How would one go about deciding that valvesprings are causing a bad harmonic at a certain RPM? I mean how would you narrow it down that far that its not say a balancer, or rod, or something else?

 

it helps to have a dyno. You will see where power drops a little then gets back on track. It sorta looks like it goes into valve float then gets out of it a little later.

 

I understand that, but what makes it the valvesprings causing it. Or is it just an educated guess?

 

Jon,

A spintron also helps a bunch in this area.....

All springs have a set resonant frequency or two, or three etc.... That's one thing that makes a beehive nice, the tightening coils have different frequencies... the one BIG WAVE (frequency) you have operating at the same frequency makes for uncontrollable valve motion at certain RPM while lots of very little waves make very small impacts on valve motion. Think of a single or dual non beehive spring as the BIG WAVE in this example and the beehive as lots of small waves... Does less damage with lots of small waves and not a Tsunami.


I think the best case of resonant frequency was with a bridge called "galloping gerdie" or something of the sort.... You know the famous film of a bridge shaking all around like a hula dancer... the wind caused it to operate in it's resonant frequency and it when to hell in a hand basket VERY quickly.

Bret

BTW good pulp fiction quote in the sig.... or maybe you are just a fan of Ezekiel and his visons of space craft or whatever he saw.

 

I had to do a double take on that first quote Bret. Yes, dual or triple valve springs have 2 or 3 resonant frequencies. At first i thought you were talking harmonics. If someone was hitting the higher order harmonics then they picked a really terrible spring in the first place

Tacoma Narrows Bridge.

-brent

 

Thanks man on that bridge....

I hope that made sense what I said... This is coming from the non engineer in the family explaining this topic.... Hope I absorbed it well.

A side note to this.... back in the days of Dale Earnhardt RCR was having a problem on plate motors with this.... breaking springs left and right. Well they kept slowing the motor down more and more to stop breaking springs and they walked into the worse RPM range for the spring and valvetrain setup they had so it made it worse. Then they talked with Comp about it and fixed the problem. It can be a very baffling thing to run into, especially with narrow RPM cup motors.

Bret

 

What color is the coating on your valves? Dark greyish and kinda thick feeling, or gold, or almost clear? How long are your valves?

The ones I found are also 2.080 for the intakes. They are 5.445" long and weigh 87 grams. I also got 1.625 exhaust valves that are 5.495" long and they weigh 79 grams. Those weights are just a little more than the hollow stem LT4 valves, but still a good bit less than even stock LT1 valves. The only problem is that they are older Manley valves with the old thick coating which has been known to flake on occasion... I am probably going to run some of Comp's beehive springs with Ti retainers. I might as well mention that I've got Jesel Pro Series shaft rockers that have had the optional lightening and bead-blasting done to them, so you can say that these heads are going to have a pretty serious valvetrain...

 

Well either way i'll attempt to further explain for the both of us.

All spring/valve systems have and infinite number of resonant frequencies. As cyclic motion (rpm) increases in frequency the first resonant frequency encounterd is described as: the natural frequency, the resonant freq, the first harmonic ect.

Usually this is the primary design concern. The additional resonant frequencies, or harmonics, are at multiples of the natural frequency. Thus the terms 2nd, 3rd, xth order harmonics.

A double valve spring has 2 springs and therefore 2 different natural frequencies occuring at different rpms. Between the 2 of them they can maintain control of the valve. The behive as bret described has a range of varying natural frecuencies as the radius changes. Both of these methods of controlling the impact of the natural frequency work similarly to cancel out or 'self damp' the uncontrolled vibration. The behive also has the advantage of a lighter retainer.

If you are reaching multiple harmonics (actual harmonics and not just zeroing in n the nat freq) with a valvespring then it was very poorly designed because the harmonics occur when you double or triple your rpm. To do that you'd have to hit the first harmonic (natural frequncy) at a very low rpm.

There are some very interesting resonant feq / harmonic issues in suspension and tire design. I haven't studied the actual math of the valvesprings in detail.

