Review of “Theoretical Underpinnings of Functional Modeling: Preliminary Experimental Studies”

Kurfman et. al. have performed a design study where students are asked to create functional models both with and without training on the formalisms inherent in the system. This study involves several functional areas of design theory. Additionally, Kurfman et. al. elaborates on the background of functional-based engineering design. Finally, the results of the study are discussed [1].

Functional modeling is an evolving method of engineering requirement elaboration used in the conceptual stage of product development. Functional modeling allows design engineers to create a model of a system from the perspective of the relationships between subfunctions to achieve a resultant function. By following a formal methodology for functional model creation, design engineers may produce reproducible results that adhere to a common engineering nomenclature.

There is a saying that good ideas are already taken. This is often the case for product designs. By normalizing design descriptions, existing ideas may be categorized and stored for future use. Additionally, abstract designs may be elaborated by using existing design catalogs. This decreases design flow time while also improving product reliability through the use of using validated designs.

The most common complaint regarding functional modeling is that it does not provide for repeatability between design engineers. This can be accounted for by providing training to design engineers to use a standard language. This is accomplished in the study by providing a control group and test group with formal training being the differentiating factor between the groups.

Throughout the history of functional modeling, numerous attempts have been made to improve the repeatability and standardization of functional modeling. Value analysis was the first major milestone in establishing a common language. This was followed by work by Collins but was limited principally to rotary wing aviation. Later versions expanded upon this work to create common vocabularies.

One fact touched upon by Kurfman et. al. is that customer requirements that do not relate to flows are typically defined as constraints. Constraints are unassociated to any flow and are considered to be an integral part of the product. The example provided by Kurfman is the requirement for a product to be low-cost [1]. Contrary to that are customer requirements that do involve flows. These flows are typically inputs to the product and outputs of the product. For example, the customer may require that a hair dryer be plugged in and produce heated air. This treats electricity as the input and heated air as the output.

Functional models must be arranged such that a logical progression of stages occurs. Causality is maintained by arranging functional links such that there is a defined cause-effect relationship where necessary. Parallel functions are allowed for simultaneous events when they are applicable. These functional chains are compiled into a single functional model ta the end of functional model development.

Kurfman et. al. propose a system whereby functional models may be used in calculations. These functional systems make use of matrices to score product functional model features. This allows for calculating functional equivalence based on functional model scoring. Functional model scoring is accomplished by quantifying functional links between function blocks. This provides a method of comparing function interrelations quantitatively for equivalency determination.

The experiment conducted by Kurfman involved the use of both trained and untrained designers to create three different functional models. The results of the study found that inexperienced students were far more likely to have mistakes in their functional models. Despite relative inexperience levels among some students, errors were less frequent than expected. This study, though limited by sample size, demonstrates that functional modeling is an easy-to-understand system that may be utilized by design engineers of all skill levels.

[1] Kurfman, M., Stone, R., VanWie, M., Wood, K., and Otto, K., "Theoretical Underpinnings of Functional Modeling: Preliminary Experimental Studies," ASME Design Engineering Technical Conferences, 2000, Baltimore, Maryland.