View Single Post
djdawson2 djdawson2 is offline
Registered User
Join Date: May 2009
Location: Evergreen, CO
Posts: 231
Send a message via Skype™ to djdawson2
Originally Posted by copbait73 View Post
Garrett have excellent turbines, BW/Schwitzer turbos have excellent turbines, BW/KKK are a step down.

This part I agree with. Many of my fellow Audi friends tried to use Garrett/KKK hybrids, simply for ease of installation. They would match up a KKK turbine housing with a GT30R, or a more typical 50, 54, or 57 trim journal bearing Garrett T04E compressor. The result were less than flattering compared to those people using full Garrett units.

Back to the compressor map topic:
I think the compressor map is a very good place to start when trying to determine the correct choice for your application. I think compressor surge has become more of a problem since the use of ball bearing turbos, as well as running higher boost. The BB turbo's ability for rapid spool-up has dramatically increased the probability for encountering compressor surge, especially in high boost applications. Why? Well, start plotting your engine's flow rate at various engine speeds on a compressor map... it becomes clear.

I've tested and dyno'd a lot of Garrett turbos, both GT series BB turbos, and more typical journal bearing turbos. I've tested turbos of both varieties with similar compressor maps. The difference is that the BB turbo spools up very fast... almost too fast for the size of the compressor. The result is too high of a PR at too low of an engine flow (lbs per minute). When this happens, we end up at a point left and above the surge line on the map... and this spells trouble. A surging compressor wil very quickly trash the bearing section of the turbo, especially if it is a BB.

Bottom line, I do believe you need to be more conservative when selecting a BB turbo. You do need to factor in a higher potential PR at lower engine flow. I think the presence of anti-surge housings on the GT BB turbos simply tells us that Garrett knew this beforehand. I have never seen anti-surge tactics employed on their journal bearing turbos... nor have I ever experienced surge with them.

Personally, if I had the time to do extensive experimenting with Garrett turbos on a CIS 930 with a 3.3, I would probably begin with the journal bearing GT37, based on the data available. This turbo, equipped with an 82mm compressor wheel, is capable of moving about 55 lbs/min efficiently. The turbine section is more appropriately sized (vs. a 3076) with a 72.5mm wheel and a housing A/R of 1.12. All of this seems ideal for a 3.3 turbo running 1 bar or less of boost.
Old 02-02-2010, 09:18 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #19 (permalink)