D – stress: General stress decomposed into membrane and
bending parts For instance: when you have typical I beam subjected to bending. In flanges D-stresses membrane are almost equal G-stresses because D-stresses bending is very small: But when you consider plate subjected to pure bending, D-stress membrane will give you stress equal 0! And G-stress will be equal D-stress bending: Looking also in G-stress and D-stress bending in those areas where out of plane load is present i.e. knuckles points on flanges, some rapid change of shapes etc.
D-Stress membrane could be used to evaluate global strength of cover –
similar to example with I beam above. But D-Stress membrane will show wrong stress in TP because local bending is not included – see example with bending plate above. However G-stress will plot complete stress in TP which are above the limit In my opinion we shouldn’t care about this because this is a local issue due to “T” connection TP with web where FEM method overstates values, and that’s one of the reason why we calculates TP thickness analytically in accordance with relevant rules.