> When the fatigue life is considered. Mean Stress and stress amplitude governs according to Goldman’s theory. Does the stress refer to von Mises or principle stress?

Neither. Fatigue calculations are *usually* based on either PRINCIPAL stresses (not PRINCIPLE) or STRESS INTENSITY (first principal minus third principal), depending on what you're looking at, the code you're dealing with or potentially lots of other things.

Cheers,

-- drej --

Another correction , it's Goodman not Goldman ! (as in the Goodman diagram)

Drej is correct, however using principals is overly conservative but safe. If you can pass the component using a stress range based on Max Principal and Min Principal at the same location for your cycle, all well and good. If this causes problems you will have to look at the range of each stress component and make a judgement about which ones to use for lifing. I could say more but it would take too long.

If we think about the multiaxial fatigue problems and we consider alpha as a biaxiality ratio then

for -1<alpha<0  Signed Tresca is very conservative
                Signed von Mises is conservative
 
for alpha=0 (uniaxial) both are O.K.   

for 0<alpha<1 Signed Tresca is O.K.
              Signed von Mises is non-conservative

for alpha=1 (equibiaxial) both are O.K.

Ref. MSC.Fatigue


http://www.geocities.com/fea_tek/
A.A.Y.
               

i'd avoid von mises, 'cause it makes everything (compression particularly) positive, therefore confusable with tension stresses.  
i'd prefer max. principal

Thanks all for your inputs.

feajob> A quick Q. How is alpha defined?

Thank u.

i'd expect alpha = sigmax/sigmay

Biaxiality ratio (alpha) is the ratio of the minimum and maximum principal stresses at a location on the surface of a component.

I had experience with Signed Von Mises, in my cases it predict (about) 5 percent longer life than Signed Tresca.

Based on MSC.Fatigue Documentations (in strain life analysis) Max. principal is a non conservative choice when 0. < alpha <= 1.

But is O.K. when -1 < alpha < 0

AAY

If you want to design base on fracture mechanic, you shoud use principle stress but if your design is base on yield stress it's suppose to use von Mises theory.

Hallo,
i have an ANSYS rst-file from a PSD analysis and have to calculate the fatigue life of the component using nCode´s FE-Fatigue. Do you know if the program (FE-Fatigue) takes in account the sigma value, as the stresses from the Ansys calculation are 1 sigma values.
Thanks

Hi,
IMHO you can't say "a priori" which criteria you should follow, because any criteria encomprises "safe" stress states that are good for some applications and bad for others. In other terms, depending on the stres state a criteria can be over- or under-conservative. The best example is if you compare VonMises with Trescà-Guest, as it has been pointed out by Feajob.
In the USA the stress intensity is commonly used (SINT in Ansys), while in Europe the convention is to use VonMises (SEQV in Ansys), as you can see if you compare ASME norms to EN. ASME are known to be extremely conservative overall (i.e., on the other hand, extremely safe) because they use both conservative criteria (Trescà-Guest "everywhere"!) and conservative limits.

Just my two-pence thoughts...

Regards

Most fatigue data are gathered using uniaxial loading. The stress amplitude you are plotting is the stress amplitude along the uniaxial direction. This is your principal direction.

I imagine the von mises stress could be a concern for ductile materials during cyclical loading. For instance, 1100 soft aluminum will exhibit plastic deformation during fatigue loading at low frequencies. In this case, the Von Mises stress can be a concern. However, for brittle materials the principal stresses should be used since tensile forces propagate the fatigue crack.

Regards
modey

Modey003, you are mixing up static failure theory with fatigue theory. You should never use VonMises for fatigue.

Use worst principal at each loading condition to be conservative or look at stress ranges for individual stress components, and even biaxial effects if you are near to the line.

gwolf,
I think that it depends on the biaxiality ratio in your specific problem. We cannot say that signed Von Mises is never correct for fatigue analysis. Please see my previous posts.

AAY

Yes feajob, I can see that you know what you are doing. Signed VonMises is not too bad.

My whole point is that microscopic level there can be plastic yeilding depending on the material. I have even seen this mechanism in plastics (which von mises is also a good criterion) and soft metals.

The failure criterion depends on many factors: material, loading conditions, and environment. For instance high frequncy cyclic loading rates can cause a fatigue crack to propagate brittlely. While low frequecies can cause a ductile crack to propagate. To sum this up when trying to design against fatigue crack propagation in FEA modeling the criterion you use depends on various factors.

Modey

but if your getting plasticity then you should not be using linear elastic fracture mechanics (my assumption !) ...
and if you're using non-linear fracture mechanics then you'll be using J-integrals ... no ?
particularly if we're splitting hairs (or heirs) about which stress to apply (max. principal or von mises).

the logic about using max. principal is that it "best" represents mode 1 (tension) crack growth.  using von mises  (as a failure criterion) includes shear into the equation and if there's much difference (between vM and maxP) then you're in a mixed mode crack scenario, and i really don't know where you'll get the data for this (crack geometry models, material crack growth, residual strength) ... but then maybe you auto guys are concerned with different aspects of FM (materials, geometry, design constraints, regulations) than i've experienced in aero.

If you are using Modified Goodman and have a biaxial state of stress, your mean and alternating stresses (for the formula) must account for the biaxial state.  Vonmises (alternating and mean) based on the max and min stress for each direction is an excellent way to do that.

ZCP
www.phoenix-engineer.com

umm, so will max. principal

zcp, rb1957,
I think you're both right. as I said before, VonMises, MaxPrincipal, Stress Intensity, etc, are all based upon ASSUMPTIONS (as every analytical theory does, anyway), and neither can completely capture the behaviour of "real" materials over all the possible stress fields: there are cases where MP becomes totally unrealistic, others where VM will under-estimate the equivalent stress. Because the crucial point is this: we always want to reconduct a stress field in a point to a single "sigma" value in this point...
Finally, I think it's important to refer to the Norms that are valid for the country you work in (or that are requested by the customer, it depends): for ex., if you refer to ASME then you will be obliged to carry all the calculations with Stress Intensity, no matter if you disagree or not in some points...

Regards