Lets Talk Torque Modelling / Virtual Torque

[Allen Vos]

yes.

Thank you for that explantaion

[Allen Vos]

They have to be balanced. That's the point of the multiple tables.

If you looked at the patent and then know that the map side is "primarily" used for throttle control, you would see that the map side directly relates to throttle as it's primary lookup as it directly relates to it's "desired position for idle" - if this is the correct patent "I'll have to double check", then it clearly states that the map "value" is referenced first to define desired throttle based off of desired eff throttle area - that then defines the torque and timing lookup for idle that is referenced on the airmass side - this is where the airmass side then plays into it. It's just as much from the airmass side as the two relate to one another, SO they must be balanced or close to it..... Lowering airmass to bring things in line does indeed work, BUT you're usually skewing the airmass side when you're doing that (making it do ALL of the work and leaving the throttle desired locked where it was, which can cause its own problems) Now, adjusting your VE table does bring the map side back in range - this is why there are those who say you don't need to adjust anything. However doing this doesn't put it back right.

Your torque goal is around 10 to 40 lbft at idle. 4th gens are fortunate enough to not have full torque control. You have other tables playing into it that you can "usually use" in place of the torque models. However there are those oddball builds that will want to idle on negative ignition timing even on gen 4's because things changed so much and tuning in the torque model will make a difference even with your more minor mods, however....

Fuel needs to be completely right, before messing with any of this and I'm not just talking about select areas of the VE tables, but all of it.

Guess it doesn't hurt to change stuff just to see what it does, but this needs to be dialed in last on one of these.

Greg, I am wondering If you could take a look at these tables for me and tell me what you think? I know I said I was shelfing it for now but I got a bit bored this week and set the Scanner graphs up as we had discussed and did a log.

I just started from scratch and flashed the stock torque follower and Virtual torque maps to the PCM and did a log from cold start and let it warm up.

Here's what I came up with

Attachment 143176
Attachment 143177
Attachment 143178
Attachment 143179

So this covers the idle only which is all I am focusing on right now. I have circled the areas that have the highest cell counts.

All Torque values are very close and fall within the 10-20 ft pounds that you had mentioned for a car with a cam in it.

How close would you like to see these numbers? My 10 degree MAP and AIR MASS were 19 and 15 respectively and 20 degree MAP and Air Mass tables were almost spot on.

Now my biggest question is regards to what you said about how important it is to keep all the air mass and MAP tables balanced.

By looking at these I feel only my 10 degree tables need altering as they are 4 ft pounds apart. Would you bother adjusting these?

If so can I uncheck the link selection and edit the Spark 10 MAP table and bring it from 19 ft/lbs down to 15 ft /lbs to match all the other torque values?

Yes I know you have reiterated VVE fueling is very important for this, but in the interest of learning the proper procedure Is where I am right now.

You had also mentioned something about a tool to correct or bring the torque tables inline. Is this something available to us?

[GHuggins]

Tool isn't done yet. Nathan's working on auto filling software for it to make the airmass corrections as he gets time. I'm not entirely sure I'll be offering it to the public after thinking more about it anyway. So far just the map adjustments and then hand working the airmass sides from the data shown are working great on the gen V's.

I can tell you that if you do like Hjtrbo or Nathan was saying on the map side instead of the airmass side it'll help you out. Then smooth or lower the 100kpa row over to about 2000 rpms from where you changed it at idle should fix a lot of your problem.

This is a 4th gen I did a while back that fixed a lot of the throttle problems like what you're talking about having. It may not look right, but they're balanced and you can kinda see how much things change model to model. This was the best power making cam I've seen in an LS3 too, so aggressive to put it nicely.

Airmass.jpg

map side.jpg

You can also use a GM crate engine calibration to probably get something really close for a mild cam. They have the correct torque models straight from GM for a mild setup.

If you really pay attention to it, you'll be able to see the relation between VE changes and the torque model changes, especially on the map side. This is why I started going after the relation of the two over a year ago. I just wasn't doing it correctly then.

[hjtrbo]

My condition for developing the software to realise your methods and ideas is that it is not to be made public.

It is a complex tool whose data output requires an experienced eye to verify and hand edit or discard the suggested adjustments. There are no documented procedures on how to use it and interpret the output, nor any ability to quality control the data that a user enters into it. The anticipated shear amount of support required will outstrip my will to live.

[SiriusC1024]

*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.

Use the beta release of VCM Suite.

[GHuggins]

My condition for developing the software to realise your methods and ideas is that it is not to be made public.

It is a complex tool whose data output requires an experienced eye to verify and hand edit or discard the suggested adjustments. There are no documented procedures on how to use it and interpret the output, nor any ability to quality control the data that a user enters into it. The anticipated shear amount of support required will outstrip my will to live.

Yes, this and I have put a lot of R&D, time and money into the knowledge for it over the years. Not to mention just skimping out on the fuel tuning and not having it at least 90% for the entire VE table dialed in would have the data big time screwed up.

