Active Muscle Control in Human Body Model Simulations - Implementation of a feedback control algorithm with standard keywords in LS-DYNA
Examensarbete för masterexamen
Automotive engineering (MPAUT), MSc
Simulation of occupant responses in scenarios involving low load levels requires the addition of active muscles to add posture control and muscular response to the human body. A model for active muscle control has previously been implemented in the Chalmers Active Human Body Model using subroutines written in the FORTRAN programming language. In order to decrease conflicts with future software updates and increase the usability of the model, the possibility to build the model using standard LS-DYNA keywords has been investigated in this thesis. The keywords need to perform signal retrieval, addition of a time delay, implementation of feedback control algorithms and solution of differential Equations for activation dynamics. In addition to the approach already existing as a subroutine to LS-DYNA, alternative ways to implement the areas of functionality has been reviewed in literature and scientific papers. A proposed solution using standard keywords called the Load Curve Method has been developed. For basic mathematical operations *DEFINE_CURVE_ FUNCTION keyword and its built in functions are used. To solve differential Equations a mechanical representation of the Equations together with the ability in LS-DYNA to simulate kinetic and kinematic properties of single and multiple nodes is proposed. All areas of functionality except a true time delay have been implemented using standard LS-DYNA keywords. Since it is not possible to retrieve data from earlier time steps a first order low pass filter has been implemented to model the effect of a time delay. Even though it does not exactly replicate the input signal, the filter provides a phase shift and can be used to model closed loop instability. The standard keyword solution has been implemented with the Chalmers Active Human Body Model (Chalmers AHBM) as well as on an elbow joint model using LS-DYNA V971, R5.1.1 or 6.1.0 standard binary. It has been validated against the implementation that uses a sub-routine to LS-DYNA; the elbow joint model in a situation similar to the ARMANDA haptic equipment (de Vlugt et al., 2003) and Chalmers AHBM in a seated position under influence of gravity and 1G forward acceleration. In simulations with the Chalmers Active Human Body Model a decrease in computational time of about 14% has been seen for the standard keyword solution compared to the FORTRAN sub-routine.
Livsvetenskaper , Transport , Hållbar utveckling , Fastkroppsmekanik , Life Science , Transport , Sustainable Development , Solid mechanics