CUP ENGINE: MORE POWER?
07-06-2012, 12:32 AM
#1
Rennlist Member
Thread Starter
CUP ENGINE: MORE POWER?
I love the Cup engine! It puts out more than 110fwhp/liter in factory stock form, runs on cheap(er) fuel, is perfectly happy at 8000rpm, lasts 100 hours, and is close to indestructible. Porsche has done an incredible job with this evolution of their 911 boxer power plant.
So, with all this performance and longevity, did Porsche leave any engine power on the table? I'm an experienced privateer engine builder and has-been racer with a strong interest in Porsche race engine development. My workplace is a fully equipped performance engine machine shop allowing in-house component design, modification, and reconditioning. There are also the proper air flow and power measurement tools to accurately verify improvements.
I undertook an engineering exercise to discover precisely what Porsche has done to develop the Cup engine, to see what improvements could be made in its legal stock form, and to test the results of an engine modified to increase power using mostly Cup components.
A 2005 Cup engine was selected as a foundation. Scientific protocol suggests starting with a fresh factory stock engine for the baseline dyno testing…. Ok, first foul: There wasn't a decent bone stock engine available, so power measurement was first recorded on a “blueprinted” 3.6 Cup run on a Dynapack chassis dyno. This was followed by a modified 3.8L based on another ’05 Cup engine and run on the same dyno.
I began by thoroughly examining the stock engine and found a number of new and rather clever power tricks. Drat, some were tricks I use in the older air-cooled engines and they were already done in the Cup! Pretty cool, Porsche Motorsports. Other areas of potential improvement were identified, and subtle modifications/corrections made. I found one piston design parameter that I couldn't make sense with, and scratched my head for some time trying to understand the purpose. I finally reached the right person at Mahle who told me Porsche had made a mistake! Really? Maybe it was an oversight, perhaps a cost cutting measure; don't know. I suspect the error applies to all 3.6 Cup engines.
The 3.6 engine went back together with all legal mods. One big improvement was the valve job (seat angles primarily) where airflow gains were developed and verified on our Superflow 600 Flow Bench. The improvement is significant. The 996 heads now flowed better than 997 Cup heads – without opening up the ports and/or intake manifolds to the 997 sizes which wouldn't be legal. I had Mahle make pistons that duplicate the factory piston in size, weight, and ring configuration, but with the engineering error corrected, and a minor tweak to improve ring seal. Measuring the compression ratio showed it to be low, and below spec. Why? After measuring the combustion chambers on lots of used 996 and 997 heads and finding major inconsistencies, I concluded the excessive chamber volume (and resulting low compression ratio) to be a result of heavy-handed valve jobs. When valve seats are cut and valves are faced during rebuilds, the valves drop in the chamber, volume increases, and compression ratio goes down. The only way to get it back is to surface grind the head to reduce the chamber volume and regain the lost compression. Care has to be taken here since this dimensional change affects other components which have to be modified for correct fit. Other stuff addressed: The titanium rods were resized to correct distortion caused by 100 hours of fury. Note that new rod bolts also typically distort the rod dimensions badly, so re-machining them is a critical step not only to keep the big ends round, but to maintain proper oil clearance. I also pinned all the epoxied water bungs with 8-32 button head machine screws, retained with Loctite. Welding, although seemingly more secure, distorts the castings and too often leaves pin-hole flaws in the weld beads.
RESULTS – LEGAL 3.6 CUP
Ok, the results… Remember that these numbers were generated on a Dynapack chassis dyno. DynoJet (which is favored by the majority of shops because it is MUCH cheaper), reads 7% higher.
These power figures are accurate and exceed dyno results I've seen posted on RL and elsewhere. Although there is no scientific comparison, I believe this engine would be extremely competitive in stock class Cup racing – and pass tech!
Now the real fun began. How well does a 3.6 Cup engine respond to modifications? Keep in mind this is no longer a stock class engine. This is a (rather inexpensive) hop-up effort and although stock appearing, can only be used in modified classes, run-what-you-brung events, open time-trials, DE’s, and such.
AREAS ADDRESSED
Displacement: I increased displacement to 3.8L with custom 102.7mm Mahle pistons and the stronger RSR cylinders. The pistons were further lightened, use 21mm DLC pins, and additional ring seal mods. The slightly larger dome provides for a 13:1 compression ratio.
Heads: Improved porting, same revised valve job as the 3.6, intake ports and manifolds opened up to the larger 997 configuration.
Cams: Custom profiles were developed with more lift and duration than the stock 3.6 cams. Hydraulic followers were retained for reliability and longevity.
Throttle Body: The stock 76mm TB was enlarged to 80mm which is the maximum the 3.6 casting would allow. This is a rather complex machining process due to the unique Cup TB configuration, but more inlet air was deemed necessary.
Exhaust: Stock 3.6 Cup headers with Cup muffler (sans cats) and some gorgeous RSR type headers with custom mufflers were tested on the dyno. Guess which one made more power. Guess again. All testing was done through the muffler(s).
