost) 2 problems with fs9:1- the mixture lever, 100% open, gives a maximum F/A ratio of 0.0875 at sea level.So a real power enrichment valve can only be simulated far above ground level, you'll have to take off from Aspen !2- the variation of the BSFC against the F/A ratio is very different in fs9 and in the real world:http://g.rouby.free.fr/Images/FS9%20BSFC%2...mix%20ratio.jpgin FS9, the BSFC is constant at F/A ratio above 0.09 !the R-2600-22 curve:http://g.rouby.free.fr/Images/R-2600%20BSF...mix%20ratio.jpgand it looks like there is no way to change neither horsepower vs F/A ratio nor Fuel flow vs F/A ratio in fs9!!Or maybe with a gauge, if it's possible to set the fuel quantity of a given fuel tank, there might be a way to cheat...guillaume