RomRaider

Here is my attempt to overlay some engine demand curves onto some compressor maps. The info is from http://www.stealth316.com. I downloaded the IHI maps from http://www.nasioc.com. The volumetric efficiency was estimated by Kyle "BlackHole" on legacygt.com. I still need to read through ride5000's thread on MRP data to estimate VE from datalogs.

Edit: Kyle "BlackHole" estimated VE from several datalogs. I've updated the spreadsheet to use his numbers.

Edit: Updated cubic inches and added some comments.

Edit: Added some Garrett maps and scaled graphs so they look better at 150% view.

Edit: Added the TD04H-13T (should be same as TD04L-13T) for the WRX guys. Don't forget to change the engine displacement to 2.0L.

wow, excellent work.

now I can see that I'd be within the compressor range of a VF22 running it at 27psi between 3000 and 4000rpms..

imagine the torque!

Unfortunately, I have no clue as to what a turbine map looks like. Is the VF22 turbine able to even spin the wheel fast enough?

If you zoom in, some of the maps have readable wheel rpm speeds.

One thing I noticed is that the VF22/23/24 IHI turbos don't favor high engine rpms on a 2.5 liter engine. These IHI turbos look optimized for 2 liter engines with free flowing exhaust.

You're attacking this from the wrong direction as far as CFM is concerned.

You should be able to calculate volumetric flow (CFM) directly using mass flow (MAF), ambient temperature (IAT) and ambient pressure (barometric pressure).

Pressure ratio, between compressor inlet and compressor outlet, is more difficult. This is because the pressure at the compressor inlet is slightly below ambient pressure, and pressure at the compressor outlet is quite a bit higher than manifold pressure.

I recall that Cobb made some compressor inlet pressure measurements in conjunction with its so called white paper slamming the SPT intake.

It should be possible to hook a boost gauge off the nipple of the compressor and see how much higher the pressure is compared to the manifold.

I was thinking about using ride5000 MRP method. Is this more accurate? Your suggestion does sound easier.



I was thinking about putting estimated pressure drops into the PR calculation. Estimated pressure drop for intake and estimated pressure drop between compressor outlet and manifold. I think that will be able to take care of this.

I actually did measure the turbine outlet pressure vs the manifold pressure when I ran stock turbo and a MBC. At pressures below 15 psi, its almost identical. Less than a 1 psi drop. Then above 17 psi I started seeing a larger drop and at 21 psi turbine pressure I was seeing only 17 at the intake manifold. Now this is at 6500. At lower rpms, like below 4500, even up to 24 psi, there was a neglegable pressure drop even with the stock intercooler. So I imagine with an upgraded intercooler, and a better turbo, there is little to no pressure drop in the efficiency range because of the very short piping run and intercooler length.

well, we both have legacies with the direct connection, cross flow IC, so those results may not be applicable.

I think that even with a 2 psi drop across the intercooler and a 0.5 psi drop across the intake, the order of magnitude of the error is not significant for the armchair tuning that we are doing. A 5% error in CR and a 5% error in CFM.

It's really just a way to eyeball the differences between compressors and see which one we want. The differences between compressors are significant enough that small errors probably won't change our decisions. I think the bigger issue is that we don't know what a turbine map looks like.

Here's an example of how I used it. Some people have taken a VF22 and put an 18G wheel into it and yet they seem to get better spool than a VF22 on an STi even though the 18G wheel is larger and probably heavier. At 100,000 rpm, the maximum PR for the VF22 is about 1.67. An 18G wheel in a TD05 housing will have a maximum PR of about 2.1 at 100,000 rpm with a slightly better efficiency. This might explain how they are able to get more boost at a lower engine rpm with the 18G wheel.

Kyle "BlackHole" calculated volumetric efficiences using his datalogs and the method outlined here:

http://www.installuniversity.com/instal ... 012000.htm

I believe this is similar to what Jon [in CT] wanted us to do in that it is based upon datalogs (MAF, IAT, and ambient pressure are used to calculate CFM). From those values, the VE at each rpm can be calculated.

I updated the excel spreadsheet with the values for a stock LGT. If you have a different engine or headers, you can calculate it using the same method.

So, the only thing missing now is pre and post turbo pressure drops.

Great work, thanks!

Justin

incidentally that's the foundation of the method i used, with boost pressure factored into the density calculation.

Don't even think of running that high a boost on a VF22. Their compressor wheels are made of cheese and will throw the blades quickly! A TD05-18G or even 20G is MUCH more reliable in my opinion.

Cheers,

Jasper.

I grenaded a VF39 at 27 psi @ 3000 on my 2.0.... hope the 22 is tougher

Has anyone else noticed that the IHI compressor wheels have to spin at a much higher RPM than the 16G, 18G or 20G wheels to get the same pressure ratio?

Maybe that's why they blow up at high PR? centrifugal forces would be much higher.

Seems like an 18G wheel on a VF22 would be an awesome turbo (similar to PE1820?).