Everyone says that exhaust flow increases for a given a/r as flange, or frame size increases. It does seem feasible given it's just an area/radius ratio, but i've been researching using 2 T3 .63 GT3582's on a 400+ na hp build, and was looking at Garrett's turbine flow numbers, and they don't separate by flange. The 3582 turbine options are T2 in .63, both T3, and T4 in .63, .82, 1.06, and v bands in .61, .83, 1.01, and here's the map. Anyone have any real world results on this, or any other turbos when changing flange size? Going by the map, it appears that flow is consistent across the flanges. #ad

 

Its not across flanges. It cant be. As you mentioned, ar is just a function of area and radius not the actual size of the hole. It one of the reasons that holsets way (and company's the measure with a similar system ie. mitsu )of measuring is in my opinion superior and more indicative of actual turbine flow.

 

Yeah, im still looking for some real world info, but have sold the 3582R's and decided to go with a single.

 

My "real world" experience for this is mostly with import cars running 4 and 6 cylinder engines. A gt30 with a big ar t2 flange makes about 100 HP less then a gt30 in a t3 flange.
Alot of the SR20 guys want to run a factory t2 flange manifold so there has been a stack of larger compressors hacked into really small t2 hot sides. My last sr made 300 even on a t2 flanged 30. The exact same turbocharger made just shy of 400 once I went to a t3 .82 housing like what should have been on it in the first place. Boost was a pound or two higher on the t3 housing but the plugs looked the same color wise and the strap mark for a given boost was significantly "colder" hence the reason to lean on it a bit more.

 

Is it the flange or the scroll making the difference? Ive run t6 flanged turbos on some very mild small cid engines, and everyone said it would be laggy once they heard t6, but in reality there was very little if any difference compared to the same turbo in a t4. I would have thought the turbine maps would be a pretty good indicator.

 

Didn't Holdener do a test between a T4 and T6 and find no difference between the two on a basic LS?

I can see the smaller flanges making a difference between one and the other though.