Carl Fausett

Instead of modifying the heads to prevent valve-to-valve cross-flow during the overlap, I have been doing that by altering the camshaft overlap instead. Our cams for boosting have much less overlap than our cams for NA motors.

Expecting your next question, "how do we do that?" it is done during the camshaft weld up and grind process. The cam overlap can be modified several degrees safely. As with all camshaft welding and grinding, quality and tight control over the process is key. When done correctly there isn't a problem. Works great.

ptuomov

One can manage the crossflow and reversion with camshaft overlap, especially if one has adjustable center sprocket cams that can be optimized on dyno. However, I think (but don't know, what do I know?) one can create a more powerful engine by minimizing crossflow and reversion with port and combustion chamber shapes and then running more overlap. But it's not like I have A-B-A dyno graphs to prove that.

With a centrifugal supercharger that doesn't really produce much boost at low and midrange rpms, the ability to run some camshaft overlap should be very beneficial. Smaller exhaust ports might help too there, before the engine is running on boost

Strosek Ultra

Tell me Tuomo, how can we calculate the optimum exhaust port CSA? If it is too small the port will be a restriction, if it is too large the flow velocity will be down. For the calculation we have to take engine size, power output and rpm into consideration. As you know I am most interested to find out the best exhaust port CSA for a NA 7 liter stroker engine having dual 36mm exhaust valves.
The circular stock exhaust port is 40mm or 1257mm2 with dual 33mm valves. I have done some minor porting job just for testing but only managed to get the flow up by appr. 9% measured at 12mm of lift (280 CFM). To flow test the exhaust port is kind of tricky. You get quite a difference in result if the exhaust port is flowed with an extension or not. I am using a 220mm long 2" diameter extension pipe which enhance the flow quite a bit. I have noticed Carl and Mr. Brown both have got lower flow figures appr. 250 CFM (28" of depression) at 12mm of lift but I do not know if and what kind of an exhaust port extension they have used. The flow bench is blowing air through the exhaust port, air which will be heated up during testing so the air temperature at each reading must be considered.
The lift diagrams show the difference between stock S4 camshafts and the camshafts I will run. You can see the difference in duration, lift and overlap.
Also do notice the much longer opening and closing ramps for the mechanical performance cams compared to the hydraulic stock cams. The very (one may say extremely) mild stock cams work best in a retarded setting but race cams usually favor an advanced setting.
Åke

Carl Fausett

To Strosek Ultra,

I wanted to give you a tip based on your post #286 two pages ago. You are talking about clearance the head to accept a .500" lift cam lobe.

Three years ago I ran an intake cam on my race car with .452" of lift and had two problems result from it. First, the lobe was long enough that it "walked on" to the cam follower. It was on top of the lifter in the center of the lobe, but both edges of the cam lobe were out into the air and they only walked onto the lobe as it turned. The second was the total lift was enough to take the cam follower out of range. We literally pounded the followers into their bottoms and caused them to fail after about three races.

This made no noises or gave us any signals whatsoever even after bits of ferrous metal started showing up in our oil filter media. We now know that the metal was the cam followers pounding inside of themselves.

I've looked for photos - I know I have some - but cannot lay my hands on them just yet.

Anyway - just a word of what happened to me and telling you so you can watch for the same and hopefully avoid it.

Carl Fausett

I have found the boost actually begins at 3000 rpm even though it may not yet appear on the boost gauge. We know this because the HP is already higher than the same motor after we simply remove the supercharger drive belt. We start seeing it first appear as a positive integer on the boost gauge at about 3500 rpm. Of course, it grows exponentially from there.

My opinion is that "mid-range" is 3500 to 5500 rpm for the 928. As boost is "in" from 3500 rpm and-up, I feel that boost is a factor in all of my head porting and tuning.

This, specifically because the racer is rarely below 3000 rpm, so it makes no sense to me to optimize porting or exhaust tuning to a rpm band they should not be in anyway. I prefer to optimize the build and the tune for the 4000 to 65000 rpm band where the race motor will live.*

You're right, I might be able to pick up a little low-end HP by adding some camshaft overlap, but at the cost of losing HP up where the engine will be spending all of its time, I think it a poor trade-off.


* Before I build a race motor, I consult with the owner to discover his FDR, gear-sets, favorite race courses (short course or long course) and tire diameters so I can establish a target rpm range to suit his application. It is not the same for each engine.

ptuomov

Carl, the first issue you seem to be describing (lobe going over the edge of the cam) is not an issue with the maximum lift per se, it's an issue with the maximum velocity of the cam profile. That's what causes the lobe to go over the edge. If you have a long-enough duration lobe profile, then you can have a higher peak lift cam profile. The given lifter bucket diameter just means that you have to lift the valve slower, over more degrees.

