in Vol. 5 - May Issue - Year 2004
The Shot Peening Process: Developing Prediction
Figure 1: Fatigue Failure and Initiation Site on a Gear Tooth (Courtesy of the Design Unit)
Figure 2: Effects of the Shot Peening
Figure 3: Relationship between fatigue strength and residual stress 
Table 1: The control variables and testing levels.
Table 2: Optimum parameters and results.
Power driven applications, performance and reliability hungry engineers are all fighting a unique enemy: Fatigue. Shot Peening, a proven process to fight fatigue, is becoming more and more sophisticated and being able to predict its effects at the design stage is the key to an increased performance.
A Short Word on Fatigue
Fatigue failure is common in most metals and is most frequently caused by tensile stresses favouring crack and micro fissure growth from or near the surface of the material. This is usually followed by the capitulation of the integrity of the material and a drastically reduced performance of the component followed by failure at an early stage of the component’s life. The magnitude and the distribution of the residual tensile stress are both critical to the general behaviour and performance of any components and should be considered at the design stage, when possible.
A better understanding
As we all know, the shot peening process is mostly used to improve fatigue life and/or fatigue strength of cyclically loaded components. This is achieved by the introduction of a layer of compressive stress into the material (see figure 2).
The predictable magnitude of the compressive stress introduced is obviously a function of the material targeted, as well as the shot peening condition and can reach values as high as 50 to 60% of ultimate tensile strength of the material. Yesterday and today’s tools to control and assess the effect of the process are the Almen strip and an estimate of the coverage.
Since 1998, two in-depth studies of the process have been carried out in collaboration with the Design Unit at the University of Newcastle. The first of these projects being the foundation for the last and still on going project aimed mostly at understanding the effects of each selected parameters on the stress distribution and to establish raw models to predict stress levels for a set of parameters and vice versa (see figure 3).
Some early work carried out by the Design Unit at the University of Newcastle, aimed at testing various peening treatments and to relate stress to fatigue strength.
It was clear (see figure 3) that higher compressive residual stresses were better for the fatigue performance of the treated components.
From this preliminary study, an estimated 60% improvement of the fatigue performance was determined and, most of all, it was clearly established that the stress profile was key.
As an in-depth follow up study, a one-year research project was carried out to fully assess the effects of identified process variables on the stress distribution and ultimately determine models leading to accurate predictions of the process effects.
The process parameters investigated were the air pressure, the mass flow, the impact angle, the distance nozzle-specimen, the exposure time and the nozzle size. Using the appropriate software, regression analyses were performed on the results obtained from statistically designed experiments and models could be established. These six parameters and their significance were investigated, aiming at relating their conjugate effects to the residual stress introduced. Each parameter was tested at three different levels (Low, Medium and/or High). In table 1, the list of control variables is shown, with their respective experiment levels and assigned values.
Each experimental site was processed with the required conditions and X-ray measurements were carried out to determine the residual stress profile generated by the process. To carry out the statistical analysis, five responses were investigated in the statistical analysis:
-The maximum residual stress (RSM)
-The depth of the maximum residual stress (DRSM)
- The shot peened outer layer (SPOL)
- The surface residual stress after peening (RSSf)
- The variation in the surface residual stress from un-peened to peened (RSSf-RSSi)
The statistical models established for the given material led to the optimum stress distribution and set of parameters shown in table 2.
The continuation of this work, currently in its final year, widened the scope of experimentation and specifically focused on developing DE models/software to predict residual stress profiles for any given geometry. Using the data calculated by the DE models, FE representations/models can be established and the expected stress distribution or profile calculated. All the models are also validated and cross checked by a broad experimental programme.
From basic estimates to accurate prediction of the stress distribution, we are designing, mastering and using the tools to pave the way to an enhanced fatigue performance and a new understanding of the Shot Peening process, where concepts such as “100% coverage” will give way to “a desirable or expected stress distribution for a necessary coverage”.
Obviously, the main interest in doing so would be to be able to understand how shot peening can help solve problems or improve components at design level. By doing so, the process will not only be the problem solver it has commonly become but, more importantly, it will become a solution provider.
ISPC Impact Finishers UK
Woodson House, Ajax Avenue
Slough, Berks, SL1 4DJ Engand