Lets Talk Torque Modelling / Virtual Torque

[hjtrbo]

I selected the 3 rpm cells at 0mg, x0.8 then highlighted all 3 columns and vertical interpolated. It still looked a little low so I went another round at x0.9.
Try it for yourself on your older tune to get the same values. Be a good learning exercise for you.
In the scanner, you need two torque charts matching the editors airmass torque table. One logs torque, the other logs spark.
I use Engine Torque as my torque channel. Delivered Torque cuts of early on WOT pulls on higher HP engines.

[GHuggins]

Delivered Torque cuts of early on WOT pulls on higher HP engines.

I've never found one doing that. They always seem to report up over 1000 lbft with delivered. Interesting...

[hjtrbo]

Clips at 868Nm on my setup

[Allen Vos]

I selected the 3 rpm cells at 0mg, x0.8 then highlighted all 3 columns and vertical interpolated. It still looked a little low so I went another round at x0.9.
Try it for yourself on your older tune to get the same values. Be a good learning exercise for you.
In the scanner, you need two torque charts matching the editors airmass torque table. One logs torque, the other logs spark.
I use Engine Torque as my torque channel. Delivered Torque cuts of early on WOT pulls on higher HP engines.

So I was able to replicate what you did with the VT-Airmass tables.

Here is a couple screen shots of the two graph tables.




Is this how I should set them up?

Ignore the first screen shot

[hjtrbo]

Change your cylinder airmass unit to mg
Increase your cylinder airmass resolution 0 50 100 150 ..... 750 800

[hjtrbo]

And delete the STFT only charts, they'll come back to bite you. With LTFT turned off you can still use the LTFT + STFT parameter as the LTFT component will be 0 anyway.

[GHuggins]

Clips at 868Nm on my setup

I know.... it clips at 640 lbft Like a lot of other things in the scanner...

To the OP, all you have to do is right click on the upper left corner of any table and either copy rows or columns then paste that into the scanner.

Don't forget to include variable cam timing into it if you're running it.

[hjtrbo]

I've never found one doing that. They always seem to report up over 1000 lbft with delivered. Interesting...

Did I read you wrong?

[Allen Vos]

Change your cylinder airmass unit to mg
Increase your cylinder airmass resolution 0 50 100 150 ..... 750 800

Oh Duh, Cant believe I missed the MG lol

Like so?

Attachment 142804

Attachment 142805

What exactly are we looking for on these tables to make our changes to the Virtual Torque, Airmass tables?

I see the changes you made to my Airmass tables, now I am just trying to understand how we go about this and why we change what we change is all.

Also There was mention about moving my idle torque parameters on the Virtual Torque MAP based tables from 30 KPA( Stock Cam) to 57 KPA (new Cam)

Are these tables used for that also?

[GHuggins]

Nope, it really does cap at 640 like you said. It was meant to mess with ya Had you double checking though didn't it

[Allen Vos]

And delete the STFT only charts, they'll come back to bite you. With LTFT turned off you can still use the LTFT + STFT parameter as the LTFT component will be 0 anyway.

Thank you for that tip

[GHuggins]

Oh Duh, Cant believe I missed the MG lol

Like so?

Attachment 142804

Attachment 142805

What exactly are we looking for on these tables to make our changes to the Virtual Torque, Airmass tables?

I see the changes you made to my Airmass tables, now I am just trying to understand how we go about this and why we change what we change is all.

Also There was mention about moving my idle torque parameters on the Virtual Torque MAP based tables from 30 KPA( Stock Cam) to 57 KPA (new Cam)

Are these tables used for that also?

Logging it is to just know if they're close to being balanced. It takes more than just shifting rows around to get things right. A dyno with correct back reporting is the only way, although if you can do some digging there are very strong arguments and proof pointing to being able to log fuel consumption and even tb port airflow (not the maf) to dial in the torque models within a few Nm's of reality. Dodge is supposed to be using tb flow for their torque modeling..... In fact a lot of what Jake talks about is very much simplified in the tb airflow modeling papers. It's supposed to be a cheaper alternative for manufacturers to not have to spend so much time and money on the modeling and make things like the "auto manuals" shift and work better.

[hjtrbo]

Oh Duh, Cant believe I missed the MG lol

Like so?

Attachment 142804

Attachment 142805

What exactly are we looking for on these tables to make our changes to the Virtual Torque, Airmass tables?

I see the changes you made to my Airmass tables, now I am just trying to understand how we go about this and why we change what we change is all.

Also There was mention about moving my idle torque parameters on the Virtual Torque MAP based tables from 30 KPA( Stock Cam) to 57 KPA (new Cam)

Are these tables used for that also?

Dyno is king, but can get close just for idle without it. Your WOT and general drive torque areas would not get any work based on just a mild cam swap if it came to me for a tune. Greg would probably dial it though .

