Forum » General Tuning Discussion » Stoichiometry at cruise

Stoichiometry at cruise

General Tuning Discussion

Discuss all things tuning in this section. News, products, problems and results. 


Page 1
Author
82 Views

Hey all,

My understanding is that the OE manufacturers goal for cruising AFR is typically 14.7:1 (pump gas) to balance between emissions and fuel economy.

Can someone share their perspective on the side effects of running leaner mixtures at cruise speeds/loads? Say 15.5:1?

I imagine long term longevity of the catalytic converter or exhaust valves may be affected due to the hotter continuous combustion.

Any input would be appreciated!

Thanks,

Phil.

If you don't care about emissions honestly the only things to worry about are high EGTs and drivability. Some cars can hang going somewhat lean, others can't.

You can often see a small improvement in fuel economy by tuning in the vicinity of 15.4-15.5:1 at the expense of emissions at the tailpipe. The catalytic convertor won't do its job at this sort of AFR but the EGT isn't going to cause damage under cruise conditions (it'll be much hotter under high load for example). If you continue to lean out the AFR further then you'll get to a point where the torque drops significantly and you'll need to open the throttle further to make enough torque to maintain a constant speed, therefore destroying any improvement in economy.

Thank you guys, that is the detailed information I was looking for. Realistically, I wouldn't go above 30% load or over 4k RPM for the leaner AFRs.

I have not experienced any driveability concerns, but at idle, it seems like the ECU is pushing the STFTs past -20. I imagine the tune is fighting itself somewhere because cruising it runs perfect. I did get a "Lean code" as well, but I imagine its just an emissions warning.

Since I am using a flash tuning application, I am not sure how this may interfere with the ECU fuel saving strategy. From my research, the (Mazda MPS) ECU advances the timing to the knock limit, and uses "Knock retard" to pull it back under cruise conditions.

A Knock retard strategy in an ECU is generally only active when knock is detected which is generally less likely to occur at lower loads/cruise. I could see it adapting timing towards the knock limit and backing it off at full throttle, though this isn't necessarily the best tuning strategy either except for maybe a turbo application as a turbo application is more likely to be knock limited rather than MBT-limited. Could you provide a link to what you've read about that? I'm not familiar with Mazda ECUs at all but what you state doesn't make too much sense to me.

Typically, when tuning a naturally aspirated engine, especially at partial loads, the goal is to advance timing until the torque output of the engine peaks, this is known as MBT (maximum brake torque) ignition timing and quite often you can go well past MBT if you are simply advancing towards the knock limit.

Your timing map should be specifying the desired timing for every load cell, just like fuel. The timing in these cells can be compensated just like fuel as well; pulling a few degrees when IATs are high is one example, pulling a few degrees when knock is seen is another example as you mentioned. Different ECUs have different knock control strategies as well, some ECUs don't so much respond to knock instantly so much as they slowly adjust the retard based off knock events as a way to adapt to lower octane fuels and will keep this retard on for quite a while. Others will instantly retard a few degrees on a knock event and work their way back up in maybe a few seconds.

A forced-induction engine on the other hand will likely behave the same in partial loads (as often in partial loads on a boosted motor there may be no boost at all) and can be tuned for MBT here. At full load, if you try to achieve MBT on a FI engine you are more likely to hit the knock limit before reaching MBT.

As for the fuel economy / AFR part, Dr. Adrian nailed it on the head quite well. I tune a fair bit of older Honda engines (BDFK-series) and many of them don't mind running a bit leaner, but they all have their own "desires" so to speak; different engines, even of the same type, while sharing many characteristics, can be quite different.

Anecdotal story of such: I have two engines, both B17A1, completely stock engines, in two different cars, one car (EG) was boosted and tuned, the other (DB2) N/A and no mods. I put a skunk2 intake manifold on my N/A engine and needed to take the injectors off the turbo engine as the stock injectors didn't fit the manifold. New injectors throws using stock ECU out of the window, so I also grabbed the ECU from the boosted car which was decently well-tuned and fired up the DB2. You would think with identical engines, identical injectors (not even a different batch of the same injectors, the EXACT same ones as they were swapped) that the tune would at least be close enough to be reasonable/driveable. I truly don't think I would've had to adjust the tune as much as I did if I had done the opposite and installed said manifold on the EG.

Thanks for the detailed reply. https://cobbtuning.atlassian.net/wiki/spaces/PRS/pages/26870132/MAZDASPEED+Tuning+Guide

You can read about the ignition advance strategy in the link above, here is the specific snippet I was referring to:

"Generally speaking, the OEM calibrators try to run as much ignition advance as possible during part throttle conditions. Timing is often beyond MBT in light load areas. Keep in mind emissions goals drive many decisions OEM calibrators make. Occasional knock response at light load is normal, intended behavior. Please understand minimal detonation at light engine load is not destructive, unlike detonation occurring at high engine load. Riding the knock limit allows the ECU to run the engine at maximum timing advance given the detonation threshold of the current fuel quality."

In regard to your anecdote, I imagine there is more at play with your engines, because OEMs and even aftermarket tunes seem to get away undistinguished between engine applications. Obviously, a dedicated tune is ideal.