Straight six crank harmonics 1

[TOHCan]

What causes or contributes to the destructive crank harmonics found on most inline six engines above 6000 rpm? Why do some survive to 8000 and beyond? Do the V-12's suffer the same?
Would crossbolted mains help reduce or dampen the harmonics? What design elements can reduce these harmonics, related to the length of the crank, stroke, material, manufacturing process? Thanks.

 

[strokersix]

It's my understanding that it has a lot to do with the length of the crank being half again longer than a v8.

I expect there is more to it though and I'm sure others more knowledgable than I will add comment.

 

[TOHCan]

Most of the domestic sixes that I've played with have this problem, and have fairly long strokes, 3.25 and above. Specifically though, I have been told that the high rpm durability of a Chev forged (non-twisted) crank with 3.25" stroke and four counterweights is not as good as a cast, 3.53" stroke, fully counterweighted crank, all other dims the same. Further, that the cast cranks survived an arrangement that lined up two engines end to end, with 1000 hp being made at the flywheel. That's a lot of twist on the rear crank!
I don't know if the cranks were phased to even out the power pulses, but assume that would be advisable.
The main question pertains to the possible refinement of the present cranks, or redesign of one that will handle up to 8000 rpm, retaining the long stroke.

 

[ivymike]

What causes or contributes to the destructive crank harmonics found on most inline six engines above 6000 rpm?
It's always a matter of mass distribution, crank stiffness, and excitation frequencies. A straight six is about as long a crank as you find in cars these days. Generally speaking, a V12 crank gets it worse, and V16, V18, V24, etc are progressively worse than that.

Why do some survive to 8000 and beyond? Do the V-12's suffer the same?

Stronger crank, more expensive damper or better tuned damper, different firing orders, etc. can all contribute.

Would crossbolted mains help reduce or dampen the harmonics?
not by a lot, if at all.

What design elements can reduce these harmonics, related to the length of the crank, stroke, material, manufacturing process?
to reduce torsional vibrations: lower recip mass, shorter crank, stiffer crank (shorter stroke helps), use viscoelastic damper(s), vibration absorber(s) (single order cancellation)

to increase crank strength: roll the fillets, nitride the fillets/pins, increase oil hole breakout radii, use undercut fillets, use stronger crank material, thicken critical sections, etc. - a stronger crank can survive worse vibes.

I was under the impression that a inline 6 cylinder engine was, under ideal circumstances, perfectly balenced
yeah, but that's unrelated to the question.

the firing order has to be designed so all the forces cancel out, but I assume any automotive engine producer has the knowledge to do this
you usually have a few choices (depending on crank layout, etc); the one that's best for vibrations may not be the best for getting the sound you want, or might have a negative impact on power output (intake/exhaust tuning, etc).

my understanding that it has a lot to do with the length of the crank being half again longer than a v8 yep, it has a lot to do with that.

I'm sure others more knowledgable than I will add comment.
I guess we'll have to wait and see. I'm certainly interested.

 

[strokersix]

The last Chevy six I built had a nodular 12 weight crank, 2.00 inch rod journals, and 4.062 stroke with a StreetDampr on the nose and a manual transmission. I ran the WEEEEE out of it, missed shifts and all. It was for short blasts so torsional harmonics probably didn't get a chance to build. It held together just fine for me.

 

[TOHCan]

I'm looking at doing some road racing, so I need to improve the long term durability. I'm wondering what can be done beyond using the best dampener and decent crank prep. I would love to roll the fillets on the cast crank if anyone knows who can provide that service, and will consider changing the firing order from the present 1-5-3-6-2-4 to a more even spacing if this is contributing. I have a torsiograph chart showing a large increase in third order or firing frequency vibration above 5000 rpm on the bare crank, and even with a good dampener, a spike at 6300-6400 rpm with again an steady increase above 6600 rpm. Also the sixth order parallels the third order. Any way to reverse the phase or some other method to add "destructive interference"?

 

[GregLocock]

Changing the firing order to one of the other configurations may help in agiven speed range.

A six, as built, is perfectly balanced for first and second order, and third order if there is no cyl to cyl variation in imep.

I am not sure that it is perfectly balanced for the other orders produced by the L/r ratio. It probably is.

But...

These all assume a perfectly rigid crankshaft. Once your crank starts to bend (and your cam starts to twist) then the phasing between different cylinders will change, and the vibrations take a finite time to travel down the length of the crank, which again upsets the phasing.

