Waterguy

Edman, I think you are getting your units wrong. The K26 maxes out at 0.24 m^3/s, which is about 35 lb/min, not 24 lb/min. The T3-50 maxes out at about 27 lb/min. The K26 compressor may be physically the size of a T3 unit (I have not checked this) but it flows a lot more air than any T3 map I have seen. Its smaller size compared to an (approximately) equal flow T04E-40 probably explains the need for higher turbo rpm to achieve boost.

I couldn't find any T04E compressor that equaled the 49 lb/min (0.34 m^3/s) airflow of the K27.2

BoostGuy951

If you plot out a 951 engine at 4500 and 6500 Rpms on the K26 Map, you will see that a 14.7 PSI, the 4500 RPM Point is right in the sweet spot of the map, where you see the letter A. At 4500 RPM, the K26 is running at Maximum efficieny. Continue to run the engine to 6500 RPM, and the plotted point will now be almost off the map on the right side. It ends up in the 50% efficiency range, which is terrible. All of this should be evident to anyone who has ever driven a k26 944 turbo. You get loads of torque at 4500 and then power just falls off.

If you had to compare the k26 to a Garret Turbo, the closest match is the T3-50 trim. It is certainly not the T04E-40 trim. That turbo is around 70% efficient at 6500 RPM, where the k26 was at 50% and almost off the map. If you plot even the Baby T3-60 trim, at 33.5 lb/min (My figure for a 951 @ 6500RPM @ 1 bar) you are at the edge of the map, but you still are around 60% efficient. If you plot the T3-50 trim, you point will be off the map, but if you continue the efficiency island outward, you will see that the point is probably right where the 50% efficiency Island would be, if Garrett had drawn the map out that far.


You can't just pick the farthest right point on the map and say, "this is what the turbo will flow". You have to pick a flow number, and a pressure ratio, if you want to compare turbos. That KKK Map extends to 50% efficiency. The Garrett Maps tend to only show you out to 65% or 60% efficiency. Which you really shouldn't go lower than anyway, lest you get the high RPM 944 effect.

Waterguy

Sorry, boostguy, gotta disagree with you on all points.

No, look closely. The outer line is 60% efficiency.

In our dreams maybe. There is no way a stock 951 has 80% volumetric efficiency at 6500 rpm. I use 71% VE at 6500 rpm for a flow of 30 lb/min at 15 psi manifold. Add in about 4 psi pressure drop across the intercooler at that rate and you have a pressure ratio of 2.4 The K26 map shows 74% efficiency at PR=2.4, 30 lb/min (0.205 m^/3) flow rate. The T3-60 is about 67 to 68% efficient at that point. The K26 flows more than a T3.

I think that, with the right wastegate, the K26 should hold 15 psi to the red line on a fully stock 951. However, I agree that at higher boost or with any breathing efficiency improvements like a testpipe, 3" exhaust, etc., it probably wouldn't flow enough to hold 15 psi to the HP peak (which is about 6000 rpm, not 6500 rpm btw.)

BoostGuy951

This is correct, It looked like a 5 at first, but I see now that is was a 6.


I had a modified 951 in mind when I picked 85% VE. But lets just forget about the 951 altogether.

Lets compare turbos by themselves.

Lets pick a desired PR, how about 2 (14.7 psi).
Lets pick an efficiency range that we don't want to drop below, how about 70%.

Now lets plot a PR of 2 on the map, and make a point where that line intersects our efficiency Island line of 70%.

Now just look at how much air the turbo will flow. We are getting a measurment of how much air each turbo will flow at a given PR and Efficiency.

The K26 will flow .212 Cubic meters/sec, as near as I can tell.
That comes out to 7.487 Cubic feet/sec.
Which is 449.22 CFM @1 bar and 70%

The T3-60 will flow 27.5 lb/min according to my measurements of the graph.
Assuming that 1lb/min is equal to 14.472 CFM,
The T3-60 flows 397.98 CFM @ 1 bar and 70%

So you are correct, the K26 Does Flow more than the T3-60.

The T04E-40 seems to flow 32.25 lb/min on the graph.
Assuming that 1lb/min is equal to 14.472 CFM,
The T04E-40 flows 466.72 @ 1 bar and 70%

So again, you are Correct, the Closest Match to the K26 is the T04E-40.

My Turbo, the T04E-46 looks like it flows 35 lb/min
Assuming that 1lb/min is equal to 14.472 CFM,
The T04E-46 flows 506.52 @ 1 bar and 70%

Damn, it makes me wish I had gone bigger. That is only an 11% flow increase from a K26.

The T04E-50 flows 41 lb/min
Assuming that 1lb/min is equal to 14.472 CFM,
The T04E-50 flows 593.35 @ 1 bar and 70%

That is a huge step up from even a 46 trim. The 50 trim is a 25% increase in flow.

So what we thought about the bigger hotside must be the major reason we see so much more power up top with the Garretts. Someone in an earlier thread said they measured the Exducer bore on the TEC Hot housing and got a measurement Very close to a KKK #8 Hot housing. Its all coming together now. That explains why a Turbo S makes so much more power but not very much more torque. It has a k26/8. What actually causes most of the high RPM power fall off is the #6 hot housing. I wish I knew this when I selected my Turbo.

