When and How to Enable Large Deflection (NLGEOM) in Ansys Mechanical

Start With the Failure Mode

Linear static analysis assumes that deformations are small enough that the original mesh geometry represents the structure throughout loading. When a part rotates or deflects significantly, that assumption breaks down. The stiffness matrix assembled from the undeformed mesh no longer reflects the actual load paths, and the solver can return displacements that are too small or stresses in the wrong location. The practical fix is to enable Large Deflection, also known as NLGEOM.

The Useful Takeaway

Turn on Large Deflection when you expect rotations greater than roughly 5–10 degrees or displacements that are a significant fraction of the model size. This tells the solver to recompute the stiffness matrix using the deformed geometry at each iteration, which improves the accuracy of both displacement and stress predictions for geometrically nonlinear problems.

How NLGEOM Changes the Solver

In a standard linear solution, the solver assembles the global stiffness matrix once from the original mesh and solves the equilibrium equation. With NLGEOM active, the solver treats the displacement field as part of the problem: it updates the element Jacobians (the coordinate mappings used to calculate element stiffness) and stiffness contributions based on the current deformation state, then iterates until internal forces balance the applied loads. In effect, the structure's stiffness is re-evaluated as it moves.

Worked Configuration

In Ansys Workbench Mechanical

1. Select the analysis branch in the tree (for example, Static Structural).
2. Open Analysis Settings.
3. Set Large Deflection to On.

The setting applies to the entire analysis unless a command snippet overrides it.

In Ansys Mechanical APDL

Issue the following command before solving:

/NLGEOM,ON

Place it after the model is fully defined but before the SOLVE command. This activates geometric nonlinearity for the subsequent solution.

Concrete Example: Cantilever Beam

Consider a cantilever beam fixed at one end with a transverse tip load. If the tip rotates by only a few degrees, linear theory is adequate. If the load is high enough that the tip rotates by 20 degrees, the beam's effective stiffness changes because the load line and geometry shift. A linear solution will under-predict the tip deflection.

With NLGEOM on, the solver accounts for the stiffening or softening effect of the deformed shape. You should expect the nonlinear solution to show a larger tip deflection than the linear result, consistent with beam theory for large rotations. This comparison is a quick diagnostic: if the two solutions are nearly identical, the deformations are probably still in the linear range.

Expected Checks and Risks

After enabling NLGEOM, inspect the solver output for the number of substeps (intermediate load increments) and iterations. A well-posed problem should converge within a reasonable number of iterations per substep. If convergence fails, reduce the load increment by increasing the number of substeps or using smaller time steps. Large load steps are a common cause of divergence in geometrically nonlinear runs.

Limits and Common Mistakes

• Increased cost: NLGEOM raises solver iterations and memory use because the stiffness matrix is updated repeatedly.
• Material nonlinearity is separate: NLGEOM does not account for plasticity, hyperelasticity, or crushing. You must also enable the appropriate material models.
• Unnecessary noise: Applying NLGEOM to a small-deformation model can introduce numerical noise and longer runtimes without improving accuracy.
• Contact compatibility: If contact is present, ensure the formulation is compatible with large deflection. Pure penalty contact (a formulation that allows small penetration) without proper adjustments can cause penetration or divergence when NLGEOM is active.
• Mesh quality: Poorly shaped elements in regions of high deformation can amplify convergence problems when NLGEOM is active.

How to Verify the Result

Run a simple cantilever beam test. Apply a tip load known to produce large rotation. Run one analysis with Large Deflection off and one with it on. Confirm that the nonlinear solution shows a larger tip deflection than the linear result.

Next, refine the mesh by roughly halving the element size and re-solve. The NLGEOM result should converge. Monitor the iteration count and residual norms (a measure of how closely internal forces balance applied loads) in the solver output to confirm stable behavior. If the displacement changes significantly with mesh refinement, the mesh is not yet adequate for the nonlinear regime.

Version and Environment Notes

These instructions apply to Ansys Mechanical in the Workbench environment and to Ansys Mechanical APDL. The exact location of the Large Deflection control may vary slightly by release, but the setting is available in Analysis Settings for supported structural analyses. Always confirm that your specific release exposes the NLGEOM control in the expected location before running production jobs.