Methods for generalization of experimental results in terramechanics

Abstract

Due to the extreme variability of soils and different tillage and settling conditions, measurements made under a single set of conditions have little practical applicability under arbitrary conditions. Results of more general usefulness require systematic measurements to be carried out over suitable ranges of the main influencing variables. The results can then be processed to obtain dimensionally homogeneous equations, i.e. similarity equations. This method was first proposed a century ago (Buckingham, 1914) [1], and one year later, Nusselt published a fundamental paper showing how it could be used to generalize the results of heat transfer experiments and to plan new experiments (Nusselt, 1915) [2]. The method is based on the postulate that all physical processes can be expressed as relationships among dimensionless parameters, and it specifies how to find those parameters. We describe new experimental results and generalize the results using dimensional analysis to obtain similarity plots and equations that are generally applicable within the feasible range of variables.