Fifteen laser-welded Olympia alloy samples were divided into three gro
ups of five samples each, with different gap distances between the wel
ded halves. The first group was welded with a 0.0-mm gap distance. The
second and third groups had Olympia shims placed in 0.3- and 1.0-mm g
aps, respectively, prior to laser welding. Each of the samples was tes
ted to failure in load fatigue at 30 Hz in a fatigue testing device us
ing an applied stress of 35,000 psi (241.4 MPa). The means and standar
d deviations for the number of cycles required to produce fatigue fail
ure for each group was 494,618 +/- 118,311 cycles for the group welded
with 0.0-mm gap distance, and 242,741 +/- 44,623 and 232,021 +/- 55,8
77 cycles for the 0.3- and 1.0-mm specimens, respectively. A one-way a
nalysis of variance showed that the 0.0-mm gap specimens had the great
est number of cycles to failure (P less than or equal to 0.05). There
was no significant difference between the other two groups. With the e
xception of two specimens that failed at the weld center, all failures
occurred at the edge of the weld.