In the case you were auto and adding forced induction, that is a WHOLE different scenario. As stated by Greg, the ECM torque reading would need to be checked against the readings from a good high end dyno. Or in leiu of that an experienced eye can know from the transmission line pressure if an increase to vtt is required. Any method all circles back to having spot on airflow data which is generated from a good tune formed from excellent injector / fuel system data.

Do 2 logs. About 8 minutes each from a bone cold start or until ECT has flattened.
Straight idle nothing else. A/C & climate fan off. Radio off. Lights off.
Probs will have to do it 24 hours apart.
Flash in your stock vtt tune the night before your first test.
After you shut the engine off following your first test, load in your mod vtt tune and let her sit till she's bone cold again, preferably the same time the next day.
Reason to flash the night before is to let any engine off timers run their course. Apples for apples testing method.
It can be done hot back to back, but better to be able to drive it so you can get injector tip temp down and let everything else settle down to nominals.
Also with the cold starts I can see if your minimum air multiplier needs a massage as a separate thing...

[hjtrbo]

Dodge is supposed to be using tb flow for their torque modeling..... In fact a lot of what Jake talks about is very much simplified in the tb airflow modeling papers. It's supposed to be a cheaper alternative for manufacturers to not have to spend so much time and money on the modeling and make things like the "auto manuals" shift and work better.

Are they using pre and post TB static pressure to get airflow?

[GHuggins]

Are they using pre and post TB static pressure to get airflow?

At it's very basic formula and this isn't even getting into the port airflow being figured into it as this is just the fuel consumption one as that directly relates to air consumption, is defined as

Indicated Torque = Fuel energy constant (which is stated as Nm/kg - otherwise 1Nm per 1kg of fuel energy output and as I understand it, is referring to the Lower Heating Value of gas which should be 44 to 46 MJ's "in the US" and as MJ/Kg which should translate to 45000000 Nm/kg as an average) * thermal efficiency multiplier (this is referring to that particular engines thermal efficiency which should be a number between 0 and 1 - most non DI gas motors are 20 to 30% whereas DI is more like 35 to 40, so these would translate to .38 for a DI and .25 respectively averaged) * fueling rate (kg/s) all divided by engine rpm as (radians/s) - the fueling rate and engine rpm are the only easy ones I'm seeing as something that can easily be logged or known here as I'm not entirely sure on the fuel energy constant and the thermal efficiency multiplier which is engine dependent...

Now your "basic" port flow uses temperature before and after, pressure before and after, current throttle area, manifold volume and MAF in it's basic formula to figure the air speed in the manifold and thus into the cylinders.

Then the formulas get more and more complicated getting into laplace domains. This all goes back to stuff written up in the 1980's and just recently put into function I guess as this was published in 2020. Honestly I start getting a little cross-eyed reading through it Their end results are torque models within 3Nm's of accuracy when compared to the engine they have on the dyno. It states they even purposely sabotaged it, killing a cylinder to verify the results stayed constant. Don't know, it's way over my head at the moment...

I have also found where if you know the weight of the rotational components you can calculate simply off of crank radians per second, but this apparently only works for lower rpms and you have to know the weight of every rotating component...

When a person says fell into a rabbit hole - hell, I got tired of trying to dig out for all the dirt that kept falling back in on my head, then said F-it and started tunneling to china Been digging into this stuff for a long time and I STILL keep coming back to using the VE table and it's changes for correcting the torque model to be working the best overall so far.......... You would think I would know something by now wouldn't you? but I just keep getting deeper and deeper and even further lost!!! I will say going by the VE tables corrections has gotten nearly all of the gen V's I do to idle right around factory timing and the tranny seems to go back to shifting and working correctly. Gotten a lot of compliments on near stock drivability. I've even used it to alter one torque model to another based on any minor engine mod that will throw it out such as anything that would change fueling - yep - the torque model is that sensitive IF you want to keep the idle "just right". If it doesn't work then what I've found is usually something in the throttle model is off and since A LOT of that isn't defined then yeah, you better know someone like yourself, Dave, Verlon or Will to get the tables out for you - gen V's are SUPER TOUCHY or can be and is why I keep saying the MAP side of the model is the most important

Just need that airmass part of the tool now


Edit for whomever may find this later....

As previously stated - at's it's basics - Indicated torque Nm = (Fuel Energy Constant * Thermal Efficiency Multiplier * Fuel Rate) / RPM in Rad's/sec

Where the following hold true -
Fuel energy constant = Lower Heating Value of Gasoline or in other words the energy output of gasoline per mole and given that all fuels have different values for this one can determine the following
87 octane in the US averages 44 to 46 MJ per Kg or an average of that being 45000000 once converted to Joules
93 octane in the US averages 46 to 49 MJ per Kg or an average of that being 47500000 once converted to Joules
You must convert to 1 Joule since 1 Joule = 1 Nm and all units must be kept the same.

