Technical Manager, Enginsoft UK
John Barnes holds a Bachelor of Engineering in Automotive Engineering from Loughborough University. He has gained 13 years of experience at TASS (Vehicle Crash Safety), Jaguar Land Rover (Body CAE), and as onsite consultant for Fiat, Daimler Crysler, JLR, Magna Steyr and Nissan. His main areas of competence cover: slow and high speed dynamic impact analysis; stress, fracture and crack propagation analysis; expert in light-weight design; urban flood mitigation and hydraulic analysis; test definition and simulation correlation; engineering design consultant. He currently works as Technical Manager at EnginSoft UK with the following duties: project manager for engineering projects; expert and trainer in process integration and optimization; mentoring engineers in various industries to help extract the benefit of optimization.
Speeches
Companies are continually looking for cost reduction by the implementation of smarter processes. This is often achieved through the purchase of new engineering tools which have to be assessed for return on investment. Often this assessment can be subjective and hard to quantify. What if current process integration and optimisation techniques could create new, bespoke, game changing tools for a company? While the benefits of standard optimisation techniques are being realised the world-over, there are new benefits to be gained by deploying response surfaces within a bespoke tool, app or widget. By doing this, complex analysis (once the preserve of highly skilled simulation engineers) can be placed into the hands of design engineers. EnginSoft UK have collaborated with Anglian Water and Atkins to develop a new methodology for creating a predictive RSM based tool that can reduce housing development cost enquires from weeks to minutes. This paper demonstrates the principle steps taken to develop the tool and highlights the key technical decisions made to ensure its predictive quality. This paper shows how ad-hoc design simulation can be replaced by an up-front full exploration of the design space using a computationally intense DoE of optimisations. Therefore, it will show how the problem can be reduced through a complex parameter reduction optimisation. It will then explore the training and validation of the RSM and how it can be published as a user friendly tool for non-technical users.