Magnetic stress analysis of structural steel

The original aim of this project (on which the thesis is based) was to design an instrument that would measure stress in steel nondestructively by means of magnetic measurements, using the known fact that the magnetic properties of most steels are significantly affected by stress. It soon became apparent that there was a paucity of relevant magnetic theory, and so the project was widened to include a (mostly experimental) study of the effect of stress on the magnetisation curves of mild steel. These were measured for magnetisation parallel and perpendicular to the stress and for tensile and compressive stresses. The highest sensitivity to stress occurs on the steepest part of the magnetisation curve, which is where the stressinstrument is used, and is also where the theory is least satisfactory. Thus another aim is to relate magnetisation measurements and theory at moderate field strengths. The thesis is in three parts: I A review of experimental magnetic stress analysis, for the period 1950- 1980; work by about 12 individuals or groups is included. II An account of the author's experimental results, both for the development of the instrument to measure stress (called the Rotation Rig) and also for the more basic measurements of magnetic properties of steel under stress. Some work is also included on the effect of plastic strain and on the relation between Barkhausen noise and the rotation rig results. III A review of magnetomechanical theory, including a contribution from the author; theories or concepts that have some relevance are by Brown, Goodenough, Ginsburg, Watson, and Jiles and Atherton. The first aim, that of designing an instrument to measure stress, has been achieved. Strictly, only the differences in principal stresses, and their directions, can be measured. The results are analagous to photoelastic modeling but working on the actual steel instead of a plastic model. The rotation rig is cheap, quick, and convenient to use, and has an error of about lOMPa in optimum conditions. iii The . second aim, that of producing a theory to explain quantitatively the effect of stress on magnetisation at moderate magnetic fields, has not been achieved. In fact the complexity of the magnetic behaviour of steel makes any theory likely to be rather complicated. However, the symmetry of the experimental results on mild steel are explained theoretically by the author. Many of the measurements, especially those with biaxial stresses, have not been made before, and are a framework on which to base further theory.