Mrmerlin

nice work Carl and crew,
I hope that you still try to get the car Ready to run, you might surprise yourself,
good luck!

Carl Fausett

Thanks MMerlin. There is a World Speed Finals event at Bonneville Oct 6-9 that I toyed with going to, but I decided not to.

Im nbot going to hurry this repair. I want to take my time, do it right, and test it before we go. I hope to be back on the dyno in about 2 months or less.

GregBBRD

Valve float can manifest itself in a few different ways. The first (and most common, in this era of cam grinding) is the "bounce" off the valve seat when the valve closes. This shows up on the valve and seat surfaces....doesn't damage the springs as badly/as quickly.

The second type of "float" is when the acceleration from the ramp of the cam overcomes the spring and the valve literally gets "tossed" off the nose of the cam. This is not as common, with newer cam grinds, which generally control the acceleration of the pieces. This "float" will smash the valve spring into coil bind (since the valve assembly gets tossed off the nose), regardless of what kind of coil clearance you started with. This clue will be detected on the spring (coils touching.) This type of valve float will destroy the springs very, very quickly.......

Interestingly enough, simply changing the stiffness of the springs by a calculated mathematical proportion will almost always not fix the second type of valve float. (The amount of spring pressure needed is not directly linear to the rpm.) Once the acceleration of the lifter/spring/valve/retainer assembly reaches a certain point, the spring rate could double or triple and still not keep the pieces on the nose of the cam.

mark kibort

http://www.youtube.com/watch?v=_REQ1PUM0rY

mark kibort

It just amazing our engines can last more than a few mins.


http://www.youtube.com/watch?v=QuWiZ...eature=related

Mrmerlin

Greg thanks for sharing that bit of info

m42racer

I have read this saga from start to finish. Very interesting indeed. Two posts in my opinion are critical in the cause. GreG B is dead on. This is very comon and the major cause of most spring failures.

Without trashing any company, there was anme mentioned in the previous posts about the cam is use. Ther lies the cause. it doesn't matter what spring you use, how good the set up is etc etc, the unwanted noise or harmonics that is created and induced into the Valve train all starts with the cam Lobe design. So often the Lifts and durations are changed in oredr to obtain graeter air flow numbers with total disregard to the actual cam control over the valve train. The design is somewhat of a cut and paste design. As Greg correctlt states, the cause can be seen in the parts in front of you. Someone who knows about Cam design could very quickly look at the lobe design and give you an answer. All the good parts you are using are wasted because of the cam design. Once you get that sorted, the power will come and you will be amazed the extra power you make. I'll PM you with a name of someone to call and ask their opinion. They can help you for sure.

Mike Simard

Very good points M42'. Nobody ever considers cam quality and there are things that even a good engine builder would never realise unless that start designing and making cams.
I have no idea about Carl's welded cams and whether they had anything to do with the spring but you would be surprised at what is going on with some lobe profiles.
You can only see outright problems but there's an infinite number of possibilities of lobe shape and problems are present in just about everyone's cams. Even some big names or succsessful race grinds have things going on that are a problem on some level. It's hard to get the lobe profile design good but even that doesn't guarantee anything becasue you still have a gaunlet of ways for problems to occur in making the master, setting up the grinder and accounting for machine wear. Even a CNC grinding machine is full of expensive maintenance issues that will show up on the cam.

Spring harmonics are something that should be considered when designing lobes but in real life, the cam company will often pick a master they already have and get on with it while recommending some hideous spring rate because they don't understand the subject.
I've seen properly designed lobes on engines similar to our 32Vs and you would not believe how light the springs are and yet they road race to 7k with reliability.

Carl Fausett

All good points - and Greg's review of nose float and seat bounce is spot-on. But remember, I already said 2 things:

That I used Beryllium-copper seats (soft) to help control bounce at closing, and that I used the softest springs I thought I could to help control the valve so as not to create undue stress on the drive belt and nose of the cams.

Already some of you are ready to throw stones at the cam grinder? I'm not. The cams look fantastic, by the way. Their wear pattern is flawless and there is no pitting or inclusions. I think I was too soft on my spring calculation and the responsibility for that is mine.

Frankly, what were the chances of mixing bits from a dozen different manufacturers together for the first time and getting it right the first time? We made good HP, and we found there is still more to be had. I'm happy with that.

