Leaning at full power?

[Nuisance]

OK, you low level flyers...

How do you lean if you are flying at sea level, lets say in a race, and you want to set the mixture at 150 ROP, at full power? Do you dare lean to peak EGT at high power? Will the engine melt pistons, etc. in a flash at 100% power?

Or, can I just lean to peak fairly quickly, and then richen from there?

TIA, John
Pagosa Springs, CO (7650 msl.)

[Bob Axsom]

This should bring in some good information.

It is not a problem in my little 8.5:1 O-360-A1A so I figure you are probably asking about a race prepared IO-360-XXXX. I had plenty of fuel so with wide open Throttle and prop just under 2700 RPM, I went up into the nominal running area of the EGT with the Mixture, then dialed the mixture it back in until I got the top speed at 1,000 AGL. When I went in either direction from the final "best speed" mixture setting or reduced the RPM the speed dropped off 3 or 4 kts rather quickly and it was hard to get it back. With mine it just starts getting rough at peak EGT with the single probe from Alcor under these conditions. I bought the 4 probe EI EGT system from Aircraft Spruce at Oshkosh so I will have a better picture in the future if there is an AirVenture Cup in 2007.

With 55 gallons I had plenty for an all out run but I was down into the low single digit numbers on the gauges on the downwind at Fond Du Lac. What was you fuel experience? Oh that's another question - never mind!

Bob Whitehouse seems enthusiastic as the new AirVenture Cup boss and Eric Whyte said he will stay on as an operations man - both plusses in my opinion. If there is no 2007 AirVenture Cup I plan to fly the US Air Race Inc. cross country air races (http://www.us-airrace.org). They are handicapped but ... Pat Purcell told me that head-to-head class racing would be made available if they have interest (entries) and sponsors. Their 2007 Race is going to start at Whicita and there is a proposed finish in Florida but that is just a possibility at this point. I don't really like the handicap because it is so hard to assure validity but if it is the only game in town I'll do it.

Bob Axsom

[Walter Atkinson]

Leaning to Peak EGT on the ROP side at full power is not advisable. You will quickly eat up the detonation margin which may result in serious problems. Leaning the FF by 15% of the max FF IS tested and *should* avoid detonation. Leaning further is risky and unless you have pressure sensors in the combustion chamber, you cannot monitor the effects.

Walter Atkinson
Advanced Pilot Seminars

[Fitz]

Probably the best way to do it is with a wideband O2 sensor in the exhaust system. This has become something of the norm in high performance automotive racing.

The sensors use a pump cell and a Nernst cell to measure 'Lambda'. A lambda of 1.0 is "stoich" or "stoichiometric", essentially peak EGT. Numbers smaller than 1.0 are rich of peak, larger than 1.0 are lean of peak. If you know the stoichiometric value for a fuel, you can calculate Air Fuel Ratio (AFR) by multiplying it with Lambda. For example, MoGas is about 14.7. So at stoich, AFR is 14.7:1 (1 x 14.7). At .85 lambda, AFR is about 12.5:1 (.85 x 14.7).

The advantages of lambda measurement are that it is very fast (it is used to 'close the loop' for EFI and emissions control on modern cars), requires no peaking (you can dial in best power without having to run near peak) , and is fuel independant. That is, if you've determined that your best power is generated at .85 lambda and best economy occurs at 1.05 lambda, those remain the values for best power and best economy even if you switch from MoGas to 100LL, etc. Also, while it has its own 'gotchas', it is not subject to some of the main gotchas with EGT (ex. EGT drops in response to preignition).

The disadvantages for general aviation are that most widely available lambda meters use O2 partial pressure of air as a fixed value so they become inaccurate with altitude. Also, the sensors do not have very long lives in leaded fuel (Bosch specs a low cost one at 200 hours, but somewhere between 100 and 200 is more realistic). And, like partial pressure, many meters become less accurate as the sensor ages because of a fixed calibration for output.

We've had a few users put our automotive wideband controllers to aircraft racing use. And, at relatively low and relatively constant altitudes we are very confident in the measurements they are getting. That is because we can be field calibrated for partial pressure and our measurement principle does not become less accurate as the sensor ages.

