3.0 twin turbo setup - turbo selection

[jpnovak]

Hi All,

I need some help deciding on a turbo to choose for a 3.0 twin turbo. The concept is a stock 78SC block running nominal 8.5:1, SC cams, EFI and twin turbo. why twin? Because a single is too easy and twins just look so cool in parallel hanging out the back.

I Calculated the output for lb/min flow rate vs pressure ratio so that I could overlay onto compressor maps.

I have been looking at Garret. the model assumes a 1.5L engine with 1 bar boost by 3000 rpm. The final point on the map is a 7500 rpm redline. This is too high for these cams but possible for the engine.

The GT17 runs out of boost way too early maxing out flow rate before 5k rpm
The GT20 can fill the air but looses efficiency and puts the overlap at the right side choke line by about 6K rpm.



the GT22 looks to be the most efficient (left) crossing the center of the island. I just don't like how there is seemingly no pressure below 3500

The GT25 stays left of the efficiency island telling me that it will likely have too much lag as the lines all plot to the left of the fficiency map until I get past 5K rpm.



Any thoughts?


Has anyone built a 3.0TT?

What turbos did you use?

[jpnovak]

I like the Borg Warner Maps for this build.

This is the EFR 6258



This is the S200SX-E.



These are a better map than what I plotted for the Garret series.

[mikedsilva]

how do you mount the turbos... off the back of a set of ssis?

[jpnovak]

Mike, that is all TBD. But yes, in concept they will be offset to sides straight off headers. Planning center dump zork with WG on outside.

[dannobee]

Maybe Juan will chime in too, but I thought Eddie Bello mounted the turbos toward the side and rear at an angle, negating relocation of the oil tank.

[juan ruiz]

I dont have any experience with 3.0.turbo set ups but here you can see how Bello had his, I hope thar helps.

[jpnovak]

Thank you Juan. I had forgotten EBello was running twins. I like the packaging of the turbos offset to the side.

I was thinking to have the turbos mounted more central like in this thread. However, I would have them spaced a little wider for clearance.

NEW UPDATED PICS : Twin Turbo / EFI Conversion



Here in TX we have have a lot of dust and often water crossings. I would pull the inlet air tubes up into the engine bay. They would not survive hanging down below the bumper for the type of usage this car will get.

[juan ruiz]

Bisimoto mounted his as you are suggesting

[jpnovak]

Yes. I am familiar with that car as well.

[Speedy Squirrel]

I usually use Garrett G25-550 turbochargers.

Supercore PN: 871388-5002S (normal rotation)
Free floating undivided turbine housing 0.72 a/r (normal rotation) PN: 740902-0069
Compressor mounted bypass valve (electric) PN: 7.01830.13.0 Pierburg

Supercore PN: 871388-5001S (reverse rotation)
Free floating undivided turbine housing 0.72 a/r (reverse rotation) PN: 740902-0073
Compressor mounted bypass valve (electric) PN: 7.01830.13.0 Pierburg

[RarlyL8]

We've done a few over the years, I have headers for the rear mount application in stock. The builds we focus on are typically around 400whp. For this we used modified K16 turbos. The spool very quick deliver the required power and pull hard through redline. Using OE turbo housings allows for the use of OE oil plumbing which is nice.

[jpnovak]

The GT25 are interesting. I thought they might be slightly too big and the GT22 might be abetter match. For a 3.3 or 3.4 the GT25 would be perfect.

I need to look closer at the K16s. Brian and I have a dialog offline.

[jonesb930]

Jamie,

I have a set of Brians headers using K16's (from a 996) on a 3.2L Short stroke motor (3.0 with bigger cyl). I had some big challenges with the oil plumbing as the 996 K16's are a bit different than the ones from the 993 as they have different flanges. I could not use the stock oil catch tanks as Brian is showing (I assume those are 993 K16's). Here is a pic with some mocked up exhaust piping from the rear. I intend to use vband clamps so that I can change the exhaust from a GT3 style strait out the back to something with a small muffler if I need to later.
Don't mind the basketballs lol.

