This morning we had a very special guest, Gary Mcdonald (shown here showing his passion for simulation!) one of the two guru directors from Trivista. Trivista specialise and consult in Finite Element Analysis (FEA) and Computational Fluid Dynamics (CFD) solutions for real world industry applications. Gary went through 4 varied yet insightful case studies from his experience with Trivista, then gave his views on the trends in this field of work. Firstly, a rotating rotor bridge design showing its evolution from solid rotating block to carefully shaped lightweight frame, using iterations of structural simulation and experimental testing. It was great to hear the clever analytical and evaluative thinking behind this process, looking at where failures may occur and how to alter designs to optimise performance. Secondly, the dirt separator/filter analysis of various designs for their effectiveness in dropping the velocity and pressure to understand the filtration behaviour for various particle sizes. Thirdly a thermal stress manifold problem, relating cyclic thermal stresses to material S-N curves to determine anticipated lifespans for product use in various load cases. Finally a real engineering design case study, where an impeller with no 3D data (requiring scanning-digitising from the casting) was analysed for its structural behaviour under high rotational velocities.
So it was all very interesting hearing about the case studies, the real world applications of these great tools. And then Gary gave us some insights into the industry, into the technology, into what’s changed recently in this world of simulation and evaluation of product performance and behaviour. Of course he started off talking about the fantastic improvements in computational speed and processing power. Things are faster!…and how we can now do multi physics problems (structural, thermal, fluid, electromagnetic) and we can couple these….and how it’s easier to move between packages, with CAD models available in transferrable, neutral formats. There’s also the added benefit with all this additional speed and power, and that’s the ability to work with nearly full (or full!!) models, that are assemblies, that include multiple contact models, that include details and features that previously we’d have to remove. In fact a huge amount of time was previously spent by analysis defeaturing models, removing fillets, holes, ribs, details that would make meshing too time consuming, too problematic, too difficult to justify. So now, much of this defeaturing work isn’t necessary. What only a few years ago would take a week to run with all details, now can be run over night. And there’s the added bonus that many hours that used to be spent on defeaturing (and remember, that’s a very time consuming job!) can now be spent analysing results, altering design parameters, optimising design solutions! And of course the final, more over arching development in the simulation community is that it’s becoming more accepted. It’s now not only an additional tool for experimental work, it’s a powerful tool that complements experimental work. FEA and CFD allow for much a more in depth understanding of the behaviour of products which have been shown in many industries to be a little too conservative in the standards that outlined how they should be evaluated. This is because these tools, if used correctly can really improve our understanding of products and how they perform in the real world. But of course we need to remember, that if used incorrectly, naively, or without proper scrutiny and evaluation, can also be dangerous and misleading!