My question is regarding SIF value from FEA tools. In result we get three sets of values for in plane and out plane namely for peak stresses, primary and secondary stresses. Which value should be in stress analysis?
Do we have creat separate model for primary stress case and secondary stress case?

Primary = plastic deformation and bursting.
Secondary = brittle material failure.
Peak = failure due to fatigue.

Generally, use peak.

Hi there,

Just for conformity and to avoid fundamental misunderstandings, here is a quote from FE Pipe Manual (1999 Full Edition):

"The peak stress intensification factor comes from Markl fatigue testing and theoretical work. Markl's approach has been extended by many into the realm of finite element analyses for common types of materials and weldments.

The secondary stress intensification factor is intended to trap the highest discontinuity effect without trapping the effect of concentrations,
i.e. stresses at the toes of fillet welds. In our classes we tell attendees that secondary stresses exist in an area that is a percentage of the diameter. In general secondary stresses attenuate with the distance root(RT) from the stress discontinuity. Peak stresses attenuate
with distance through the wall thickness.

The peak stress intensification is intended to trap the highest peak stress in the body, while the secondary stress intensification is intended to
trap the highest discontinuity stress in the body. The peak stress intensification factor is often lower than the secondary stress intensification factor however.
This is primarily true for the following reason: The peak stress to be computed in the B31 piping Codes as per Markl uses a girth butt weld as the baseline peak stress, i.e. a girth butt weld has a stress intensification
factor of 1.0 in the B31 Codes. In reality a girth butt weld in a carbon steel piping system has a stress intensification factor of between 1.7 and 2.0. Thus Markl type peak stress intensification factors are reduced by about 0.5 from what their actual values would be.
All of the allowables are adjusted for this and so there is no problem, the only result is confusion.

For the majority of cases with FE/Pipe the peak stress intensification factor will be seen to be smaller than the secondary stress intensification factor. It is because it is only accounting for half the stress that this occurs.

The primary stress intensification factor is even smaller than the peak stress intensification factor because its value is based not on the peak
stresses, or the highest discontinuity stresses, but rather on the highest membrane stress at the discontinuity. And so the primary stress intensification factor will be smaller than both the peak and secondary stress intensification factors by the ratio of the peak to membrane
stresses at the point of interest."

Conclusion:

Secondary Stress Intensification Factor does not have any relation with the Brittle Failure of material.
Secondary Stress Intensification (or Concentration!) Factor is related to ASME VIII-2 Code Secondary Stress PM+Pl+Q identified for a gross structural discontinuity and it represents the ratio between the maximum Secondary (Pm+Pl+Q) Stress Intensity (Coulomb-Tresca Theory) or Von Mises Equivalent Stress Value, on the one hand, and the NOMINAL Stress value in the absence of that discontinuity, on the other hand.
Note: Von Mises Equivalent Stress is currently used by ASME VIII-2 post-2007 Editions. ASME B31.3 Code is still using Stress Intensity concept based upon Tresca Theory.

For pipe stress engineer immediate interest, the Peak Stress Intensification Factor (SIF) represents the ratio between the the actual peak stress in the component (e.g.fitting for instance) and the nominal stress in the same sized straight pipe, both stresses developed by the same external load - e.g. bending moment, torsion moment, axial force.

As emphasized in the quoted text above, ASME B31.3 SIFs are HALF (50%) from the corresponding theoretical ASME VIII-2 Peak Stress (Pm+Pl+Q+F) Concentration Factors, because ASME B31.3 SIFs incorporates the "intrinsic" stress intensification factor of the girth butt weld, which was postulated to be 2.00.

As pointed out by PRG Documentation and PRG Software Help explanations, for the Weld Local Stress Intensification Factor's typical assumed value K = 1.35, the correlation between ASME B31.3 (Peak) SIF (i) and Secondary Stress Intensification/Concentration Factor (C) is:
i = (1.35/2) x C.

So, the SIFs values that should be introduced in Caesar II models are the "Peak SIFs" values as given by FEA Reports - one value for each individual load: in-plane bending, out-plane bending, axial load, torsion moment.

The ASME B31.3 Sustained Stress Indexes (SSIs) are NOT equal to Primary Stress Intensification Factors. The ASME B31.3 SSIs are calculated based upon limit load analyses.


Other details may be found in PRG Software manuals, PRG Website articles (download section) and Coade/Intergraph webinars.

Regards,

Dorin Daniel, Thanks for detailed explanation.

