I am using modified williamson hall method to calculate dislocation densities in martensite. As long as the material remains cubic (less %C, no visible splitting of peaks in xrd), the method works fine. But as the %C exceeds and splitting is visible, the method fails to give good results. One of the reasons of the failure is, I think, that the dislocation contrast factor used in the method is defined for cubic structures, not bct. Any suggestions are welcome!