We've talked about adding altitude correction based on some experiments I ran in a Skylane and a Mooney 231. This would make for a pretty nice GA lambda meter, but it has remained a very low priority. I'd buy one, but in a world where a heck of a lot of engines are leaned without even the benefit of a single EGT probe and where engine makers give recommended leans that put CHTs near their highest, I might well be alone.

Still, it is an area that interests me personally. So, aside from the question at hand I'd be very interested in talking to anyone else who has experimented in this area.

-jjf

[gmcjetpilot]

Originally Posted by Fitz

Probably the best way to do it is with a wideband O2 sensor in the exhaust system. This has become something of the norm in high performance automotive racing.-jjf

Fitz, that is interesting and most of it was over my head, but I agree with Walter, NO to leaning at full power. Just a side note O2 sensors don't last long with leaded av fuel (100LL has lots of lead despite the Low Lead moniker). Reason for saying NO, severe engine damage due to severe detonation and pre-ignition. Air cooled engines with huge bores are not like small bore water cooled engines, which can tolerate detonation better. 75% power (based on manifold pressure and RPM) is a safe bet ROP. At 75% power consider 100F ROP. (LOP operations see Walters excellent post on the topic, AVweb articles and Advance Pilot Seminars web site.) G

[Fitz]

Originally Posted by gmcjetpilot

Fitz, that is interesting and most of it was over my head, but I agree with Walter, NO to leaning at full power. Just a side note O2 sensors don't last long with leaded av fuel (100LL has lots of lead despite the Low Lead moniker). Reason for saying NO, severe engine damage due to severe detonation and pre-ignition. Air cooled engines with huge bores are not like small bore water cooled engines, which can tolerate detonation better. 75% power (based on manifold pressure and RPM) is a safe bet ROP. At 75% power consider 100F ROP. (LOP operations see Walters excellent post on the topic, AVweb articles and Advance Pilot Seminars web site.) G

No one is saying LOP. You want to be at best power, generally about 125-175 ROP. It is a race. You can't easily do this with an EGT gauge because you would need to 'peak' first to find a temperature relative to it. This potentially puts the engine just rich of peak, were pressures and CHTs soar (lots of heat, lots of fuel, means ripe for detonation).

But if you are looking at the relative pressures in the exhaust, you can find max power with great precision (say .001 lambda increments) without ever having to go near stoichiometric (peak). Throwing still more fuel into the mix beyound max power is just a waste. Charge cooling is basically a myth. Fuel does not really act as a coolant, changes in mixture just alter the speed of the flame front. Slowing the flame front beyond a certain point is counter productive because timing is typically fixed on an air cooled aircraft engine.

O2 sensing would be even more beneficial for finding and tuning best economy because the of the EGT curve of avgas LOP. But I hadn't really brought that up because many GA engines do not distribute fuel evenly enough to run well at peak, let alone lean of peak - at least not without upgrading the injectors.

Yes, the sensors die in leaded fuel, but consider the application proposed here. You want to fine tune best power, at wide open throttle, without ever exceeding the engine specs. It would take a lot of $40 sensors to equal one overhaul!

What I found with the 231 is that if you stay with a set TIT, the detonation margin actually shrinks with altitude. Again, here might be a case where measuring the composition of the exhaust, instead of just it's temp, might be an advantage. Or think of climbing like you would a race. You don't want to throw fuel and efficiency away, but it is dangerous to see where the peak is...

I'm not saying that the GA public is ready for O2 sensing - I've been following John Deakin's crusades to get people to properly lean with EGT for years! I'm just saying that one of the limitations with EGT is that it is a relative scale and the only way to find the reference point is to run the engine at stoichiometric, very near peak pressure and temperature. IE, the problem isn't leaning itself, but the instrumentation available.

Look at some of the AdvancedPilot samples on their website. They have a great presentation where you find the absolute EGT reading on your normally aspirated plane that equates to best power at a low altitude, then advise you to re-lean to that absolute temperature when operating at a much higher density altitude. This theory is correct. Wideband sensing is just a much more accurate and responsive way to find that point. It also can do so without ever having to run at peak and can continue to find the point even if you change the blend of the fuel.

-jjf

[rv6ejguy]

Originally Posted by Fitz

So, aside from the question at hand I'd be very interested in talking to anyone else who has experimented in this area.

