OpenSees Cloud
OpenSees AMI
Be More Specific than "Fiber Element"
Original Post - 01 Oct 2019 - Michael H. Scott
Visit Structural Analysis Is Simple on Substack.
I hear the term “fiber element” a lot, as in “We used fiber elements to model the response of this reinforced concrete moment frame”. Usually, when people say something like this, they are referring to “force-based distributed plasticity frame elements with fiber-discretized cross sections”.
That is a mouthful. I can see the desire to shorten the description to “fiber element”. But let’s not shorten it so much that the meaning becomes ambiguous.
After all, displacement-based distributed plasticity frame elements can use fiber sections, as can frame elements with a mixed formulation and all the variations in between. The element formulation makes a big difference in the simulated results.
You can even use fiber sections in a zero length element to perform standalone moment-curvature and axial-moment interaction analyses. But that’s beside the point–no one is going to say “fiber element” to refer to these use cases.
More importantly, saying “fiber element” to refer to the “force-based distributed plasticity …” mouthful implies some sort of exclusive relationship between force-based elements and fiber sections. In fact, it’s an open relationship and force-based elements see other people all the time. Force-based elements work just as well with stress resultant section models.
So, let’s stop using the term “fiber element” to refer to a particular element formulation with a particular type of section constitutive model. I don’t expect the entire mouthful–I would be happy if people said “force-based fiber element”.