Got me a SVX motor......

[93forestpearl]

So are you not sayin, but sayin the 35R doesn't necessarily like higher pressure raitos, as in 20-25psi?

When I build some EJ25D heads that was going to be my next move.

[BAC5.2]

Um, I can't think of any turbo that would operate at a pressure ratio of 20 to 25....

But since you mean 20ish psi, that's around a 2.4:1 PR. That is just on the narrowing end of the fat part of the compressor map.

People running 3:1 pressure ratios are only seeing a viable window of a few hundred RPM of usable boost. There are other turbos that would outflow the 35R for those requirements. It pushes the turbo to the low side of efficiency when you get to levels that high.

Really, a PR of around 2.25:1 or 2.20:1 puts the 35R right in the efficiency range. But a smaller motor lacks the ability to flow the air that turbo would produce at those pressure ratios. So it's really suited for lowish boost on larger motors. Even in full-race setup, you are leaving a LOT on the low-end table by using the 35R.

[93forestpearl]

I don't have my heart set on any certain pressure ratio, as that is only one axis on a 3D map. Its all about airflow, and I know better than that, I just haven't looked at any compressor maps yet since I'm not really in the market as of now.


I'll just be looking for something in that range, the 6-700hp capable turbos. I haven't decided if I want to ditch the boxer rubles for a twinscroll setup. If I did, I wouldn't feel bad about a 4088R.


My buddy is trying to push me towards a Borg-Warner turbo. He's got a S400, but that's a little different ballgame. I'm not trying to make 800whp anytime soon.


If you have other suggestions fofr something in the 35R's airflow range that would be better suited to a 2.2L I'm all ears.

[BAC5.2]

Identifying a target pressure ratio gives you the information you need to properly read a compressor map. If you have a target pressure ratio, you can use that along with volumetric efficiency and a few other environmental variables and figure out the mass-flow-rate of the engine at a particular pressure ratio.

Then, plot that against the compressor map. This will, in essence, show you where to expect to hit full boost, and where it will taper off too. For example, a 2.35L at my altitude, at a 2.39:1 corrected pressure ratio (22psi corrected for intercooler efficiency and such) with a .95 VE value (the motor will have Ej20 heads with cams), it will flow a range of air mass.

Plotting my numbers on a compressor map for a GT3582 and you'll see that the turbo cannot push a 2.39:1 pressure ratio below 30lb/min. So you match that to the chart you generate with the mass flow rates of your setup, and with mine you'd see that it won't be capable of flowing 30 lb/min until 4700RPM. So, you could expect to hit full boost at 4700-ish RPM. You can also see that, at 2.39:1 the 35R will push in excess of 60 lb/min. But again, referencing the table from before you can see that, at a HARD redline of 8000 RPM, the motor is only capable of pumping 52.1 lb/min of air. So you can't actually use what the turbo is capable of. On top of that, you can see that you don't even enter the most efficient PR/Mass Flow range of the turbo.

But, compare this to a GT3782. This turbo is capable of a 2.39:1 PR, flowing as little as 20 lb/min. Back to the chart, 20 lb/min hits at 3400 RPM. It will hold a 2.39:1 pressure ratio until 55 lb/min, just past our 52.1 lb/min at 8000 RPM. If you look at the line, you'll also see it cross directly through the MOST efficient part of the compressors operating range.

So you can see that with the 3782 it'll make peak boost below 4000 RPM and it will hold peak boost until 8000 RPM. You can plot the flow rate in the centermost point of the highest efficiency island and you will be able to guesstimate your power curves. With that motor on the 3782 you are in the biggest efficiency island between 5300 and 6600 RPM, so you can expect to make peak HP in that range. You'll make peak torque around 4000 or so, I would think.

SO, that said, I would recommend the 3782. It looks like a sick turbo and I bet you could easily crack 500whp with it.

[93forestpearl]

Thanks Phil, I don't complain when people do my homework for me


I would have gotten there eventually, but I was in no hurry.

[93forestpearl]

I can't seem to find a water cooled variant of the GT3782, let alone a ball bearing version. Maybe I just don't know where to look.

[BAC5.2]

Nope, the 3782 is not a ball-bearing turbo and I can't say that I've seen a water cooled one.

I wouldn't worry about it. Oil cooling only is perfectly sufficient, just let the car idle down for a little while before shutting it off. Installing an oil-cooler might be helpful though.

Meh, I didn't do any homework FOR you. I gave you the information from a 2.35L build. Those numbers won't relate necessarily to what your doing. You'll have to figure out the calculations on your own

Either way, I just made an excel file with the calculations on one sheet and then I've just been plotting it all on every compressor map I can find. I spent about an hour working on it last Saturday and logged some 10 turbos. I found out that the 20G sucks, and that the holset HX40 is a big bitch, but the 8-bladed variant would be really nice for something displacing a lot of cubic inches. In fact, if you have a, say, 5-liter, and run say, 6 pounds of boost, you could potentially see 78% compressor efficiency!

Holset's are hard to find compressor maps for though, so I'm still looking for the HX/Y 30 and 35 maps.

[93forestpearl]

This reminds me, my buddy got a GT3782 out of an '06 Powerstroke with the variable vane turbine section. He's working with the firmware guy for his MAF Translator Pro to have a table to control the vanes.

[BXSS]

I want to know how your buddy's powerstroke turbo works out (EJ20K-?) - the turbo outlet looks bigger than the t-body!

[93forestpearl]

Its a Wiseco/Eagle 22T shortblock with EJ25D heads. I'll let everyone know how it does when all the bugs are worked out.