Steady state VE tuning results dont translate to the street

Hey guys!!

Ive been doing a lot of work on my VE map on the street with great results. Im lucky enough to have access to a Dyno on occasion so I decided to go cell by cell like every webinar and course on this site recommends. I made sure i was dead center on each cell and let everything equalize before adjusting the VE numbers. The results was a very jagged looking drivabilty area, and when i got the car out on the road my closed loop fuel correction was constantly correcting 5-10%. I guess im just confused as to why the dyno tuned VE table doesnt translate to the street??? Ive attached an image of the unfinished VE table, excuse the large spike, I was just beginning to feel out the Cam advance stuff.

3.4L 10:1 compression 2JZGTE vvti w/ BC 276/276 cams

Probably Air temp. Check your logs.

Dyno with the hood open and a fan on vs hood closed driving down the road.

Cell by cell tuning does leave a jagged looking map.

once your on the streets your transitional corrections come into play, accel enrichment etc.

what are your trims like if your driving steady?

Air temps were actually fairly consistent between the dyno and the street (90-110F). The closed loop was adding fuel everywhere after the dyno. At Idle and cruise its was averaging an 8% positive trim.

My first thought was air temps also. I know the M1 does a background trim adjustment to the fueling based on air temp. Theres not much adjust I have other then the Charge Cooling Gains and Temp Correction Factor.

Another thing I found interesting was that at idle (after stopping for a red light or something) it'll be pulling 4% or so, and as the intake air temp rises (while idling at the light) the fuel correction slowly moves to zero (the fuel trim and intake air temp delta eachother). I would think that higher air temps in the intake would naturally need less fuel?? In my case its backwards.

Cooling is the usual culprit and I expect to see a variation of around 2-3% potentially between what I see on the dyno and what I see on the road. This is one reason why I'm a big advocate of confirming your tune on the road where possible. The M1 in our 86 tracks pretty consistently between the dyno and the road though and the VE table has a relatively smooth and consistent shape. 8% error is quite significant so I'd be looking very carefully at what is different between the operating conditions on the road and dyno.

As a separate note on our original Toyota 86 package I found the air temp comp (which is handled automatically) was a little too aggressive and resulted in the lambda moving lean as IAT rose. I ended up adding a charge temp approximation table to my custom package but this is now a standard feature in the production packages and can help deal with this scenario.

Andre,

I dont see any sort of IAT estimation table in GPR. The only thing I have is a correction factor table which is mildly confusing.

Bob

Did you use the Dynos AFR meter or the one you use on the road whilst tuning on the dyno? I've heard many a story like this where the car is tuned on the dyno then when driven on the road using a different gauge the readings are off.

Granted, I'm only an enthusiast and don't know if the stories are fact but I'm aware that different controllers can give different readings.

The wideband used on the dyno and street are one and the same.

How was the fresh air supply on the dyno? I've seen some chassis dynos where the exhaust extraction isn't sufficient and the dyno area is like a big exhaust gas re-circulation device. Your tuning ends up being too lean, and fuel needs to be added when more oxygen at a given intake pressure is available.

On the race cars I tune, I sometimes run a 8" flex duct air hose to near the air intake, this is driven by a fan located with a fresh air intake.

There was just a squirrel cage fan pointed at the front of the car. And a tube hooked up to the tail pipe that’s part of the exhaust evacuation. I’m sure the air out on the street was cooler and more oxygen rich.

Assuming you're on a recent package then you will have the option for 'Inlet Manifold Temperature Mode'. If you enable this it will give you access to a charge temp estimate table. You can find this on the 'Engine Systems - IAT' worksheet.

First as already mentioned, the conditions on the dyno must be as closed to real world conditions as possible. Before I had my own dyno I hired dyno time on another company. I never feel comfortable there and found as longer I was on the dyno as richer the car run and as more I felt pain in my head. It turned out the exhaust extraction was much to less, which leaded to a big external exhaust gas recirculation. A good dyno room should change the whole air in the room minimum six times per minute. If there is no separate exhaust fume extraction it must be even more.

Second: I found MGP or MAP as load axis doesn't work very well in part load. The reason is the follow:

Let's assume max boost is 1.0bar at full throttle at 4000rpm steady state. If you hold 40% throttle at steady state the turbo spools slowly up to 1.0bar. If you now tune this cell with 40% at 1.0bar, the engine is going to run lean if you make a full throttle. This is because the part opened throttle (40%) leads to some kind of turbulent flow inside the manifold and this reduces the Volumetric Efficiency.

Conversely if you tune your AFR to target for example @ 30% throttle and 4000rpm steady state, the turbo produces 0.5bar.

If you drive on the street through a corner and floor the pedal all to the metal out of the corner, the turbo spools up from zero to 1.0bar an the ECU goes trough the same cell, you tuned steady state on 30% TPS, now with 100% throttle. Means the VE is higher now in the same cell. Consequently you will see lean AFR's during transient spool up.

I found tuning the main table spanned with TPS vs RPM with boost off, and then do a 4D overlay table with Boost vs RPM compensation covers the load conditions much better. Well tuned, you get very good and consistent transient AFR's. Besides the tuned 4D table gives a great picture of the turbo efficiency changes depending on boost level and RPM. I use this strategy regularly on VE based ECU's with great success. I used it also once on a M800 (Milisecond based) and found it isn't working that good there. Think this is because the ms base ECU doesn't take the Target AFR into calculation.

Another wrinkle in this is the VVTi. OP what is your cam advance map looking like? Are you using it in non boosted areas?

Tuning in AplhaN is not an option for me unfortunately. There isn’t going to be much going on with this until spring now also. But what I can say is that the dyno conditions are certainly less than optimal.

Bob