FEA and Design, An Overview by Kotur Raghavan
In the year 1956, Professor Ray Clough and his associates at the University of Berkley published their work leading to the development of the Constant Strain Triangle (CST) element. That was the first ever element developed for continuum (2D plane stress in that case). Many consider the CST to be the mother of all finite elements. It will not be wrong to consider the year 1956 to be the birth year of the modern finite element method as is known today to all CAE practitioners. We can say that FEA is more than sixty years old today.
The impact of the finite element method can hardly be over-emphasized. The method has taken the field of structural analysis to heights which were never known before. This is due to the progress theory, solution algorithms and growth in computational technology. An added catalytic aspect has been the availability of ready-to-use analysis packages which cater to almost all structural analysis requirements. In today’s scenario no structural analysis problem is too complex to handle.
Any analysis should have design applications to be of practical use. It can be a new design or may involve improvements to arrive at optimum design or to overcome design shortcomings. In this regard, the author of this article has two concerns.
The first is that that FEA has not made noticeable inroads into the realm of engineering design. This is so in spite of six decades of phenomenal progress. There are a few domains, such as aircraft engines and space applications, which use finite element method extensively. But these are exceptions rather than the rule.
The second concern is that FEA has had little impact on the way Machine Design in taught at the undergraduate level. Textbooks on Machine Design have remained essentially the same for eighty years or more. The subject of Machine Design, as is taught, involves many simplifying assumptions. Such assumptions were needed in earlier days because of limitations in the state of the art. The enhancements in knowledge, which have been witnessed, should translate into improvements in the methods of teaching.
The present author has addressed this issue over time and made presentations in several forums expressing the concerns and the need for remedial measures. In order to focus on the basic limitations and shortcomings in machine design methodology, a large number of problems have been revisited.
Such problems are going to be published during the coming days, one at a time in this platform.
Another important aspect according to me is that the available packages ,unfortunately, allow users to handle large complex problems without having too know Theory of Elasticity ,and Finite Element method itself- Too much of user friendliness is having adverse effect. Solid model packages provide ability to model every nook and corner which in most of the cases are not necessary for the purpose of analysisReplyDelete
Hai Dr Raghavan. I remember FEA was talked about a lot during 70s at IITK. I did not know you specialized in that. I had an impression that you specialized in vibrations. Good to see your popular articles. Very inviting and engaging.ReplyDelete
Strange, what is taught is often out-dated in some subjects. Stranger more is that this happens where research and new developments take birth and grow.
Hello Narasimham. Happy to read your observations. I worked in vibrations and structural dynamics. In my doctoral thesis I used only FEA. I developed my own programs. That started in the year 1971. Till date I worked only in that field.ReplyDelete