When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
For can completeness and for future generations, here are photos of the difficult "up chute" part in the stock intake manifold, specifically in the driver side of the throttle body:
I paid big $$$$ for wet flowing the heads before and after with stock heads and my modded heads. no one wanted to share the cost when I posted here, so the numbers go with me to the grave. Wet flow is so much more accurate than dry flow.. and they are usually done by a shop that does aerospace engineering. lets just say that I spent more on that project than you could pay cash for a low mileage GTS 5 speed. I have wasted so much money and time on development with the variocam, and the motec tuning and the flow work on the 928 heads that I should be done with this car and just buy something that performs. not going to happen and maybe one day I will get it running. The numbers are worth it but the stress and loosing your *** over people who agree to share the cost is not.
Interesting stuff. In theory, Variocam should have applied to a 928 being that a 928 is just a 944 engine x2.
On my 944S2, (3.0L NA), I took a 968 head that had variocam, and retrofitted it to my motor. Coupled with the 968 DME, the setup worked 100% great and made my motor into a 968 motor overnight. It was a bit of a novelty, and lots of work for that extra 25 hp.
Ultimately I got rid of that motor and put an 86 turbo motor in there. With a basic Lindsey racing package, the car made 300whp with no drama, no fuss, and no tuning headaches. Walked all over the Variocam car.
Sometimes the way to make more power is more simple, more brutish than a more complicated method.
928SG, Interesting stuff. In theory, Variocam should have applied to a 928 being that a 928 is just a 944 engine x2. On my 944S2, (3.0L NA), I took a 968 head that had variocam, and retrofitted it to my motor. Coupled with the 968 DME, the setup worked 100% great and made my motor into a 968 motor overnight. It was a bit of a novelty, and lots of work for that extra 25 hp. Ultimately I got rid of that motor and put an 86 turbo motor in there. With a basic Lindsey racing package, the car made 300whp with no drama, no fuss, and no tuning headaches. Walked all over the Variocam car. Sometimes the way to make more power is more simple, more brutish than a more complicated method.
In my opinion, you should have turbocharged that 968-headed variocam 3.0L motor!
I think Porsche's view for a long time was that they'll do the bare minimum with the engine and rely on image, image, image to sell the car with high profit margins. The bare minimum for a turbo motor was two valves (not four), static cam timing (no VVT of any kind), and static intake (no active resonance intake). Many other car factories that are more in the business of selling stake and less in the business of just selling sizzle, bare minimum for a turbo motor wasn't enough.
I paid big $$$$ for wet flowing the heads before and after with stock heads and my modded heads. no one wanted to share the cost when I posted here, so the numbers go with me to the grave. Wet flow is so much more accurate than dry flow.. and they are usually done by a shop that does aerospace engineering. lets just say that I spent more on that project than you could pay cash for a low mileage GTS 5 speed. I have wasted so much money and time on development with the variocam, and the motec tuning and the flow work on the 928 heads that I should be done with this car and just buy something that performs. not going to happen and maybe one day I will get it running. The numbers are worth it but the stress and loosing your *** over people who agree to share the cost is not.
Not trying to get your results for free, but did you learn anything important from that wet flow experiment? My impression (not fact or experience) is that wet flow flow bench results are very informative in carburated cars, but not so informative with multi-port injection cars.
Correct; once port injection is employed only the last couple of inches between injector and valve matter with regards to wet flow (as in the case of a very low pressure drops encountered with large individual runner throttle bodies).
But still wet flow analysis can be used to view fuel mixture effects in the combustion chamber as a result of swirl measures employed in these last few inches of this critical area of port design
"How the air flows through the ports can also affect fuel distribution and mixing. A traditional flow bench doesn’t really tell you anything about what’s happening to the air/fuel mixture as it passes through the ports and enters the combustion chamber."
"Wet flow testing can reveal things about port flow that are totally invisible with ordinary flow testing. Mixing a fluid that contains UV dye with the air as it enters the port simulates what happens to the air/fuel mixture as it enters the combustion chamber. Viewing the chamber from underneath, you can often see the liquid puddling or streaking as a result of turbulence and misdirected swirl. Playing around with the port configuration can often reduce the puddling and separation that results in a loss of potential power."
"The fluorescent dye allowed us to see the behavior of the air/fuel mixture in fine detail. We could watch as a vortex would form, grow, and move around the chamber like a miniature tornado as we adjusted the valve lift. We could spot areas where the vortices joined to form a cyclone of fuel and air. We could see where flow was turbulent, and where it was stagnant. We felt like blind men who had been given the gift of sight."
