Bone screws are used in orthopaedic procedures to fix implants and stabilise fractures. These procedures require care, as improperly torquing the screws can lead to implant failure or tissue damage, potentially requiring revision surgery or causing further disability. It was proposed that automated torque-limit identification may allow clinical decision support to control the screw torque, and lead to improved patient outcomes. This work extends a previous model of the screw insertion process to model complex thread geometries used for bone screws; consideration was made for the variable material properties and behaviours of bone to allow further tuning in the future. The new model was simulated and compared with the original model. The model was found to be in rough agreement with the earlier model, but was distinct, and could model thread features that the earlier model could not, such as the fillets and curves on the bone screw profile. The new model shows promise in modelling the more advanced thread geometries of bone screws with higher accuracy.Clinical relevance: This work extends a self tapping screw model to support complex thread shapes, as common in bone screws, allowing more accurate modelling of the clinically relevant geometries.