| Engineering products for the customer is difficult. The criteria are subjective, and often quality and cost are competing objectives which must be balanced. This dissertation proposes a model which structures the design process, and a methodology to engineer for customer preference.; The methodology is illustrated with an example from automobile door design. Customer preference for solidness of closing sound is addressed. A model was developed to quantify this preference, and also the preference for closing sound/cost trade-off. The model also links customer preference to specific system responses, allowing the designer to design for these physical responses. An improved door design was suggested and compared to the current design. The preferred design was selected, and the probability of a correct decision was quantified.; The customer preference model was developed using a jury representing customers. A number quantifying preference was assigned to closing sound for several doors. These doors were measured for closing sound pressure/time history. Several factors in the time histories were quantified, and the preference value was regressed on them. A correlation was found with one factor, high frequency content. A subsequent assessment for closing sound/cost pairs quantified trade-off preference. Using regression, a multiattribute value function was estimated.; A model predicting the significant system response was created. Experiments were performed to identify the closing sound paths, and to identify the specific impact which was the vibration source. Based on this insight, an analytical model was formulated using the impedance approach. A statistical characterization of door panel behavior was employed, as details of the panel geometry are rarely specified during the preliminary design stage.; The preference and system models, and an economic model were employed to decide between two design concepts. The reliability of the decision was quantified; this was necessary because of the degree of uncertainty in the model functions and in the preliminary design description.; In addition to the specific demonstration, suggestions for generalization to other attributes and products are included. This work is interdisciplinary, using techniques from marketing, mechanical analysis, engineering economics, decision analysis, and statistics to form an integrated design methodology. |
