Analysis of valve cam lift curves comparing high and low RPM performance, showing deflection and discussing resulting valve failures.
| Identifier | ExFiles\Box 158\5\ scan0020 | |
| Date | 1st March 1939 | |
| 110 S.A.E. JOURNAL (Transactions) Vol. 44, No. 3 Graph 1: [A line graph with 'Lift in Thousandths' on the y-axis from 0 to 400 and 'Camshaft Deg' on the x-axis from 100 to 320. The title of the graph is 'Lift Curve of High Acceleration Cam'. There are two curves, a solid line labelled '500 rpm' and a dashed line labelled '2050 rpm'. Both curves start at 0 lift around 95 degrees, peak at approximately 380 thousandths around 45 degrees, and return to 0 lift around 340 degrees. The dashed (2050 rpm) curve shows slightly more lift and deflection, particularly on the closing side.] Fig. 1 — The dashed curve shows the deflection that takes place with a cam of high acceleration when run at high speed when compared with the lift at comparatively low speed. This curve very forcibly brings out the reduction in deflection at the closing point. Fig. 2 is an enlarged scale section of the closing event, and it will be noted that the valve contacts the seat at, or very close to, the desired velocity, which is at the rate of 0.0005 in. per deg of cam travel, or, at 2400 rpm, with a velocity of 0.3 fps as compared to 3 fps obtained with the higher-acceleration cam shown on Fig. 2. Fig. 5 shows an enlarged section of the closing side of both types of cam, and it will be noted that the high acceleration deflects more at 2400 rpm whereas there is very little change in the low-acceleration type. All of the foregoing data were taken with the valve actually contacting the seat. Fig. 6 represents the closing part of the lift curve of another type of overhead valve, the data being taken with negative clearance so that the valve could not contact the seat, and it will be noted here that the curves follow the same general tendency, indicating that it is not all rebound due to the valve striking the seat but due to reaction of the parts. These curves are not shown with the primary thought of developing methods of overcoming false motion, but simply to illustrate the action that actually occurs under operating conditions and to which we believe can be attributed many of the failures that, at first glance, appear to be purely design or material failures of the valve proper. In addition to the original consideration given at the time of the layout and design study, there are always the variables that creep in during manufacture and operation. In all of the tests made and covered by the lift diagrams, the cams were made very close to the desired figures. There is always the possibility of the acceleration being increased due to incorrect grinding since the velocity varies as the difference in lift, and the acceleration (being the second derivative of the lift) will vary as the difference of the velocity per degree; and, consequently, a slight error in grinding, if the accelerations are already on the high side, will exaggerate the deflections shown. A certain type of engine had been in service for some time with little or no difficulty when reports were received that the valve steel of later engines was inferior to the original. With very little knowledge of the complete design, the valve manufacturer was given the problem of solving the difficulty by attempting to provide better material and heat-treatment, although nothing could be found wrong with the original stock. The strength in this particular part was increased by a redesigned heavier section, but without overcoming the failures. The situation was becoming rather serious when careful investigation disclosed the fact that slight modifications had been made in the engine proper which permitted somewhat higher operating speeds and, under these conditions, sufficient false motion was developed in the gear to stress the valve stem severely, resulting in a major failure. Attempts made to study the effect of seating velocity with respect to the valve and cylinder-seat life are oftentimes made extremely difficult due to the many variables entering into this problem. Tests made by C. G. {Mr Griffiths - Chief Accountant / Mr Gnapp} Williams of the Institution of Automobile Engineers¹ indicated that, by doubling the seating velocity (from 1.2 to 2.4 fps), the rate of seat sinkage increased on the order of 8:1. These tests were made under Graph 2: [A line graph with 'Lift in Thousandths' on the y-axis from 0 to 90 and 'Camshaft Deg' on the x-axis from 354 to 326. The title of the graph is 'Closing Side of High Acceleration Cam'. There are two curves, a solid line labelled '500 rpm' and a dashed line labelled '2050 rpm'. Both curves show the valve lift decreasing as the camshaft degree decreases, illustrating the closing event. The dashed (2050 rpm) curve shows significant deviation and bouncing ('false motion') compared to the smoother solid (500 rpm) curve.] Fig. 2 — The extent to which false motion is imparted to the valve with the high-acceleration cam of Fig. 1 is illustrated in this enlarged view of the closing side. ¹See Journal of the Institution of Automobile Engineers, April, 1937, pp. 28-41: 'Factors Influencing Wear of Valve Seats,' by C. G. {Mr Griffiths - Chief Accountant / Mr Gnapp} Williams. | ||
From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).