Sach951

Wow... great thread ;D Given that I like running in the top end I know what turbo I'm heading for eventually...

Thanks,
Sach

BoostGuy951

I'm gonna have to get the 50 trim Compressor installed on my Garrett.

TonyG

Put on a 60-1HiFi...

That's the compressor! :-)

Sach951

Always like to hear more about your setup Tony... I love what it sounds like you've accomplished

Cheers,
Sach

Laust Pedersen

“Does anybody have the actual size of the K27 compressor wheel? That way we could compare its actual size to the Garrett wheels.”

Edman951

Here are the dimensions on the compressor side of the K27: Inlet ID 2.1”; Outlet ID 1.7”; max OD of wheel: 3.00”.

I know this since I have one in front of me, which is for sale on eBay … hint, hint …
For sale because I have found ways of making plenty of boost on the K26-6, more about that, when I have fuel to match.

Laust

Danno

"The K26 maxes out at 0.24 m^3/s, which is about 35 lb/min, not 24 lb/min. The T3-50 maxes out at about 27 lb/min."

"If you plot out a 951 engine at 4500 and 6500 Rpms on the K26 Map, you will see that a 14.7 PSI, the 4500 RPM Point is right in the sweet spot of the map, where you see the letter A. At 4500 RPM, the K26 is running at Maximum efficieny. Continue to run the engine to 6500 RPM, and the plotted point will now be almost off the map on the right side. It ends up in the 50% efficiency range, which is terrible."

This is assuming a VE-volumetric efficiency of 1.0; that all of the air pumped makes it into the engine. If so, we'd see zero boost on the gauge. Since the air has to be forced through the intercooler and intake plumbing, through the intake-manifold and partially open valves (and some closed ones), the VE will be in the 0.4 to 0.6 range. Multiply your flow numbers by that and you'll see that it's still within the flow-ranges of the K26. The most efficient range in the 951 engine is actually around 3500-4500rpm and after that VE goes down with increases in RPM. That's why we have such a dramatic torque-drop in the upper-RPMs.

FSAEracer03

Ok, I'm new to the KKK turbo thing... can someone explain to me why the Turbo S used the K26 instead of the K27?? I thought the K26 was larger and flowed better?! Now I'm confusing myself here...

Waterguy

My opinion: because it was (is) the correct size for the horsepower objectives of the car.

Think about it: when the Turbo S came out at 250 hp in 1988, it was faster than almost all the stock cars then on the road. Faster than Corvettes, Ferraris, Porsche 911 and (in fact, but not on paper) the flagship Porsche 911 Turbo. Even in Turbo Cup form, the factory never went beyond 305 HP (260 rwhp.) The K26 is right in its efficiency zone at these power levels. At the time Porsche was using a K26 (3LDZ) on the 300 HP, 3.3 L 911 Turbo.

If Porsche had wanted to go higher than 330 hp (280 RWHP), it would have used a K27. But to do it right (from a factory engineering perspective) they would have had to also upgrade the exhaust, fuel injectors, fuel pump, AFM, intercooler, etc. They were constrained by space limitations for the intercooler, and could only fit in an undersized unit that has a 1 psi pressure drop at 205 cfm. The final result would have been a car that would make crazy horsepower by simply adding a manual boost controller! Imagine the warranty claims for head gaskets, CV joints, trannys, bottom ends, etc. Plus, this was a company that had recently dealt with a series of lawsuits from families of people that had killed themselves in the "ridiculously overpowered" 260 HP 911 Turbos (with unforgiving boost and handling characteristics, another story.)

Instead, they built a 250 HP car that would last 150-250k miles and 15+ years and STILL be in good enough shape to modify to make crazy horsepower.
If you want to ask the factory a question, ask them why they used such crappy wastegates!

Edman951

Waterguy
Yes i did.

Laust,
Thanks alot
ok so you are saying that the K27 housing has a Inlet ID 2.1”; Outlet ID 1.7”;
And the wheel has a 3.00'' OD. major
But can you mesure the exact size of the wheel. on the small side witch is the inducer.
Please fill in the blanks
K27= compressor housing with 2.75'' OD/ 2.1'' ID inlet, 2'' OD/ 1.7'' ID outlet, compressor wheel inducer ____"OD, and major 3'' OD.
Thanks alot in advanced.

Laust Pedersen

“K27= compressor housing with 2.75'' OD/ 2.1'' ID inlet, 2'' OD/ 1.7'' ID outlet, compressor wheel inducer ____"OD, and major 3'' OD.
Thanks alot in advanced.”

Edman951
The compressor wheel intake OD is almost there already, namely the inlet ID minus the clearance, which then is 2.07”.
To be even more detailed the inlet starts with a 2.44” ID, then conically narrows down to a cylindrical section with a 2.10” ID in which the compressor wheel starts.
Anticipating your next question, I measure the R to be about 2.25”
I hope this helps more than confuses.
Laust

Konstantin

@ Tony
according to turbonetics the 60-1 HI FI is smaler and flows only 90-95% of the real 60-1 compressor.
So why not use the real 60-1 compressor?

Konstantin