If secondary SIFs result in discontinuity stresses, it in general is useful for brittle applications in general (assume a crack size, and you have your discontinuity), but as FE/Pipe does not offer any crack design, its Secondary SIFs are not useful for brittle applications. Therefore, please amend my statement as Dorin has described.

Michael,

I'm afraid you still persist in confusion.

The Stress Concentration Factor (SCF) or Stress Intensification Factor (SIF) concept as used in "classical" pressure vessel (ASME VIII-2) and piping (B31) engineering IS NOT USED in Fracture Mechanics.

Fracture Mechanics operate with the STRESS INTENSITY FACTOR (K) concept, which is fundamentally different than SCF/SIF.

SCF/SIF is a dimensionless quantity and is associated with a structural discontinuity. It may be a gross structural discontinuity (such as tubular junction, shell-nozzle junction etc.) and we deal with Secondary SCF/SIF in this case, or it may be a local structural discontinuity (weld toe, local thickness decrease, or quite a crack), and in such case we deal with Peak SCF/SIF.

However, if the local discontinuity is a crack, the corresponding peak SCF/SIF DOES NOT GIVE ANY PREDICTION REGARDING THE POTENTIAL MANNER OF CRACK EXTENSION.
The Peak SCF/SIF indicates ONLY the secondary stress (Pm+Pl+Q) increase up to the maximum Peak level (Pm+Pl+Q+F), but the crack may extend by ductile/stable mechanism (e.g. crack size increases if load/stress increases) or by unstable/brittle mechanism (e.g. crack increases with high speed even if load/stress decreases), REGARDLESS the value of Peak SCF/SIF.

Stress Intensity Factor (K) used in Fracture Mechanics gives a clear information about the crack extension/growth mechanism.
In the simplest definition form, for a filiform crack of "2xa" (mm) length in an "infinite" thin plane plate subjected to uniform tensile load perpendicular to crack plane/direction, with a normal nominal tensile stress S (MPa), the Stress Intensity Factor corresponding to Mode I of crack growth is:

KI = sqrt(PI*a) * S,

so that Stress Intensity Factor (KI) is expressed in N/sqrt(mm3).

The given crack will grow by unstable/brittle mechanism if Stress Intensity Factor (KI) is equal to or higher than the CRITICAL VALUE OF STRESS INTENSITY FACTOR (KIC), this latest quantity being a MATERIAL Characteristic (e.g. Material Toughness Characteristic) which depends mainly of Temperature - when temperature decreases KIC decreases and the risk of crack unstable growth is higher.

Conclusion: SCF/SIF as used by us, piping stress / preessure vessel engineers, when we perform "classical" design in accordance with ASME B31 and ASME VIII-2 codes, DOES NOT HAVE ANY CONNECTION with Fracture Mechanics prediction assessments.

Regards,

Hello All,

This is indeed a very nice discussion.

I would like to share my idea on this one.

ASME B31.3 does not put distinction between secondary stress (PL+PB+Q in ASME VIII Div. 2 Part 5) and peak stress (PL+PB+Q+F in ASME VIII Div. 2 Part 5). I think both of this requirement is addressed in one expansion stress requirement which is based on 2Sy limit with certain safety factors and which includes also a stress reduction factor (f) to take into account the number of cycles. Basically, Piping code have a more simplified approach than Pressure Vessel Code. Generally, we know that secondary stress addresses the prevention of ratcheting (incremental plastic collapse) which is approximately limited to 3Sm (or 2Sy) which can be considered a Low Cycle Fatigue criteria (where a material can have plastic strains). While the peak stress (PL+PB+Q+F) addresses the High Cycle Fatigue (where a material can fail even the stress amplitude is below the yield strength) and the allowable criteria is based on S-N curves produced from Fatigue Testing which includes different unaccounted factors during test.

Going back to the original question, I agree to use the peak stress SIF to be used for Piping Calculation especially if the system will experience High Cycle Fatigue.

Please correct me if I have made some wrong statements fellow Stressers.

Cheers!!!

For Primary stresses the ASME B31.3 Code Default is to use 0.75xSIF.

But now, we have the SSI for primary type of stress, I will quote from Dorin Daniel reply:

"The ASME B31.3 Sustained Stress Indexes (SSIs) are NOT equal to Primary Stress Intensification Factors. The ASME B31.3 SSIs are calculated based upon limit load analyses."

Any other opinion is greatly appreciated.

Cheers!!!

Whatever Dorin is saying, I bet is right.

Regards,

I'll buy it, too. I shouldn't have hip-fired.

I accidentally responded to the incorrect post when looking up my old instructions on how to upload to the forums.

Nothing to see here...