TWI Industrial Member Report Summary 220/1983
By T R Gurney
Background
Fatigue tests have been carried out on longitudinal non-load-carrying fillet welded specimens under several different stress spectra, which varied in respect of block length, minimum stress and spectrum shape. The majority of the tests gave sigma (n divided by N) less than 1.0 but the rule defined by equations [1] and [2] gave a safe prediction of life in all cases except when the minimum stress in the spectrum was very high. For spectra with a minimum stress range equal to 0.2 x the maximum stress range the values Fatigue tests have been carried out on longitudinal nun-load-carrying fillet welded specimens under several different stress spectra, which varied in respect of block length, minimum stress and spectrum shape. The majority of the tests gave r-Nn less than 1.0 but the rule defined by equations 111 and [2] gave a safe prediction of life in all cases except when the minimum stress in the spectrum was very high. For spectra with a minimum stress range equal to 0.2 x the maximum stress range the values sigma (n divided by N) were much higher than in the other tests: this behaviour cannot, as yet, be explained.