LT1 Throttle Oscillation Question

[smokeshow]

A smaller engine can be expected to make more torque per unit air than a larger one... Lugging around extra weight at part load isn't free. I think you should read up on the fundamentals, see if that doesn't clear up some questions..

https://patents.google.com/patent/US7021282B1/en

[DSteck]

The way HPT has the view formatted causes significant confusion. When you view in the right way (different graphs at the RPM breakpoints that are APC/MAP vs spark), it makes a lot more sense. RPM isn't actually a variable of the math.

The "crash" you're seeing outside of normal control is the result of math extrapolating out to areas you don't operate in. This all circles back to why back calculating off of tables where people just fill things in is a bad idea. It is far better if your coefficients calculate based off of the actual data you get while driving so those operating conditions are accurate. This is the same screwed up behavior that can be seen in virtual VE tables where the numbers just get ridiculous when you drift farther from the region that the engine actually operates in. These equations are a "best fit" as opposed to perfect replications of the engine.

The screen shot is a stock LT4 torque model at the 550rpm breakpoint against APC (in mg) and spark advance. At each APC, torque goes up as timing goes up as expected.

[GHuggins]

The way HPT has the view formatted causes significant confusion. When you view in the right way (different graphs at the RPM breakpoints that are APC/MAP vs spark), it makes a lot more sense. RPM isn't actually a variable of the math.

The "crash" you're seeing outside of normal control is the result of math extrapolating out to areas you don't operate in. This all circles back to why back calculating off of tables where people just fill things in is a bad idea. It is far better if your coefficients calculate based off of the actual data you get while driving so those operating conditions are accurate. This is the same screwed up behavior that can be seen in virtual VE tables where the numbers just get ridiculous when you drift farther from the region that the engine actually operates in. These equations are a "best fit" as opposed to perfect replications of the engine.

The screen shot is a stock LT4 torque model at the 550rpm breakpoint against APC (in mg) and spark advance. At each APC, torque goes up as timing goes up as expected.

So an LT5 doesn't run above 4600 rpms? Or are you saying that they literally didn't bother to dial any of it in outside of the main spark groups that "it should" see that kind of ignition timing in? Or are you just saying how it is in hpt is wrong, which I've also often wondered.

[GHuggins]

A smaller engine can be expected to make more torque per unit air than a larger one... Lugging around extra weight at part load isn't free. I think you should read up on the fundamentals, see if that doesn't clear up some questions..

https://patents.google.com/patent/US7021282B1/en

I'll read through it. I understand how a tighter squeeze can net more torque, but to say a smaller engine, especially a NA one, makes more torque when we always see a bigger piston area make more torque on a dyno, well I guess I would have to see it graphed out correctly to understand the "true" torque models.

[GHuggins]

The way HPT has the view formatted causes significant confusion. When you view in the right way (different graphs at the RPM breakpoints that are APC/MAP vs spark), it makes a lot more sense. RPM isn't actually a variable of the math.

The "crash" you're seeing outside of normal control is the result of math extrapolating out to areas you don't operate in. This all circles back to why back calculating off of tables where people just fill things in is a bad idea. It is far better if your coefficients calculate based off of the actual data you get while driving so those operating conditions are accurate. This is the same screwed up behavior that can be seen in virtual VE tables where the numbers just get ridiculous when you drift farther from the region that the engine actually operates in. These equations are a "best fit" as opposed to perfect replications of the engine.

The screen shot is a stock LT4 torque model at the 550rpm breakpoint against APC (in mg) and spark advance. At each APC, torque goes up as timing goes up as expected.

I might be speaking too soon, but even air per cylinder torque based off of rpm and spark doesn't make sense to me. Why would torque increase so high at idle just because your wot at idle? You can still only suck in and pump so much at a low rpm. Would physics not still dictate that it would make more torque at the peak torque point of engine rpm's?

[DSteck]

So an LT5 doesn't run above 4600 rpms? Or are you saying that they literally didn't bother to dial any of it in outside of the main spark groups that "it should" see that kind of ignition timing in? Or are you just saying how it is in hpt is wrong, which I've also often wondered.

I'm not sure what you're asking. They aren't going to force engines to run at 6000rpm at 1500 APC at -10* (seriously, I can't enter a degree symbol?) most likely. They're going to gather data from the core region they operate in, whatever that is at 6000rpm, and then run the surface off of that.

We can all agree we're never going to run an engine at 1000mg on APC and 40* timing at 550rpm. Yet the way the current editors/tools work, they're going to make you stick a data point in that cell before they'll calculate the coefficients. This is where problems come in, and this is where it'll throw off your core operating area.

My personal opinion is that the display approach HPT uses (where you get multiple graphs based on spark) is wrong. They should give you 12 graphs, one at each of the RPM break points, let you pick the current cam phasing you're in, and then populate tables that are APC or MAP vs spark timing. RPM is not a variable of the equation; it's just a constant at that particular RPM (engine speed is not part of the math).

I might be speaking too soon, but even air per cylinder torque based off of rpm and spark doesn't make sense to me. Why would torque increase so high at idle just because your wot at idle? You can still only suck in and pump so much at a low rpm. Would physics not still dictate that it would make more torque at the peak torque point of engine rpm's?

And this is why there's a MBT table. Again, there are conditions you will never hit. That doesn't mean you can't manually calculate some arbitrary torque value at that point. I can get a torque value for 100* of spark advance at six million mg of APC at 6000rpm. It's definitely not going to be accurate, but it's a math equation. You can plug in anything and get something back out. The coefficients were generated through regression, and regression works best when you're ONLY considering the data that you're actually getting.

