Requesting Help Validating a FEA Simulation
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report
Background (Can be ignored unless you're curious)
I'm designing a tool to cut a threaded wire. The tool is a threaded tube and works by tightening on the threaded wire as the tool abuts a stationary nut, which is also on the wire. The wire then fails because of the resulting tension between the tool and the nut (think counter-torqued nuts).
To save time, the threaded tube is split and the two resulting leaflets splayed out so they can pass over the threaded wire faster. An outer tube with an internal taper on it deflects the two leaflets back toward the thread so that the female thread on the tool then engages with the wire, then the torque can be applied to cut the wire.
The Problem
I'm trying to model this in an FEA Static Stress Linear study. Essentially, I'm modeling a cantilever with a prescribed displacent and a torque at the end of it. The overall tool length is 70 mm, the length of the leaflets is 30 mm and the amount of deflection is 0.156 mm.
In the study, I have split the model to simplify it. I have fully constrained the near side of the tool and created separate frictionless constraints on each of the split surfaces. I assigned a prescribed displacement of 0.156 mm to the end of the leaflet. Then I assigned a torque to the end of that same leaflet.
I am using adaptive mesh refinement to ensure mesh convergence. I set the target convergence rate to 1% but only achieved ~10% and the chart seems pretty jagged. I have at least one singularity at the tip of the leaflet. Does this seem untrustworthy then? What can I do to improve accuracy?
I've read most of the information available on the site for FEA simulations but am not an FEA expert. I would appreciate help to see if this simulation set up appears as valid as is possible within F360. Is there anything I can do to make it more accurate? With the length of the leaflet being 30 mm and deflection being 0.156 mm, I figured a static linear study is still appropriate but maybe it isn't? Is it valid to use the prescribed displacement and the torque loading condition simultaneously or should they be dealt with another way?