when I was still in engineering at Boeing, we were not allowed to use FEM for composites or metallic panels and parts. Only hand calculations using basic strength of materials equations using proprietary Boeing applications. Every single part's strength check notes were vetted and reviewed before being signed. And because it was all hand calcs using equations that a person could print and run through, it was all above board. In that way you're right that aerospace engineering is very conservative by nature and has healthy margins built in. For repair manuals, we had to once again do the hand calculations and then run it past a Boeing appointed FAA representative. Their job was to look at it as if they were the FAA and needed convincing about a particular design or repair. That is what a lot of people have been citing as one of the problem areas- FAA having delegated the responsibility of the final design review to Boeing engineers rather than having their own person do the review. But FAA doesn't have the budget to employ people to do this level of review for each and every system.JayS wrote:I know examples of mission critical requirements certified by FAA based only on simulation results. Obviously all other civilian agencies mostly copy paste FAA certification, its true for most of them if not all. Structural simulations are particularly mature technology by now for many test cases. There are some test points, which are still too complicated for simulations, they need validation test e.g. FBO or are used as ultimate verification point for a whole set of test points, eg wing ultimate strength. The amount of test points covered by simulations are only gonna grow in future.Dileep wrote:Simulation is not trusted for any safety or mission critical aspects. As per my understanding, none of the aerodynamic aspects are certified based on simulation. However, electrical simulations are sometimes accepted as means of compliance now. Of course, they are much simpler than aerodynamics. I have seen only a few line items that accepted (electrical) simulation among the thousands I have seen in my rather short experience in this field. We do not deal with aerodynamics, so don't have data on them.
Even RF simulation is not accepted.
CEMILAC/RCMA never accept simulation for domestic stuff as I know.
As I said previously, it depends on how much confidence the certifying agency has on the OEM's methods or how well the OEM can convince the agency. Its definitely a gray area, but Aerospace Engineering is quite conservative in nature as you would know already. So things are not taken lightly. We praise global OEMs all the time but then they get a lot of leeway from FAA. If FAA puts its foot down on each and every certification point to be verified in some real life test, I wonder if developing and certifying a Civil Airliner will even remain a economically feasible activity at all.
Typically first of the family takes longest. Later derivatives only need to recertify design changes, not full set of test points.
But FEM was permitted for panels for 1 specific check that all panels had to pass and that was aerodynamic deflection. That was a very basic static analysis, nothing fancy.