Numerical Prediction of Seating Position in Car Seats

Abstract

Two of the main design objectives for car interiors are comfort and safety. These aspects are both determined by the seating position of the occupant. Seat manufacturers use the SAE Three-Dimensional H-Point Machine™ to measure seating positions to design, audit, and benchmark seats. The seating positions measured with the H-Point Machine form the basis of a seat design, including comfort and safety aspects. Currently, the seat design process is largely based on prototype testing, which makes this process time-consuming and expensive. Consequently, there is a large demand for efficient design tools that enable an optimal combination of seating comfort and safety aspects. Numerical modeling provides an efficient means to optimally combine various seat design characteristics prior to prototype testing, thereby reducing design costs and time-to-market. Therefore, the ability to predict the H-point and the corresponding comfort and safety aspects in the early stages of the seat design process is extremely valuable to car and seat manufacturers. This paper presents a numerical model of the commonly used SAE Three-Dimensional H-Point Machine™. The model was developed using a combination of multibody and finite element (FE) techniques. The applicability of the H-Point Model was verified in two steps. Firstly, tests in which the H-Point Machine was placed on a block of foam were simulated. The measured position of the H-Point Machine was compared to that predicted by the model. After this, the H-Point Model was used together with a full FE seat model to predict the H-point of an actual car seat. The predictions with the H-Point Model were also compared with similar simulations with a FE model of the human pelvis. The FE pelvis model provides a more realistic prediction of the seating posture of a real human, since the deformation of the flesh is incorporated in this model. Based on the simulations, it can be concluded that the H-point can be accurately predicted with the numerical approach presented in this paper. The level of accuracy was found to greatly depend on how accurate the seat foam is modeled (e.g. mesh density, foam characteristics).