An Experimental Investigation of the Effect of Functional Ordering on Morphological Chart Exploration

摘要

Morphological Charts are tools used to systematically explore the design solution space. It consists of a table with first column listing the functions that define the design problem and subsequent columns enumerating means (solutions) to achieving each function. Combining one means for each function produces an integrated design concept solution. The potential number of integrated design concept solutions which can be generated from a morphological chart increase when adding rows or columns. However, limited guidelines are available to aid in curbing this combinatorial explosion while identifying high quality concepts. In addition, instructions to systematically arrange functions and means in the chart are missing.;In this research, an experiment is conducted to understand how morphological charts are explored and what impact functional arrangement has on it. The motive behind this experiment is to develop guidelines and recommendations which can help designers in generating novel high-quality concepts. The experiment consisted of two problem statements, each with five different functional arrangements: 1) Most to Least Important Function, 2) Least to Most Important Function, 3) Input to Output Function, 4) Output to Input Function, and 5) Random. Sixty-seven junior mechanical engineering students were provided a prepopulated morphological chart and asked to generate integrated design concepts. The generated concepts were analyzed to determine how frequently a given means is selected, how much of the chart is explored, what is the sequence of exploration, and finally the influence of function ordering on them. Experimental results indicate a selection frequency of 27-35% from column one, 17-26% from column two, 15-21% from column three, 12-19% from column four, and 8-18% from column five. Hence, results suggest a tendency to focus more on the initial columns of the morphological chart. The effect of function ordering on selection frequency along rows seems insubstantial, though further experiments are necessary for validation. Results also suggest lower chances of design space exploration with both Input-to-Output and Output-to-Input functional arrangements.