speed secret for 356/912 race engines

[snowman]

The following is a "speed secret" for 356/912 race engines.

The intake stack length, from the back of the valve to the top of the intake horn determines the rpm range that the engine will make max power. There are multiple lengths that will work but the shortest one is the most beneficial and works over the largest rpm range. This is well known.

The secret is that if you have the valves closed and fill the intake with water all the way to the top of the horn, carbs and all it must be exactly 110% of the cylinder volume. (note for the mechanically challenged. DO NOT DO THIS WITH THE ENGINE IN THE CAR as you won’t be able to remove the water). The reason for doing this is that you can get about 110% VE with tuned intake and exhaust. The 110% volume is exactly what’s required to charge the intake. For example if your engine is 1720cc each cylinder is 430cc. The intake must hold 430cc plus 43cc or 473cc.

This means that if you port the intake and heads you must be careful. If you increase the diameter to much you may need to shorten the intake stack, increasing the rpm where the peak performance occurs. This is a problem that must be solved simultaneously for optimum performance.

The second part of the secret that is not so well known is that there are multiple intake lengths that work. Unfortunately the longer lengths will not produce as strong or as wide a power peak.

[das908kind]

Wow! fascinating. Where did you find that information?

Anything else?

[das908kind]

BTW, you wouldn't happen to be the same "snowman" that hauled an 18-wheeler full of bud across the Texas state line, sending Smokey on a wild goose chase, are you?

[Mr Beau]

Snowman,

You have to be careful when stating items as 'facts'. Do some research on Helmholtz intake and you'll se it's not quite so easy. Being an EE, you'll like the electrical analogies used to describe the mechanical processs. Order a copy of the SAE paper cited for more indepth explanation.


Link 1

Link 2

Link 3

[snowman]

This info was obtained from some of the most experienced gurus of the industry. Its very close to being right, a whole lot closer than most other info out there for the amateur. Its not "easy" as several things have to happen to make it true, but a good guide to get to first base. After that, the last couple of percent, is for the pros who have the money and their life to dedicate to the solutions. For old engines like the 356/912 its a fact as they have been beat to death and then some. You arn't allowed to design a new engine, only work with what you have.

The referances you cited are very interesting. A good analogy is stepped waveguide design for broad band operaiton. Also, since an engine is a Pulsed System and sort of periodic you can apply very advanced theory to broaden the response over a larger rpm range, just like you can with microwaves. Just got to love those fourier series.

I am not an automotive engineer, but I haven't seen anything published using advanced broadband theory, only attempts at flat responses and absolutely nothing about giving up more after the peak to help enhance the peak. Some small ammount of ripple in the response would certainly broaden it considerably, without giving up anything significant. I would suggest any automotive engineer that really wants to opitmize the tuning read up on everything a Dr. Ralph Levy has patented and written on broadband filters and wave guides, especially the periodic waveguide structures. Make sure the electrical equivalents are correct, and possibly generate some new models using distributed elements as they will likely be needed. You will need at least a masters level of math, some background in pulsed periodic systems like radar and a whole lot of time. I am absolutely certain that much better results can be obtained over much wider rpm ranges using this theory than simple ones used in the papers refered to. I would suspect that somewhere, especially in the big auto companies, that this may have been done, Probably based on acustic theory that is more directly applicable than electrical. On the other hand some of this theory is very advanced, even for the microwave industry, and may not have been looked at yet.

[id10t]

Wonder why nothing like this was ever posted by Harry Pellow or others on 356talk...

[snowman]

Maybe he didn't know about it or didn't want to tell. I got the tip from one of the preeminent cam grinders with a whole lot of Porsche experience. I also got over 145hp on the ground with my 912. The guy does tell others but only after they meet certain conditions of his so I will not repeat his name without his permission.

The info also makes a lot of sense. With proper tuning you should have at least 110 percent the cylinder charge waiting in the intake or it (ram charging) just can't happen. A 356 is very simple as each intake runner is independent and there is no plenum. The biggest problem is the intake size varies along the way and it contains fuel as well as air. Calculations that do not take these things into account do not work. The combustion chamber is also the bathtub type which has a dramatic effect on the parameters. So you go with what a million people have already done, and it works.

As to why it works, its simple to visualize in the frequency domain, (rpm). Each firing of a cylinder is an impulse. If you look at an impulse in the frequency domain you see a sin(x)/x presentation for each bang. Since there is a series of bangs you can line up the sin(x)/x waveforms next to each other. If everything is right the side lobes of the waveform can add or subtract from the adjacent ones. If they add you have what we want and ram tuning. Very easy to visualize in the frequency domain, very hard to visualize in the time domain. That’s the power of higher level math. This representation also shows why there are multiple solutions, and why they are not as good as the fundamental one.

[912UK]

interesting stuff