FEM logo

How to correct your Von Mises criterion to also check buckling?

Plate, stiffener and girder strength analysis.

Industrial FEA software like FEMAP or ANSYS don't contain practical solutions for the check of plate structures other than the Von Mises criterion only. An evaluation if the structure also complies with the required standard can be very time consuming.

The buckling Eigenvalue method is an incorrect buckling analysis method and leads, conservatively applied, to heavy designs. To include built-in tolerances in your FE model is a hassle as also explained in 'How to verify your design'. Therefore, this should be included in the justification method.

Plate structures, including stiffeners and girders, should be verified according to the international standards provided by e.g. DNVGL and ABS. The FEMDS software module is based on the guidelines in the following documents:

Both standards are covered in the hand calculation tool provided by FEMDS.COM:



In the software the same limitations apply as stated within these codes. Note that DNV-RP-C201 is equal to the Eurocode EN 1993-1-5 for classes 1-3 as specified in EN 1993-1-1.

The software contains three modules:

  • Plate buckling check
  • Stiffener buckling check
  • Girder buckling check

Each module is programmed as separate modules, as also given in the standards.

Since each code uses its own approach, ABS did a comparison between buckling codes of DNV, ABS and API. This comparison can be downloaded here. Actually this comparison only presents clear results concerning panel buckling but is vague on the comparison of stiffener and girder buckling.

API Bulletin 2V, C6.2 stated the following:

“…. there are no widely accepted methods to deal with the problem of combined biaxial compression and shear, and the same applies to the case where lateral pressure acts together with these loads. Engineering judgement must be used to address such cases.”

Although the ABS comparison references DNV CN30.1 which is partially replaced by DNV-RP-C201, the same methodology is followed to verify and compare the two methods for panel buckling in the compression - compression quadrant.

For plate buckling both codes use the same equation for bi-axially loaded plates with edge shear:

In the software this equation is used in all quadrants resulting in the following failure envelope (in this case with edge shear set to 0). Note that in all other quadrants than compression - compression, φ=1.0

From the figure above it is directly clear why the Von Mises criterion is too optimistic when compression is applied in a plate construction.

For stiffener and girder buckling, both codes use different approaches. This is made clear in the graph below, which shows the different buckling curves for DNV, ABS but also API. The reason to show API is that from correspondence with ABS it appeared that they adopt parts in their code from API. However, as shown in the graph below all codes show completely different curves.

These different curves are due to different methods and philosophies, which consequently lead to differences in calculation results. Especially when combining the biaxial compression loads, shear and lateral pressure, large differences can be discovered between the codes. More reasons for these discovered differences between the buckling codes of DNV and ABS are given in the code comparison;.

You can try this yourself by using the hand calculation tool of FEMDS.COM. Click on the below 'Panel buckling check' button.

Customer area

Sign in




Forgot your password?
Hand calculations

Latest software release

Sign up for a free trial <