| This dissertation presents an ecological model of human prehensile behavior utilizing the technique of dimensional analysis to develop pi-numbers that express body scaled relationships between hands and objects of prehension. Hand patterns were classified relative to generic action categories of immobilization, displacement, projection, and manipulation. These generic action categories do not embody all prehensile patterns but, do indeed capture a majority, as well as patterns of specific interest in the design of products that interface with hands. The present study presents evidence that prehensile behavior is body scaled and that subjects can accurately predict a priori, without visual or tactile feedback, grip architecture and grip capacity, as a function of generic action category. For each generic action category critical points are provided for shifts from bilateral to multilateral grip architectures, and boundary points for one-handed graspability. Critical point values decrease as a function of action category, when ordered as above, which is seen as an indicant of task difficulty. Finger use frequencies were used to define subcategories unilateral, bilateral, and multilateral, on the basis of finger to object contacts and opposition planes present in a grip. The traditional interpretation of the hand possessing 14 links affording 21 degrees of freedom is reassessed and an alternative is proposed based on fingertip to object contacts in which control and coordination in the hand can be simplified to 5 links affording 10 degrees of freedom. Experimental results are marshalled to support an ecological model for prehensile behavior in which there is the global action category of grasping, and four categories of generic grasping action within which grip architectures are classed relative to the number of opposition planes and fingertip to object contacts, for dominant and nondominant hands. As a summary, the benefits and application of body scaling in design is discussed and compared to the utility of current ergonomic data. |
