Measuring vibration characteristics in seating

Measuring vibration characteristics in seating

A long commute is often stressful and tiring, so comfortable seating becomes increasingly important for drivers as well as passengers. But how do we test the overall comfort quality of seats?

A comfortable and pleasant ride is an important factor for drivers and passengers alike when they decide which vehicle to buy. Vibration induced by powertrain, ancillaries and road can dramatically affect how passengers judge the quality of the ride. The design of a comfortable seat is, therefore, paramount, as it is a means of reducing the vibration excitation affecting drivers and passengers. In order to efficiently evaluate seats in terms of vibration ride, an objective algorithm that can predict the subjective perception of seat vibration is required.

Many studies have been done to predict the subjective perception of riding vibration comfort. ISO 2631 is the most often used standard for this purpose. The relevant studies and standards are based only on the estimation of vibration energy with a frequency weighting function applied and often do not take into account the amount of temporal variation. In collaboration with Toyota, two subjective experiments for seat vibration quality were conducted on road as well as in Brüel & Kjær’s full vehicle NVH simulator. While the on-road subjective assessment provides more realistic ratings of seat vibration quality, it is not cost-effective and does not provide the subjects with the possibility of back-to-back comparisons, that is, comparing two or more different seats, one after another, within the same experimental session.

The current joint investigation proposed an algorithm called ‘Vibration Roughness’ correlating with overall subjective pleasantness by considering the temporal variation of vibration signals together with the overall vibration energy. Vibration Roughness explained our subjective data better than the traditional metrics using only vibration energy, and the predicted results were less affected by the sensor locations and directions.

> Read the full white paper here