Initial Development And Prototyping Of A 102 mm GM LS Throttle Body On A Supercharged Tundra

Grant from TRP stated:
"Going from 76 to 95mm: 47.5^2 / 38^2 = 1.56 So you have potential for 56% more airflow given the same pressure ratios.
Not sure you want to alter your Throttle vs Pedal tables that much though. Note the warning in HPT. I know with many GM ECUs changing those tables more than 5% can permanently brick the ECU."
He also said he ran dual 90mm on his GM and got close tob50% with no issues.
Anyone know or seen if this applies to Toyota ECU's @ViktorG ?

 

I've been looking through some HPTuners forum posts and seeing that there might be 3 areas to tweak when changing TB sizes, at least there are on Dodge vehicles (Throttle Body Airflow table, Desired Throttle Small Range and Desired Throttle Large Range). Most suggestions are to make small changes, in the 5-10% range of existing values and not to try to scale to match. There is also an assumption that the existing numbers aren't derived from real-world testing or scaled calculations, but more of a "feels good" method of tuning them.

https://forum.hptuners.com/showthread.php?62333-Bigger-Throttle-Body/

 

Question for a side bar: why for are the torque values in the graph different from the stock file? I realize it is supercharged but looking at a few other tunes, these values remain unchanged from the stock values up to 400ish. Where did those numbers come from? Dyno'd? scaled up?

 

I believe my stock trd tune had pretty high numbers. @snivilous changed changed them in my tune. But I sent my base tune in to TRP and Grant stated they were the same as @armyoffoo tune . I'm not sure if they were his base trdctune or his new one.
My new values.

 

It seems there are some tables that are scalable and others to play around with until you hit that right mojo...if I knew more I'd venture to give it a go.
I emailed @snivilous when I had the 102mm installed and brought up the light switch or on/ off affect when giving it throttle to see where you can get part throttle drivability back.

 

Retracted

Edit: Those values aren't from my truck, must be scaled up...my tune file shows 399 as the highest value.

 

The more I read about other makes/models having all this fine-tuning ability for different TB sizes, the more frustrated I get.

Saw others using this formula on the HPTuners forum, Pi * (r * r) * (% open) = volume

WOT on a stock TB would be 3.14 * (38 * 38) * .86 = 3899mm squared
WOT on 84mm RCF TB 3.14 * (42 * 42) * .86 = 4763mm squared
WOT on 90mm GM TB 3.14 * (45 * 45) *.86 = 5468mm squared
WOT on 102mm GM TB 3.14 * (51 * 51) * .86 = 7023mm squared

 

Isn't the 0.86 that you're using in these formulas degrees of TB opening rather that percent? Apologies, I'm not at home to look in HPT.

 

Looks like degrees.

 

Don't forget degrees is different than percent, so your calculation example is showing ~74deg throttle (maybe that was your intent, just seems like a random number). I think that's where @Wynnded confusion came from.

 

@Silver17 posted a Holdener video about TBs recently. At 50% throttle opening the tested TB setup had 93% flow. Sounds like it isn't linear which is potentially causing some issues.

Relative "Airflow" = 1 - cos( T * 0.5 * Pi ) , T = throttle position, from 0 to 1.

Throttle relative opening = ( 2/Pi ) * arccos( 1 - Xr ) , Xr = Relative "Airflow", from 0 to 1.

Formulas from https://www.hpacademy.com/forum/general-tuning-discussion/show/throttle-position-and-airflow/

 

Here is the CAD calculated values.



Which all of that work to say it does appear to be a linear change which I didn't expect. If you sum the total areas calculated for each body size, they are the exact same difference as just multiplying the throttle body areas:



All of that to say, in regards to opening at least, you can just straight multiply the area increase of whatever throttle body.

 

I think @wynned was asking about @armyoffoo Calculations of .86 for degrees. I was just posting the table from Hpt.
Hopefully we can get it dialed in for the part throttle drivability.

 

Looks like it's not ENTIRELY linear, but extremely close. Extrapolating that data out one step further:

 

The area will be linear but the flow isn't. I'll see if I dig up the holdener video explaining it.

 

Oh no doubt, I would imagine airflow efficiency goes up with more throttle because you disturb the air less. Take it to the limit and your aerodynamics are horrible fully closed, and fully open it's effectively a tube with a plate in the middle so is extremely unhindered and I would imagine mostly laminar. Someone could model that in a CFD program. But I also would think the stock map would have similar efficiencies already baked into it, so scaling it based on area may not be too far off.

 

Also here's my fleshlight looking model to calculate area. I used the throttle body ID, then drew a plate 1mm thick and 0.1mm smaller diameter, then varied the angle, then extruded the area you could "see through" the throttle body and measured that area:

 

Would it be worth looking at the tables for the 2UR-GSE w/ 76mm vs 2UR-GSE w/ 84mm?

 

Thanks for the work up. With my 95mm TB I'm at 56% increase from stock.

 

Thinking about it more, if you had a pressure sensor after the throttle body and before the blower, I would think pressure could be measured and that could POSSIBLY(?) be correlated to airflow? If the TB is fully open there is no vacuum, if it's closed you have full vacuum, and plotting those datapoints at various throttle positions could give a curve to then calculate optimum torque from.

