driver demand at wot

I've been doing experimentation trying to understand how the torque model and driver demand works while in WOT. I purposely set the driver demand torque quite low at WOT. I see the indicated torque being where I expect based on the driver demand I set, but the load appears to be much higher than I was expecting. I was expecting that the throttle would have closed to reduce the load to the corresponding load cell in the torque model to match the indicated torque.

What am I missing here? Maybe my understanding how the torque model works at WOT with driver demand is not right?

This is a 2011 Mustang GT.

Attachment 91684Attachment 91685Attachment 91686

Quote:

---

Originally Posted by dbb2797

I've been doing experimentation trying to understand how the torque model and driver demand works while in WOT. I purposely set the driver demand torque quite low at WOT. I see the indicated torque being where I expect based on the driver demand I set, but the load appears to be much higher than I was expecting. I was expecting that the throttle would have closed to reduce the load to the corresponding load cell in the torque model to match the indicated torque.

What am I missing here? Maybe my understanding how the torque model works at WOT with driver demand is not right?

This is a 2011 Mustang GT.

Attachment 91684Attachment 91685Attachment 91686

---

Driver Demand table settings are bypassed at WOT unless you have purposely changed that functionality in the tune. This is why a FI car, or any modded car for that matter, can still get full power (higher than stock) without ever touching DD.....

I don't see throttle angle in your channels list. Are you logging it?

You need to set WOT start / end above 600 to disable forced WOT at 100% pedal.

I have experimented a lot with this and actually run my WOT start/end at 595 / 599. So I am almost exclusively in DD

But how would you change the functionality in the tune to follow indicated torque at WOT?

Also this is what doesn't make sense to me. The ECU is using the driver demand table to command a certain torque so that the throttle feels smooth. And then looks that up in the torque tables to determine the load it needs to target right? So assuming that at WOT it ignores this, even at partial throttle how does the ECU determine how much the torque the car is making?

Look at my screenshot below from a different run at the exact same RPM and same gear where I have the driver demand at WOT set normally. Desired Brake torque, ETC torque and indicated torque are all shifted up by about 150 ft pounds because of my dubious numbers I purposely put in my tune from the post above.

Does the ECU really have a way of determining a somewhat accurate estimate of torque? Am I not logging the correct PID? Or are the torque values not real and just a "derived" number completely off a base of what is in the DD table? Which would mean that the ECU never really can estimate torque, and at partial throttle it's really just looking up in the torque table from DD and figuring out what the load it needs to target?

Basically what I am trying to understand is can the ECU actually estimate the torque that the engine is making or can all it really do is determine load from the MAF and the torque numbers are just a "derivation" of a hard-coded value in the DD table?
Attachment 91696

Are you saying that the only difference in tune between those two screenshots is just the DD table?

That data suggests otherwise.

Not to mention, the ambient temp is 25* cooler on the second snapshot.

that is literally the only difference in those two logs. It is just the DD. The tune is the only difference. Yes I've noticed that every once in a while the ambient temperature is not accurate in the log. I'm not sure why. I have hundreds of logs from the track. Let me find another one where the ambient temperature is accurate.

here is a log with the ambient accurate and within a few degrees for comparison. As you can see, same thing. The torque values are much higher (about 150 higher) than my tune where I purposefully set the DD very low at WOT.

Attachment 91697

Interesting.

I was under the impression that indicated engine torque would be unaffected by the DD table. Maye you stumbled on something previously unnoticed. You literally changed no other settings between the runs?

I haven't read a complete account on here of how the Ford DD system works, that I consider fully accurate.

That is correct. I am 100% confident that the only thing different between the runs is DD table. In fact to make sure when I changed the DD that I was working on the same exact tune I had previously been loading, I read the tune from the car and modified the DD. and loaded that back in. Did my log I pasted above. And then after that run and and log, I loaded back in that exact same tune I had previously read out to make sure again that my next log was again with the DD back to normal and I was comparing apples to apples with the logs from the tunes.

