Quantifying Comfort

Natalia Camprubi, Fernando Rueda, Ivan Alonso,
Advanced Design & Analysis, IDOM, and Centro Tecnologico Grupo Copo, both in Spain

It can be complicated and time consuming to satisfy all these requirements while working with complex foam material. Additional design difficulties arise from measuring and evaluating results — particularly in the subjective area of comfort.

Most of the variables to be considered when designing car seats relate to geometry and materials. The mechanical properties of foams depend highly on their strain level, which requires building many test samples to evaluate seat response. However, it can take weeks to prepare a physical specimen for testing. First comes the process of building the mold with the correct geometry. Samples come next, and, finally, the seat must be readied for testing.

Numerical simulation using finite-element analysis (FEA) is a valuable tool for shortening this complex design process. The process of modeling seats in a virtual environment links CAD with material databases, letting users type in and evaluate various loads and stresses without the time constraints of real-world testing. It’s well established that FEA can predict the response of designs under specific circumstances and supply data useful for improving geometry and materials. But how can simulation help in the evaluation of comfort? This requires translating the sensation of comfort into quantifiable variables to measure it from a mechanical point of view.

The measurement of comfort can take place under static and dynamic conditions. Typical variables relate to occupant position, such as hip or seat-reference points, as well as to pressure distribution, or the response of a seat with an occupant model on it and subjected to a certain range of vibration. Variables come from the CAD geometry of the vehicle-seat-occupant assembly, or from the mechanical response of the seat to stress tests. Since it’s possible to numerically simulate tests, it is possible to assess comfort via FEA. In light of this, we recently studied the feasibility of developing a virtual environment for foam-seat testing. Abaqus, from Dassault Systems’ SIMULIA in Providence, R. I., provided the FEA software.

Part of the study involved simulating static indentation — the mechanical response of a seat cushion when an occupant sits on it. First, we lab-tested a cushion positioned on a rigid support. Technicians placed a test form shaped like an occupant’s thighs over the cushion. We gradually applied a vertical load downward, simulating the action of an occupant sitting on the seat. Then technicians measured the penetration of the test form on the cushion for certain loads.

In general, there are two steps involved in putting tests such as this into numerical terms. First is to define the specific material model for elastomeric foams. Next is the simulation of the static-indentation test itself.

Flexible polyurethane foams of the sort used for seats are hyperelastic, meaning they undergo large strains, up to 90% in compression. They also have excellent energy-absorption properties. In a nutshell, the materials are highly nonlinear and exhibit material softening in the first load cycles (Mullins effect).

Fortunately, Abaqus code contains the Ogden material model for highly compressible hyperelastic materials. The model exhibits isotropy at a macroscopic scale (force causes the material to move evenly in all directions) and nonhysteresis (the material bounces back after a load is removed). Values of material parameters can be determined by means of a least-square calculation from stress-strain measures of simple experimental tests.

The material database used for the study contains a wide range of polyurethane foams. Also included are specifications of polymers required to manufacture foams, as well as different material and mechanical properties. The database greatly speeds design because it lets users easily evaluate the performance of the same geometric design with different materials.

In our study, once the generic material model was particularized with experimental data of simple tests, we simulated more demanding tests such as that for static indentation. The Abaqus simulation curve of the vertical displacement showed good agreement with the curve from real-world tests.

Simulation also gives a lot of other useful information. Examples include contact-pressure distribution on the upper surface of the seat cushion, vertical-stress distribution in the seat cushion or contact area, and distribution of the load between seating plane and side wings. Overall, the study more than met our expectations of using FEA to assess seat-cushion comfort.

Make Contact
SIMULIA, simulia.com