Fatigue Time Step with Preload

[thecadguy]

HI,

I am running a fatigue analysis in n-code and was trying to find out a way to run the analysis using time steps and eliminate the pre-load after the first cycle.

I run the time sequence in nastran for one fully reversed cycle including the pre-load. I post process in n-code and get good results. However, I just realized that it is using the pre-load for every cycle.

Does anyone know how to run the fatigue using the pre-load for the first cycle and just the sine wave without the pre-load for the remaining cycles.

Thanks,
CAD GUY

 

[rb1957]

why would the preload relax after one cycle ? compared to the cyclic damage, how much damage can one cycle cause ? (can you neglect it ?)

you could always use miner's rule ... model the non-preload loadcycle, combine with damage from 1 cycle of with preload loadcycle.

I don't know ncode in detail but I would've expected that the loading was described as different cycle forms repeated for different numbers of cycles. Maybe the problem is that it expects preload to be either on or off, and once on (defined) it's on for good ?

another day in paradise, or is paradise one day closer ?

 

[realtime2]

Something to check:

Sometimes the initial half cycle gets repeated when one puts the "0" starting point into
the test or simulation time sequence.
history e.g.: 0 100 50 100 50 100 50
would repeat the 0 when the cycle set is reapplied or Rainflow counted in a simulation.

Usually the solution is to leave out the 0 for the fatigue history, otherwise it
will be part of the simulation's rainflow cycle count. Thus the appropriate test time history
would be 100 50 100 50 100 50
and one would let the test or simulation computer find its own way to get
from 0 to the first 100 value and it would only be applied once.