Think of it like this: when you populate a histogram for VE, you've got MAP (or pressure ratio) versus RPM. You're not going to hit the corners, and in a turbo car, just because you can hit 200kPa at 6000rpm doesn't mean you can do any better than atmospheric pressure before 3500rpm. So now you have a table that you've been filling in with a bunch of dead space, and you need to generate coefficients for a predefined equation from your data. Wouldn't you rather generate those coefficients off of what you actually logged instead of trying to manually extrapolate and re-run the coefficient generator?

[GHuggins]

Next question since this would GREATLY help to dial things in. Is this table using some sort of air per cylinder value based off of throttle position or is it using a sonic airflow equation?

Throttle area.jpg

Now I really wish I could just see the "actual" models instead of what hpt is showing.

[GHuggins]

This is what I was referring to on the LT5's. Yes i completely understand shaping things and how that continues to flow outside of the normal realm, but nonetheless some this still doesn't make sense unless it's just how hpt converts the data.

LT5 Airmass.jpg LT5 Map.jpg

[smokeshow]

This is what I was referring to on the LT5's. Yes i completely understand shaping things and how that continues to flow outside of the normal realm, but nonetheless some this still doesn't make sense unless it's just how hpt converts the data.

LT5 Airmass.jpg LT5 Map.jpg

Read what Dave said about calibrations only reflecting data from areas that the engine actually operates. The answer is there lol

[GHuggins]

Yes, I read that and answered it. So it's never going to see light throttle high rpms from someone leaving it in a lower gear playing? Or it's never going to see 10 to 20 degrees of timing from full boost? Again, appears to me they made the lower loads super negative to offset for throttle controls unless I'm just not understanding the axis correctly. They are cylinder airmass in mg and map in kpa right?

LT5 full load.jpg

[smokeshow]

I would say cam angle at zero is the most likely reason that area is not represented.

[GHuggins]

Well now, I did not even consider that and you are 100% correct. I keep forgetting that GM used variable cam timing on some boosted motors and not on others. The LT5 indeed used it. It does however still nose dive off, but nowhere near as bad. Stupid me either way...

Next question as I understand the operating routines and logic more so now that I have a different patent to look at

Map is used first and foremost. Defines throttle and then Airmass is used to define target MAF flow and target timing more or less from that. This is how I discovered it to work anyway to a degree and the patent seems to back that up.

What defines "target" air per cylinder based off of throttle as that seems to be the 3rd thing it's looking at? Is it the above throttle area table? Or is it something else yet to be mapped?

[DSteck]

Next question since this would GREATLY help to dial things in. Is this table using some sort of air per cylinder value based off of throttle position or is it using a sonic airflow equation?

Throttle area.jpg

Now I really wish I could just see the "actual" models instead of what hpt is showing.

Without knowing where you got that from, any answer is a guess.

I would say cam angle at zero is the most likely reason that area is not represented.

Was just about to say this. Look at the factory cam phasing and then change the intake cam position in the torque models.

One saving grace for time with aftermarket cams is that more than likely... they're going to get locked out to a fixed position which takes your iterations down to 20% of what they'd need to be since you only need to focus on one cam angle.

[GHuggins]

Without knowing where you got that from, any answer is a guess.


Was just about to say this. Look at the factory cam phasing and then change the intake cam position in the torque models.

One saving grace for time with aftermarket cams is that more than likely... they're going to get locked out to a fixed position which takes your iterations down to 20% of what they'd need to be since you only need to focus on one cam angle.

That should be the throttle area/rate table out of a LT1. They change for cars, trucks, boost or no boost and don't seem to be necessarily tb oriented as you can have 2 running the same tb use different tables.

Yes my stupidity as I said. I dial in the models for cam movement, but that's me and doing it the only way I currently know how to until I'm better enlightened.

[DSteck]

That should be the throttle area/rate table out of a LT1. They change for cars, trucks, boost or no boost and don't seem to be necessarily tb oriented as you can have 2 running the same tb use different tables.

Yes my stupidity as I said. I dial in the models for cam movement, but that's me and doing it the only way I currently know how to until I'm better enlightened.

Is that a guess, or who gave you that? There's a lot of random tables in the files that can look like a smooth curve. Post a file and a memory location.

[GHuggins]

Is that a guess, or who gave you that? There's a lot of random tables in the files that can look like a smooth curve. Post a file and a memory location.

What I was fixing to do Got sidetracked with emails. All I know is it makes a BIG difference on boost additions as far as throttle control goes, so was thinking sonic airflow, but it also just looks like throttle area via a tps setting that is dictated by the max throttle area setting.

Can the air per cylinder per throttle be defined? I assume it's using something to control that setting?

[DSteck]

I fixed the one you mentioned and added a paraphrased annotation.

[GHuggins]

I fixed the one you mentioned and added a paraphrased annotation.

Yes, this was the other way my coder had it. Well not axis wise. Actually works better as small changes make bigger differences adjusting it that way Although I prefer the scaling to go horizontal - yeah I'm picky - I know...

Ok, this goes further to explain why it needs to be lowered for bigger tb's even though you would think it would need to be raised. Don't ever raise it. Bad things happen.

Thank You.

[DSteck]

Yes, this was the other way my coder had it. Actually works better as small changes make bigger differences adjusting it that way Although I prefer the scaling to go horizontal - yeah I'm picky - I know...

Ok, this goes further to explain why it needs to be lowered for bigger tb's even though you would think it would need to be raised. Don't ever raise it. Bad things happen.

Thank You.

Horizontal vs vertical, whichever. I set it up the way it is by the OEM.

This table shouldn't change that much if the throttle area table is set, but none of the aftermarket throttle bodies have this data available. There is a missing throttle area table that is important.

[smokeshow]

It's rare for Dave to give someone special treatment like this. Lol