Valve Spring Truth?

[David Redszus]

The purpose of a valve spring is to make sure the valve follows the path determined by the cam, during acceleration and decceleration.

The rate of a valve spring is constant (unless we start to bind coils) and does not change with engine speed. The resistant force provided by the spring is the product of spring rate and valve displacement. At low lift, the spring force is low, at high lift, spring force is high.

The forces acting upon the valve are the product of acceleration (provided by camshaft design) and valve train mass. Valve acceleration will increase with engine speed. Spring resistance does not change with speed. The faster we turn the engine, the greater spring force resistance is needed. But it can't provide it.

Now ask, at what cam angle is valve acceleration highest? Max valve acceleration does not occur at the nose. Max valve acceleration will occur as soon as the follower leaves the ramp and rides onto the flank of the cam. So is there a problem? You bet there is.

Maximum valve acceleration occurs when the spring force resistance is quite low and the valve barely off of the seat. The same events occur upon closing, but in reverse order. So in order to control the valve motion when subject to high acceleration, it becomes necessary to use a higher rate spring, even though the spring rate across the nose is now excessive.

What we need is a spring with high resistive force at low lift which becomes softer as more lift is added. Not yet invented. All we can do is run the softest spring we can without valve float or bounce at higher speeds.

The proper solution is to lighten the entire valve train as much as possible, make each component as stiff as possible, and hope that our little old cam grinder knew what he was doing. Because the answer to our problem lies in the shape of the cam lobe, not lift or duration, but shape. Lobe shapes are defined mathematically; each lobe has its own formula that predicts its behavior in the engine.

You can write your own predictive model. Simply multiply your valve train mass by the valve acceleration which will result in valve force. Now subtract the spring force (at that lift point). If the spring force is greater than the valve force, the valve will follow the cam. If the valve force is greater than the spring force....Houston we have lift off. Airborne, all the way.

There are programs that will perform all the necessary calculations and any competent cam grinder has one available (or should have).

[David Redszus]

Sorry guys, but I just need to vent. No offense to anyone here on the board, but "once an engineer always an engineer".

"Load" and "pressure" are not interchangeable terms. Valve springs exert loads on the retainer (and of course on the spring seat). Loads are measured in units of force.

Pressure is load divided by area as in pounds per square inch (PSI). This is useful in calculating stress in the spring wire, but doesn't really apply to the loads exerted by the spring on the valve.

Along that very same line, we engineers often see the terms inertia used in place of momentum.

Inertia is derived from "inert", without force or motion. It refers to the force necessary to move a stationary mass.

Momentum refers to the energy contained in a moving body. It is the product of mass and velocity.

A closed valve has inertia. A moving column of air has momentum.

[MadBill]

...The rate of a valve spring is constant (unless we start to bind coils)..

But while we're being precise, the above statement excepts progressive wound (although they do undergo a form of successive coil bind) or beehive springs...

[David Redszus]

...The rate of a valve spring is constant (unless we start to bind coils)..

But while we're being precise, the above statement excepts progressive wound (although they do undergo a form of successive coil bind) or beehive springs...

Progressive wound springs (either with changes in coil spacing or changes in coil diameter) will still return an increasing spring rate with displacement if there is coil bind. Otherwise they behave the same as other springs.

[iadr]

So,with the same valve weights etc and the same rates/pressures on springs,a smaller OD spring will hang in there a little bit further than an equvivalent spring of bigger dia??

For sure. there used to be a cool little graph in the back of the Crower Catalog that compared their "little" but high rate spring to some of their (presumably more durable) larger double of similar or slightly more pressure (load ).
Seemed the little spring was often the way to go.

Along those lines, they also offered .049" wall pushrods.
Seems like they came up with a direction that worked in some cases, that has now been more or less phased out. It must have worked in some cases, though, for them to have offered it all.

[Cammer]

Thanks to all for your posts.

We have just scratched the surface of this important subject.

As more members become affluent in Spintron testing and other dynamic testing this subject will find new life!