Lighter valve train

[digger]

imagine a case where you have adequate force "over the nose" but the spring is installed without any preload and therefore the force on the seat is effectively zero. How well do you think that will work for preventing bounce?

The parameters that affect valve spring selection and tuning are:
spring free length
spring installed height
spring rate
valve lift & acceleration

If installed height is equal to free length, then seat force will be zero. But a nose force will still
be present resulting from spring compression due to valve lift. Whether it will be adequate will
depend on valve acceleration and valve train mass.

If installed height is reduced by use of shims, then seat force will be increased and nose force will be
increased by the same amount.

now ask your self what is the theoretical acceleration of the valve as it comes off the closing ramp (assume constant velocity) onto the seat ? assume parts are ideal and perfectly rigid

Valve acceleration numbers are typically as follows:
base circle to ramp = +0.25 in/deg^2
cam flank = +2.0 in/deg^2
over the nose = -1.0 in/deg^2

While the term "nose force" is commonly (and incorrectly) used, it's force is greater than ramp acceleration
but not as demanding as flank acceleration. Maximum spring force to control valve float and consequent
valve bounce is necessary during flank acceleration. Unfortunately, spring forces produced by spring rate
and valve lift are sometimes inadequate to control valve launch. Increasing the seat spring force, and the
nose force by reducing the installed height is a more tunable approach.

Since maximum valve spring force needed to control valve launch occurs about 30 degs after/before the
ramp, higher seat forces are needed which may also result in excessive nose force.

Acceleration as valve is placed onto the seat is infinite theoretically due to the discontinuity of the profile (two lines meeting at an angle) . In reality it’s not due to stiffness of material effectively smoothing out the extreme peak. This is why the seat preload matters as you can’t effectively throw the valve closed at the closing velocity without something to control it.

[David Redszus]

imagine a case where you have adequate force "over the nose" but the spring is installed without any preload and therefore the force on the seat is effectively zero. How well do you think that will work for preventing bounce?

The parameters that affect valve spring selection and tuning are:
spring free length
spring installed height
spring rate
valve lift & acceleration

If installed height is equal to free length, then seat force will be zero. But a nose force will still
be present resulting from spring compression due to valve lift. Whether it will be adequate will
depend on valve acceleration and valve train mass.

If installed height is reduced by use of shims, then seat force will be increased and nose force will be
increased by the same amount.

now ask your self what is the theoretical acceleration of the valve as it comes off the closing ramp (assume constant velocity) onto the seat ? assume parts are ideal and perfectly rigid

Valve acceleration numbers are typically as follows:
base circle to ramp = +0.25 in/deg^2
cam flank = +2.0 in/deg^2
over the nose = -1.0 in/deg^2

While the term "nose force" is commonly (and incorrectly) used, it's force is greater than ramp acceleration
but not as demanding as flank acceleration. Maximum spring force to control valve float and consequent
valve bounce is necessary during flank acceleration. Unfortunately, spring forces produced by spring rate
and valve lift are sometimes inadequate to control valve launch. Increasing the seat spring force, and the
nose force by reducing the installed height is a more tunable approach.

Since maximum valve spring force needed to control valve launch occurs about 30 degs after/before the
ramp, higher seat forces are needed which may also result in excessive nose force.

Acceleration as valve is placed onto the seat is infinite theoretically due to the discontinuity of the profile (two lines meeting at an angle) . In reality it’s not due to stiffness of material effectively smoothing out the extreme peak. This is why the seat preload matters as you can’t effectively throw the valve closed at the closing velocity without something to control it.

We have no infinite forces at work in any part of an engine. Every metal part has some degree of flexibility which results in damping forces to contain movement. Parts that do not flex will quickly break.

Preload is important not just to help close the valve, but also to exert sufficient seat PRESSURE (spring force/seat contact area) to seal properly.

But even with no preload (zero seat force) any slight amount of valve lift will produce a closing force, small as it may be.
As an example: using a spring rate of 150 lb/in, no preload (zero seat force), 5mm of valve lift will produce a closing force of 29.5 lbs. Not much but perhaps sufficient in a light weight, slow running engine.

That is not to say that pre-load and seat force are not very important; they most certainly are. Reviewing data from
a BMW camshaft, while there is a G force spike at the base circle to ramp junction, it is small and of short duration.
But when we are on the flank of the cam, the acceleration force is more than double at a valve lift of only 0.055".
Across the nose, the negative force is about half of the opening/closing forces. What is the spring open force
at 0.055" of lift? Not much.

Spring rate = 100 lb/in ..*Free length - installed height (mm)
..................open spring force (lbs)
Inst *............0......-4.......-8
Lift.......2......8......24......39
...........4.....16......41......47
...........8.....31......47......63

Spring rate = 200 lb/in ..*Free length - installed height (mm)
..................open spring force (lbs)
Inst *............0......-4.......-8
Lift.......2.....15......46......77
...........4.....31......62......93
...........8.....62......93.....124

We would really like higher seat forces to assist in sealing and to control cam contact without excessive nose spring force which has little benefit. But since spring force increases with lift, we have few options.

Your point regarding stiffness of materials is very important. In a valve train, every component will flex with load.
But the least stiff component is the valve spring. If the spring natural frequency approaches the forcing frequency, the
result is a large amplification of force due to harmonics. It will break everything. The answer is to use very light but very stiff components, whenever possible.