Flip de Wet from the University of Pretoria asked ESTEQ for help to teach their under graduate Machine Design students on the use of SimXpert with MD Nastran.

The University of Pretoria signed up on the ARISE program from MSC.Software which gives their students access to these latest packages while being users of MSC.Software since before 1993.

SimXpert and MD Nastran is the new generation analysis software, specifically developed for Multi-Discipline (MD) analyses. MD Nastran incorporates all the best in class technologies from previous point solutions such as Nastran, Marc, Dytran, LS Dyna and Easy5, and uses SimXpert as its Pre- and Post Processor, replacing Patran.

The students were shown an overview of available modern day technologies to solve all sorts of complex analysis problems and comments like “I didn’t know you can do that!” were expressed a few times.      

In this class, <! var prefix = 'mailto:'; var suffix = ''; var attribs = ''; var path = 'hr' + 'ef' + '='; var addy65677 = 'www.tandm' + '@'; addy65677 = addy65677 + 'esteq' + '.' + 'co' + '.' + 'za'; var addy_text65677 = 'ESTEQ?s Test and Measurement'; document.write( '<a ' + path + '\'' + prefix + addy65677 + suffix + '\'' + attribs + '>' ); document.write( addy_text65677 ); document.write( '<\/a>' ); /> ESTEQ’s Test and Measurement <! document.write( '<span style=\'display: none;\'>' ); /> This e-mail address is being protected from spambots. You need JavaScript enabled to view it <! document.write( '</' ); document.write( 'span>' ); /> business unit set up a cantilever test problem, instrumented with a strain gauge and deflection meter and showed the real-time strain results on screen while the students added weights as a tip load to the structure.

ESTEQ decided to mimic the normal day challenges faced by engineers every day in practise as far as possible by not providing the students any information regarding the problem, other than what they could determine or measure themselves and by deliberately ignoring their mistakes and withholding any assistance to which parameters would be critical for their analyses.

The students had to make their own assumptions and a few volunteers measures all the relevant dimensions. The rest of the class were given a few of these different sets of data (clearly showing operator error) which proved to be very handy for their learning further on. The students were then given the assignment of converting the

 tested strain results to stresses (using their own assumption for Young’s Modulus) and compare both stresses and deflections to that of hand calculations and FEA using SimXpert.

Before handing in their assignments, students could upload their inputs and results to a dedicated webpage for discussion in the next class.

The vast spread of input data and subsequent spread of results made a very interesting topic for discussion. While refusing to give the students the answers to the problem, they were taken on a typical engineering journey in the everyday challenges faced by an engineer when confronted with conflicting sets of results.

Initially, all the questions from the students revolved around the usage of SimXpert, and not about the real engineering issues at hand. It was clear that the use of the software was thought to be the reason for the differences between the FEA results and the measured results, rather than their own inputs and assumptions.

After much deliberation, the technique of critically evaluating your inputs and assumptions were illustrated by using a sensitivity study to identify critical input parameters to further scrutinize. Obvious mistakes in the measured data were then also highlighted as well as others that were not so obvious at first, but evident by the inherent relationship between stress and deflection in a linear elastic problem (an assumption few cared to check!).

What proved to be very valuable for the students was the fact that a lot of their disjointed knowledge were starting to come together in one simple problem. A problem which at first glance were extremely simplistic and easy to solve, but proved to be complex when test measurements, hand calculations and FEA results did not match!

In interviews with a few students afterwards, the following comments were made: 

• “I heard of these technologies but have never seen it before now and didn’t realise it was so advanced”.
• “I didn’t realise how difficult such a simple problem (cantilever beam) could be and how many assumptions affected the end results”.
• “I thought that design is what you do with a CAD package, but now realise it is only a drawing aid to fit stuff together! No CAD tool can do all this that we just did.”
• “We did not realise how accurate FEA is and that the errors are actually introduced by our own assumptions.”
• “I am so excited about everything that I’ve learnt, I can’t wait to tell my Dad.”

After the class a few students came forward with interesting problems they wish to take on for their final year projects, not knowing whether or not technology was available to aid them, and excited about the prospect of new opportunities.It was a great privilege for ESTEQ to aid the University in underwriting the importance of what they are teaching their students every day by bringing a lot of it together in a practical manner and showcasing the importance of Engineering knowledge to bring technology available from MSC.Software to its full potential.

The author wishes to express his thanks to Flip de Wet and Gerhard Booysen from the University of Pretoria for their invitation and assistance with this task and for Wimpie Olivier and Karl Du Plessis from ESTEQ for their preparations and help in the practical sessions.

Article by Paul Naudé, Technical Manager at ESTEQ.