Stress intensity (K) predictions are presented using crack opening displacements with¼ points elements at the crack tip to simulate the crack tip singularity. These results are compared to available literature values under remote applied tension for a 2D thru-crack and a 3D semi-elliptical surface crack geometry. Crack-opening displacements with ANSYS were found to calculate K for the 2D thru-crack geometry to within 1 % of an available reference solution and were not strongly influenced by the crack tip mesh parameters evaluated. Predicted stress intensities with crack-opening displacements were then considered for the surface flaw geometry for a range of crack aspect ratios (0.2 ⩽ a/c ⩽ 2.0) and crack depths (0.01 ⩽ a/T ⩽ 0.8). K for the surface flaw geometry was also not significantly influenced by the crack tip mesh refinement and matched literature solutions to within ±5% of predicted K for the crack depth position. Predicted stress intensities at the surface position using crack-opening displacement approaches were: (a) not strictly valid given the nature of the crack tip singularity, (b) dependent on the stress state assumption, and (c) dependent on the degree of mesh refinement. These difficulties were avoided by selecting a 2° angular position below the surface with a plane strain assumption to calculate K from predicted crack-opening displacements. This approach produced a stress intensity that was reasonably assumed to be representative of K at the surface position and was not significantly influenced by the mesh refinement or stress state assumptions. K with this approach was somewhat higher than results using alternative approaches in the literature. The implication of these results compared to other finite element based results with respect to Raju-Newman interpolation equations are discussed with probability plots. Examples with crack opening models in specific aircraft engine components are also provided.