22.08.2015   Neutron Scattering Applications and Techniques

Measurement Validation

Due to experimental uncertainties, there is considerable variation in residual stress measurements; a common strategy is to validate measurements by using two or more methods. Some neutron residual stress measurements have been validated (Figs. 4.4 and 4.5) by modelling [8], deep hole drilling [7], or contour methods [8, 9].

image040

Fig. 4.4 Comparison of model and neutron diffraction measurements. Contour plots of predicted transverse (all) and longitudinal (a33) residual stresses. Predicted stresses compared with (a-d) neutron diffraction-measured residual stresses along the D lines. Reprinted with permission from (O. Muransky, M. C. Smith, P. J. Bendeich, T. M. Holden, V. Luzin, R. V. Martins, L. Edwards, Int. J. Solids Struct. 49, 1045 (2012)) [9]. Copyright (2012) Elsevier

Due to the time and complexity of residual stress measurements, computer modelling of residual stresses has been pursued by many groups. The development and validation of models relies heavily on residual stress measurements. As there are significant uncertainties in both modeling and in residual stress measurements, the two activities inform each other’s results. Agreement between modelling and measurements, or between different measurement techniques, improves confidence in both sets of results.

The Versailles Project on Advanced Materials and Standards (VAMAS) Technical Working Area (TWA) 20 ring-and-plug strain round-robin specimen has been used to validate neutron diffraction measurements with good correspondence in results [11].

Подпись: Fig. 4.5 Hoop residual stresses in E-beam welded uranium cylinder measured by neutron diffraction and by the contour method. Reprinted with permission from (D.W. Brown, T.M. Holden, B. Clausen, M.B. Prime, T.A. Sisneros, H. Swenson, J. Vaja, Acta Mater 59, 864 (2011)) [10]. Copyright (2011) Elsevier
image042

The European network on neutron techniques standardization for structural integrity (NeT) round robin had a number of samples for both residual stress measurement (by neutron diffraction, with some deep hole drilling and also using the contour method) and modelling [12-15]. The results show good correspondence (Fig. 4.6), although there were some systematic shifts in modelling and contour-method results, when compared to other methods. This led to changes in material descriptions used in modelling, and an appreciation of the effects of localized yielding on contour-method results during cutting.

There are difficulties comparing results as the neutron method averages stresses within the gauge volume. Finite-element analysis (FEA) results are discrete so the results should be volume-averaged to produce values over similar gauge volumes to neutron diffraction results. As real welds often have significant distortion, the spatial position of neutron results should be considered, if they were taken in a straight line they will often be at different distances from the surface while FEA results made on a ‘straight’ weld will be from different areas.