I was wondering if anybody could share some light into my query. I am running a random vibration analysis within an assembly. I have bolted joints which conventionally should be configured using "No Pen" local contact conditions. However the Radom study supports "Bonded" or "Allow Pen" local contact conditions only.
I have the luxury of modelling a configuration that has failed real life MIL STD shaker testing and am getting comparable results with the actual failure using "bonded" local contact sets. The area that failed is an Aluminium 5251 Sheet metal component and the sim is showing approx. 93MPa of stress which exceeds the material endurance limit for fatigue.(3 hour vehicle mounted longitudinal profile). When running the same study using "Allow Pen" my values for stress drop down to a negligible amount which I know is wrong!.
My current situation is that I have replaced this component with a turned Aluminium 6082 T6 version and therefore removed unwanted stress concentrations, however I need to check to see if this stiffer/stronger part has translated the energy back down into another area of my assembly. I am comparing stress and displacement values in another known failure area but I could do with a bit more guidance on the correct use of contact sets for this Dynamic study.
My question is
1. What difference does applying a local bonded contact condition have over keeping with just the default global bonding?
Also has anybody done any work on using Miners rule for Random Vibration Analysis. I have only used this for sine profile frequency stuff in the past.