torquesplit

Torquesplit is a term used in mechanical engineering to describe the distribution of an applied torque among multiple load paths connected to the same input shaft or drive source. It is commonly applied in drivetrain and transmission design, where a motor or turbine’s output is divided among parallel shafts, gears, or couplings. The concept helps engineers predict how much torque each branch will transmit under a given operating condition and can be used for load sharing, redundancy, or torque-vectoring strategies.

In a simplified linear model, parallel branches are assumed to experience the same angular displacement. If branch i has a torsional stiffness k_i (including the stiffness of shafts, couplings, and any flexible elements, with gearing reflected as appropriate), then the torque in branch i is T_i = k_i θ, and the total input torque is T = sum_i T_i = θ sum_i k_i. Consequently, the share of torque carried by branch i is T_i = (k_i / sum_j k_j) T. This relationship holds in the small-deformation, quasi-static regime. For nonlinear, dynamic, or high-speed conditions, the distribution must be computed using full torsional dynamics that include inertias, damping, and time variation.

Design considerations for torquesplit include achieving the desired torque shares, accounting for friction, backlash, temperature effects, and manufacturing tolerances. The model can be extended to include gearing by using effective stiffness values k_i_eff that reflect gear ratios, so T_i = (k_i_eff / sum_j k_j_eff) T. Applications appear in modular drive systems, all-wheel-drive torque distribution, and multi-actuator robotics.