Thermal Efficiency is the engine's ability to turn burnt fuel into energy output. This is where it can differ greatly engine to engine. Basically a normal port injected gas engine is typically 20 to 40 % efficient and more likely to be 20 to 30 % efficient
A normal DI engine is somewhere between 35 to 40 % efficient
Throw a cam into things and all of this changes usually lowering the efficiency due to pumping losses.
For instance a cammed DI motor may drop to 33 % efficient.
Thermal efficiency is always used as a decimal between 0 and 1 so for 33% this would be .33

Fuel Rate and Airflow Rate are said to go hand and hand as more air typically requires more fuel you would need more air, so for fuel rate we can substitute Airflow rate. This in turn changes into the following.
Throttle Body Airflow Kg/s = (MAF kg/s / Intake Air Density Kg/m^3) * (Baro Pa / Map PA) * (Throttle opening / 100)

Intake Air Density (kg/m^3) then = (baro Pa * Molar Mass of Air) / (Gas Constant * Intake Air Temp K)
Dry air normally has a Molar Mass of 28.97 and a Gas Constant of 287.1 J/(kg*k)

So if you have a cammed DI engine for example, this would be the basic scanner function -
((47500000 * .33 * (( [16.70] / (( [2340.90] * 28.97 ) / ( 287.1 * [15.240] ))) * ( [2340.90] / [2335.90] ) * ( [2111.156] / 100 ))) / [12.55]) + 40

You might notice the (+40) at the end of the equation. Here's the problem with this. You have to add accessory torque and friction torque to this for it to be anywhere near accurate. The problem with that is friction torque increases with rpm, so it's not a straight forward addition. Accessory torque, well there's momentum there so it won't necessarily increase with rpm.

Accessory torque for GM vehicles from all the digging I've done = 5 to 15 Nm's
Friction Torque Varies 10 to 50 Nm's for the most part, but can go higher.

The math formula works, but only seems to really work for these reasons at or around idle. Using the VE table on the other hand seems to be working out rather well not only with coming up with almost identical numbers to the formula at idle but getting more realistic numbers everywhere else.

[hjtrbo]

I'll set aside some time. Email me. Can you do me a solid and write me a detailed step by step (with screen shots) of what you need it to do in your email. Sometimes you go over my head with stuff I haven't joined the dots on.

[TragicMike]

*popcorn*

I have been too lazy to figure out why my torque tables can't be edited.
I know it's probably just an update.

Just bookmarking for later.

[Allen Vos]

Dyno is king, but can get close just for idle without it. Your WOT and general drive torque areas would not get any work based on just a mild cam swap if it came to me for a tune. Greg would probably dial it though .

In the case you were auto and adding forced induction, that is a WHOLE different scenario. As stated by Greg, the ECM torque reading would need to be checked against the readings from a good high end dyno. Or in leiu of that an experienced eye can know from the transmission line pressure if an increase to vtt is required. Any method all circles back to having spot on airflow data which is generated from a good tune formed from excellent injector / fuel system data.

Do 2 logs. About 8 minutes each from a bone cold start or until ECT has flattened.
Straight idle nothing else. A/C & climate fan off. Radio off. Lights off.
Probs will have to do it 24 hours apart.
Flash in your stock vtt tune the night before your first test.
After you shut the engine off following your first test, load in your mod vtt tune and let her sit till she's bone cold again, preferably the same time the next day.
Reason to flash the night before is to let any engine off timers run their course. Apples for apples testing method.
It can be done hot back to back, but better to be able to drive it so you can get injector tip temp down and let everything else settle down to nominals.
Also with the cold starts I can see if your minimum air multiplier needs a massage as a separate thing...

Thank you for providing the steps I need to log and figure things out.

I have to be honest, I feel a bit lost reading all the posts from you guys on Virtual Torque.

Like, I understand the very basics, but how to go about dialing it in is where the struggle is for this fella.

I'm very much a guy that needs a bit of guidance when setting stuff like this up and when it works I can look back and sort of reverse engineer it so to speak so I can under stand it.

I will flash the the stock VT tables in the car tonight and do a cold start tomorrow night.


On a side note you mentioned my minimum air multiplier needs tweaking?

Can I ask what gave it away? lol

Are we talking about the idle airflow mult ECT Table( ECM 2198)?

[Cringer]

*popcorn*

I have been too lazy to figure out why my torque tables can't be edited.
I know it's probably just an update.

Just bookmarking for later.

Something on VCM Suite version 5.x broke virtual torque...so you'll need to use version 4.x.

[Allen Vos]

Something on VCM Suite version 5.x broke virtual torque...so you'll need to use version 4.x.

The newest Beta version will do Virtual Torque.

I had the same issue lol