Josep Maria Font Llagunes

A Computational Benchmark for 2D Gait Analysis Problems

The aim of this paper is to present a computational benchmark for gait analysis that has been developed in order to share real data captured in a biomechanics laboratory and the results of the inverse dynamic analysis. This work belongs to the library of computational multibody benchmark problems that the Technical Committee for Multibody Dynamics of the International Federation for the Promotion of Mechanism and Machine Science (IFToMM) is developing. The work presents the kinematic and dynamic study of human motion by means of multibody system dynamics techniques. The subject selected to perform the experiments walks on a walkway that encloses two force plates. The motion is captured by 12 optical cameras that acquire the position of 37 passive markers. The inverse dynamic analysis (IDA) is carried out using a 12-segment 2D model with 14 degrees of freedom. Displacement signals are filtered using an algorithm based on Singular Spectrum Analysis (SSA) and the natural coordinates of the model are calculated using algebraic relations among the marker positions. Afterwards, a procedure ensures the kinematic consistency and the data processing continues with the approximation of the position histories using B-spline curves. The velocity and acceleration values are then obtained by analytical derivation. The double support indeterminacy is solved using the Corrected Force Plate (CFP) sharing method. The IDA provides the joint drive torques that the musculoskeletal system generates during human locomotion from acquired kinematic data, foot-ground contact forces and estimated body segment parameters (BSP). All this information is available online in http://iftomm-multibody.org/benchmark. Therefore, it can be viewed by other researchers, which can submit their own results using the same input data and proposing new solutions.

A Friction Model for Non-Singular Complementarity Formulations for Multibody Systems With Contacts

The dynamics of multibody systems with many contacts are frequently formulated as a Linear Complementarity Problem (LCP), for which several direct or iterative algorithms are available to solve it efficiently. These formulations rely on discretized friction models that approximate the friction cone of the Coulomb model to a pyramid. However, they produce rank-deficient LCPs even though the physical problem does not have constraint redundancy and has a unique solution. Here, a new discretized friction model is presented which results in an LCP formulation with a full-rank lead matrix. This model relies on an inertial term to couple the equations of the model, which behaves as close to the Coulomb model as the other discretized models. Moreover, it is shown through some simulations that some algorithms can be used with this formulation, which could not be used with the other rank-deficient LCP formulations.