X-Ray Investigation of Residual Stresses of Cylindrical Steel

Abstract

Sachs' boring-out method is used for determining the residual stresses in a cylinder. When a test piece that is separated from a longer cylinder -solid or hollow- is too short, the boring-out test will not yield the same residual stresses that would have been observed in the original cylinder. Bühler has stated, from his experimental investigation, that the residual stresses in the separated specimens are roughly the same as those in the original longer specimens, provided the ratio of length to diameter remains greater than about 2.One of us studied theoretically the influence of end faces of the separated specimen, proving that for any arbitrary distribution of longitudinal, the tangential and radial stresses over a cross section, Bühler's experimental conclusion is always valid.In the present study, experiments have been carried out to find the influence of the end effect of detached specimens by X-ray stress measurement, that is, the residual stress distributions of the longitudinal (σ1) and the trangential (σ2) stresses on the specimen surface are determined by shifting the specimen from end to mid-length.In the X-ray measurements, “Twin-counter method”is employed together with an ordinary film method. The twin-counter method enables us directly to read out a small shift in the position of X-ray reflection line which is caused by lattice strain; then (σ1+σ2) are continuously recorded from end to mid-length. The difference of σ2 between end and mid-length zone is negligibly small, and σ1 is zero at the end, then σ1 in the mid-length can be determined without using the sin2ψ method which is a well known method in the X-ray stress measurement, the result showing a good agreement with the ordinary Sachs' boring-out method.The residual stress distributions are suddenly changed within the distance comparable to the radius of the cylinder from the end. However, they are gradually changed in the same direction for the distance of the same radius. Beyond the point at the distance of the diameter from the end, the stress distributions are almost kept constant, and it appears that the residual stresses of this zone are the same that would have existed in the original cylinder.These experimental results can lead us to the conclusion that the residual stresses in the original cylinder remain unchanged in the vicinity of mid-length of the separated specimen, provided the ratio of length to diameter exceeds 2.