This presentation was made at the NAFEMS European Conference on Simulation-Based Optimisation held on the 15th of October in London.

Optimisation has become a key ingredient in many engineering disciplines and has experienced rapid growth in recent years due to innovations in optimisation algorithms and techniques, coupled with developments in computer hardware and software capabilities. The growing popularity of optimisation in engineering applications is driven by ever-increasing competition pressure, where optimised products and processes can offer improved performance and cost-effectiveness which would not be possible using traditional design approaches. However, there are still many hurdles to be overcome before optimisation is used routinely for engineering applications.

The NAFEMS European Conference on Simulation-Based Optimisation brings together practitioners and academics from all relevant disciplines to share their knowledge and experience, and discuss problems and challenges, in order to facilitate further improvements in optimisation techniques.

Resource Abstract

Today’s high-performance road bike framesets are made of carbon composite materials. A well-designed frame typically weighs around 800 g while a fork has a relatively high weight in comparison, of about 400 g. The Swiss bicycle manufacturer BMC identified this mismatch in weight as an opportunity and is redesigning the all-carbon fork of their flagship bike. Besides the reduction of weight, additional goals are an aerodynamically beneficial design as well as tuned stiffness targets.



The use of composite materials gives engineers and designers great freedom to tailor and vary the material properties in a structure. Yet, it also presents a daunting task as there is a very large number of design parameters. In order to create a sound composite structure, engineers need to choose appropriate ply materials, ply sizes, fiber directions and number of layers in conjunction with the geometric dimensions of the part. All this can add up to hundreds of parameters. On top of this, engineers must consider different loading scenarios as well as balancing the contradicting design targets of stiffness, compliance and strength. How can one find a solution in what seems like an unattainable task?



The entire design process of the fork is simulation driven. An automated workflow which comprises multiple CAE tools is used: starting with an independent CAD tool for the shape changes, followed by a sophisticated composite pre-processor for the laminate design and a finite element solver to compute the design load cases.



This workflow allows the distribution of design variations to multiple compute resources. The combination with an Evolutionary Algorithm (EA) delivers a very powerful tool to search vast design spaces. The distribution of simulation work is in fact essential as using EAs on such high-dimensional optimization problems also means that a huge number of possible solutions need to be evaluated, perhaps up to several 10’000s. A single computer is not capable of solving this task in a reasonable time. The presented platform overcomes this issue by distributing the computational work to many compute resources. Monitoring functionalities allow the tracking the optimization process by means of a fitness definition and changes in parameter values. Whenever necessary, engineers can adapt the model and EA entities such as parameterization, optimization objectives and constraints during the optimization runs.



The benefit of the presented EA is the fact that the parameters must not be mapped and re-interpreted for manufacturing by designing engineers. The applied EA handles continuous and discrete parameters robustly. The input parameters can take the shape of real numbers, integers with lower and upper bounds but also lists of discrete entities including strings where no ordering can be applied. The parameters types relate directly to the engineering design choices such as length- and angle-ranges for the geometric entities, discrete values for ply orientations and lists of materials. Thus, the parameterization can be set up very closely to the actual design and hence the optimized solutions are very close to the manufactured part.



The presented optimization platform is used as engineering tool to find an improved structural design of the fork. This work shows the applied workflow to design BMC’s next generation fork, it explains how the optimization process is controlled and how the multitude of simulation tasks is managed.