VALVE DESIGN

VALVE DESIGN

THERE is matter of considerable moment to racing-car engineers in the paper “Valve and Valve Seat Technique for Automobile and Aero Engines” which F. R. Banks, of The Ethyl Export Corporation, published in the December journal of the Institution of Automobile Engineers. Mr. Banks said that it is no longer possipossible to get away with valves made from ble per cent. nickel-steel” or “bumped up iron railing.” Even in these enlightened times there are still many examples of poor exhaust valve design and that of its surroundings in the cylinder block. A hot-running exhaust valve—one having a head temperature in the region of 800° C. or higher—will have a short life and may limit engine performance. The valve and its operating mechanism should require little or no attention between decarbonising periods-20,000 to 30,000 miles, or 350 to 500 hours in the case of

high-duty aero engines. The valve material should have high strength at elevated temperatures, be resistant to scaling and hot and cold corrosion attack, and possess a certain degree of stem hardness. Improper seating is caused by extending the guide into the port and counterboring the boss, by leaving too much material between the seat and the water jacket and by distortion of the seating

in the block. Some restriction of the exhaust port is unavoidable but the exhaust side can put up with a certain degree of throttling without serious power loss and the greater guide area will result in a cooler valve. The port may be spread laterally (looking at the port end) making it more or lesa oval in plan from the bossing to the outlet, so retrieving a certain amount of area, but water passages must not be restricted around this end of the port. In small s.v. engines greater seat distortion is usually encountered, and, in conjunction with small stiff valves, trouble can arise because of faulty contact. Because the valve position the hottest part of the cylinder is not the highest and if the water boils locally round the exhaust seating it is more difficult to get rid of steam pockets than with an o.h.v. design. In many cases, bolting down an s.v. head stresses the block to such an extent that distortion of the valve-seatings occurs, and in order to lap in the valves a dummy head has to be first bolted in place. Turning to valve-shapes, the fiatheaded valve is cheap to produce and universal. The dished or modified tulip type is highly satisfactory, with plenty of material backing the seat to prevent overheating. To-day’s materials have destroyed the advantages the pure tulip valve once had for the exhaust side but it is sometimes employed as an inlet valve

in conjunction with fancy port shapes. The neck has a large exposed surface area and therefore it usually runs hotter and feeds back heat to the seat. The mushroom headed valve is used mostly in lorry engines. It is probably somewhat heavier for a given diameter and perhaps stiffer mechanically, which may not necessarily assist in good sealing, as it will not easily conform to seat distortion. It offers a relatively large surface for heat collection.

Blind holes or slots in the valve head are to be deprecated as they tend to cause deformation of the head and a fruitful source from which cracks commence. S.v. valves can be ground in with the aid of a tool possessed of a rubber sucker.

The author considers that the stem should be of as large a diameter as possible, because, in spite of weight being undesirable, considerable heat can be removed via the stem and guide, as well as by the seat. Considerable importance is attached to the proportioning of the stem and neck to the head. The author favours the 45° seat for the exhaust valve as it is easier to maintain a good seal than in the case of the

30° seat. The 30° seat gives a better valve opening coefficient, but this is of greater importance for the inlet valve. In the case of very narrow seats. the use of 45′ or 30° is an extremely moot point. Seat widths vary fairly widely. For instance, in one ease for a valve diameter of 1.5″ the width is approx. 2 mm. and in another approx. 3 mm. It is often believed that a wide seat improves heat

dissipation. In practice this is not necessarily the case, because unit loading exerted by the valve spring will be lower and also there is more chance for particles Of carbon to become trapped between valve and seat. If a soft seating material is used a wider seat than usual may be needed to prevent seat shrinkage.

In connection with seat wear and shrinkage, the closing velocity of the valve on its seat is in the region of 1.5 to 2 ft. per see. and this holds good for a wide range of engines of various designs. It is important to maintain correct tappet clearances, otherwise enlarged clearances will increase the dosing velocity and may increase wear. However, this is better than too small a clearance, which is often demanded by the slaves of silence and is the beat way of quickly ruining the valve.

The valve stem Should be as hard as possible and have a good Surface finish. The clearance between stern and guide should be as small as possible as not only does this assist in removal of heat, but modern experience shows small initial stem and guide clearance results in lower rate of wear between the two. The author recommends inserted valve seats, but says that for engines having capacities of 750 c.c. to 2,000 c.c. it is probably impracticable to fit inserts. Alloy irons, heat-treated high-speed steels and Silcrome No. 1 make good insert materials. Probably the best arrangement is a steel ring with a seating of Stellite No. 6. Various methods of fitting inserts are satisfactory but the author prefers the plain ring insert. There is no reason why a steel insert should not be satisfactory for an aluminium car head—aero engine designers have discarded aluminium-bronze and Monel metal in favoui of steel. N.M.C. or D.T.D. 49b and Silcrome No. 1 are satisfactory. The modern car valve has a life of 10,000 to 15,000 miles and the aero engine valve a life equivalent to 45,000 to 50,000 miles, before overhaul. There are nine-cylinder two-valve radial aero motors in service developing about 1,100 b.h.p. or about 120 b.p.h. per valve. Naturally, they operate only for a few minutes per flight at these figures and the cruising power is about half these values. Whereas the average b.m.e.p. of a 2-litre car on a run of 100 miles will be about :30-40 lb. per square in., reaching 80 to 00 lb. for short periods, the aero engine operates continuously at about 110 to 140 lb. per square in The author here makes the very interesting point that in low-duty car engines valve trouble can sometimes be traced to the valves running so cool that they are not plastic enough to conform to local seat distattion and so burn at the point of leakage. He remarks that British aero-motor designers only fill the valve stem with sodium, allowing the heads to run hotter than those in American engines. This is because we use four valves in big cylinders and small valves remain too stiff if the heads are cooled. It is fairly certain that most valves in British aero engines have some degree of plasticity. In tests in which stem-filled valves were replaced by hollow-headed valves in which sodium came within 2 to 3 mm. of the seats, just as bad and in some cases worse burning occurred—in an engine known to suffer bad head and insert distortion. Rolls-Royce use Brigh tray seats and heads to defeat

corrosion. The author believes valve cooling by sodium to be unnecessary and unduly costly for car work, save, perhaps, for special sports and racing engines. He does not think the poppet valve will be banished from car engines for many years to come, but recalls Argyll’s prewar and Arrol-Aster’s post-war attempts to develop the Burt single sleeve valve [what of Vauxhall?—Ed ). The valvespring does not always receive proper attention. Spring ends must be flat, to obviate side thrust.