Deputy President's Column November 2022

15/11/2022

Category: General, President's Column

Today, we all rely on computers, whether it's our basic pocket calculator, our mobile phone, or our home PC. But do we have the experience to know if these computers are giving us the right results?

In certain instances, we most certainly do. For example, if you tapped 13 x 13 into your calculator and it presented you with the answer 1234, then you would immediately know that this result was wrong and you may then decide to find that elusive pencil and paper and go back to the basics we were all taught at school.

Unfortunately, the same logic doesn't apply to more sophisticated computer programmes. Using these we can now analyse and design huge 3D structural frames, produce CAM data for automated manufacturing, tell us what thickness of fire protection to apply, or determine the capacity of large and complex (or even small and simple) steelwork connections.

So, should we be relieved or concerned?

Like the maths question above, we can sometimes test the computer results by using more traditional methods such as comparisons to full-scale tests. And if your workshop plant suddenly starts to drill holes in the wrong place then we can assess and fix the computer fault. We can even check the design results of simple computer outputs by comparison to existing (although likely conservative) design codes. But what if these design codes don't exist or require a level of interpretation given our increasing desire to design taller, longer, greener, and eye-catchingly iconic steelwork structures?

Is this when we turn to FEA and place our full trust in the computer programme? Or should we be turning back to more tried and trusted methods based around hand-calculations that are derived from the existing design codes that we understand and are more familiar? After all, we haven't made too many disastrous mistakes in the past.

I am in no doubt that our industry's future lies with the use of computer analysis, but I strongly believe that, even though we have now been using computers for many years, we are still in a transition period and have to tread very carefully. After all, it's not just the computer output that we should be wary of, it's equally down to the computer input. So, as we move from an industry filled with engineers who once relied on a pencil and paper, to an industry filled with engineers who have grown up with nothing but computers, then we must ensure that we maintain the experience and knowledge that are needed within our combined ranks.

To clarify, I am not saying that computer-reliant engineers won't have the necessary experience and knowledge to design and build our iconic structures of the future, this will surely depend upon their qualifications and individual route to the design team. What I am saying is that we simply shouldn't accept what the computer throws out without having an approximation of what results were expected. For me it's simple, if you don't know what the ballpark answer should be then you shouldn't be pressing the button.

And speaking of ballparks, as sports arenas around the world become more adventurous and aesthetically challenging, they are the perfect example of where the computer analysis can help to create the unique combination of structural excellence and visual pleasure. So, it seems that I'm not a complete computer non-believer after all.

Gary Simmons, BCSA Deputy President

Share