Agreed about not fully reading an account of how the DD works. I'm beginning to suspect that the ECU is not in fact calculating torque based on any reality. What it seems to me is that it's using the DD as just a baseline to start it's derivation of indicated torque. Again, unless I'm completely off base here, it almost seems as if you could scale your DD and your torque tables to anything you wanted. As long as in partial throttle the torque tables matched up the scaled versions with load values that matched.

But again also, I don't understand why at WOT it wouldn't also operate the same. I expected the ECU to command the throttle to close. But it really appears to be ignoring DD and indicated torque values in WOT at least.

Did you check your WOT start/end like I suggested?

With low maximum DD settings, that is the only reason the throttle would open all the way. It might even be skewing the indicated TQ.

ahhh I totally missed that post from you but I see it now. Thanks! I'm going to go try upping the WOT end in a few minutes. I'll report back with the results.

I updated with a third log that's accurate. The air temp is actually within a few degrees. I've seen a number of times the ambient not being logged correctly and it always defaults to that 59 when it's not accurate for some reason. The first and third log screenshots I posted show accurate ambient and they are within a few degrees of each other (84 vs 81) . Look at the intake temperature. It's higher on the third screenshot yet that log (with DD set normally) is showing indicated being about 150 higher. If anything that should be showing lower indicated. Trust me it's not related to weather or engine conditions.

While you're out, log throttle angle channels. I am wanting to know why your effective area is so different between screenshots. Between that and the air temps, this could be the cause of some of the variance. Air flow and air temps certainly can affect the torque model and what the engine 'thinks' it's making.

If your pedal is in WOT territory it's going to blend away from DD. You don't want to be at full WOT following the DD table since it will likely be a limiter and cause other problems. Many tuners don't even bother with that table lol.

Awesome, thanks CCS86, adjusting the WOT start/end is now getting me the results I expect with the throttle closing.
blackbolt22, take a look at he below log too. I had the throttle angle channels in my previous logs but didn't realize I didn't get them into the screenshot. So these below will have them.

I've attached the log below. This run was with the WOT DD fields being the artificially low values as before. And the indicated torque does still appear to be based off what is in the DD table. I am a little surprised that even at 63 degrees the Load is still well over 100 though. But that may be because of the large CJ mono throttle body.

Attachment 91706

Below you can see the log I immediately did afterwards where I loaded back the normal DD values. Again notice the indicated being 555 vs 366 in the artificial low DD log above. Also notice the ambient not being correct on this run though. Between logs the only thing I did was shut off the car after log 1, load the second tune and immediately start the second log. So that ambient in the second run I think is being logged incorrectly for whatever reason and I believe plays no part in these logs.

Attachment 91707

What I feel like is happening here is that the ECU really has no clue what the actual torque the engine is producing. It feels like to me that the DD is supposed to be somewhat realistic values for wheel torque, and then the ECU uses that as a baseline to start estimating what torque the car is producing (friction losses, spark, load, temperature, etc). It makes me wonder if I hypothetically were to scale all the torque tables to let's say 200 as my max DD torque at WOT. That I would see the ECU estimating indicated torque as somewhere in the 300's. And if I were to scale the torque tables to match the corresponding load values for those torque values (along with every other torque limiter, range, etc), again hypothetically everything should work identically to how the car performs today.

So again, it feels like that the DD table *is* the estimated wheel torque the car in reality should be matching. And if the DD is wrong and the car is producing drastically different torque values that's what makes tuning the rest of the torque values tricky. Because you won't be getting an accurate indicated torque. You're getting a false value. Which this may be why tuners work around it by preventing DD from coming into play at WOT. But if you were to put the car on a dyno and find the real accurate wheel torque values, and you plugged them back into DD, you'd then have a working torque model. I could be completely off on all of this of course ;) but I can't explain why else I can easily spoof such drastically different indicated torque values using the DD table.

DD is ignored @WOT is what everybody says - not quite according to my findings especially with manual transmission.
Example below - had to adjust the DD to remove those small dips @WOT and there was no other way around it.

DD is taken into consideration @WOT but I'm not sure how and why. When ETC request is smaller than Brake Torque PCM is in super sensitive mode.

