Strosek Ultra

The 944 guys do install an oil jet to lubricate and cool the cam chain.
Post #63: https://rennlist.com/forums/924-931-...r-400hp-5.html

I have been working on motorcycle engine power modifications all my life and the long camshaft chain for DOHC engines were always a problem. We used reinforced cam chains from IWIS and others.

Åke

ptuomov

I've got those stronger chains in one engine.

Out of curiosity, what spring loads were you running, seated and open? I'd like to know that data point, just to get some sense how close to the limit the stock chains are.

As a side point, with well matched camshafts and turbine, a turbo engine doesn't require meaningfully higher spring loads than a normally aspirated engine with the same camshaft and redline rpm. If the cam and turbine are not matched for the rpm that the car is run, sometimes one may need a tiny bit more spring load on the exhaust side. That's what I believe, from experience with the stock S4 engine, logic, and talking to people who build high-performing turbo engines.

Carl Fausett

Believing I wanted no more valve spring pressure than I absolutely needed, I sneaked up on it over several years and builds.

You can see the progression here:
http://www.928motorsports.com/parts/...ve_springs.php

Things to keep in mind: I like using beryllium-copper valve seats to help give the valve a soft landing, a better seal, and reduced hysterisis. So we believe when we did see valve "float" we were throwing it off the tip of our lobes at high R's, not valve bounce. That's the second point: my information is also a little different than notmal because of the high-lift cams involved.

But the examples I/we saw are these: the stock valve springs did fine with stock cams and 10 pis of boost. There wasn't a problem.

In the next iteration, moderate lift cams and about 14 psi of boost, we had valve float at about 5800 to 6000 rpm. That's when we went to the middle set of springs on this chart and that was enough to keep the valve train happy at that time. Raced that setup successfully for about 2 seasons.

The next build was higher lift cams again, and now we were building for Bonneville, not road racing, so I was building for 18 to 20 psi of boost. We had valve float again. You can really hear it on the dyno - its an unmistakable sound. So the valve springs on the bottom of this chart, the strongest ones, were sourced. And we're good again.

This year, we have gone from the stock 928 lifters to the lighter VW part, and we wonder if we could go back down a spring rate as a result. Some of us think "maybe" and some think "no" (I'm a no) so I'm leaving them as-is.

ptuomov

I agree that one should use the minimum load springs one can get away with without valve float.

It would be nice to know the acceleration profiles for the cams one uses. I don't think its the lift per se that requires stiffer springs. It's the negative acceleration of the cam, which has to be exceeded the spring force at each lift.

In terms of reducing the valve train mass, I think the required spring load goes up and down proportionally to that mass. I think you can just compute it whether you're fine for sure with lower load springs: Take the same percentage of from the springs that work that you lower the valve train mass.

Carl Fausett

Thanks for the link! Interesting!

Carl Fausett

I agree. Unfortunately I don't have those slope and profile numbers. My cam manufacturer has them, I can ask.

mark kibort

how does stronger valve springs control boost? valve leaking under float, bleeds off boost pressure?

as a side note, if you are getting valve float at 5800rpm, there is something wrong with the springs, or the ramp rate of the cams are radically different than stock... true in your case?

Carl Fausett

Boost pressing on the back of the valve has the effect of making the valve appear heavier to the spring.

Our big intake valves are 39.5mm in diameter, or 1.557 inches.
Link here: http://www.928motorsports.com/parts/32v_valves.php

If I have done my math right, the surface area of the back of that valve is 1.91 sq inches (less the reduction for the stem diameter and the force on the bell section, which I am leaving out for simplicity).

On that intake valve, at 20 psi of boost, this means the boost is pressing down with 38.2 pounds of force. In other words: that's a virtual 38.2 pounds added to the physical weight of the valve that the spring must lift and close against the seat. When the valve only weighs 2.3 ounces, you can see that the boost is actually the larger of the forces the spring must overcome.

Of course, this doesn't even speak to the inertia of the valve, which was traveling down at speed and must be stopped and reversed in direction...


BTW: Kibort, As an engineer, I know you know this. Meaning no disrespect - just trying to explain it for the other readers.

ptuomov

The problem with that logic is that it's the pressure differential between the cylinder and the intake port that matters. The intake port can't be considered in isolation.

Suppose that we have sensible camshafts. During the intake stroke (downstroke), the total pressure differential is going to be something like 1-2 psi, with the port at 14.7+20 psia and the cylinder 14.7+18 psia. That's not going make a big difference between the max positive cam velocity point and the max negative velocity points where the spring is needed.

