How to calculate valve spring requirements? 2

[BobSP]

Hello everyone,

I'd like some help to determine how strong the valve springs need to be for a given application. I know the easy way is just to look at manufactures recommendations for the camshafts, but if i want to actually know how much i need, how would i do that?

If i know the mass of the valve, spring and retainer, the lift of the valve and the RPM, is that enough? Do i need to factor in camshaft acceleration, and if so, how do i determine that? Is there a formula that takes all that and gives me the required spring force?

 

[BrianPetersen]

The cam profile will establish what the acceleration of the valve is as it goes over the nose of the cam lobe. This requires detailed analysis of the motion profile of the cam lobe.

Complicating matters is that the spring itself has inertia, and due to the speeds that are potentially involved, it doesn't compress uniformly like it does in a static-compression scenario.

One way of at least somewhat dealing with that, is to have the spring nearly approach coil-bind (within a few millimeters) at maximum cam lift. There are others. But it can't be ignored.

The rotation of the crankshaft is not uniform. The rotation of the camshaft is not uniform. The belt/chain that connects them acts like a spring/damper.

This is not a simple matter.

No doubt you can get a ballpark number through motion profile analysis, but the factor that has to be applied on top of that is the thing that you don't know.

 

[BobSP]

So basically it is trial and error?

 

[tbuelna]

So basically it is trial and error?

It's only a process of trial and error if you lack the extensive technical knowledge and analytical capabilities to do a thorough job of designing the valve spring. As BrianPetersen noted, designing a valve spring for a high-performance application is no simple matter. The most difficult thing to get right with a metal valve spring is usually its fatigue life performance. But I don't think you were considering fatigue life when you asked how "strong" a valve spring needs to be for a given application.

 

[Panther140]

Some things are better left to trial and error. if you aren't designing it from scratch, do some basic math and estimations to get you close, then test them. Even the people with the analytical skill to design this kind of thing always test them. The thing about physics is nature understands it better than humans do and it never misses a variable. There is a point of diminishing returns here, especially if dont have 100% of the data on 100% of the variables in this application I would look at how far the valve moves and how long it takes to move that far. Then look at the mass of the valvtrain's moving parts actuated by the cam. This is all going to operate at half of crankshaft RPM

 

[EHudson]

The mass of the valve, spring and retainer are not the whole mass to consider. Any lifter and pushrod must be returned to the closed position and rockers must be rotated about their respective axes.

Another complicating aspect is the resonance of the spring-mass system. Taylor mentions making the resonant frequency about 13X the operating speed if memory serves.

Yes you can model the entire system mathematically but without accounting for the flexibility of the parts things might not match up with real world results.

Look at it this way: If you are going to take the manufacturer's word that cam XYZ is the right one for you, why would you try to figure out a way around said manufacturer's spring recommendation?

 

[tbuelna]

I agree with EHudson, I would have no hesitation following recommendations for matching components (like valve springs)from reputable aftermarket suppliers of camshafts.

But if BobSP wants to use this situation to learn about valve spring design, then he has a couple options:

1) For the cheapest approach, he can start by getting a couple good technical references on valvetrain & cam design and become familiar with the basic principles involved. I would recommend "Introduction to Engine Valvetrains" by Yushu Wang and "Cam Design Handbook" by Robert Norton. Using what he learns from the text books, he can create an analysis spreadsheet to evaluate his proposed valve spring design. After settling on a suitable design he can check it against the spring recommended by the camshaft supplier.

2) A slightly more costly approach, but one that is likely more accurate and far less time consuming, would be to purchase a software tool like Engine Analyzer Pro w/ Cam Profile add-on. You'll need to determine the cam profile specifics which is best done with a digital analyzer. You'll also need to figure out a baseline spring design to use as a starting point. After that you can iterate the spring characteristics to optimize the design. The cost of the software needed is probably less than $1K.

3) If you want to do a very accurate analysis of your spring, you can use a multi body dynamic simulation application such as this one designed specifically for your task. The software is pricey to purchase, but you could also pay someone that has the software or similar capabilities to do the modelling work for you. I'm sure they would be more than happy to take some time to answer any questions you have about the results.