1423 Powhatan St., Suite 1
Alexandria, Virginia 22314
Phone (703) 836-6727
Fax (703) 836-7491
Email: asnehq@navalengineers.org

ASNE Day 2016 - Technical Paper Session 4 : Thursday, March 3, 2016 1100-1230

Additive Manufacturing


Authors: Colin P. F. Shields, Joshua T. Knight, and David J. Singer

Title: Flexible Ships, Future Programs


Design is traditionally product centered, with the goal of finding the best product to meet a set of constraints and requirements. Applications of this perspective force design toward product generation and optimization. The premise is that one can back out, or infer, design knowledge from a product. Each new unit of knowledge can then be used to guide the generation of the next product. This has important corollaries for how design knowledge is improved. Chiefly, one can:

1. Generate, or optimize, more variants of the product.
2. Conduct more analyses per product.
3. Increase the fidelity of analyses and/or product representations.

In the U.S. Navy, a product centric design process is evidenced by investments in tool automation, high performance computing, and various degrees of physics-based simulation capabilities. Tool automation allows designers to look at more variants of the product, and high performance computing makes more computation – more analyses – possible. Efforts to increase fidelity of methods, such as computational fluid dynamics, are continuous. While these efforts have yielded undeniable benefits, the product focus can still be limiting. Product focus often results in intensive simulation and modeling in order to explore the product design space at a useful fidelity, consuming large resources. This may inhibit the creation of novel product alternatives and ignores other fundamental properties of design, potentially leading to design failure.

In a general sense, a product focus is logical and appropriate given that the construction of a vessel may eventually occur. However, within the context of concept design, the generation of knowledge is of equal or greater importance to the product model(s). The authors postulate that there are ways to fundamentally broaden our representation of the design and knowledge generation process for early- stage ship design beyond physical descriptions of a product, like properties of mass and geometry. This paper proposes a knowledge-decision design framework which facilitates an investigation of complex design behavior and introduces a new perspective of how to leverage this behavior.

The knowledge-decision design framework is mathematically defined through governing mechanics of knowledge generation and decision-making. Knowledge generation captures product and context relationships in a network representation. Decision-making is based on the theory of path-dependence and creates a set of time-independent conclusions about the design product. This defines decisions in terms of the knowledge created and allows the framework to be applied to theoretical design processes as well as specific design cases. The framework does not impose a specific structure of design knowledge and may be adapted to match established formats for representing design processes.

Existence of fundamental, complex behavior in design is proven by applying the proposed framework and criteria for a system to exhibit complexity. This provides a new perspective on the observed behavior of large engineering design programs including failure and the impact of design automation. Classification of design as a complex system indicates that complexity cannot be avoided by simplifying the product, and instead must be managed and leveraged toward beneficial outcomes.

Current approaches are unable to leverage complexity in design because they do not consider design as a system. The proposed framework enables this by considering the organization and emergence of knowledge and decisions. Design analysis is introduced as an analysis paradigm that evaluates design system behavior to guide product development towards a productive use of complexity. Design analysis attempts to understand the complex, path-dependent relationships between design knowledge, decisions, and process which ultimately define if a design will be successful. This approach is complementary to analysis of designs - generation and evaluation of products - as it does not create products, it studies how product decisions affect the design system. Possible theoretical and naval applications of design analysis are discussed. The paper concludes with a discussion of the implications of complexity in design and how future research may address the difficulties it presents.

Return to ASNE Day Program