All the other areas addressed in the stock class engine were also addressed in the 3.8.
RESULTS – 3.8L MODIFIED CUP
CONCLUSIONS
Legal 3.6 Cup: A stock engine can be improved over that which the factory or PMNA provides and still pass scrutiny by any sanctioning body not requiring a factory or PMNA seal.
3.8 Configuration: Kowabunga! The 3.6 Cup engine responds dramatically to a moderate increase in displacement and the ability to breathe better. Further, the significant increase in power comes at a modest monetary cost particularly if the engine is due for a rebuild anyway.
NOTE
All dynos are different, and provide different readings. On the Dynapack dyno, we make sure NASA bound engines read no less than 7% under the mandated max power allowed in the NASA class. When the cars get verified on NASA’s Dynojet dyno during scrutiny, power is consistently at the NASA maximum.
So, with all this performance and longevity, did Porsche leave any engine power on the table? I'm an experienced privateer engine builder and has-been racer with a strong interest in Porsche race engine development. My workplace is a fully equipped performance engine machine shop allowing in-house component design, modification, and reconditioning. There are also the proper air flow and power measurement tools to accurately verify improvements.
I undertook an engineering exercise to discover precisely what Porsche has done to develop the Cup engine, to see what improvements could be made in its legal stock form, and to test the results of an engine modified to increase power using mostly Cup components.
A 2005 Cup engine was selected as a foundation. Scientific protocol suggests starting with a fresh factory stock engine for the baseline dyno testing…. Ok, first foul: There wasn't a decent bone stock engine available, so power measurement was first recorded on a “blueprinted” 3.6 Cup run on a Dynapack chassis dyno. This was followed by a modified 3.8L based on another ’05 Cup engine and run on the same dyno.
I began by thoroughly examining the stock engine and found a number of new and rather clever power tricks. Drat, some were tricks I use in the older air-cooled engines and they were already done in the Cup! Pretty cool, Porsche Motorsports. Other areas of potential improvement were identified, and subtle modifications/corrections made. I found one piston design parameter that I couldn't make sense with, and scratched my head for some time trying to understand the purpose. I finally reached the right person at Mahle who told me Porsche had made a mistake! Really? Maybe it was an oversight, perhaps a cost cutting measure; don't know. I suspect the error applies to all 3.6 Cup engines.
The 3.6 engine went back together with all legal mods. One big improvement was the valve job (seat angles primarily) where airflow gains were developed and verified on our Superflow 600 Flow Bench. The improvement is significant. The 996 heads now flowed better than 997 Cup heads – without opening up the ports and/or intake manifolds to the 997 sizes which wouldn't be legal. I had Mahle make pistons that duplicate the factory piston in size, weight, and ring configuration, but with the engineering error corrected, and a minor tweak to improve ring seal. Measuring the compression ratio showed it to be low, and below spec. Why? After measuring the combustion chambers on lots of used 996 and 997 heads and finding major inconsistencies, I concluded the excessive chamber volume (and resulting low compression ratio) to be a result of heavy-handed valve jobs. When valve seats are cut and valves are faced during rebuilds, the valves drop in the chamber, volume increases, and compression ratio goes down. The only way to get it back is to surface grind the head to reduce the chamber volume and regain the lost compression. Care has to be taken here since this dimensional change affects other components which have to be modified for correct fit. Other stuff addressed: The titanium rods were resized to correct distortion caused by 100 hours of fury. Note that new rod bolts also typically distort the rod dimensions badly, so re-machining them is a critical step not only to keep the big ends round, but to maintain proper oil clearance. I also pinned all the epoxied water bungs with 8-32 button head machine screws, retained with Loctite. Welding, although seemingly more secure, distorts the castings and too often leaves pin-hole flaws in the weld beads.
RESULTS – LEGAL 3.6 CUP
Ok, the results… Remember that these numbers were generated on a Dynapack chassis dyno. DynoJet (which is favored by the majority of shops because it is MUCH cheaper), reads 7% higher.
These power figures are accurate and exceed dyno results I've seen posted on RL and elsewhere. Although there is no scientific comparison, I believe this engine would be extremely competitive in stock class Cup racing – and pass tech!
Now the real fun began. How well does a 3.6 Cup engine respond to modifications? Keep in mind this is no longer a stock class engine. This is a (rather inexpensive) hop-up effort and although stock appearing, can only be used in modified classes, run-what-you-brung events, open time-trials, DE’s, and such.
AREAS ADDRESSED
Displacement: I increased displacement to 3.8L with custom 102.7mm Mahle pistons and the stronger RSR cylinders. The pistons were further lightened, use 21mm DLC pins, and additional ring seal mods. The slightly larger dome provides for a 13:1 compression ratio.
Heads: Improved porting, same revised valve job as the 3.6, intake ports and manifolds opened up to the larger 997 configuration.
Cams: Custom profiles were developed with more lift and duration than the stock 3.6 cams. Hydraulic followers were retained for reliability and longevity.
