Why Premature Failures Occur
Page 44
If you've noticed an error in this article please click here to report it so we can fix it.
How Albion Investigate . Mechanical Failures: Tests Reveal Reasons for • Breakages and Excessive Wear in Certain Components
SUCCESSFUL investigations of premature mechanical failure in commercial vehicles were described by Mr. G. H. Lee, B.A., A.M.I.Mech.E., manager of the experimental department of Albion Motors, Ltd., in a paper
read . before an Institution of Mechanical Engineers meeting, at Leeds University.. .
.Mr. Lee said that it would be incorrect to assume that troubleshooting was a major activity of a research and development department. To keep up with competition in home anq fOreigri markets, there was• much more to be done than curing troubles with existing vehicles and no manufacturer. could afford merely to improve his present designs.
Mr. Lee said that if a case of mechanical trouble with a vehicle was referred to Albion Motors, Ltd., for investigation, the record cards con cerned were first consulted. A card was kept in respect of every vehicle manufactured, and if the complaint were not isolated, it was possible to see whether the trouble was spread or concentrated. • Record cards were • also maintained for each type of vehicle so that if there were several cases of piston breakage in which one particular model was involved, they would come together on the same card. The number of entries determined the action to be taken—for example, three steering failures on one mo01"would require immediate investigation.
Wrong Information In considering a complaint it was often difficult to locate the real cause of the trouble or discover whether it even existed. The investigation of a problem had often started with wrong information, because drivers were seldom reliable witnesses.
They knew something had broken, and their version of the events leading up to this event might be coloured by their own diagnosis of the cause. In other cases, the user would admit that the vehicle had been grossly overloaded or the engine overspeeded.
Indicating the wide selection of scientific instruments for investigations involving stress, vibration, temperature or pressure, Mr. Lee remarked that where it was necessary to use indirect means for measurement embodying fairly complicated electrical circuits, there were many possible sources of error. It was most important to have a reliable method of calibration, whereby a check could be kept throughout the tests.
As a general principle he gave this advice: "never use a thermocouple where a mercury thermometer can be employed, or an electronic instrument where a trace on a moving strip of paper will give all the information which is required."
The latter method figured in an
investigation which concerned breakage of oil-engine exhaust-valve tappets. The trouble was not widespread, but where breakage :did. occur, the ultimate damage could be. serious.
In ascertaining why the tappets broke, traces to indicate valve motion were obtained, using a strip of cardboard over which was folded a length of paper with carbon paper underneath it, in conjunction with a• pip brazed on to a valve cap on an engine having the camshaft •electrically driven. . Sliding the card past the pip enabled the:valve to trace its own motion accurately on the paper. • . Valve Bounce
From the resultant traces at variouscamshaft speeds,. it -could be .se.eii:that valve bounce. started:at .3,100' tp.m., of the crankshaft, and that it was severe at 3,400 r.p.ra. -at which point the tappets 'had failed in service.
. A section of the, latter trace was scaled and plotted against the true valve motion and the path of the piston.over the top of its stroke. This comparison showed that when the exhaust valve bounced the piston had moved out of the way, but the tappet did not follow the camshaft form and made no contact until the valve was hit by the rising piston.
The inlet valve was hit by the piston towards the end of its first bounce. Evidence of this was shown by a heavy marking of the pocket in the piston crown. The tappet did not break because it took the blow at its centre, on the base circle of the cam, whereas with the exhaust tappet the blow was taken off centre. This problem of exhaust-tappet breakage was solved by redesigning a valve spring.
Engines Overspeeded This showed that engines were severely overspeeded in some types of operation. In the early days of oilengined vehicles it was considered that designing for 10 per cent. excess speed would be adequate, but it now appeared that an excess of over 50 per cent, was required.
It was found that the skew gears driving an oil pump wore rapidly when both were made of case-hardened steel. The material of the driven gear on the pump was altered to a bronze, but this was then found to suffer from pitting. The materials of the two gears were then interchanged, with greatly improved 'results.
The reason was that bronze was the better material for the gear which had the greater length of contact and had, therefore, less chance to develop local heat, whilst the steel was better for the short contact length. In effect, the highly stressed point of contact travel was slower.
Reports were received of connectingrod big-end failure because of bolt stretch in the fixing of the big-end cap on the connecting rod of a fourcylindered oil engine. The defect occurred after an engine overhaul, and it was immediately suspected that the bolts had been over-tightened. The specified torque for these bolts was 400 lb.-in., but checks made indicated that this could be exceeded without difficulty. Therefore, as a torque spanner was not universal equipment, it was considered advisable to raise the safety margin by increasing the bolt diameter.
A very different stress on the bolt could be produced for an indicated torque according to .whether the bolt thread was dry, oiled or graPhited. Much more stress was imposed if a thread was graphited instead' of oiled, consequently it was dangerous to use „graphite when working' to a specified torque.
, Drive Overloaded.
Mr. Lee mentioned that under certain conditions of operation, the rear-axle worm drive could be subjected to heavier stresses than it was originally designed to carry, resulting in heavy pitting of the bronze wheel. Failures of this type were usually associated with tipping vehicles, which might spend a large proportion of their lives in crawling out of difficult places in bottom gear and well overloaded.
High tooth loading caused by this kind of work resulted in a tendency for the top surface of the bronze teeth to be pushed along by the advancing contact of the steel worm, producing fatigue failure of the bronze beneath the surface, the primary reason for the deep pitting. Pitting did not start on the bronze outer surface, because this had become partially work-hardened. This type of wear was quite different from abrasive wear, which occurred under more lightly loaded conditions, and could proceed safely until the bronze teeth were worn thin.
Under conditions of very light load, a mileage of 300,000 might be put up before the bronze wheel was worn out, with only light pitting during that time. Under slightly heavier loading, this life might be halved and heavier pitting observed; and under still heavier loading the mileage might be reduced to 50,000, failure being caused by heavy pitting. Under the heaviest loading conditions, worm wheels had been known to fail after 20,000 miles, the pitting being accompanied by severe scuffing of the surface, which resulted in high temperatures because of friction losses.