When the valve is closed, there's typically more pressure in the cylinder than in the intake port (compression, power, and exhaust strokes) so the boost is not going to push the valve open either.

My conclusion from this is that on the intake side 20 psi of boost is going to require maybe 4 lbs more spring load over the normally aspirated motor, which is within the margin of safety for most springs that work in the normally aspirated motor.

mark kibort

Good description. also a very interesting set of dynamics going on here. the boost pressure on the back of the valve would always be there but as soon as it opens and fills the cylinder, there would be near equal pressure, but as it closes, it would have to fight this 38lb/sq" force to seal certainly giving an air cushion to the valve as it closes and seals. the extra inertia would actually help the valve open more if it is greater than the spring pressure, which im not certain it would be. I guess at that point, you would hear some noises as the valve lost contact of the lifter or the lifter lost contact with the cam lobe.
so, by stronger valve springs are you assuring boost wont be lost in without a good valve seal... but that's confusing, as it didn't close all the way on intake, you would lose some cylinder filling and thus power,

interesting.

Carl Fausett

I appreciate your input, but do not agree with your conclusion that an increase of 4 pounds of spring load would do it. I say this because I have real-world data from increasing the boost, getting some valve float, increasing the springs to stop it; yet when I went from only 14 psi to 18 psi the valve float was back. See post 48 above.

That tells me that the 37 lb increase was just enough, but not too much, to answer for the 14 pounds of boost.

ptuomov

My experience is more consistent with my theory, as are the simulations that I've run. Then again, I have a turbo motor and I have not done back to back valve spring change experiment. What I do know is that the stock S4 cam, intake valves, and intake springs don't behave any differently at 6500 rpm or so whether the intake manifold is ambient pressure or 20 psi from the turbo.

Are the valves that you hear floating in your engine on the intake or exhaust side? I haven't thought thru the supercharged engine exhaust side.

mark kibort

How would the pressure on the back of the valve do anything, if anything, as it opens, it would provide this "pressure" for the first 1mm of travel, as it opens, then both sides would have the same exact pressure, right up to peak lift and as it closes, same thing, pressure would be equal on both sides of the valves right up to the point it closed, so that boost pressure cant hold it open. in fact, after bottom deadcenter, the piston is on the way up, so the pressure theoretically in the cylinder is now rising vs the absolute boost pressure in the intake and on the back of the valve.

How are you detecting "float"?

ptuomov

If there's flow, then there's a pressure differential between intake port and cylinder. That's the only way fluids flow. The pressure difference should be small, however. If the cylinder pressure is ever when the valve is open much lower than the intake port, then you're choking the motor with a poorly flowing intake port. If the cylinder pressure is ever higher than the intake port pressure when the valve is open, then the valve probably shouldn't be open in the first place and the valve events aren't optimal for that rpm.

Thinking of supercharged engines with headers, here's one intake valve float scenario that doesn't really apply to turbo motors. This is a little bit far fetched when you plug in the numbers.

Suppose that near the redline, the suction pulse from the header makes it to the combustion chamber slightly before the intake valve opens. Now, the combustion chamber could have pretty low pressure and we have the whole boost pressure pressing on the intake valve. The intake valve opens on the boost pressure alone. You'd have to have really high boost and really strong suction pulse to get that done against any reasonable seat load.

Of course, if this is the case, one should either make the header pipes longer or open the intake valve earlier, instead of swapping in stiffer springs.

Haven't thought thru the exhaust side completely. I suppose that it could be possible that a high-pressure pulse coming from the header could blow the exhaust valve open? Higher boost means stronger exhaust pulses. I'd have to think this thru.

mark kibort

Yes, we all know there is flow as a result from differential pressure, but that even strengthens my assumption (case) In that the there is a pressure gradient too and its going to be larger deeper in the cylinder, vs near the valve. on either side of the valve, the pressure differential will be very close. worst case, we are talking about a very low pressure differential, especially as the piston hits the bottom and stops (where there probably is minimal cylinder filling. then, the piston starts to head up as the valve closes. there is NO pressure or flow that could result in any effect on the valve backside surface. if so, I would really love to know how. the only pressure that I can imagine Is the aerodynamic drag of the valve in the air stream . again, at BDC, im sure there is not much of that either. so closing the valve at that point is working against no flow or pressure differential.