E-Archive

Shot Peening in the Automotive Industry

in Vol. 21 - November Issue - Year 2020
Shot Peening, Quality And Machine Learning, A Dream?
Mario Guagliano

Mario Guagliano

Here we are for another MFN issue and another column about shot peening in the automotive industry. This time, the idea of the subject of the present column came into my mind after a short talk with an engineer working in one of the main automotive manufacturers in Europe.
Indeed, the subject of our friendly conversation was how to better exploit the benefit of shot peening in cars produced in large volumes. He told me that his opinion about that was not strictly and mainly related to the performance increase we can get by a proper and optimized application of shot peening; this is valid for race cars, but it is not the main objective of cars for everyday life. 
It is true, we can study and optimize the parameters to get the maximum (or almost the maximum) increase of the fatigue strength of a peened part: in this way we can also think to redesign the part to make it lighter. But a complete and balanced redesign of the system means that everything should be reviewed and changed, and this clashes with the multiple constraints of industrial production. In other words, a way to proceed like that is long and cannot be limited to shot peening and its application.
A more immediate and useful way to improve the effectiveness of shot peening and its ability to be a part of the design process is to define appropriate ways to check the quality of shot peening in industrial production. This means looking for efficient ways to control the final results of shot peening, to ensure that the result is not varying over large production sets.
This could lead also to use more challenging shot peening parameters with exciting expected results, thus leading to a gradual improvement of the process and of its final performance. 
Indeed, this means we have to develop quality control methods able to be accurate, reliable and fast at the same time, so that the production rate is not affected and reduced.
If we focus on shot peening, we should check especially the residual stresses at critical points, the roughness and the surface finishing after the treatment. 
As regards the residual stresses we all know that today there are many accurate methods to measure them. However, many of these require material removal, thereby making the measured part unusable, even to get results on the surface. X-ray diffraction is not destructive, at least if we limit it to surface measurements, but it is very slow, and not very suitable for an on-line application. Maybe the Barkhausen noise method is the one that suits better the points of this kind of applications.
How about surface finishing? In this case the use of optical no-touch methods could be good even if they could not be assessed fast enough. In both cases we need qualified technicians able to correctly interpret the results and decide what to do. 
The final part of the conversation with this automotive engineer was about this last point, that can be crucial, but that today can be faced by using digital technologies, big data and everything related to 4th Industrial Revolution tools. 
By introducing sensors, algorithms, simulations and machine learning tools, also the application of shot peening can be improved in terms of performances and quality, guaranteeing a constant improvement of the process and speeding up the production rate.
He concluded that maybe he is a dreamer. Well, maybe he is dreaming but I would be not surprised if his dream will become reality soon.

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