I know that they were doing this at Brooklands in the early '80s. Tornado taileron skill panels IIRC. I have checked my apprentice log book and have no notes on the subject, either on process or inspection/testing

Maybe it is more practical to do flap peening rather than shot peening. There are flap peening kits you can buy – not so sure about shot peening to – you'll probably need a tool which would not be available to the home shop user.

At work we use flap peening often when reworking aircraft aluminium parts (especially on landing gear components). In theory it should improve the resistance of the material to resist crack propagation. This is achieved by inducing multiple small stress points on the material surface ( by the flap peen).

The flap peening kit should be easy to use i you have an air die grinder. I think the peens we use are from 3M – but I'm sure they are expensive. Maybe there are some other cheaper brands.

Just bray it with a hammer a few times and the jobs a good un

If you do try this idea, you need to use shot which is whole (no cracked pieces etc) and try to ensure that the impact s are all the same 'weight'. Cracked shot will cause sharp edged notches – just what you don't want for fatigue resistance, and uneven application will ensure that some of the dents are deeper and wider than others, and this is not the 'stretched evenly all over' condition you need.

An alternative – perhaps not as effective but it depends on your application – is to clean up the surface with abrasive working in line with the major stress (so the scratches are in line, not across, the main loads), and work up to a fine polish, always working along the scratches, not across them.

Hope this helps. Of course, fatigue depends on many other factors, including the crystalline nature of the material, and the presence (or not) of repeated shock loads in use.

Cheers, tim

Barrie.

Rather than try use a procedure critical process in home environment, have you considered changing the grade of Aluminium or changing features in the design to reduce any stress concentrations. Indeed has the component actually been stressed? If you are using say 6082-T6 you could change to 2024-T or T6. In this cases you get the double whammy of higher stress limits at yield and puts you further to right on the fatigue curves. Mil HNDBK 5 has the all the data you need and can be downloaded from the interweb.

Dave