Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Residual Stresses

  1. Mar 13, 2006 #1
    I'm having some trouble understanding residual stresses. I need to model a residual stress pattern into a finite element model. Can someone recommend books or references that has computations with residual stresses?
     
  2. jcsd
  3. Mar 13, 2006 #2

    Astronuc

    User Avatar

    Staff: Mentor

    See this paper
    http://techreports.larc.nasa.gov/ltrs/PDF/conf-rpqnde-92-p1863.pdf
    FINITE ELEMENT MODELING OF THE BULK MAGNITIZATION OF RAILROAD
    WHEELS TO IMPROVE TEST CONDITIONS FOR MAGNETOACOUSTIC
    RESIDUAL STRESS MEASUREMENTS
    J. P. Fulton and B. Wincheski
    Analytical Services and Materials, Inc.
    107 Research Drive
    Hampton, VA 23666
    M. Namkung
    NASA, Langley Research Center

    Google "residual stress","finite element analysis" and see what pops up.
     
  4. Mar 15, 2006 #3

    PerennialII

    User Avatar
    Science Advisor
    Gold Member

    I got stuck with your "into" .... are you interested in applying residual stresses as a predefined field in FEA or carry out a residual stress simulation (for example of welding or whatever 'process' results in the existence of residual stresses) of something particular?
     
  5. Mar 17, 2006 #4

    Astronuc

    User Avatar

    Staff: Mentor

    As PerennialII mentioned, it would be useful to know the area of particular interest. For example, welds do have residual stress, particular fusion welds in the area of final solidification. Hot tearing can be a problem.

    One challenge in FEM of solids or systems involving phase change from solid to liquid is the transformation of thermal strain into mechanical strain. Thermal strains 'do not' produce stress, mechanical strains do - and this is a key factor in modelling dynamic thermal systems with rapid temperature changes which cause a rapid redistribution of mechanical load - as well as changes in the yield and ultimate strengths - i.e. shift in the plastic or yield curve.
     
  6. Mar 17, 2006 #5
    I am interested in effect of residual stresses on fatigue cracks. I have a parametric model of a fracture specimen, and I am trying to model residual stress pattern (due to prior yielding, probably due to fluctuating loads) as a predefined field as PerrenaII mentioned. From literature I gather that this could be done by modeling the residual stress pattern as a thermal strain field and I have been trying to this for a while now without much success. I am using ANSYS. I would be grateful for any help anyone can give me about this.
     
  7. Mar 18, 2006 #6

    PerennialII

    User Avatar
    Science Advisor
    Gold Member

    Sounds doable, however can be done with various degrees of complexity like Astronuc's answer already implies (seems to apply quite generally to residual stress problems). I'm thinking is the thermal strain approach you're after in this case just a substitute problem to get the analysis (and Ansys especially) to accept (to take in) the initial stress field (or/and are you using it to transfer a solution from a model without a crack to a one with a crack)? Such steps are quite often done in these types of problems, alternatively would have to consult Ansys manuals, whether the software accepts an element by element (or node by node, integration point by integration point) input of an initial residual stress state to an analysis (if you'd have a literature/previous solution available you could use it then via such a procedure). I'd say it does, but am not 100% sure.

    Next, you could naturally try to do the analysis which would actually lead to the existence of the residual stress field - for example an appropriate nonlinear elastic-plastic analysis of your configuration under fluctuating loads and see what results, and transfer the results of your 1st analysis as input to your 2nd, fatigue crack analysis. Are we getting anywhere towards the right track?
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?