05-29-2009, 08:19 PM
Join Date: Mar 2001
Location: Austin, DFW, Portland, SOCAL
Cosworth Manifold/ GT30R Testing...
As most of you know the Cosworth Intake Manifold has been out for a month or so. While it has been part of a larger build in a number of cases, I have yet to see a single side by side direct comparison to the stock manifold.
To that end, we decided to perform such a test. For a platform we chose my 2008 STI:
Pre testing it had the following mods:
Perrin GT3076R turbokit with .63 A/R turbine
Perrin FMIC and Intake
Deastchworks DW1000cc Injectors (NEW*)
Cobb Accessport Version 2
Perrin Turboback Exhaust with single cat
Shortblock with Weisco Pistons, Pauter Rods, ACL Bearings
To make the test as accurate as possible, the car was ran on our Dyno at two boost levels with the stock manifold. The manifold was removed (on the dyno) and replaced with the Cosworth manifold. The car was then ran again with the exact same ECU and boost settings. This all occurred in about a 4 hour period.
Before we get to the results, here are a few comments about the install:
Construction and fitment:
The Cosworth manifold is a direct bolt in. It was VERY EASY to install. Fitment was perfect, and after the install the engine by looked much better as the routing of vacuum fittings is cleaner with the Cosworth manifold. If you were paying attention and had the right tools at hand, you could swap to it in 1 hour. I left the TGV housings on the engine and just replaced the upper manifold.
I was running two boost levels prior to the swap: 20psi and 23psi The tune represents what I would call 'agressive'. AVCS has been fine tuned (extensivly), and timing is as much as I could get on this fuel. I netted a bit more stable power at the richer 11.2-11.4 AFR.
First the lower boost:
The dotted line represents the Cosworth manifold, while the solid line. Spoolup was a bit slower, but I would not read anything into that. The temperature of the exhaust manifold makes a huge difference in spool up, and it wasn’t possible to have the exact same exhaust temps each time. (although I tried).
However, midrange torque is lower with the new manifold. As you can see from the boost trace, boost was also a bit lower with the Cosworth manifold. I did several runs, and I suspect the upper limit of the wastegate duty was not enough to reach the same boost level with the newer manifold. As the car approached redline the power output was almost equal even with a bit more then 1psi less boost in the Cosworth setup. Based on this, it would seem power is to be gained only at higher boost and power levels, and perhaps at higher RPM levels as well.
Second the higher boost:
At the higher boost levels I had enough duty cycle to get to the target boost in both tests. Spool up differences are hard to measure, but it did seem like the Cosworth spoolup was a tad slower. Above 6500 RPM power was higher with the Cosworth manifold, and this gain was sustained across multiple runs. Peak HP with the stock manifold was 413whp, and with the Cosworth manifold it was 426, a gain of 13whp.
While extensive further tuning may have netted an even greater gain, this test was focused on the bolt on effect. Since I really didn’t push the engine that hard before the swap, I don’t have a good base to compare. However seat of the pants tuning revealed the car was a bit easier to tune at the higher RPM/ Boost levels with the Cosworth manifold. I did turn up the boost just a bit and leaned it out some, and managed a run at 438whp.
It is possible results will be different for a much larger turbo, or even for the .82 A/R GT3076. We will be continuing this testing on a 35R equipped car, so I’ll have more results in a few weeks. The summary from my testing so far is that the stock manifold is not the significant limiting factor at the 400whp level, however there are some gains to be had at higher power levels. A back to back test on Tim's car with teh 35R will tell a lot more.
Last edited by Cobb Tuning; 05-29-2009 at 08:26 PM.