Maybe even the normal boost port, if we're modifying the optimum torque values then may as well look at positive boost. Though my thinking with before the blower is the variable of the supercharger would be removed. But maybe slowly stepping up throttle body sizes and plotting vacuum/boost vs throttle angle, then go and update optimum torque based on that curve, then run a larger throttle body and update it again, and keep repeating slowly moving up. I don't know if either of those methods are more accurate, if even correct at all?

 

Not sure what is throwing things off to be honest. I need to digest more.

My money is on the logic of the TB signal converter not perfectly mapping to what Toyota expects the TB position to be so the calculations are off. EDIT: The GM throttle body may have its own logic too.

Why do I say this? We can run the RCF TB with no issues presuming the TB area is simply scaled up and it works.

Possible the underlying CAD calculations are making some weird assumptions. May not matter but there is some leaking on the small end (closed) and the obstruction by the throttle blade on the big end (WOT); I don't see how these would change things so significantly to cause idle and other issues. Previous point partially debunks this but as the opening and blade get bigger it may matter more?

I think your graph supports this too but looking at this paper the TB area calculations aren't perfectly linear (page 6) and diminishes as the angle approaches WOT at least for the points selected it still doesn't look like it matters enough to cause issues:
http://www.diva-portal.org/smash/get/diva2:23184/fulltext01.pdf

 

Yes, it's TB degrees open, probably not what we're looking for...but that's what others were using to determine the ECT scalar number based on the GM tables available.

 

If you need a 102mm fleshlight....

 

Lots of great info on TB's!
We have perfect lambda and TRP that is scaling the signals for the different TB types. My idle with the 102mm and 95mm was spot on after install. The GM and Dodge forum members had drivability issues going to a larger TB. It seems they are able to fine tune different TB parameters, airflow, large small TB plate etc, where the Toyota TB functions areimited to torque, throttle and pedal positions. I'm sure there may be more. But playing with the different values and what's commanded for part throttle should be a possibility. Wot is wot so not much needed there.

 

That eliminates that theory then. Sorry.

 

I'm only looking to make sure shifts are on point and not hitting too hard, or hanging. I can control the hanging by letting off the throttle slightly and it'll shift. I did have one cold start P050a and hit a P0101 at part throttle, so just need some fine tuning as it were. Overall I'm very happy and wouldn't swap back to the RCF unless I had to.

 

What's interesting as I re-read @armyoffoo post concerning hellcat tuning for larger TB's. As @snivilous table shows it should be linear and be able to plug corresponding values to get part throttle response inline with Toyota parameters.
They used the smaller TB tables, voltages, to get the larger TB's to respond for everyday driving. But that's for dodge ecu's.

 

Sorry y'all not trying to derail y'alls findings and discussion. I'll say my peace and I'll shut the fuck up.

I know you were tracking what I was getting at especially based on @snivilous and @Mdl 's comments in responses.

I'm still skeptical that TB surface calc is linear since the blade edge moves in an arc. The linear approach is probably used as a quick and easy calculation without needing to use trigonometric functions.

My understanding of calculating disk surface at an angle would use the equation (Pi*r^2)*cos(throttle angle).

First part is area of a circle. Second part sets it up so TB surface is 100% at angle 0 indicating closed, 0% at angle 90 indicating open, and everything in-between as the blade sweeps it's arc.

So at 45 degrees, or 50% throttle opening, it would be off by ~20% if you use 50% vs cos(45).

Why? Cosine at 45 degrees is 1/sqrt(2) or 0.7071067812 (meaning ~70% surface) not .5 for 50%. A value of .5 x TB area would be the value of the blade surface at 60 degree of open throttle.

If the above is correct, the more important question is what is the impact of a 20% error propagated through airflow calculations? I'm still looking for a good explanation of airflow characteristics since it isn't linear and if it matters if it is boosted or not (I suspect it doesn't matter).

You addressed TPS as a possibility of an error. If the vendor already says it is on the money the only way to test is to bench both TBs and log the inputs and output or via observation like in this test:
https://youtu.be/LRdP9p8XEWg?si=CvW1y0nXGiX-xFmg

I'll exit with my horse now and am ready to be educated more on this subject.

EDIT: I referenced this answer from Paul Grimshaw assuming it is correct:
https://www.quora.com/How-can-you-c...30-degrees-off-of-facing-directly-towards-you

 

No need to stop.as your on the money. And I didn't mean to come across that it's plug and play where its good to go. We were hoping it would be simple and it was suggested it is not linear and the other variables need to be addressed. I was testing the relatioship commanded throttle, throttle position, pedal position and rpm speed. There were other functions I wanted to log but hpt doesn't have them for Toyota or I picked the wrong ones.
I haven't got into the nitty gritty the device. The main goal was just to get the Gm TB signal to respond to the Toyota signal without throwing codes.
Our problem is in hpt we need to be able to make further adjustments for all the other vvariables. Thanks!

 

I sent in my data log and will try some suggestions for shifting.
Response: "Looks like your shift RPMs match what is in the tune that Army/Bryan sent me.
For example, in first gear 1000 RPM output = 1000 x 3.333 = 3333 engine RPM. So at around 30% pedal it would be shifting at around 3300 engine RPM.

If your shift tables are the same, you can reduce the values in the table if you want it to shift earlier.

Desired Output RPM to shift = Desired Engine RPM / Gear Ratio
If your shift times are too long/flared, you can bump up Base Line Pressure."

I will attempt to make some changes and see where it's at.