Throttle Body: The stock 76mm TB was enlarged to 80mm which is the maximum the 3.6 casting would allow. This is a rather complex machining process due to the unique Cup TB configuration, but more inlet air was deemed necessary.
Exhaust: Stock 3.6 Cup headers with Cup muffler (sans cats) and some gorgeous RSR type headers with custom mufflers were tested on the dyno. Guess which one made more power. Guess again. All testing was done through the muffler(s).
All the other areas addressed in the stock class engine were also addressed in the 3.8.
RESULTS – 3.8L MODIFIED CUP
CONCLUSIONS
Legal 3.6 Cup: A stock engine can be improved over that which the factory or PMNA provides and still pass scrutiny by any sanctioning body not requiring a factory or PMNA seal.
3.8 Configuration: Kowabunga! The 3.6 Cup engine responds dramatically to a moderate increase in displacement and the ability to breathe better. Further, the significant increase in power comes at a modest monetary cost particularly if the engine is due for a rebuild anyway.
NOTE
All dynos are different, and provide different readings. On the Dynapack dyno, we make sure NASA bound engines read no less than 7% under the mandated max power allowed in the NASA class. When the cars get verified on NASA’s Dynojet dyno during scrutiny, power is consistently at the NASA maximum.
07-06-2012, 01:12 PM
#6
Rennlist Member
Thread Starter
Dyno Results
The Dynapack is a brake dyno, whereas the Dynojet is an inertial dyno. Both are repeatable, but measure power in different ways. It's generally accepted in the industry that an inertial dyno will read higher than a brake dyno.
07-06-2012, 01:14 PM
#7
Rennlist Member
Thread Starter
Trending Topics
07-06-2012, 01:19 PM
#8
Rennlist Member
Thread Starter
ECU's
I didn't think it would be this long before someone asked that. The 'stock' engine used the MS4.1 Cup ECU that was in the car (05 6Cup). The 3.8 used the same Cup ECU that had the parameters changed for a 3.8L, but without any further real-time tuning.
As you know, ECU's are a big issue when modifying the Cup engines. I've been looking for a solution for quite some time (me and others). My search continues!
As you know, ECU's are a big issue when modifying the Cup engines. I've been looking for a solution for quite some time (me and others). My search continues!
07-06-2012, 01:22 PM
#9
I didn't think it would be this long before someone asked that. The 'stock' engine used the MS4.1 Cup ECU that was in the car (05 6Cup). The 3.8 used the same Cup ECU that had the parameters changed for a 3.8L, but without any further real-time tuning.
As you know, ECU's are a big issue when modifying the Cup engines. I've been looking for a solution for quite some time (me and others). My search continues!
As you know, ECU's are a big issue when modifying the Cup engines. I've been looking for a solution for quite some time (me and others). My search continues!
07-06-2012, 01:36 PM
#10
Rennlist Member
Thread Starter
ECU
Here's a short version of a long, ugly story... I found a local guy that swore he could hack into the Cup ECU. Sadly, he killed it. I had to replace it, and found only one used ECU ($8K+ new), which had been configured for a 'mild' 3.8. Ya-hoo. Turned out that the experience hurt my pocketbook, but it solved the problem for the short term.
07-06-2012, 01:42 PM
#11
Instructor
I can tell you that you Dyno Dynamics will read right at 3% lower. I also own a Dynapack and a friend of mine has a dyno dynamics. Nice build, i squeezed 419 whp out of a 3.9 liter with 2004 street car heads with 06 cup cams.
07-06-2012, 01:42 PM
#12
Rennlist Member
Thread Starter
Oh, more info Clay,
I found the right guy in Germany that says he can 'duplicate' the 3.8 load in other standard Cup ECU's. I haven't tried him yet, but I will in an attempt to repair the smoked ECU that I have.
I found the right guy in Germany that says he can 'duplicate' the 3.8 load in other standard Cup ECU's. I haven't tried him yet, but I will in an attempt to repair the smoked ECU that I have.
07-06-2012, 01:49 PM
#13
Interesting. I was going to do all the motec stuff on my gt3r but ended up selling it to cheyne! I decided i would keep the car "stock" but now i regret it because i hate being beholden to pmna or the 1 or 2 people who can hack the ecus....
07-06-2012, 07:15 PM
#14
Rennlist Member
Thread Starter
Mm
07-06-2012, 08:41 PM
#15
Burning Brakes
Purely in the interest of science, I am willing to donate a 05 996 cup motor to be sent to PMNA for rebuilding.
It can then be sent to you for dyno testing, upon completion of your test, just return to me the motor.
We can ask the forum for donations to cover the rebuild costs and shipping costs so that they can definitively have a baseline number in which to compare your build to
It can then be sent to you for dyno testing, upon completion of your test, just return to me the motor.
We can ask the forum for donations to cover the rebuild costs and shipping costs so that they can definitively have a baseline number in which to compare your build to