As a related note, Simard's 7L motor is reliably running .500" cam on stock-size lifter buckets. That's because his cam profile (Jones cam) has a lower peak velocity. Ake's engine will have larger diameter lifter buckets, so he can run a higher velocity cam profile and lift the valve quicker to the .500" level.

Out of curiosity, what do you mean by "pounded the followers into their bottoms"? What part hit what other part?

ptuomov

That's a fair point, I am thinking street cars which need to drive well at low rpms.

There's still the issue that running overlap and fast-enough exhaust ports is the only way to evacuate the combustion chamber from the exhaust gasses. You still need the camshaft overlap for that at high rpms. And you can run more of it productively if you port the heads not to crossflow from intake to exhaust but instead to use intake charge to push exhaust gas out. By my theory anyway.

Carl Fausett

You are right, a little valve overlap can still be useful to the boosted motor. I did not mean to imply that we run no valve overlap on our boosted cams. We just run less than on our NA camshaft grinds.

Carl Fausett

Pull the hydraulic cam follower apart. There is a pintel that rests upon the top of the valve, and the other end of that pintel was hitting the inside of the bucket. Those two pieces of metal would normally never touch. I hope I am describing that OK for you. I took .020" out of the cam lift and the problem completely went away.

that's a good plan, we also considered increasing the diameter of the followers to keep the tall cam lobes we had.

ptuomov

What caused that? The side loading on the lifter, due to the lobe going off the lifter surface?

In any case, I recommend everyone reading this spending a little bit of time with a pencil, paper, ruler, and compass thinking about whether it's the maximum lobe lift of the maximum velocity of the lobe that causes the lobe to go off the lifter.

V2Rocket

sounds like a good case for solid lifters...a few 944 DOHC engines have made use of them. ill see if i can find the source...
(edit: https://www.lindseyracing.com/LR/Parts/SOLIDS16V.html)

they don't have to be crazy for maintenance either - my Subaru has DOHC solid-lifter "shimmed bucket" arrangement and they only need to be inspected every 100k miles, the shims almost never need to be changed out even at that mileage.

ptuomov

I don't know. 800hp 7L 928 engine with that big cams is a little bit of "there be dragons" territory on the map... ;-)

If I'd have to guess, I'd go with PipeMax.

With 33mm exhaust valves, it would center the recommendations around about 1600 mm^2 exhaust port and suggest the following headers:

--- 3-Step Primary Pipe Specs --- ( Low to Mid-Range Torque and Hi RPM HP )
1st Step Dia. inches= 2.053 Length= 12.4 to 13.8
2nd Step Dia. inches= 2.178 Length= 6.2 to 6.9
3rd Step Dia. inches= 2.303 Length= 6.2 to 6.9

With 36mm exhaust valves, it would center the recommendations around about 1700 mm^2 exhaust port and suggest the same header specs.

But you have your own formulas for sizing these, so I'm guessing you've already got a pretty good idea what you want to do!

Strosek Ultra

Carl, I do not understand what you mean by "the lobe walked on to the cam follower". Do you mean the lobe was riding on the edge of the follower?
My cam profiles are fairly kind to the valve train and theoretically they can be run on followers as small as 31mm (stock follower 35mm). More aggressive profiles with higher lift velocity need larger diameter followers. Also note I am using mechanical lightweight bucket lifters. In order to make room for larger valves and enhanced intake flow the tilted angle of the exhaust valves has been changed as well as the C-C distance between the intake valves creating the need for oversize 37mm followers. The photo is showing some of the followers I have to my disposal.
Åke

Strosek Ultra

Well Tuomo, I was interested in what your PipeMax would say. I have decided to increase the exhaust port size from a circular 40mm port (at the flange) to an oval port 40 x 50mm which will increase the CSA to 1657mm2. Think this is pretty much right for the dual 36mm exhaust valves.
The stepped header dimensions are about what I have in mind but I believe PipeMax is focusing more on high rpm peak power. I think tuning the headers for an rpm in the middle between peak torque and peak power will enhance the power curve. Primary pipe total length would then be in the range of 31-32 inches.
Åke

Strosek Ultra

Here are some cams not having too much overlap which I think Tuomo find interesting for the turbo engine. Depending on high lift velocity the intake grind #274 theoretically need oversize lifters of a minimum diameter of 36mm. For the exhaust grind #278 the minimum safe lifter diameter is 34mm. With an adjustable chain sprocket on the intake cam the overlap and cam timing can be adjusted in any desired way.
These are hydraulic grinds.
Åke