There is a very good post by Extreme Tuners that talked about some wet flow testing of a port injected F1 engine they had access to (lots of pretty 3D flow maps) but a bit too esoteric for this discussion (not to mention 200 pages of posts I would have to go through to find it)
Last edited by bertram928; 02-24-2017 at 05:00 PM.
In my opinion, you should have turbocharged that 968-headed variocam 3.0L motor!
I think Porsche's view for a long time was that they'll do the bare minimum with the engine and rely on image, image, image to sell the car with high profit margins. The bare minimum for a turbo motor was two valves (not four), static cam timing (no VVT of any kind), and static intake (no active resonance intake). Many other car factories that are more in the business of selling stake and less in the business of just selling sizzle, bare minimum for a turbo motor wasn't enough.
Really hard to do. Not sure if you are aware, but most all of the (successful) 944 turbo builds are 8v motors. There are a few turbo'd 16 valvers out there, but even Porsche, when it built the 968 RS Turbo made it an 8 valve car. I have dicked around with Variocam a little, and it's close cousin, Varioram (found on the 993), but at the end of the day, it is, to me, a gimmick. You can get the sort of power Variocam or Varioram make by using other tricks.
It is most appropriate in an emissions regulated street car, where it makes the most sense.
Really hard to do. Not sure if you are aware, but most all of the (successful) 944 turbo builds are 8v motors. There are a few turbo'd 16 valvers out there, but even Porsche, when it built the 968 RS Turbo made it an 8 valve car. I have dicked around with Variocam a little, and it's close cousin, Varioram (found on the 993), but at the end of the day, it is, to me, a gimmick. You can get the sort of power Variocam or Varioram make by using other tricks. It is most appropriate in an emissions regulated street car, where it makes the most sense.
In my opinion, the best 944 turbos run 16v heads. Now, that may be just my bias. I like Turbo Tim's torque curve, for example.
I agree that for a turbo race car, only at a fairly high level and under specific rules do you want to take on the complexity of variable geometry intake and/or variable valve timing. Under most rules, you'd probably never go there.
The four-valve head is a no-brainer for a turbo race car, though. I think 968 Turbo RS is some sort of a compromise between constrained budget, existence of reliable 8V parts, and the rules that required detuning anyway. Even if the rules mean you have to detune it no matter what, I think it's still a lot more efficient in every possible way to run a four-valve head, as long as you have the budget. Don't know for sure, though -- these are opinions and not facts.
Almost all factory street cars all have four-valve head and VVT now, turbo or not. Emissions, fuel consumption, low rpm torque, etc. are all reasons for that. The variable geometry intake is mostly waste of money after you have full VVT system, I think.
Overall the wet flow showed that the 928 heads flow very well but have nasty pooling issues that are resolved by reshaping the bowl and the transition from one open port to an individual bowl area for each valve. Significant reshaping of the valve back cut on the stock 968 valves helped as well. The data that was used in previous wet flow changes and the improvements showed that the HP improvement was a measly 17%... the pictures and video from the changes showed the lowest pooling that they had achieved to date and they wanted to map the 928 intake ports just to mimic this behavior if possible in other engines. They may small changes to the exhaust ports but over all no one has made a 928 engine similar to mine.
The factory 45mm x 50mm oval ports with the cut out for the injector were ported to 50mm round ports on the adapter plates & THE HEAD into the transition area then to the bowls. So while the science and the numbers look spectacular, we will all find out when it is tuned and dyno'd.
Originally Posted by bertram928
Correct; once port injection is employed only the last couple of inches between injector and valve matter with regards to wet flow (as in the case of a very low pressure drops encountered with large individual runner throttle bodies).
But still wet flow analysis can be used to view fuel mixture effects in the combustion chamber as a result of swirl measures employed in these last few inches of this critical area of port design
"How the air flows through the ports can also affect fuel distribution and mixing. A traditional flow bench doesn’t really tell you anything about what’s happening to the air/fuel mixture as it passes through the ports and enters the combustion chamber."
"Wet flow testing can reveal things about port flow that are totally invisible with ordinary flow testing. Mixing a fluid that contains UV dye with the air as it enters the port simulates what happens to the air/fuel mixture as it enters the combustion chamber. Viewing the chamber from underneath, you can often see the liquid puddling or streaking as a result of turbulence and misdirected swirl. Playing around with the port configuration can often reduce the puddling and separation that results in a loss of potential power."
