Tag Archives: finite element analysis

Reasons I became an engineer: #3

Decorative image of photoelastic fringe pattern in section of jet engine componentThis is third in a series of posts reflecting on my path to becoming an engineer.  In the previous one, I described how I left the Royal Navy and became a research assistant at the University of Sheffield in the Department of Mechanical Engineering [see ‘Reasons I became an engineer: #2’ on May 3rd, 2023].  My choice of research topic was dictated by the need for a job because I had to buy myself out of the Royal Navy after they had sponsored my undergraduate degree and I needed a salary to allow me to make the monthly payments.  So, I accepted the first job that was offered when I went back to the University to talk about my options.  I worked on investigating the load and stress distributions in threaded connections with a view to designing bolted joints that would be lighter, stronger and with a longer life.  We used a combination of finite element modelling [see ‘Did cubism inspire engineering analysis?’ on January 25th 2017] and three-dimensional photoelasticity, which is an experimental technique that has fallen out of fashion [see ‘Art and Experimental Mechanics’ on July 17th, 2012].  I was fortunate because all of my work as a research assistant went into my PhD thesis which although not ground-breaking resulted in several journal papers [see ’35 years later and still working on a PhD thesis’ on September 16th 2020] and, with the help of personal contacts, a post-doctoral fellowship at the Medical School at the University of Calgary, Canada.  In Calgary, I worked on the design of experiments to measure the stress in the pericardium, which is a sac that surrounds the heart – still engineering but a major shift in focus from industrially-focussed mechanical engineering toward biomedical engineering.

Image: Fringe pattern in section of photoelastic model of jet engine showing distribution of stress from Patterson EA, Brailly P & Taroni M, High frequency quantitative photoelasticity applied to jet engine components, Experimental Mechanics, 46(6):661-668, 2006.

Did cubism inspire engineering analysis?

Bottle and Fishes c.1910-2 Georges Braque 1882-1963 Purchased 1961 http://www.tate.org.uk/art/work/T00445

Bottle and Fishes c.1910-2 Georges Braque 1882-1963 Purchased 1961 http://www.tate.org.uk/art/work/T00445

A few weeks ago we went to the Tate Liverpool with some friends who were visiting from out of town. It was my second visit to the gallery in as many months and I was reminded that on the previous visit I had thought about writing a post on a painting called ‘Bottle and Fishes’ by the French artist, Georges Braque.  It’s an early cubist painting – the style was developed by Picasso and Braque at the beginning of the last century.  The art critic, Louis Vauxcelles coined the term ‘cubism’ on seeing some of Braque’s paintings in 1908 and describing them as reducing everything to ‘geometric outlines, to cubes’.  It set me thinking about how long it took the engineering world to catch on to the idea of reducing objects, or components and structures, to geometric outlines and then into cubes.  This is the basis of finite element analysis, which was not invented until about fifty years after cubism, but is now ubiquitous in engineering design as the principal method of calculating deformation and stresses in components and structures.  An engineer can calculate the stresses in a simple cube with a pencil and paper, so dividing a structure into a myriad of cubes renders its analysis relatively straightforward but very tedious.  Of course, a computer removes the tedium and allows us to analyse complex structures relatively quickly and reliably.

So, why did it take engineers fifty years to apply cubism?  Well, we needed computers sufficiently powerful to make it worthwhile and they only became available after the Second War World due to the efforts of Turing and his peers.  At least, that’s our excuse!  Nowadays the application of finite element analysis extends beyond stress fields to many field variables, including heat, fluid flow and magnetic fields.