Investigation of surging in helical valve-springs

Abstract

This investigation is concerned with surging in helical valve-springs in internal combustion engines. The main object of the work was to observe and measure surging in an actual engine, benchmounted, whilst the cam-shaft was motored.

A considerable amount of development work was carried out in order to estabAsh a reliable method for measuring the amplitude of surging and the cam-shaft speed at which it occurred. The method ultimately used was to pull .a smoked-glass slide through a small rig so that a trace was made by a small pointer attached to the central coil of the valve-spring. A second pointer vibrating at a constant known frequency 1 left a further trace from which the cam-shaft rotational speed could accurately be found.

Surging became increasingly pronounced as the cam-shaft speed increased and~ response curve was drawn -showing surging amplitude plotted against cam-shaft speed. Harmonic analysis showed how the amplitude of surging is roughly proportional to the amplitude of the exciting harmonic.

Here, as in all practical vibrating systems the damping factor controls the amplitude of surging for a given cam-valve-spring configuration.

The magnitude Qf the damping factor was obtained :from the peaks of the response curve. Because of its relevance to this phenomenon damping was further investigated in a separate experiment, in an attempt to discover how it was affected by the precompression of the spring.

The results showed that the damping was higher at large and small precompressions than at intermediate precompressions. It was apparent that no advantages would be obtained by increasing the precompression of the spring within the engine. However early traces taken from the engine with increased spring precompression,(use having been made of the tappet-adjusting screw,) indicated that the damping had increased -for that part of the trace drawn when the valve was closed, showed a considerable dying-away of the surging amplitude.

Further traces showed how, at a different surging speed the amplitude of vibration increased in that part of the cycle when the valve was closed. It was realised that two frequencies causing longitudinal motion were being excited and that beating occurred. Careful replotting of the enlarged traces showed that the two frequencies were separated by about 12 c~p.s.

A further experiment was devised to explain this additional phenomenon. The spring was vibrated at various precompressions and the three natural modes of vibration, torsion, bending and pure longitudinal, were noted. It was apparent from this experiment that during bending vibration longitudinal motion occurred. The results revealed that the original work on the engine had been performed with the longitudinal and bending frequencies virtually coincident and in the later work at increased precompression the two frequencies were separated by 15 c.p.s. Thus careful observation of the form of the vibration throughout a complete cycle of the cam rotation revealed that two frequencies cause longitudinal motion and each contributes to valve-spring surging.

Quite a large proportion of the work has been concerned with drawing out theoretical and actual surging cycles for one complete cam-rotation at various surging speeds. It is shown theoretically how by increasing the damping factor ten-fold surging can be virtually eliminated.