-jjf

I've had a WMS wideband installed on our Sube Turbo 6A for a while along with 4 probe EGT. It is very interesting to have this info available as we can tweak ignition timing and mixture inflight with our EFI and look at rpm / TAS gain/loss at many different settings. This allows us to get optimal settings for speed and economy. Widebands may be the best way to tune for best power. The WMS meter uses the actual Bosch chip for internal calibration so does not need the free air hassle.

Using Decalin TCP has our O2 sensors lasting in excess of 40 hours now.

[Walter Atkinson]

Some facts of physics:

Best Power is found at between 75 and 80dF ROP.

It is only going to be about 2-3 degrees F cooler on the CHT's than 40dF ROP where CHTs are peaked.

The difference between Best Power and 100-125dF ROP at max rated power is very little on CHTs and ICPs.

Leaning to Best Power at max rated power is a really good way to experience detonation.

We used to Race at Reno. We only needed the engine to last for six laps. We couldn't do it at Best Power. It took a lot more FF than that to keep the engine together for six short laps.

Take those facts for what they are worth and make your own decisions.

Walter

[Fitz]

Originally Posted by rv6ejguy

I've had a WMS wideband installed on our Sube Turbo 6A for a while along with 4 probe EGT. It is very interesting to have this info available as we can tweak ignition timing and mixture inflight with our EFI and look at rpm / TAS gain/loss at many different settings. This allows us to get optimal settings for speed and economy. Widebands may be the best way to tune for best power. The WMS meter uses the actual Bosch chip for internal calibration so does not need the free air hassle.

Using Decalin TCP has our O2 sensors lasting in excess of 40 hours now.

FWIW, the Bosch chip uses a fixed calibration curve - based on a group sampling. So it does become less accurate with altitude and age. On the Megasquirt web page there is a long white paper on this. They are trying to develop a different measurement principle but I have my doubts about its practicality.

I've also tested the impact of partial pressure of fixed current measurement with calibrated gases. I don't have my lab notebook handy but the difference between the beach and my house (at an elevation of about 2,000') was pretty signifcant. But, it is important to keep that in perspective. We are trying to provide accuracy in the .001 to .002 Lambda range, even on our lowest cost controller. Most controllers offer on the order of .025 - .1 lambda accuracy, and only really promise that near stoichiometric. So loss of accuracy, etc. is relative and may not matter for many applications.

I did have one thought about where our speed and accuracy might be useful for aviation. I once saw a pro racing engine builder configure one of our controllers for full speed (we always measure hundreds of times per second but users typically set our outputs to slower updates based on average). He then viewed our analog output on a o-scope and used the peaks and valleys to tune valves. This inspired me to experiment with multiple cyl measurements. I tapped the ignition pulse of one cyl to give me a reference, which I used to trigger my own scope (with offset), then compared the readings from the manifold to individual widebands on each cyl. The results were very good. I also found that I could detect individual miss fires based on the puff of O2.

All this only worked at lower RPMs, which limits its use in automotive. But all the GA engines that I have used typically operate below 3000 RPM. Getting cyl by cyl lambda with one or two sensors, instead of 4-6 seemed like a possible benefit when working with leaded fuels. I averaged quite a bit better than 40 hours in my tests, probably because of the measurement principle used, but life was still limited.

Sorry to sound like an ad! I actually joined the forum because I've recently been flying a friend's 10 year old RV-6A and have been giving some serious thought to building an RV-8. Of course, as soon as I saw an area for engines I just had to take a peek...

Your use of EFI intrigues me. I'll do a search for other threads, but do you have any other web links, descriptions, etc.? Also, if you don't mind my asking, how much Decalin TCP are you adding to your AvGas?

-jjf

[rv6ejguy]

We are using about 1/4 oz Decalin per 10 gallons of 100LL.

With our EFI and the WMS wideband, we can datalog AFRs at each rpm range to be able to reprogram the fuel values to the desired ratio.

Some of our clients are running our system on Lycos and able to get very good economy at 50 to 75F LOP, 20F spread between hottest and coldest EGT. One fellow leans in the climb to around 100F ROP. He has over 350 hours now on an O-320 in a Glastar.

Your experiments with individual O2 probes is interesting. I always wanted to do that but I have enough hardware on our mule already.

You should build that 8!

http://www.sdsefi.com/aircraft.html