Avoiding damaging spring vibration
In race engines that contain them, the valve spring is the most highly stressed component. In the best springs, the materials, heat treatments, hardening processes, manufacturing techniques and surface treatments combine to allow them to operate at incredible levels of stress. If they are of the best quality, correctly installed and operating within calculated limits then valve springs are very reliable; they fail because they are either over-stressed due to driver error (over-revs on down changes are normally the reason for this) or spring resonance.
When excited at certain frequencies, the spring will vibrate at one of its natural frequencies, and in such situations the stresses in the spring may be far higher than those in which we intend the spring to operate. It is therefore imperative that engine designers consider techniques and methods to reduce the tendency of the spring to operate in this resonant condition, which is often known as spring surge.
The first thing that should be done is to ensure that none of the major harmonics of the cam profile coincide with the natural frequencies of the spring. A cam profile is not a simple sinusoidal wave, and can be analysed to show the strength of a number of harmonics. Basically, the profile is made up of sine waves of frequencies that are integer (whole-number) multiples of the actual frequency of valve opening. The strongest of these are the low-number harmonics, so we need to avoid these if possible.
The exact number that needs to be avoided is not universally accepted, and depends on the experience of the engine designer. I’ve heard of people avoiding as few as the first five harmonics, and others who choose eight or more. This essentially means that the natural frequency of the spring and mass system (the mass being that of the components, the action of which the spring controls) needs to be N times the basic excitation frequency. So, if the engine revs to 18,000 rpm and the cam turns at 9000 rpm, which equates to 150 opening and closing events per second, we need the natural frequency to be at least 150 N. This is the strongest fundamental defence against surge.
Where two or more springs are used to control each valve then we can have them specified and supplied with a light interference fit. It is usual to specify the springs with different natural frequencies so that, if surge should occur, only one spring will resonate at a time. The damping action due to friction between the resonating spring and its partner reduces the damage done by converting some energy to heat and, where space allows, some spring suppliers also supply special dampers that fit between pairs of nested springs.
Progressively wound springs, which have tighter coils at one end, are also effective. As the spring compresses, the most tightly wound spring coils come into contact, changing the effective number of coils and increasing the spring rate, which is a measure of the spring’s stiffness. The natural frequency of the spring-mass system is proportional to the square root of the spring stiffness, and where the change in stiffness throughout the opening event is large enough, the fact that the excitation no longer coincides with the natural frequency of the spring effectively controls vibration.
Where a single spring is used, beehive or tapered springs are also another way of providing a spring with a changing natural frequency throughout the lift curve, achieving the same effect as a conventional progressively wound spring.
Written by Wayne Ward