Tensors?

[bink]

I was reading this thread on LS1Tech -> http://www.ls1tech.com/forums/showthread.php?t=710290&page=1&pp=20

Is GM using tensor equations/matrices in VE calculations?? Seems like it would be logical....but advanced beyond me! :D

[Maggie]

Hmm..."tensor" as in multilinear function? Or, maybe as a table or field where each point constantly varies with respect to location of activity?....wow, I'd love to meet the engineer who pulled that off! :Eyecrazy:

[bink]

Marcin/RedHardSupra is looking for someone to work through the calcs with him....using MatLab.

Anyone with a working knowledge of MAtLab and/or maybe tensors??

[ToplessTexan]

I am up to my ass in work right now but I'll shoot him a PM.

[bink]

I am up to my ass in work right now but I'll shoot him a PM.Wow, that would be great!! maybe sometime you guys could explain it to me.....in simplified terms. :D

[redhardsupra]

i have some other ideas on solving it besides tensors, but if we could get a nice general method using tensors, that'd be great. we'd have one method to solve the problem with 6 or the full 17 or whatever it is parameters.

[redhardsupra]

cmon guys, i need some math help, there's gotta be one engineer in here...

[Acer]

redhardsupra,

I am an engineer & willing to have a go (maybe foolish ;-). Please PM me & we can work out how we proceed together.

Cheers,

Acer

[joecar]

It's been more than 20 years, and never had to apply tensors in all this time...
I'm digging thru my old textbooks, but I'm away on a linux device drivers class this week.

[redhardsupra]

hey guys, do you have skype? it allows you to make conference call up to like 10 people, and i could show you how far i got and where i'm stuck. the cool part about it is that i can also send you files and type equations at the same time to all of you, which is way better than doing regular fone. what day/time would be good for you?

[joecar]

I don't have skype...

[redhardsupra]

it's free, all you need is a mic, and they're just few bux, a lot of new monitors/laptops just have them built in.

anyway, i figured out a way to tune the BIAS table, so the biggest obstacle so far is done. now all i need is data, anyone with a E40 car willing to be my data logging mule? :)

[mistermike]

The more coffee I drink, the tensor I get.

[Acer]

redhardsupra,

I drive an HSV GTO LS2 with E40 PCM & tune it using Flashscan V2. What data do you want to collect & what are you doing with it? Just PM me if you feel more comfortable.

Cheers,

Acer

[redhardsupra]

here's the list:
IAT, ECT
MAP, RPM

GM.VETABLE_DMA

DYNAIR

AFR(both commanded and wideband), BEN

IPW, GM.INJFLOW (IFR lookup basicaly)

here's few goofy LS2 specific things that i havent seen at work at yet, but would like to:
GM.APCYL_E40_DMA
GM.DYNAIR_LS2
GM.ENGZONE1_DMA
GM.AIRFLZONE_DMA

and i'm also not sure what's the difference between:
GM.DYNAIRTMP_DMA
GM.CHRGTEMP_DMA
but i definitely need one of them, just now sure which one is the appopriate one for LS2 SD calculations.

i would also need your B0101 and the B4901 tables.

i think that's it. the _DMA stuff doesnt' take any ODB2 bandwidth so feel free to go hog wild and scan them all ;)
I'd really appriciate it if you guys could log all this stuff so we could make sense of all of them.

What i'm doing with it? I'm working on a whole new way to tune VE, GMVE, and also the parameters that replaced VE in the E38/E67 PCM's.

[Acer]

redhardsupra,

No worries, I'll log some data by next weekend & send you the EFI Live log file. Haven't logged some of these "goofy" LS2 variables before so will see how they log.

If I understand it correctly, you are working on an algorithm/equations for back calculating the VE using airflow readings etc. ;) . I have been using a similar technique on LS1's for 3-4 years & have developed a similar approach for my LS2. Would be good to compare notes though & see if we can solve a few funny things I have noticed as well.

Have sent you an email as well.

Cheers,

Acer