Hi all
in an HCF analysis for a Shell body, how do I use probability in Stress level in wohler curve ?
Probability can be of real applied stress, stress level in Wohler curve is deterministic in experiments of material, but curve has probabilistic boundaries due to material properties deviations. If you have uncertainty in applied stress, then compare tails in distributions of stress (right) and material (left).
You can build two statistical distributions - for stress and for expected scattering of material properties (e.g. yield stress, MPa). When you see them both on the same probability graph with X axis in MPa, you can see that right tail of applied stress may overlap with left tail of strength. Failure probability is calculated under their intersection.
Dear Farzad Torabi,
it's a very good question, but from my perspective the answer is still open wrt to reliability with the existing methods. The SN Curve itself maybe based on a specific occurrence probability what is test specific and of course you must know it. If you know the standard deviation of your material, what is mostly not the case since you have the SN curve only, you can transform your SN Curve to other probability levels. The real difficulty comes from the HCF because the most SN curves end a 1E6 for endurance levels. Exactly the range Nf >1E6 with low stress amplitude must be assessed. Also for HCF and vibration response you have see to how is your stress occurrence probability to assess via 1-/2- or 3-Sigma level. Parallel you can try a strain life approach and apply a probability variation for the plastic (LCF) and elastic (HCF) part separately.
Kind regards,
Robert
I think you could find this paper interesting:
Nicolas gayton, "Approfi Project - Probabilistic Methods for the Reliability Assessment of Structures Subjected to Fatigue", Fatigue Design 2011, CETIM, Senlis.
Thank you very much but I can not find the paper, shall you please share the link?
Hi Farzad, could we consider bootstap fir this problem? Let s try
- Badges
- Science topic