Statistical Representations of Human Populations in Ergonomic Design 2007-01-2451
Since product development cycles are continuously rationalized, the usage of human models in product design becomes more and more necessary and popular. In particular, they enable designers to evaluate criteria related to anthropometry, such as accessibility and sight conditions, at an early design stage.
Usually, it is the designer′s task to design satisfactory accommodation for an entire population or at least a certain subset thereof rather than for single individuals. In practice, most situations preclude covering the entire population at reasonable cost. Consequently, tools are required which provide discrete samples representing the anthropometrics of the target population with respect to design requirements. These samples should be as small as possible in order to limit the evaluation efforts.
From the nineties on the human model RAMSIS has been applied in the design process of the automotive industry. Taking into account the automotive customer target population worldwide, RAMSIS provides anthropometrical databases from several nations.
This paper presents multi dimensional statistical functions of the digital human model RAMSIS which generate optimal test samples according to specific design and population requirements. They make use of the multi percentile approach and a specific physique typology clustering. These approaches are discussed and compared with the well-known boundary manikin concept. The functions use anthropometrical databases. Several applications and solutions are given, ranging from one-/two-sided problems, where anthropometrics are directly correlated with the design, to complex design problems, where this correlation does not apply anymore.