This question is pinpointed toward ANSYS Workbench users. I'm looking at 2017 ASME Sec. VIII, Div. 2, Part 5.4.1.2(a). Workbench has an Eigenvalue Buckling analysis to ascertain the bifurcation load of an assembly. It has always been my understanding that an Eigenvalue Buckling analysis in Workbench is a low fidelity, first pass at a model to determine a rough ballpark bifurcation load. In other words, an engineer should not put a lot of faith in the outcome of said eigenvalue analysis in Workbench. So I hesitate to use that analysis to satisfy 5.4. However, I don’t know what alternative I have in Workbench. Using ANSYS, how would one determine the bifurcation buckling load?
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations TugboatEng on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
ANSYS Eigenvalue Buckling Analysis for Sec. VIII, Div. 2, Part 5.4 2
- Thread starter LS_SMS
- Start date
-
1
- #2
centondollar
Structural
The eigenvalue analysis is the so-called "bifurcation analysis". Eigenvalue buckling analyses may be reasonable if the structure is imperfection-insensitive (plates, stiffened plates, many frames etc.), while other analyses must be undertaken if the structure is highly imperfection sensitive (shells, arches etc.).
Determining the collapse load for simple structures (not a ship, necessarily, but maybe your assembly) with reasonable accuracy can be done in modern FE software by running a fully geometrically and materially non-linear analysis with either moderate rotation and small strains (von Karmán non-linearity) or large rotation and strains. The analysis proceeds roughly by:
a) determine the eigenvalues and eigenvectors in bifurcation analysis (buckling load and buckling mode shapes).
b) apply a suitable buckling mode (the first one usually, but not if the geometry is very complex and the eigenvalues are closely spaced) as an initial imperfection to the model and scale it with e.g., code-mandated bow imperfection values.
c) Define the elastic-plastic material law and kinematics (von Karmán or large displacement); if the process is expected to be highly dynamic (involving snap-through), also define mass properties to account for inertia.
d) Run a non-linear analysis (with a robust solver, e.g., modified Rik´s method, that can handle non-monotonic force-displacement and snap-back) which traces the force-displacement curve accurately. If inertia effects are included, step d) also involves an explicit or implicit time-stepping algorithm..
Note that superposition does not hold in non-linear analysis, so all loads in step d) must be applied in the order and time they are expected to occur at in reality. You should consult your colleagues about the analysis procedure. A description of "non-linear buckling analysis" (as opposed to linear buckling analysis, which is eigenvalue analysis) can be found by googling.
PS. Remember to take all advice on the internet with a grain of salt - even my advice! - and verify it independently as thoroughly as you can muster.
Determining the collapse load for simple structures (not a ship, necessarily, but maybe your assembly) with reasonable accuracy can be done in modern FE software by running a fully geometrically and materially non-linear analysis with either moderate rotation and small strains (von Karmán non-linearity) or large rotation and strains. The analysis proceeds roughly by:
a) determine the eigenvalues and eigenvectors in bifurcation analysis (buckling load and buckling mode shapes).
b) apply a suitable buckling mode (the first one usually, but not if the geometry is very complex and the eigenvalues are closely spaced) as an initial imperfection to the model and scale it with e.g., code-mandated bow imperfection values.
c) Define the elastic-plastic material law and kinematics (von Karmán or large displacement); if the process is expected to be highly dynamic (involving snap-through), also define mass properties to account for inertia.
d) Run a non-linear analysis (with a robust solver, e.g., modified Rik´s method, that can handle non-monotonic force-displacement and snap-back) which traces the force-displacement curve accurately. If inertia effects are included, step d) also involves an explicit or implicit time-stepping algorithm..
Note that superposition does not hold in non-linear analysis, so all loads in step d) must be applied in the order and time they are expected to occur at in reality. You should consult your colleagues about the analysis procedure. A description of "non-linear buckling analysis" (as opposed to linear buckling analysis, which is eigenvalue analysis) can be found by googling.
PS. Remember to take all advice on the internet with a grain of salt - even my advice! - and verify it independently as thoroughly as you can muster.
-
1
- #3
centondollar - excellent advice. That is basically what the 2023 ASME Code rules are going to be.
m_rizdon (and anyone else who wants it) please contact me offline and I will send to you the 2023 Buckling Rules. The current rules, especially with respect to the eigenvalue buckling rules are potentially unconservative and should not be used. Or, if you would like, you can follow what's in this paper.
m_rizdon (and anyone else who wants it) please contact me offline and I will send to you the 2023 Buckling Rules. The current rules, especially with respect to the eigenvalue buckling rules are potentially unconservative and should not be used. Or, if you would like, you can follow what's in this paper.
- Status
Similar threads
- Locked
- Question