[Speedy Squirrel]

I was referring to Garrett G25, not GT25 (not their finest marketing moment). Here is the compressor map. 80% is about as good as it gets. The rotating assembly is almost exactly the same size as a K16.

The K16 has a lot of pluses. It is still a journal bearing turbo, so basically oil cooled. The latest turbos have ball bearings and are really intended to be water cooled, though people seem to be getting away without it.

Newer turbos have stainless steel housings that can be welded to stainless exhaust tubing. That can come in handy.

EFR turbos are the best out there, and I use them exclusively on single turbo builds. Titanium Aluminide turbine wheel EFR turbos can put a few twin turbo’s to shame.

The G25 though comes in mirror image designs which makes designing the exhaust a little easier and just looks fantastic. I think the human mind craves symmetry.

I like the Eddy Bello layout. It looks like the oil tank has gone elsewhere to make that happen. Nothing wrong with that.

[jpnovak]

Great into. Thanks. I need to do some thinking about this.

I agree that I think the EFR series would be a great match. Especially in a single. But as you said, The mind craves symmetry.

[JakobM]

Compressor efficiency is a bit of muddy water marketing tool these days IMO

Small compressor wheels in large compressor housings (large volute volume) give high efficiency numbers by nature -> you only want high efficiency and large volute volume to harvest last bit of peak power from a certain wheel diameter at choke rpm and you pay no or less attention to boost-build-up. E.g. +70's% efficiency is engineered for quick boost build-up and typical for street and small diesel applications where +80's% efficiency is typical peak power related or more static industrial operation. If a spinning shaft with a compressor wheel is put in a small room without a compressor housing, you get closer to 100% efficiency out of the door... (rough example)

Journal bearing turbos flow a lot of oil and act as a small heater for the oil sump as the turbo core is more heated than the engine. Having two journal bearing turbos will heat up oil a bit more than a pair of ball bearing turbos with the oil mist flow

Running ball bearing turbos without water is best with mid- to large sized turbos as the heat inside bearings is lower from the significantly lower shaft(s) rpm

[jpnovak]

Jakob, thank you for continuing the discussion.

I agree that being toward the bottom of the efficiency curve, meaning lower turbo tip speed, is a good thing. And no, I do not want to wring out every bit of power. But I do want to make sure that total airflow for engine speed is at least on the map.

Most important factor to me is smooth boost delivery. I do not want on/off boost making the car difficult to handle. This car will spend a lot of time on very tight roads and decent amount of track. I prefer to handle on the edge rather than point-and-shoot speed.

Great thought process about the oil feed and cooling. I think my planned cooling system will be up to the task but this is a concern. I also have concerns (and a possible solution) about the heat buildup under the bumper. It will for sure need to be vented.

[tx996tt]

JP,
An area to consider is Turbine A/R and the powerband impact against desired driving style. Following are the turbine A/Rs:

K16 (996T variant) 0.80
G25/550 0.72
GT2560R (aka GT28R) 0.64

Garrett posted an article detailing turbine A/R impact on a high HP Civic:

https://www.garrettmotion.com/news/newsroom/article/turbo-tech-how-to-turbo-and-the-difference-an-a-r-makes-for-optimizing-your-turbo-system/

Here is a data point using stock K16s (996TT variant):

3.0L SC, large port (38mm) heads, 8.5:1 JE custom turbo pistons, 964 cams, Bitz racing style fabricated aluminum intake, 55lb injectors, Ford Racing 70mm TB, Haltech standalone, GM LS1 coils, water meth injected, short bell housing 930:
Dynojet 415 whp / 395 wtq at 1.1 bar

The K16 and GT2560R are internally WG reducing fab work if using SSI or RSR style headers

For spirited street / canyon driving the GT2560Rs with the small turbine A/R look promising. Feedback welcome as I am also considering swapping to 2560s...

[2mAn]

I have a similar planned project that has stalled out... I have some nice modified 930 manifolds and promotive modified 993 turbos.

I will hang out here to absorb some knowledge, looks like this is on the right track

[jpnovak]

Just a quick update. I ended up with matched pair (symmetrical) G25s. The car is still being built. It will be later this spring before we attempt to get all of this going.