Optimization of friction stir welds of aluminium alloys
The effect of rotating and welding speed on the mechanical properties of AA6056 joints made by Friction Stir Welding are investigated. Different welds with rotating speed of 1000 and 1600 RPM and travel speeds of 230, 325 and 460 mm/min were produced. An ultrasonic (UT) non-destructive testing (NDT) procedure was applied to characterize the presence of possible weld defects prior to mechanical destructive testing. The joint mechanical properties were evaluated by means of static tensile tests and fatigue tests. Both static and fatigue properties show an appreciable improvement as both travel and rotating speed increase. However, in the case of maximum travel speed and maximum rotating speed, the welded specimens may display weldment defects that reduce the evaluated fatigue limit for such processing conditions. The correlation between process parameters and weldment static properties was carried out through an analysis of variance (ANOVA) of the data.
by: A. Squillace(1), T. Segreto(1), U. Prisco(1), R. Teti(1), G. Campanile(2)
(1) Dept. of Materials & Production Engineering, University of Naples Federico II, P.le Tecchio 80, Naples, Italy
(2) Alenia Aeronautica, Metallic Materials Technology, V.le dell’Aeronautica, Pomigliano D’Arco, Naples, Italy
| Attachment | Size |
|---|---|
| Iproms_Friction_Stir_Weld.wmv | 4.85 MB |
| IPROMS_Segreto_Squillace.pps | 7.12 MB |
login or register to download the paper.
Dear Authors,
It is very interesting to read your papaer. I have some experience in analysing experimental results using ANOVA. Often the experiments are carried out within a certain range for the parameters under investigation. Hence the ANOVA results are valid and reasonable as long as the working range does not exceed or reduce too much outside this limit. In other words, although interpolating within the range was acceptable the experimental results didn't always match the ANOVA results when extrapolating outside the range. I wonder if you have done any trials outside the range to verify the ANOVA results. I would be interested to hear your experience with extrapolating the ANOVA results outside the experimental range.
Cheers,
Michael.