Feed Forward Tuning

[glausb23]

Hello,
When tuning the feed forward tables, what exactly does the Idle Mass Flow Integral value tell you in terms of how close you are to being well-tuned in the feed forward table?

Is the goal to change the value in the feed forward table until the Idle Mass Flow Integral value in the channel drops to 0? If my understanding is correct, this would indicate that the feed forward itself is close to a perfect value, because the integral control is essentially doing nothing to correct idle mass flow value at that point. Is my understanding of this correct?

Thank you,
Benjamin

[Stephen Dean]

Hi Benjamin,

The aim is to have minimal PI intervention in the Idle control operation, but when I setup the Idle on vehicles that I work on, I do not concentrate greatly on the PI values, more the idle quality and entry and exit into and out of Idle. If you do this, then the PI numbers tend to come back to minimal values.

[glausb23]

Got it, thank you.

In order to try and get the throttle area table correct, I decided to put the throttle body on a flowbench and measure the flow rate of air through the TB at all the servo % positions listed in the table. After compiling the results into the table, I ended up with this is a resultant graph.

Throttle Area Graph flowbench results.JPG (75.22 KiB) Viewed 6668 times

I am doubting the accuracy of this method, as the overall curve does not look logarithmic like I believe it should. I thought I would try to tune with this new throttle area table anyways, just to try it. I don't believe it could be accurate, as I needed to have the feed forward values set to 100% at 68 deg F coolant temperature, and only down to about 75% at full (hot) operating temperature.

It would seem apparent that the values in the throttle area table are over-stated and would need to be decreased to lower the feed forward values to a more reasonable level of control

Is there a better way to tune each throttle area % for each throttle servo %? I cannot wrap my head around how physically flowing the throttle body like I've done could be wrong.

Any help is appreciated,
Benjamin

[Stephen Dean]

I know a few teams that have done a similar process with the Throttle Area table, and have ended up reverting back to the base table as the drivers have complained able the feel and driveability of the throttle with this change. Some throttle bodies are also scalloped to make a more linear mass flow through the body as well.

The Idle Mass Flow Feed Forward Main numbers are a percentage of the Idle Actuator Throttle Aim Minimum parameter, what is this set to in your Package?

[glausb23]

We currently have the following values for the feed forward tables, after doing a couple of startups with this new custom throttle area table. I've really only tuned the 1500-2000rpm columns since changing over to the new table, while the rest of the values were roughly guessed/interpolated from the other well-tuned sites.

Feed Forward Snippet.JPG (58.51 KiB) Viewed 6654 times

This is with our Throttle Aim Maximum setting at 12%. I fear moving the throttle aim maximum value up much higher, as doing so will likely worsen throttle pedal feel and idle mode transition, wouldn't it?

Obviously if I were to increase it to say, 15% it would allow me to decrease the values in the feed forward table to a more reasonable/tunable value, rather than being right at the limit of 100% at 20 degC. In my current state, I would actually want greater than 100% feed forward values in the 0 deg C rows. This is obviously not possible, so I'd need to either increase the Throttle Aim Maximum value, or find a correct throttle area table, correct?

Thanks so much for the continued support.

Benjamin

[glausb23]

Also, after re-reading your previous response, what is the base table that you are referring to, which other teams have referred back to? Are there throttle area tables specific to each throttle body? I would think that the diameter of the throttle body does not matter much, so long as the geometry is the same. With the area being circular, a 40mm version of a throttle body should have the same % area as an 80mm TB at the same servo opening percentage, correct? Or am I thinking of this wrong?

Thanks,
Benjamin

[David Ferguson]

The base table refers to what you would find in one of the Packages you download from https://moteconline.motec.com.au/Package

[Stephen Dean]

Hi,

A 40mm throttle open to 10% will flow a lot less mass through it than an 80mm throttle at the same percentage opening. A very quick calculation shows that the 80mm throttle will flow around 4 times as much mass as a 40mm at the same angle. This is a reason why just going larger on a throttle can sometimes be a backwards step in the behaviour of an engine.

I have found that a higher value in the Idle Actuator Throttle Aim Maximum can actually improve the transient response of the throttle into and out of Idle Control, as well as the operation of the Idle control, as it provides a larger window for the system to operate in. If the number is too small, it creates harshness in the responses.

The transient behaviour of the throttle out of Idle is controlled by the Idle Throttle Pedal Blend and Idle Throttle Pedal Offset Limit parameters, these control how the Throttle Aim is blended from being controlled by Idle to Pedal.

[glausb23]

Coming back to this thread after taking a brief break from engine tuning over the last week. Thank you guys so much for the responses and sharing your technical expertise on this.

Playing around with tuning the feed forward table, searching for appropriate Throttle Aim Maximum values, and making adjustments to the throttle area table have led me to find the general trend that in order to keep the feed forward tables at reasonably low values (so as to not exceed 100% feed forward values with the idle mass integral control still trying to add more air).. it seems as though dropping the values in the Throttle Area table (at least up to about 25% Servo position) keeps the system happy.

I've read a few threads online now and talked to a couple of local tuners who use MoTeC that have actually suggested that I just set the Throttle Area table to a 2-point linear graph 0 to 100 (linearly interpolated). They have then said that the pedal feel can more easily be tuned using the pedal input to servo output translation table, rather than trying to do this with the Throttle Area table. What is your opinion on this?

Would there be any noteworthy drawbacks to setting the Throttle Area table linear like this, specifically would it cause any negative effects on idle mode transitions, and engine deceleration back into idle control? Any input is greatly appreciated.

Our Throttle Aim Minimum is currently set to 0. I honestly did not even know this parameter existed up to this point. After reading through the context help, I can see how this value could certainly be the culprit to our issues with the engine not decelerating back into idle mode without dying (transitioning from overrun into idle control). Is there some math that can be done to figure out what this value needs to be set to on our engine?

Thanks,
Benjamin

[glausb23]

In our current configuration, the throttle aim minimum does not have any enabled axes for table setup, and is rather just set to a constant single cell value of 0.

To set this up correctly, should I enable the engine speed axis, and set the table up so that there is a value of 0 for all engine speeds within the bounds of idle aim + Idle Aim Ramp Down Limit, and then progressively start increasing these values for engine speeds higher than those above mentioned?

Thank you,
Benjamin