Anybody who thinks they Know It All should rush to Joe Gibbs racing team and apply for a job. After loosing 2 engines this last weekend (one due to valvetrain failure) I'm sure they would love to hire. And you know those NASCAR teams... they pay for crap and hire the least experienced. (I'm trying to point out it isnt as easy as you think)

ptuomov

What software are you guys using to spec valve springs, given the valvetrain mass and cam profile? This to everyone, not just Carl.

mark kibort

I dont think you would have ANY of the harmonic issues at engine RPM less than 6500, and thats where Carl should be running this engine anyway. if he thinks he has valve float and thats the reason for the power drop off, that would be surprising. look at the videos of 6-7-10,000rpm valve characteristics dyamically. Its a boosted engine, and the fall off is pretty characteristic. Carl, what is your boost level at 7000rpm now? If he cant get more power past 6000rpm, just use the stock stuff and be done with it. (or something near stock , but a little more high performance, probably what you have in there now). I would think the cause was a defective spring, but it sure would be worth someone to look at the cam design to take a swag at any possible major issues that could be present. again, I think you are chasing your tails if you think you have float and harmonic issues that caused the break. sure harmomics are a part of any spring mass system. you hit a major resonance, and all heck breaks loose. However, I dont think the engine speeds are that great to have it actually "break" a spring, even if it was operating through a resonance point. Just some general thoughts I had.

ptuomov

Mark --

He's got short runner intake manifold, huge intake ports, big valves, .442"/222@0.05" intake cam, etc. Why are you saying he should be running this engine at less than 6500 rpm? With those parts it should be run at higher rpms, IMO.

The supercharger is a belt driven centrifugal supercharger. The boost will continue to grow with rpm, unless there's some cold side boost control valve or something. Unless the supercharger rpm is getting excessive or it's going way off the map, I don't understand why the torque curve should rapidly "fall off" past some rpm point.

On springs resonances: I like the beehives because they don't really have the resonance problem. The beehives in my low-compression engine built by Jim morton have 38138 cycles per minute resonating frequency, which will hopefully mean no problems in the intended rpm rnge of that engine.

Best, Tuomo

mark kibort

Because he is already at 730rwhp and its happening at 6000rpm. He doesnt need any higher rpm, unless he plans on getting 800+hp out of his set up. I dont know the details of his supercharger, but the dyno graphs of most superchaged large displacement engines, always seem to fall off at moderate RPM. Yes, the boost will go up with rpm, but so does the loses. they go up with the square of RPM, so this is the main reason why it will fall off at some point, and do so in a dramatic fashion. His dyno run curve, looks like this is happening. as the RPM increases, so does the air flow. the runners, valves, ports are all rapidly shrinking in "effective" size.

as far as resonance, even with a 38Khz resonance frequency, the sub harmonics can be brutal. And, when the input speed equals, or is a factor, the resonance frequency, problems can arrise. I think the 6000rpm speed this engine is turning is pretty slow for any adverse problems, especially springs breaking.

just a few random thoughts. MK

GregBBRD

My thoughts were simply "food for thought" and not directed at the cam grinder....just general information. Especially since Web Cam's Porsche lobe designs are probably "old school", Mike Simard makes points worth considering. As long as the cams are out, you might think about having the lobes looked at on some sort of current lobe software. Pretty easy thing to do and won't take much time or money. I actually have my lobe designs checked by a couple different people, to eliminate any "bias", once the cams come back from grinding.

The valve spring restrictions, in the very limited space that a 928 offers, is a challenge. It occurred to me that the basic idea of using a "dual spring" is to avoid engine damage, should one or the other springs fail. Since this obviously did not prevent damage, you might consider using one of the "beehive" springs that are offered. Might give you a much larger selection of springs to chose from.

Again, just food for thought. I'm certainly not an SAE mechanical engineer, as you apparently are.

ptuomov

I am not a SAE mechanical engineer, or even an internet engineer, but I know that the PAC 1223 beehive springs for 4V modular Ford engine fit in the 928 S4 and later heads. Peterson American Corporation will reset the length for a lot charge of $200.

I got once set left of the four that I bought, in case anyone is interested. Available at my cost of about $625 plus shipping, which includes the tiny 8g steel retainers and Ferrea LS-7 locator discs. My springs are reset to the following loads/length, one can shim them to a slightly different installed heights (mine are between 1.400 and 1.350).

PAC-1223L (L is for custom load)
Height inches, load lbs
1.550, 38
1.500, 52
1.450, 66
1.400, 80
1.350, 95
1.300, 109
1.250, 124
1.200, 139
1.150, 154
1.100, 170
1.050, 186
1.000, 203
0.950, 222
0.900, 242