It would be great if stock springs are enough.

Now question remains how much larger lobe stock heads can take without modifications. If no one has answer to this I can check in few days. GTS heads and S3 cams are going together soon, I hope.

 

Stock springs are not enough for the stock S4 cam! But I guess it all depends on how high one revs the engine,

Thanks, those measurements would be great. It's my understanding that webcam profile 274 with 0.480" lift still fits in the bore, but haven't tried it.

 

Tuomov, do you know webcam's duration numbers at .050"? if you do it would be better to post those numbers to keep us all in the same playing field. And the stock springs arent great for running the sharp nose on the cams. But if the nose can be left in more of an egg shape then the stock springs are OK.

 

Search for 944 S2 at http://www.webcamshafts.com/index_bl...le_search.html

The results include

VALVE LIFT .480/.442
DURATION 254°/240°
DURATION @ 0.050" 234°/222°
GRIND NUMBER 274/279

Designed to increase mid and upper end performance. May cause fair idle in fuel injected engines. Check ALL clearances!

 

Here's my simulation of the stock S4 engine with stock cams and with the Webcam 274/279 profiles on S3 cores (114 LSA). The simulation assumes flappy open. With stocks springs, the simulation indicates valve toss with both stock cams and Webcam 274/279 grinds before the stock redline.



By 6600 rpm, the stock springs give up for both the stock intake cam and the Webcam 274 intake cam in the simulation. You can see the Webcam 274 cam lift and actual lift diverging on the way down. The springs lose the control at the nose, and the cam runs away. Then, about 50 degrees after the BDC, the lifter and valve make contact with the cam and stage a massive bounce. The stock cams S4 don't behave as badly, but at about 6700 the springs loose the plot completely, never catch the cam, and the valve bounces off the valve seat. These are of course just simulations, but pretty much consistent with the pitted cam lobes, worn valve guides, and beat up valve grooves in all engines that have been opened.

The sim assumes stock springs shimmed to 45 lbf seated load for intake and 50 lbf for exhaust.

 

What lifter weight is used in those simulations? How much VW lifter helps or does it make any difference?

 

I used 174.3g reciprocating mass for these sims, so they already inlcude the effect of the lighter lifter but not the beehive springs or bigger valve. With stock lifters, things get even worse.

Here's my breakdown. The first column is stock and the second column is light-weight lifter, beehive spring, and 968 intake valve. This is not super accurate but I think it should be in the ballpark.

Recip mass of valve train.
grams grams % reduction RPM factor
Lifter 92.5 68.0 -26.5% 16.6%
Keepers (2) 2.3 2.3 0.0% 0.0%
Steel retainer 18.3 8.0 -56.3% 51.2%
(1/2) Valve spring 27.0 20.5 -24.2% 14.8%
Exhaust valve 62.4 62.4 0.0% 0.0%
Intake valve 62.4 70.9 13.7% -6.2%

grams grams % reduction RPM factor
Total for intake valve recip weight 202.5 169.7 -16.2% 9.2%
Total for exhaust valve recip weight 202.5 161.2 -20.4% 12.1%

 

Here are the cam grind specs for the 32v Ford from Sean Hyland Motorsports. Note the separate turbo cam and blower cam, and the conservative lift on the road race cam:

http://www.seanhylandmotorsport.com/...m_profiles.pdf

Sean Hyland's heads and camshaft page:

http://www.seanhylandmotorsport.com/...mponents.shtml

Great thread!

 

Cool, more data. The mod Ford exhaust valves are much smaller than those of 928 S4 (plus my exhaust valve will be +1mm over the S4 stock). I think that one can cut back a bit on the exhaust duration from that mod Ford turbo cam, but other than that it looks just what the doctor ordered.

 

One more thing. If one gets uses the Webcam 274/279 profiles, then one can up the compression ratio a bit.

Knock index is measure that the software computes, above 1 the engine may knock and above 2 it's likely to knock. The stock S4 has a compression ratio of about 9.4 and produces the maximum knock index value of 1.6 in the simulations. With the Webcam 274/279 profiles and stock CR, one gets maximum knock index value of 1.2. One can up the CR to 10.125 before exceeding 1.6 knock index. This in turn ups the simulated power a bit to about 360hp, as you can see from the below simulation graph:

 

swaybar,

Thanks for that first link.

It is interesting that their road race cams have not much lift but tons of duration. I believe that the reason for this is to give more of an egg shape to the lobes, this reduces the likelyhood of valve float. As well as increasing the life of the cams.

It is also interesting that their turbo cam is larger on the exhaust, probably trying to obtain a flow variance closer to 85%. Good information there for us to work with.

 

My guess would be that they are trying to produce the flattest possible torque curve.

I think you are right. The flow bench results show that the mod Ford head flows much less on the exhaust side than our 928 S4 heads. The Sean Hyland Motorsports Stage II heads flow about 70% exhaust/intake and their full race heads flow about 80% exhaust/intake.

Modular Ford head Intake CFM @ .500" Exhaust CFM @ .500" Ratio
4.6 4V, 1996-98 STAGE 2 287 194 68%
4.6 4V, 1996-98 FULL RACE 304 241 79%
4.6 4V, 1999-01 STAGE 2 273 195 71%
4.6 4V, 1999-01 FULL RACE 288 237 82%

The 928 S4 heads are closer to 90% exhaust/intake flow. Which, by the way, is exactly what I want for a turbo engine!

Porsche 928 head Intake CFM @ .500" Exhaust CFM @ .500" Ratio
Stock S4 head 301 267 89%
RMI ported S4 head 321 278 87%

This all supports your hypothesis that the long exhaust duration is because of the relatively restrictive exhaust port. With the S4 heads hooked up to a turbo, I am guessing that one could run about the same cam profile on the exhaust and intake.

 

http://www.tomei-p.co.jp/news/Report-EJ25-V1.html

Here's Tomei's technical report on building up the Subaru WRX STI EJ25 turbo engine. Recall that EJ25's cylinder shares the major dimensions with the 928 S4: Four valve head with basically the same valve sizes, same bore and stroke, etc. Rod is a bit shorter, though.

In any case, the Tomei report also runs some camshaft tests:



According to this, the 252/252 duration camshafts seem to be the best compromise.

EJ25 is an excellent comparison for one more reason. The intake manifold is a major restriction at the top of the rev range, much like the 928's intake manifold. Tomei report reads: "All tested versions exhibited a peak of 6,000 RPM boundary limit, which is believed to be caused by the lack of intake air at that RPM range. The duration cannot be increased without the modification of the intake components (surgetank or intake plenum, intake port/polish, etc)."

 

Both responses are probably correct, but here's something to support #2, from Sean's book on page 61 of 'How to Build Max-Performance 4.6 Liter Ford Engines':

"Road-race profiles check in at .385 inches of lift, 228 degrees duration. This produces a flat torque curve from 4100 to 7200 RPM."

And from page 62:

"You will note that out biggest camshaft only features .472 inches valve lift. This is because the 4-valve engine is far more sensitive to changes in duration than to increases in lift."

 

Thanks for your input on this topic Swaybar, it has come at a very good time.

The stock GT cams have .393" and .353" of lift. But the duration on the exhaust cam is a real choke.