No abstract

. (2018). Compteraided design model parameterisation to derive knowledge useful for manufacturing design decisions.

The transformation in manufacturing capability being driven by new processes, such as additive manufacturing, offers huge potential for product innovation and opportunity to create bespoke designs tailored to individual specifications or needs. However, current design systems and tools are not yet capable of fully capitalizing on these new technologies and new approaches are needed. Many current methodologies are top-down and sequential, offering limited flexibility and an overly constrained design space. Post-processing is needed to ensure that a design can be manufactured. This work presents a novel bottom-up methodology to generate designs that can be tightly integrated with the additive manufacturing environment and that can respond flexibly to changes in that environment. Focusing on overhang as an exemplar manufacturing constraint, the method engenders changes in the design either by locally adjusting the geometry to stay within limits or by adding an appropriate support structure. The method is bio-inspired, based on strategies observed in natural systems, particularly in biological growth and development. The design geometry is grown in a computer-aided design-based, bio-inspired generative design system called ‘Biohaviour’. This process is similar to plant growth, and the design’s final configuration, shape, and size are informed by both the manufacturing capability and internal design stresses. The approach is demonstrated for overhang limit and build orientation and is extensible to any general situation.

No abstract

Current practice in three-dimensional (3D) part construction and product assembly within a digital mock-up (DMU) is based on the nominal sizing of geometry. This approach does not reflect real part dimensions when they are manufactured or the assembly conditions as they come together during the final build. The value of virtual assembly planning (VAP) using clash and gap detection is therefore limited to modelling error identification within the computer-aided design (CAD) domain for nominal conditions only. VAP could become a significantly more powerful tool if the 3D representation of components within the DMU were more representative of 'as manufactured' dimensions. This paper presents the evolution and validation of an enriched DMU (EDMU) introducing the concept of the hybrid assembly which represents components in sizes which can range between upper and lower dimensional limits in addition to the nominally defined form. The resulting capability has been exploited to enable a designer to perform tolerance analysis in the CAD domain in advance of manufacturing planning, allowing tolerance consideration earlier in the design process as the DMU evolves. This has the potential to reduce the time, effort and computational load associated with the application of statistical tolerance analysis by methods engineers later in the product development cycle when the costs associated with change are higher.