Choosing Turbos
#1
Instructor
Thread Starter
Choosing Turbos
Time to make a decision on turbos for my 996 race engine. I've been aware of turbocharged engines and how to select turbos since the early 80's. My car (Cayman) was a 4.0L M97 and initially we used Turbonetics turbos as our goal for the engine was rather modest to slot into the GT2 class ~430hp at the wheels. When we upped the boost the results were not good so it was time to do some serious engineering and upgrade the turbos.
Most people think that a lot of boost pressure is the way to big power numbers but actually it's not boost but airflow through the engine that makes power. All 4 valve Porsche engines flow very well out of the box so it's a big advantage to start with. Anything that can be done to increase airflow by limiting restriction makes the system more efficient and powerful. So we designed the physical layout and ducting to keep losses very low.
Choosing a turbo is much easier now if the manufacturer supplies compressor maps and turbine flow numbers. On line tools like Garrett boost advisor and BorgWarner MatchBot allow an excellent analysis to be made. When I ran the analysis for my 4.0L M97 boost advisor recommended a very small turbo G25-550. I ran a more detailed analysis on MatchBot it confirmed the choice. My race shop wanted to go with much bigger turbos but I persuaded them to use the G25's and the results where outstanding. Throttle response was amazing and we made 710 WHP at 1.3 BAR. Unfortunately the engine started to lift the heads and leak combustion gasses into the water jacket.
So now we have a strong 996 Mezger TT built for the car and the design process starts again. I've heard good things about Xona Rotation turbos and my race shop has a good relationship with Evospec an engine builder that is a vendor for Xona. They build a lot of very high output Mezgers for straight line racing (1200 WHP!) They recommended XR6564 or XR6157, but there are no maps available for these turbos. He did not have any dyno charts for engines using these either. Doing my due diligence with units having complete engineering data Garrett G30-660's look like the best choice. The compressor flow is about the same as the XONA 6157 but with smaller wheels on both sides. Smaller wheels mean faster spool.
So my choice is Garrett G30-660 with a .83 AR hot side. A 996 Cup in the shop with a 3.8 L 996 TT is getting the larger XONA 6564 and it will be interesting to see how the two engines perform on the same dyno.
This dyno sheet is from my M97 4.0L with the tiny G25-550 Garrett turbos. We are shooting for 800 WHP with the 996TT
Most people think that a lot of boost pressure is the way to big power numbers but actually it's not boost but airflow through the engine that makes power. All 4 valve Porsche engines flow very well out of the box so it's a big advantage to start with. Anything that can be done to increase airflow by limiting restriction makes the system more efficient and powerful. So we designed the physical layout and ducting to keep losses very low.
Choosing a turbo is much easier now if the manufacturer supplies compressor maps and turbine flow numbers. On line tools like Garrett boost advisor and BorgWarner MatchBot allow an excellent analysis to be made. When I ran the analysis for my 4.0L M97 boost advisor recommended a very small turbo G25-550. I ran a more detailed analysis on MatchBot it confirmed the choice. My race shop wanted to go with much bigger turbos but I persuaded them to use the G25's and the results where outstanding. Throttle response was amazing and we made 710 WHP at 1.3 BAR. Unfortunately the engine started to lift the heads and leak combustion gasses into the water jacket.
So now we have a strong 996 Mezger TT built for the car and the design process starts again. I've heard good things about Xona Rotation turbos and my race shop has a good relationship with Evospec an engine builder that is a vendor for Xona. They build a lot of very high output Mezgers for straight line racing (1200 WHP!) They recommended XR6564 or XR6157, but there are no maps available for these turbos. He did not have any dyno charts for engines using these either. Doing my due diligence with units having complete engineering data Garrett G30-660's look like the best choice. The compressor flow is about the same as the XONA 6157 but with smaller wheels on both sides. Smaller wheels mean faster spool.
So my choice is Garrett G30-660 with a .83 AR hot side. A 996 Cup in the shop with a 3.8 L 996 TT is getting the larger XONA 6564 and it will be interesting to see how the two engines perform on the same dyno.
This dyno sheet is from my M97 4.0L with the tiny G25-550 Garrett turbos. We are shooting for 800 WHP with the 996TT
#3
Smaller wheels mean faster spool.
John has these same turbos on his 4.0l and is well on his way to 800+ wheel.
Specs on mine are 51.5/73 x 68/62mm . John's might be a bit different as Tial previously used 57/73 x 60/56mm prior to 01/2019.
#4
RL Community Team
Rennlist Member
Rennlist Member
#5
It is fairly common on 996/997 turbos. I've never seen a great explanation to it beyond intake harmonics.
I'm waiting to see if it goes away on John's car when he dyno's with his new Y pipe from speedtech.
This thread cover it some
https://www.6speedonline.com/forums/...chart-q-2.html
I'm waiting to see if it goes away on John's car when he dyno's with his new Y pipe from speedtech.
This thread cover it some
https://www.6speedonline.com/forums/...chart-q-2.html
#6
RL Community Team
Rennlist Member
Rennlist Member
I saw the same dip on one of John's dyno plots too - thanks for the link, I too will be interested to see if it goes away and/or why it occurs in the first place.
#7
Instructor
Thread Starter
Thanks for posting the Xona dyno charts. Looks like a pretty good solution for your cars.
A little more on my design methodology: After calculating the mass flow I need to meet the power level goals, I look at compressor maps to find the most efficient unit at the mass flow and pressure ratio I'm running. This minimizes the heat of compression so the air entering the intercooler is colder and also reduces the power required to drive the compressor from the turbine. Then I look at the turbine mass flow and pick an AR that will match the exhaust flow requirements. Sometimes the AR is a bit of a SWAG. In this case an AR of .83 looks to be spot on to about 1.4 BAR, above that pressure a larger AR would be better. Typically a smaller AR spools more quickly, HP ramps more gently and back pressure limits top end power. A larger AR spools later, HP ramps quicker and top end power is greater.
