Care &Maintenance: Fatigue failure
Page 38
If you've noticed an error in this article please click here to report it so we can fix it.
In order to recognise metal fatigue staff must understand the cause
FAILURES are often blamed on drivers' abuse of the vehicles, but workshop staff can be equally guilty.
A careless mechanic may forget to refill the radiator with coolant, or a gearbox with oil, with very expensive results, but an apparently much less serious omission in the workshop can lead to a fatigue failure.
Most faults that occur in vehicles approach fairly gradually, but fatigue failure is sudden and can be dangerous.
At every inspection, the mechanic looks for signs of imminent failure, most of which result from wear, of bushes, brake linings, universal joints and gears.
Then, there are the fairly sudden failures of hoses, which may perish or burst, and this can be a form of fatigue, and then we come to metal fatigue.
If you pull on the ends of a piece of metal firmly enough, say on a bar of steel, eventually it will break. The stress at which the material breaks is called the ultimate tensile stress, and is the basic indication of the strength of the material.
An interesting point about this type of failure is that the material elongates so that the section is reduced in the region of the break just before failure.
On the other hand, if you strike the bar on the side so hard that the end is cropped off, the break is clean.
Vehicles are designed so that in service the components are never stressed highly enough to break in this manner, so the only time you are likely to see such a failure outside the laboratory is after an accident If a half-shaft shears in service, or if a spring leaf breaks, it is almost certainly a fatigue failure.
There is nothing mysterious about the cause of fatigue failures, although it is difficult to predict when they will occur, especially on vehicles, where the stress levels can vary so widely. For a fatigue failure to occur, the part must be subject to an alternating stress -in other words, the stress level may be alternated continuously from 75 to 45 Mf\lisq m (5 to 3 ton /sq ink.
If the alternating stress is constantly varied between two levels, it is fairly easy to predict the life of the component, and in fact, with steel once the alternating stress levels are known, the component can be designed so that it will last for ever, for all practical purposes: In one typical case, for example, if the range of stress was 280 MN /sq m (18 ton sq in), the component would fail after the load had been applied 10,000 times; if the stress were reduced to 200MN sq m (13 ton / sq in), the component would last indefinitely.
You can, of course, test components by applying varying stresses, and this is what vehicle manufacturers do when they are developing special components such as axle shafts.
But, even then, you can only estimate the loads that will be applied -you don't know how often the driver will drop the clutch with a bang when tne engine is developing maximum torque, and so some sort of guestimation is needed in setting up the test programme.
So what has all that got to do with workshop staff? Well, to start with you need to under
stand what causes a fatigue failure so that you can recognise the failure when you see it, and not just assume that whenever anything breaks the driver has just been so hamfisted that he has broken it off like a carrot.
Fatigue failures start from a small crack in the surface of the metal, and then the crack spreads right through the material. In some cases, the crack can spread so quickly that it will not be visible, even a short time before the failure occurs.
Whereas a break resulting from an ultimate tensile stress failure is clean across the whole surface of the break, with a fatigue failure, shell-like markings tend to radiate from the place where the crack started, while there are no signs of any elongation, or stretching of the material.
On exposed vehicle components, there is likely to be a little rust at the origin of the crack.
One reason why manufacturers spend so much time testing components that are likely to fatigue is that the safe stress depends largely on the shape of the component.
For example, the best shape is a plain bar with parallel sides, but anything with changes in, section, such as a shaft with a shoulder on it, will fatigue much more quickly.
The reason for this is that , any sharp edge or change in section acts as a 'stress raiser', so reducing the fatigue life of the component.
A good example of efforts being made to reduce the stress raisers is in a typical half-shaft, where the change in section between the flange and shaft is a large radius — not a sharp corner.
Every so often, a careless action by someone in the factory results in there being an extra stress raiser on a part, usually in the form of a small nick or scratch in the surface.
Since the failure starts with a crack, if there is a crack there already, or even a scratch, failure will be that much quicker.
Equally, careless handling in the workshop can cause the same stress raiser to appear. Even the marks from where a part has been held in the hardened jaws of a vice can be enough to start a fatigue failure, although scratches from filing, or grooves and chisel marks are more likely to cause trouble.
Great care is needed when handling any of the highly stressed parts of the vehicle, especially those that affect safety, such as the steering arms and drop arms.
Connecting rods, crankshafts, gudgeon pins, valve gear and all shafts come into tnis category.
Therefore, at all times, soft plates should be fitted to the jaws of the vice before any one of rrieSe components it inserted.
Care is also needed in filing off faces, or when trying to remove ball-joints from the steering arms.
In other words, remember that if you want to avoid that dreaded fatigue, make sure that there is no chance of the surface being scratched or nicked.
If a scratch is detected, it can be polished out, but care is needed to ensure that the surface really is polished, and mat it is not just a mass of fine scratcnes.