The short answer – It depends on what you need. The big question is how to decide between the two.
As far as time goes, both methods can be quick or slow, and resource-dependent.
Finite element analysis can do everything that classical analysis is capable of doing, but the reverse is not true. Occasionally, I have been tasked with performing a classical, or manual, analysis in order to support the conclusions that finite element analysis has arrived at. Doing this has never been a problem, other than taking the time to look up the equations and crank out the numbers.
The results from a finite element analysis simulation can raise questions as to its accuracy, viability, acceptability, etc. The recipient of the analysis may ask “how do I know this is right?” There is no “absolute” method to check any analysis result. The best insurance is close attention to detail and thorough documentation.
Personally, I have run a significant number check cases verifying finite element analysis results. This says nothing of the several hundred test cases that are available online. Early on in the finite element analysis game, I went through quite a number of them in order to boost my confidence in the software. comparisons to test results provides an excellent backup for both FEA and classical methods.
Recently, I ran a sample problem with a CFD code called Fluent, an ANSYS product. It was a fairly simple run that simulated the airflow over a NACA 0012 airfoil. Being a viscous simulation, the results could be checked against a number of published wind tunnel test reports. The results matched closely – within a few percent.
I was a bit suspicious of this, as the simulation can be tweaked to correspond to the tests. I followed this simulation up with runs that compared drag figures with standard wind tunnel calibration objects, this one being the tunnel results from the drag on a sphere, which matched closely.
Several other simulations combined with checks against test results demonstrated that I could use Fluent- as well as CFX – for fluid dynamic simulations as well as heat transfer simulations – with confidence.
Calculation errors can rear their head anywhere, regardless of the analysis method. The challenge is in sniffing them out. At times, it could be something extremely simple, like slipping a decimal point.
Many moons ago, I had a conversation with an engineer who was involved in the original DC-8 design work. The calculations in question involved the internal wing volume available for storing fuel.
Several engineers were scratching their heads at the results of the estimates, as there was no way that the wing was going to carry the fuel required. As it turned out, somebody had slipped a decimal point on the estimate of pounds per cubic foot of fuel volume! A simple error, yes, but until it was found the designers were stuck.
Applying a little comparative reasoning to the Boeing 707, which was approximately the same size and used the same engines as well as being of similar range and gross weight, yet stored all of its fuel within the wings made it pretty obvious that the Douglas engineers had made an error somewhere.
Over the years I have developed confidence in my ability to produce good results from both methods by keeping a critical eye on the basic tenet of analysis work, which is picking through the details, comparing similar analysis results from other investigators, and never becoming complacent.
Norman T. Neher, P.E.
Analytical Engineering Services, Inc.
Elko New Market, MN
www.aesmn.org