-brent

 

I should have added to the above that a ti spring being lighter would raise the natural frequency to a higher RPM just as Mindgame said earlier in the thread, maybe avoiding the resonance problem all together.

-------------------------------------------------------------------------
http://www.dreambike.com/images/loopvamoots4.jpg

Yes titanium is quite resiliant. I was first exposed to its wonderful properties over 10 years ago in the bicycle industry.

Did anyone know that titanium is the 9th most common element in the earths crust? Now if only there was a more economic way to seperate all the junk from it and it was a lil easier to weld and machine.... (don't anyone steal my plan for the nobel prize)

-brent

 

I remember seeing video of that bridge, was fine until like 30-40mph winds or something and turned into a big noodle I wonder if that engineer(or team of engineers) ever got work again after that

That sounds more scientific now. So is it something that you could measure with a ??"sound frequency measurer"??? (dont think thats the tech. term ). Since the valvespring manufacturer might know what the springs natural frequency is?

jon

Yeah, wasnt sure if people would recognize it, I figure ill finish the quote out week by week

 

To better analyze the valvetrain and eliminate the effects of noise and vibration caused by combustion and reciprocating components, you'd use an eletric motor to cycle the engine. Ala Spintron.

Many years ago, and working at Algor, we delved pretty deeply into a computational modeling project geared towards valve spring dynamics and developing a simple modal approach for linear valvesprings. Most important was acquiring a discrete description of a spring..... varying coil diameter, pitch, modeling implications of coil clash, etcetera.

I remember going over a great deal of post graduate level materials for that project cause even with a staff of very bright engineers, the mathematics were still in a state of on going evolution. Too make a long story short, I soon came to the realization that (at that time) computers simply hadn't evolved to a state necessary in dealing with those types of calculations. At least, not on the PC level. Either way, we built on two models... a modal model for progressive springs in a frequency domain... aka, "fun with Fourier techniques for time variant systems" and a complex discrete model. Both had their strengths and weaknesses but given todays technology, especially in the realm of solid models/accurate spring "descriptions"... we could have done something really snazzy.

But I digress... the point of all that is.... simulation can be a big help. One test is worth a thousand simulations, but a sim will get you moving in the right direction. Rest assured in knowing that the BIG guys are utilizing both.

edit: what's really intersting is seeing what happens to a camshaft at 6000 rpm and up.
Will make you think twice about trying to save a few bucks by staying away from a billet.

-Mindgame

 

Sounds interesting. I guess something like a spring would seem very difficult to mathematically model, since there are so many variables involves with metal hardness, the memory the metal has, etc. etc. But knowing even a fraction of what places like NASA are capable of researching seems like it would be a piece of cake for them.

 

There is never anything easy about simulation. Statics are somewhat easy... but the world is dynamic and that's where you fall into a mountain of high level mathematics in a big hurry.

This is why I chuckle when I hear my nephew, aspiring engineer, say... "I can't wait to get into thermodynamics" or something along those lines.
The only person who thinks they'll really enjoy it are those who've never taken the class. Either that, or they're the types that'll never get laid.
These concepts are hard to get your head around and you can truely appreciate that when you've been there.

NASA is too busy filtering their own urine so they can drink it later.

Valvespring dynamics modeling is very specialized though. Not a whole bunch of people doing it on an R&D level.

-Mindgame

 

Do you know (or think) that OEMs take any of this into acct, or is it something that doesnt need to be worried about at the lower RPMs? I would think somebody like GM, or BMW/Benz more likely, would have the ability to deal with stuff like this.

 

I don't doubt that there's a good deal of in house testing going on, but I would be willing to bet that companies like GM are using Ricardo's dynamic modeling software to come up with intital designs. That, a good deal of experience, lots of testing along with consultation will get the job done.
The groundwork is out there, so we're not tallking about reinventing the wheel... just making it better.

On that note, anything you can find on Sir Henry Ricardo is well worth reading. I have a book at the office (can not remember the title) that is 5-star reading material all the way. I'll post the title when I get a chance tomorrow.

http://software.ricardo.com/products/

-Mindgame