CONTACT WELDING IN WIHK SPUING HELAYS 



921 



interfere with the mitiation of the arc during- the next weld cycle. 

 Limited tests made with nitrogen and iieliuin atmospheres gave no 

 indication of improvement in weld ciuality. 



In the course of the many studies employing diricrcnt kinds of in- 

 strumentation, an electronic counter was used to measiu'e successive 

 arc durations. Variations from 20 to 230 microseconds were observed 

 for the prevailing conditions during early tests. The contacts welded 

 were measured for strength of weld and a correlation found between 

 short arc durations (up to 65 microseconds) and weak welds, Fig. 20. 



Circuit impedance was found to be important in producing uniform 

 weld strength. The assistance provided by the Laboratories in the early 

 work led to better control of arc duration. Hawthorne continued this 

 study and arrived at the circuit, Fig. 21 which greatly improved the 

 uniformity of welds. The improved circuit resulted in less arcing in th(^ 

 jaws, thus prolonging their life and precision. Uniformity of burn-off 

 and of weld strength both showed marked improvement. Fig. 22 shows 

 a typical weld strength distribution for this circuit based on standard 



PER CENT OF SAMPLE 



3500 



3000 



3 2000 

 O 



500 



in 00 — 



(\J (0 <o 



CM fO ■<! 

 ro — O 



f) fO (\i 



30 40 50 60 70 80 



WELD STRENGTH IN POUNDS SHEAR 



P'ig. 22 — Weld strength distril)ution for 13,200 samples tested Ironi 600,000 

 welds made in production. 



