Skip to main content

For my group's research, we need to perform a modal analysis on a square aluminum bar with small pillars attached to the top of it (one unit cell has a base cube of 6x6x6 cm and a pillar of 1.5x1.5x3 cm, and there will be 15+ unit cells in our baseline model), and due to some limitations with money and time, we are having a debate over whether or not we should glue the pillars to the bar, or take the time and money to produce a solid piece.

My question is: Will the resultant FRF graphs from a modal analysis where the structure is simply glued together be acceptable to within a reasonable margin of error from if we were to perform the analysis with a solid piece?

Note: we are NOT looking into ANY strength properties, we are simply looking at the FRF graphs to attempt to create some band gaps in a few specific frequency ranges. Our range of frequencies that this will be measured at are between 10 and 8000 Hz. The point of this specific research is to test a theory and produce some baseline results, so we are not expecting perfection yet. If we can prove the theory, and then can get a grant, then we will be producing the solid model with a lot of precision.

I hope you guys can help us solve this debate, I appreciate any and all help you can give! Smiler

Replies sorted oldest to newest

I totally agree with Testtech. If you picture a structure in 3D as series of interconnected mass-spring elements and assuming that glue vs. metal element stiffness are drastically different, it will become clear that modal parameters will also be different.

To verify this concept, you can do a quick modal test on a metal bar; then cut it, glue the halves together using, say, locktite glue, and test again. I predict hugh difference.
RM

Add Reply

×
×
×
×
Link copied to your clipboard.
×