Lean Tuning and Ignition Timing

It is my goal to lean tune an engine for a specific economy cruise function. This would be leaned out specifically for a narrow cruising RPM range and medium to low load.

I hope I could achieve 1.1 lambda as the power drop-off would not be that substantial (hopefully)....I had thought possibly a small amount of timing advance could offset the loss in power from the lean mixture.

My questions are:

With a rise in combustion temperature resulting from a lean mixture, would ignition advance even be needed?

Does the speed up in combustion effect cylinder pressures similar to running more advance? Or does it merely become knock limited earlier in the advance range compared to a leaner mixture.?

I do realize some retard may even be needed but assume this engine is very conservatively tuned from the factory, large displacement and 9:1 compression ratio. The goal of this tune would be the economy even at the sacrifice of power (to a degree) this is not a racing application.

A few thoughts from my experience:

Leaning air/fuel mixture more than lambda 1.03-1.05 in cruise mode does not help fuel economy but increases fuel consumption.

Leaner mixture does not necessarily burns hotter - https://youtu.be/vcgmEKhCFTs

Increased Compression ratio significantly helps fuel consumption providing right camshafts are chosen.

I have no ideas about the answers to your questions. But I do have ideas about how to tune for maximum fuel economy...

I think the first thing you should do is find a way to measure fuel economy on a load-bearing dyno. Just thinking about it -- perhaps you do this using ECU logged injector pulse width (or duty cycle both are OK if RPM is constant), and a test where you run the engine at a constant speed / torque (essentially constant horsepower) and take the average pulse width or duty cycle. Choose a speed / torque that is similar to what you see with logged data cruising at your target speed (so, maybe 10% throttle @ 2100 RPM -- on the dyno, with your current tune, 10% throttle @ 2100 RPM makes say 50 ft-lbs torque or about 20 hp (HP = Torque x RPM / 5252).

Now you can test with with different Lambda Targets, each of which would have the Ignition tuned for best torque (MBT). Let's say that your target was already Lambda 1.0, so you run the car at 2100 RPM holding 50 ft-lbs of torque (or an indicated 20 hp), for say 1 minute. Then you look at the logged data and determine the average injector pulse width over the time at 2100 RPM. Next you tune the ECU for Lambda 1.1 (you only need to do a small area say 1800-2400 RPM and 10-15% throttle), adjust the Ignition timing for MBT, and repeat the test. Now 2100 RPM and 20 hp may take 12% throttle -- but you run it for 1 minute and see if the average injector pulse width is less. If it's less then you are using less fuel for the same power, or increasing your fuel economy.

Now, what happens if you change the target Lambda 0.92, adjust the ignition timing, and now the engine runs at 2100 RPM / 20 hp and 8% throttle, what is the average injector pulse width / duty cycle?

What is the trend? Is your fuel economy increasing as you run a leaner mixture? What happens if you target Lambda 1.2 or Lambda 1.35? If the trend is the reverse, try Lambda 0.88 or 0.85. Keep going until you find the peak.

Dave, makes sense...you've prioritized the MBT and let that guide the injector pulse width to accommodate target lambda. And that appears to be a sound way to tune for economy with MBT.

In my case I am willing to sacrifice a bit of MBT for economy, based on drop in MBT and how much I can live with. The pulse width averaging does seem like a pretty reliable fuel consumption indicator.

I have heard these engines are dramatically under tuned so there is quite a bit of advance that can be added in this range...so just hoping to come ahead on both MBT and economy.

Shota, compression increase is not an option in this case. I have heard similar Lamdba figures of 1.05 being about max. I am shooting for the moon but will settle for less of course :)

Dave, my reasoning is solely based on the assumption that IPW would be less for leaner AFR targets under identical RPM and load with the MBT varying between the different targets.

I have no experience to back this up of course....just an assumption.

Your method seems the exact way to get the most efficient power, however.

If you reduce the timing to less than MBT (at a given lambda target), you will be making less torque for the same injector pulse width, meaning it will require more throttle to get the same torque/hp for a given RPM. I am sure that this will increase the injector pulse width and will result in less fuel economy.

I think I've realized the confusion. The Lambda Target and the throttle position are what determine the necessary injector pulse width. The ignition timing is just tuned for this mixture to produce the best torque. So MBT is best torque for a specific throttle position (load) and RPM. Not necessarily the highest torque possible with any amount of fuel.

To maximize fuel economy we have to minimize injector pulse width at a given RPM / hp level. Different Lambda Targets will mean that a different throttle position will be required to get the same torque/HP -- that is what we need to compare -- the injector pulse width at two different lambda targets and throttle positions, but the same RPM. In order to get the maximum power for a given throttle position, you need to tune the ignition timing for MBT.

You would also need to consider the engine architecture etc. for example a 2 valve, lazy v8 will struggle with poor combustion efficiency at cruise speeds so will need to be richer than stoich just to initiate combustion let alone produce torque so true lean burn will be difficult to achieve.

Try to bottom out the problem by fuelling to lambda 1 and tuning the ignition for MBT in your cruise area then reduce fuelling until torque drops off and then re tune the ignition to MBT at that specific lambda paying attention to EGT, knock and NOx/UHC with a 5 gas analyser.

You might also find it beneficial to tune the injection timing in that cruise condition as there will be gains to be had by doing anything that will increase combustion efficiency.

Scott....that was pretty well my original plan minus the gas analyzer. The engine is already factory tuned at lambda 1 so I was going to start lean to begin with. What would the gas analyzer help with?

Appreciate all the suggestions, this won't happen for a few months but I am trying to develop a game plan for the least amount of dyno time as this will be a specialized dyno capable of at least 20,000 lbs.

If it's got sufficient specialisation, it may have an option for measuring fuel useage fuel in - fuel out. If so, there may be an option for a simple torque Vs fuel use. Failing that, if it can interface with the ECU, you may have an option for pulse width Vs torque as a direct read out?