Attachment 91710

DD torque or ETC torque plus the torque needed for the engine to idle is the scheduled/ indicated torque that comes from the torque model(MBT). Idle has a reserve so it's not always using MBT in its calculations. Idle has to always be factored in to prevent stalling when going from WOT/partial to zero pedal. It's usually 60 Ft lbs or so. 15+ get the CLIP/ADD system to help with small pedal changes actually meaning something, but leaves a 738ft lb limit. Disabling it effects the idle torque to jump from a high to a low constantly. I'm not sure, but 18+ will probably have some of its own quirks in driver demand control. I have limited experience with them. This is just the base of how it ALL works.

I've posted the Ford patent, and the flow chart from the patent multiple times. It does not get more accurate, sorta from the horse's mouth. Only Because of the many slight variations in calibrations it's up to you to figure out the HPT tables and "what does what". 11-14, 15-17, 18+ and auto vs manual are all slightly different. Then there are OS changes for specific applications and many strategy changes in each OS that are different. Its very hard to share information when it's so easy to say "this is 100% wrong because it doesn't fit my application perfectly." With out acknowledging all the differences in all our applications. Fundamentals versus specifics gets lost in the HPT translation template.

look at a 2013/2014 Shelby as it also bypass the WOT start end for reference

Quote:

---

Originally Posted by dbb2797

Awesome, thanks CCS86, adjusting the WOT start/end is now getting me the results I expect with the throttle closing.
blackbolt22, take a look at he below log too. I had the throttle angle channels in my previous logs but didn't realize I didn't get them into the screenshot. So these below will have them.

I've attached the log below. This run was with the WOT DD fields being the artificially low values as before. And the indicated torque does still appear to be based off what is in the DD table. I am a little surprised that even at 63 degrees the Load is still well over 100 though. But that may be because of the large CJ mono throttle body.

Attachment 91706

Below you can see the log I immediately did afterwards where I loaded back the normal DD values. Again notice the indicated being 555 vs 366 in the artificial low DD log above. Also notice the ambient not being correct on this run though. Between logs the only thing I did was shut off the car after log 1, load the second tune and immediately start the second log. So that ambient in the second run I think is being logged incorrectly for whatever reason and I believe plays no part in these logs.

Attachment 91707

What I feel like is happening here is that the ECU really has no clue what the actual torque the engine is producing. It feels like to me that the DD is supposed to be somewhat realistic values for wheel torque, and then the ECU uses that as a baseline to start estimating what torque the car is producing (friction losses, spark, load, temperature, etc). It makes me wonder if I hypothetically were to scale all the torque tables to let's say 200 as my max DD torque at WOT. That I would see the ECU estimating indicated torque as somewhere in the 300's. And if I were to scale the torque tables to match the corresponding load values for those torque values (along with every other torque limiter, range, etc), again hypothetically everything should work identically to how the car performs today.

So again, it feels like that the DD table *is* the estimated wheel torque the car in reality should be matching. And if the DD is wrong and the car is producing drastically different torque values that's what makes tuning the rest of the torque values tricky. Because you won't be getting an accurate indicated torque. You're getting a false value. Which this may be why tuners work around it by preventing DD from coming into play at WOT. But if you were to put the car on a dyno and find the real accurate wheel torque values, and you plugged them back into DD, you'd then have a working torque model. I could be completely off on all of this of course ;) but I can't explain why else I can easily spoof such drastically different indicated torque values using the DD table.

---

I'm not surprised that 63* of a monoblade gets you over 100% load on a NA car. I have nearly 13 psi, fed with a VMP Gen 2 into my Coyote, and the difference between a twin 60mm and twin 65mm is essentially nothing until over 6500 rpm. Similarly, at lower RPMs, say 3000, when the blower can still make full boost, there is essentially no difference in airflow between 40* ETC and 82*.

What between these two runs surprises you? Now, with WOT start/end disabled, the lower commanded DD values net you a lower ETC opening and lower indicated TQ. That is what I would expect.

I also see 3 or 4 DTCs in those logs. probably worth sorting out, because DTCs could affect things in weird ways.

If your ambient temp values are jumping around like that, I would look at the sensor and wiring. That is not normal.