There are always assumptions around pressure losses in the external piping and volumetric efficiency of the engine so dyno data and pressure and temperature data from the engine will tell the tale if I picked the compressor wisely and my SWAG on the AR is good.
A little more on my design methodology: After calculating the mass flow I need to meet the power level goals, I look at compressor maps to find the most efficient unit at the mass flow and pressure ratio I'm running. This minimizes the heat of compression so the air entering the intercooler is colder and also reduces the power required to drive the compressor from the turbine. Then I look at the turbine mass flow and pick an AR that will match the exhaust flow requirements. Sometimes the AR is a bit of a SWAG. In this case an AR of .83 looks to be spot on to about 1.4 BAR, above that pressure a larger AR would be better. Typically a smaller AR spools more quickly, HP ramps more gently and back pressure limits top end power. A larger AR spools later, HP ramps quicker and top end power is greater.
There are always assumptions around pressure losses in the external piping and volumetric efficiency of the engine so dyno data and pressure and temperature data from the engine will tell the tale if I picked the compressor wisely and my SWAG on the AR is good.
Trending Topics
#8
Since you have a motec, have you thought about adding turbine speed sensor along with exhaust back pressure sensors to get the full picture?
I get your methodology but its hard to compare apples to apples without a compressor map for the xonas.
I get your methodology but its hard to compare apples to apples without a compressor map for the xonas.
#9
Instructor
Thread Starter
We have turbine speed sensor, exhaust backpressure, pre and post intercooler temps on my car. We will have solid data to work with. I was willing to try the Xonas but no maps = no confidence in the results
#10
Agreed, I had reached out to them for a map as well pre-purchase. Given their background and how long they've been around I'm surprised they haven't taken the time to document this stuff.
Have you flow modeled your intercooler? I gotta say that piping looks pretty small and the end tanks are pretty anemic for a 3.8l with a single intercooler. Especially if you're going to be running in the southwest with high track temps.
Have you flow modeled your intercooler? I gotta say that piping looks pretty small and the end tanks are pretty anemic for a 3.8l with a single intercooler. Especially if you're going to be running in the southwest with high track temps.
#11
Instructor
Thread Starter
Agreed, I had reached out to them for a map as well pre-purchase. Given their background and how long they've been around I'm surprised they haven't taken the time to document this stuff.
Have you flow modeled your intercooler? I gotta say that piping looks pretty small and the end tanks are pretty anemic for a 3.8l with a single intercooler. Especially if you're going to be running in the southwest with high track temps.
Have you flow modeled your intercooler? I gotta say that piping looks pretty small and the end tanks are pretty anemic for a 3.8l with a single intercooler. Especially if you're going to be running in the southwest with high track temps.
Last edited by diverdog; 07-31-2020 at 04:35 PM.
#12
I'm guessing the 0.72 A/R on the G25-550s? That's a really well matched turbo. I'd guess the G30-660 in the 0.83 will spool roughly 400-500rpm later. A couple hundred rpm for the larger wheels and another couple hundred for the larger A/R. If you need a bit more top-end, there's always the G30-770 which I'd guess would add another 150-200rpm of lag. I think you'll hit your 800whp at about 1.65bar at about 50lbs/min which puts you pretty close to peak compressor efficiency with the G30-660. The Xona 6157 is closest flow-wise to the G30-660, but it needs a 76mm exducer compressor in order to match the flow of the 67mm exducer Garrett. Xona uses a 64mm inducer turbine vs 60mm for the G30. The G30-660 will destroy the Xona 6157 in transient response and it will spool up faster too. The Garrett G-series is very impressive and I've seen nothing to make me think the Xonas can compete. The G-series is essentially a publicly available Garrett Motorsports turbo. There's a reason Garrett is the mostly widely used turbo in professional motorsports and they are the largest OEM turbo supplier in the world. The shop should do their test with the Xona 6564, do some track driving, and then swap to the G30-770. They'll make the same power and pick up a ton of response and spool-up. 71mm comp vs 79mm, 60mm turbine vs. 72mm. Or, go with G35-900 which has 76mm comp and 68mm turbine vs the Xona 6564 79mm comp and 72mm turbine. They'll have a ton more power capability and have similar spool.
The following users liked this post:
moburki (08-06-2020)
#13
Instructor
Thread Starter
Pretty good analysis. The G25's are .92 AR. G30's with a .83 have a similar exhaust flow. 51 lb/min mass flow is good for about 900 wheel HP but at a higher boost that will require a bigger AR turbine housing to limit exhaust back pressure. My vendor has built many 3.8 TT Mezger's but none with G series turbos and the Protomotive GT3 manifold and cam combination I'm running. I want to start out with a very manageable boost curve and more modest HP until I get used to driving the car again. The waste gates will have a 1 BAR spring and be adjustable in car to 1.5 BAR or so depending on what exhaust back pressure looks like.
Running at lower boost even the tiny G25-550's would work well. I think 900 WHP is the practical limit for the G30-660. Above that G30-770 would be a better choice
Running at lower boost even the tiny G25-550's would work well. I think 900 WHP is the practical limit for the G30-660. Above that G30-770 would be a better choice
Last edited by diverdog; 08-01-2020 at 03:27 PM.
#14
Ah, interesting. I think the G30-660s in the 0.83 should only spool ~200rpm later then. What's your target for engine dP? Are you trying to keep it positive (intake mani pressure - exh mani pressure) the whole way or allow a negative dP of say 6psi at redline?
#15
Instructor
Thread Starter
Positive all the way to redline. I want the engine to run cool as possible.