[DirtyDave86]

Great reading material in here. I was under the impression I should be matching my VTT to the logged delivered engine torque.

[GHuggins]

Great reading material in here. I was under the impression I should be matching my VTT to the logged delivered engine torque.

That works great for gen 4's. You just have to make sure and add to it in the higher loads and rpms to compensate for friction losses. I also figured out the math equation for calculating torque a few pages back, but it too requires frictions and losses to be added back to get the correct answer, so I don't see the point in it myself. Someone who can pull the coding apart and then come up with the rest for their particular motor, well that's a different story.

[DirtyDave86]

That works great for gen 4's. You just have to make sure and add to it in the higher loads and rpms to compensate for friction losses. I also figured out the math equation for calculating torque a few pages back, but it too requires frictions and losses to be added back to get the correct answer, so I don't see the point in it myself. Someone who can pull the coding apart and then come up with the rest for their particular motor, well that's a different story.

The math is wayyy over my head

If I am understanding you correctly on this
Let?s say I?m driving at 1800 RPM .50 on the map
Scanner is showing around 280 lb ft delivered engine torque. I would want to match my VTT tables to what the delivered engine torque is, at any given rpm and map cell? On my LS2 G8 I show 500ish delivered engine torque at WOT,

[GHuggins]

The math is wayyy over my head

If I am understanding you correctly on this
Let?s say I?m driving at 1800 RPM .50 on the map
Scanner is showing around 280 lb ft delivered engine torque. I would want to match my VTT tables to what the delivered engine torque is, at any given rpm and map cell? On my LS2 G8 I show 500ish delivered engine torque at WOT,

You need to plot each individual spark table vs airmass and map using delivered torque, but basically yes with what you're asking and then add some back for the losses. This works good on gen 4's only because gen 4's use other players. If they were fully torque controlled on the other hand, this wouldn't work.

You can apply the delivered values 3 to 5 times then see how it's running. You also need to hand shape things to get delivered to be something n the 5 to 25 lb/ft range at idle in park while hot. You might be surprised with how differently it'll run after doing this for a few passes. Wait till the end to add the losses back as sometimes it'll increase the higher values on it's own.

[NotSure]

Wait till the end to add the losses back as sometimes it'll increase the higher values on it's own.

What losses are being added back in, drivetrain? Are the torque figures in the VTT supposed to be axle? I need to revisit my VTT. At one point I was only updating the stock values only in cells if Delivered was higher, but then smoothing and/or calculating was messing up (raising) other areas which ended up with harsh WOT shifting. Reviewing a log showed that the ECM was not even trying to reduce spark during the shift, apparently because the torque figures I had in the lower spark tables were too high and the ECM just decided, “F-it, pulling spark wont reduce torque so I wont bother”. I put vtt back to stock and things went back to normal. Gen4 SBE with blower.

[GHuggins]

Losses = anything friction or parasitic associated in relation to what's bolted onto the motor and is directly engine related only.

VTT is engine torque, typically in Nm's, as measured on a dyno with everything unhooked from it including but not limited to the water pump, alternator, ps pump and so on plotted mathematically via a quadratic formula to best fit multiple timing tables. That part in particular is what I got hung up on for several years as I always thought the OE set the timing and "fully" dialed in the models - I've since learned they don't.... Even on the engine dyno, losses have to be calculated back in. They measure them as I was informed by using electric motors to spin the engines up while not running. Losses can also be measured on a dyno by killing the engine and plotting on decel as well. I would randomly do something like that to measure drive train loss running a chassi dyno by pushing the clutch in after a pull and hitting the read button on the dyno again. Same thing for the engine only they shut it off. The electric motor method would be a whole lot more accurate and easier, but I don't know of anyone that has one

The OE goes as far as including pumping losses into the equation along with rotational frictions and even something as simple as the torque required to run the oil pump at different oil temperatures. HJtrbo posted some of these a few pages back. Typically with a cam you'll see it drop in the lower loads. With a blower, it'll smooth out instead of having the usual big last bit of rpm jump.

Right now Dave (DStek) is the only one I know coming out with a true method for plotting the torque models straight off of a chassi dyno, but he has the way to get all of the loss values out of the cals to plug them back into the tables as plotted.

I haven't seen any issues with timing cuts during shifts plugging the values directly in on the gen 4's. Most likely you weren't correcting the map model, which is what directly controls that kind of stuff and in adding to the airmass side you just got things really skewed. I will say if you have a high stall converter the torque numbers will be greatly thrown off while logging. You have to make sure you lock the converter for accurate readings in that regard.

For gen V's I run a completely different HJtrbo created program per my own build request that he has taken several steps beyond to reshape the torque models back to OE specs. So far, it's working wonderfully on the map side. He's working pretty dang hard on the airmass side right now, but that's for gen V's, which run complete torque controlled systems. Then if a customer request something special like more pronounced downshifting coming to a stop for more engine braking or whatever, I hand shape them from there.