Gears are among the most important automotive elements to be shot peened. This is due to the fact that this treatment allows weight and volume reduction while increasing the performance of the car. Gears are mainly subjected to two kinds of damage, bending fatigue at the tooth root and surface or sub-surface damage on the flank, due to contact fatigue.
In recent years, many studies have been devoted to bending fatigue strength at the tooth root; the analyses are generally both numerical and experimental. Concerning numerical analysis, they are mostly aimed at determining the stress profile at the root of the fillet of the gear tooth, to assess the point where the stress is maximum and the in-depth stress gradient under the most critical load condition (the outer point of contact when there is only a couple of gear teeth engaging).
As regards the experimental tests for bending fatigue investigations, it isn’t usual to respect the real working conditions of gears, but it is common to develop tests where only bending fatigue damage is considered. It is important to mention that in this case gears are fixed on a general purpose uniaxial fatigue test machine while dedicated devices are designed and manufactured to load the gear tooth by respecting the critical position of the load with respect of the bending fatigue damage. The main problem is the design of the device for the load application: in fact, this latter should have a geometry able to correctly represent the direction of the load. Besides, the contact with the gear should be correctly lubricated: friction changes the direction and the value of the resultant force and, finally, disturbs the results of the tests.
In this way we are able to assess only the behaviour of the gear under bending fatigue loading without any possible influence of damage caused by other failure modes.
The second kind of damage of gears is surface deterioration due to the cyclic contact. Due to the geometry of the gears, the load is transmitted through a very limited contact area. Thus, very high contact pressure distributions and surface/subsurface stress take place and, due to their cyclic repetition, cause the development of microscopic surface cracks that propagate, causing the removal of a small portion of material that makes the gear unusable. This is the so-called pitting damage, probably the most common cause of failure of gearboxes. Another kind of damage related to contact fatigue is the so called "spalling". In this case the crack develops internally and propagates toward the surface: also in this case, the final result is the removal of small parts of materials that cause the substitution of the gear.
The effect of shot peening on contact fatigue damage concerns mainly pitting. In fact, the residual stresses induced by shot peening are able to delay or to stop the propagation of pitting cracks, thus increasing the torque that the gear is able to transmit. This effect is less evident for spalling, since the crack starts at a depth where the effect of shot peening is less evident and when the crack approaches the layer of material affected by shot peening, it is too long to be stopped (I am simplifying the phenomenon but this gives the idea of what happens).
Anyway, in the case of contact fatigue, it is important that the experimental tests used to study the behaviour of gears reproduce the real in-service contact conditions well. Otherwise, the results cannot correctly describe the development of the process. It is possible to perform contact fatigue tests by using rotating rolls with a prescribed relative sliding. But what happens during gear mating, where sliding changes during a contact cycle, is different! And even if the same materials and the contact pressure distribution are considered in the tests, the results will be different from the one you can get by using full-scale gears operating with conditions similar to the effective ones (i.e.: at FZG in Munich it was shown that if you use the wrong experimental test method, you could demonstrate that beer or milk are the best lubricants for gears, but it isn’t!).
So it is common to use dedicated machines, most times back-to-back machines, where only the mechanical losses are provided by the electric motor.
So, if you don’t want to underestimate or to overestimate the effect of shot peening on gears, it is important to dedicate money and time to reliable and accurate experimental tests with the right devices. Otherwise you risk wasting time and money.

Shot Peening in the Automotive Industry
by Mario Guagliano
Contributing Editor MFN and
Associate Professor of Technical University of Milan
20156 Milan, Italy
E-mail: mario@mfn.li

Author: Mario Guagliano