New enhancements and validation of force based posture and discomfort predictions

Many digital human models provide the crucial method to calculate postures in virtual environments. Traditionally these methods refer to experiments and make use of statistical modeling. They provide sufficient results in their specialized domain, but cannot be used in general applications. In addition they do not consider important design aspects as forces and discomfort. Hence a new force based approach has been introduced in order to overcome these shortcomings. Based on biomechanical models of active maximal and passive receding joint torques, a mechanical optimization generates static stable postures and related discomfort ratings. The results were promising, but showed some model deficiencies and were not validated in detail. The present paper continues this work and provides necessary model enhancements. Finally the new prediction models are validated on real experiments. The validation results are presented and discussed regarding to usability aspects and future development work.