"The fluorescent dye allowed us to see the behavior of the air/fuel mixture in fine detail. We could watch as a vortex would form, grow, and move around the chamber like a miniature tornado as we adjusted the valve lift. We could spot areas where the vortices joined to form a cyclone of fuel and air. We could see where flow was turbulent, and where it was stagnant. We felt like blind men who had been given the gift of sight."
There is a very good post by Extreme Tuners that talked about some wet flow testing of a port injected F1 engine they had access to (lots of pretty 3D flow maps) but a bit too esoteric for this discussion (not to mention 200 pages of posts I would have to go through to find it)
Overall the wet flow showed that the 928 heads flow very well but have nasty pooling issues that are resolved by reshaping the bowl and the transition from one open port to an individual bowl area for each valve. Significant reshaping of the valve back cut on the stock 968 valves helped as well. The data that was used in previous wet flow changes and the improvements showed that the HP improvement was a measly 17%... the pictures and video from the changes showed the lowest pooling that they had achieved to date and they wanted to map the 928 intake ports just to mimic this behavior if possible in other engines. They may small changes to the exhaust ports but over all no one has made a 928 engine similar to mine.
The factory 45mm x 50mm oval ports with the cut out for the injector were ported to 50mm round ports on the adapter plates & THE HEAD into the transition area then to the bowls. So while the science and the numbers look spectacular, we will all find out when it is tuned and dyno'd.
Thanks for sharing this.
Do I understand this correctly that these tests do not correspond to the variocam + stock intake manifold casting but for your newer ITB motor project?
I'm still stuck in the stone age and using the stock intake manifold in my turbo car! ;-) This isn't super relevant for my turbo project in that sense, but nevertheless extremely interesting.
Are you planning to use one injector per port, or two with one injector close to the port for low rpms and another injector further upstream for high rpms? What's the spray pattern on your injectors? I'm assuming a single cone based on the injector boss location on the throttle body element.
I used a stock 928 s4 87-88 head for the flow testing.
with the same cams.. they are simard billet cams tuned for my bore diameter.
The interesting thing was that with a factory head, and factory cams, they found a 37% increase in HP and torque from reshaping the bowls and back cutting the valves to prevent the pooling and nasty wicking of fuel on the ports. Using factory heads and valves fuel was actually pooling on closed valves and dripping into the combustion chambers... in liquid form... and blowing out the exhaust...... in a heat created vapor.. uh yeah that is good for performance... add spark... um yeah...
Originally Posted by ptuomov
Thanks for sharing this.
Do I understand this correctly that these tests do not correspond to the variocam + stock intake manifold casting but for your newer ITB motor project?
I'm still stuck in the stone age and using the stock intake manifold in my turbo car! ;-) This isn't super relevant for my turbo project in that sense, but nevertheless extremely interesting.
Are you planning to use one injector per port, or two with one injector close to the port for low rpms and another injector further upstream for high rpms? What's the spray pattern on your injectors? I'm assuming a single cone based on the injector boss location on the throttle body element.
Overall the wet flow showed that the 928 heads flow very well but have nasty pooling issues that are resolved by reshaping the bowl and the transition from one open port to an individual bowl area for each valve. Significant reshaping of the valve back cut on the stock 968 valves helped as well. The data that was used in previous wet flow changes and the improvements showed that the HP improvement was a measly 17%... the pictures and video from the changes showed the lowest pooling that they had achieved to date and they wanted to map the 928 intake ports just to mimic this behavior if possible in other engines. They may small changes to the exhaust ports but over all no one has made a 928 engine similar to mine.
The factory 45mm x 50mm oval ports with the cut out for the injector were ported to 50mm round ports on the adapter plates & THE HEAD into the transition area then to the bowls. So while the science and the numbers look spectacular, we will all find out when it is tuned and dyno'd.
Actually the factory 928 4V oval intake port is not 45mm x 50mm. The port is appr. 48mm wide and 38mm high (smallest dimensions at the manifold mating surface). You have to measure the height of the port perpendicular to the air flow. Assume that huge 50mm round port is measured perpendicular to the manifold mating surface which put the true height of the port to about 45mm.
Reshaping of the bowl area, of the short side radius and of the back side of the intake valves is common practice in order to enhance flow as I have shown here. https://rennlist.com/forums/928-foru...ughout-15.html (start at post #218).
I have to get going with hopefully the last head porting experiment where the 42mm race intake valves are spread apart by 3mm (C-C increased from 44mm to 47mm) in order to reduce shrouding at the area between the valve heads.
Åke
02-23-2017, 02:15 PM
For completeness, here's the stock part