Your big problem is firing order excitation of first torsion and first bending of the crankshaft, typically at around 300 Hz and 400 Hz, respectively.

Either will destroy engines.

To deal with the first you need a torsional damper. To deal with the second you need a bending damper.

I wouldn't mind betting the average racer does not run a bending damper.

Full counterweighting reduces the stresses in each main bearing, it turns the engine into a row of 6 single cylinder engines. We once took this literally and designed a 12 counterweight crank - the prototypes were great, the production ones weren't for some reason to do with the manufacturing process.

Cast iron cranks have a bad rap, but I would point out that the internal damping of cast iron is much higher than that for forged damping, so the resonances won't spike quite so badly. Aslo, between the CI and the extra mass from the counterweights the resonant frequencies will be lower and so may be out of the running range you are using.

The two cranks linked up may be OK if there is a decoupler (flywheel or compliance) between the two cranks. I doubt the torque itself is killing cranks, judging by the turbo boys experience.

The rising trend above the resonance is inevitable - but I bet if you measured 10 different cranks one would have a trend back towards zero above the resonance. That's probably due to the distribution of mass along the crank, there are balancing methods to do that (balancing of flexible shafts above critical speeds) but they look like trial and error to me!



Cheers

Greg Locock

 

[TOHCan]

Thanks, Greg, I should have mentioned that the crank that was used to generate the graph was the shorter stroke forged crank with four counterweights. Is there a way that I can check for the effect that you mention on a conventional crank balancing machine?

And thank you, too, Mike, for addressing and confirming some of my suspicions. Dwelling on the mass distribution for a moment, I notice that the counterweights are the same along the length of the crank. Could the counterweights be reduced at strategic points on the shaft to change the mass distribution and add some dampening in a specific rpm range? Enough to extend the usable range from 6500 to around 8000? The rest of the engine can be made to survive if the crank will.

 

[ivymike]

I'm guessing that a relatively uniform mass distribution is your best bet, but I could be wrong about that. I'll bet that if you know somebody with some free time who knows how to use holzer tables to arrive at mode shapes and frequencies you could check pretty quickly - I can't do that anytime soon though, 'cause I don't currently have access to my tool for that (working away from home).

 

[GregLocock]

You'd need to run the crank at its true speed range, and you'd need dummy big ends etc. Even then, i think you'd probably need a realistic block structure as well - in other words an in-situ experiment.

How I'd do it is to put an accelerometer on each main bearing, run the engine throughits speed range, and then work on reducing the first order by balancing the crank webs neaest the worst main bearing, and work along the crank like that.

But before that I'd read up on "balancing of flexible shafts above critical speeds", there are probably some good shortcuts.




Cheers

Greg Locock

 

[TOHCan]

Thanks, guys, you both have given me the direction I need to investigate this further. Merry Christmas!

Oh, and it sounds like I should put two V-6's end to end instead of making a common-crank parallel dual inline six to get the twelve cylinder engine... but that's another thread!

 

[jrnsr]

If you want an exquisite V12, dig up an old LaFrance fire truck V12... 527 cubic inch, 215hp, gas, overhead valve and technically overhead cam, and only about 20" wide. Intake and exhaust come straight up from the top. They're usually low mileage and pretty cheap. It was designed by Pierce for the Pierce Arrow in the 1920's and was stuck into firetrucks into the late 1950's. I plan on putting one in my '47 DeSoto, then convert the firetruck to turbodiesel/electric/steam hybrid.

 

[Hasbro]

IN the 1970's there was a Ford engineer that Drag Raced a 300 Ford '6' using a three piece sectioned 'Boss 302' cylinder head furnaced brazed into one piece. He had worked through the crankshaft brakage/power robbing vibration issues I heard, by dissconnecting the flywheel from the crankshaft using an elastomer. I don't know if it was done like a front damper or what. RPM? It was to the moon! He was a killer with that combination for some years untill he retired. Look into Bruce Sizemore, this may give some input.

John Haskell
Aire Research Engr.

 

[TOHCan]

Thanks for the reminder, I have that article somewhere in my "archives" (boxes in the basement). I'll see if I can find it. If I remember correctly, he also used three 660 cfm carbs on an IR intake, something that also raised a few eyebrows.