Articles
Graham Hawkes, fluids group leader at Frazer-Nash consultancy
Many people are put off using computational analysis because of a misconception that it always requires big computers churning over complex numerical solvers. But analysis can vary widely from a few hours through to weeks of calculation.
This combination of flexibility and insight make it a valuable part of the design and assurance cycle across many industry sectors.
Computational analysis adds granularity to the design codes and standards that engineers follow. When the boundaries of conventional engineering practice are stretched, this process can improve productivity, make the uneconomic economic, or justify life extension of assets.
But, to seek out these gains, you need the skills to do the analysis and a sound understanding of the physics, combined with familiarity with the codes and standards.
While analysis can help you, it can also be an expense. Just because a highly detailed analysis is possible, it doesn’t mean it has to be done that way. Sometimes a much simpler solution can help you to make a decision far more quickly and cheaply.
In any analysis programme, the first step is to assess the codes and standards to see where they are being challenged or where the gains may lie. The next step is hand calculations, to quantify the gains and scope out the analysis requirements and level of detail necessary. This step also provides you with useful information to ‘sanity check’ the analysis results.
Decision makers who are new to analysis can be concerned about the technique’s validity, and rightly so. When engineers talk about validation, they tend to focus on the performance of the analysis software against standard test cases or well-controlled experiments.
When leading an engineering project, you will also have uncertainties in your inputs, and in the way the results will integrate with other parts of the project. You need to account for the whole picture of a project, not just pieces of it, and lessons learnt from the use of analysis on previous projects can be invaluable.
Although the latest analysis tools and methods can be alluring, you need to be pragmatic. A good analysis should balance the required confidence in the outputs, time and budget with these uncertainties. In a way, you have to do a cost-benefit analysis of
your analysis.
You don’t need all the pieces of a jigsaw puzzle to see what the picture is. Astute analysis will let you work out which bits matter and what they look like.
Analysis shouldn’t be undertaken lightly, but nor should it be feared. As long as your approach captures the important physics at play and doesn’t get bogged down in the unnecessary detail, it can add value.
For more information on how analysis can help you, visit www.fnc.co.uk