Pure Penalty vs. Augmented Lagrange Contact in Ansys Mechanical

The contact setting that can hide a bad pressure result

You solve a bolted flange assembly, the job finishes without errors, and the contact pressure plot looks smooth. But if you used Pure Penalty contact, the surfaces may be interpenetrating by a fraction of an element, and peak pressure can be under-reported. That matters when pressure drives a gasket seal check, a bolt yield calculation, or a fatigue assessment.

Takeaway: check the Contact Formulation property before trusting contact pressure. Pure Penalty is faster and easier to converge, but Augmented Lagrange usually gives less penetration and more accurate contact pressure at a higher solve cost. Choose based on whether pressure accuracy or runtime is your constraint.

How the two formulations handle contact

Pure Penalty: a stiffness-based contact spring

Pure Penalty treats contact like a stiff spring between mating surfaces. When nodes penetrate, the solver applies a resisting force proportional to penetration depth and a contact stiffness value. Ansys often estimates this stiffness from adjacent element stiffness, though you can override it.

The method adds no extra degrees of freedom and usually needs fewer global iterations, which helps early sweeps. The downside is that penetration is controlled only by stiffness. Too soft and surfaces visibly interpenetrate; too stiff and the global matrix can become ill-conditioned, amplifying numerical errors and causing convergence stalls or stress oscillations. Because pressure is derived from penetration, pressure accuracy tracks directly with penetration error.

Augmented Lagrange: penalty plus constraint correction

Augmented Lagrange keeps the penalty spring but adds a Lagrange multiplier field—an estimate of contact pressure that the solver updates each iteration. The multiplier drives penetration closer to zero, so results rely less on an artificially stiff penalty.

This typically reduces penetration and improves contact pressure accuracy near bolt holes, gaskets, and sliding interfaces. The cost is extra inner iterations per contact pair, more memory for multiplier history, and sometimes harder convergence when contact opens, closes, or sticks and slips.

Worked comparison: bolted flange submodel

Run this on a small submodel before changing a large production file.

Build the benchmark

1. Model one bolt and its surrounding joint, or use an existing submodel with a bolted flange interface.
2. Mesh the interface finer than the surrounding structure.
3. Apply bolt pretension and an external tensile load representative of service.
4. Insert a Contact object on the flange faces and set friction if sliding is possible.

Run Pure Penalty, then Augmented Lagrange

1. In Ansys Mechanical, select the Contact region and open Details.
2. Set Contact Formulation to Pure Penalty. Leave Contact Stiffness on Program Controlled unless you have a calibrated override.
3. Solve. Record maximum penetration, peak contact pressure near the bolt hole, solve time, and equilibrium iterations.
4. Change only Contact Formulation to Augmented Lagrange. Keep the same mesh, loads, and stiffness. Solve and record the same metrics.

Do not treat any specific pressure or time as universal. Results depend on element type, mesh density, time step, and Ansys release. The useful output is the difference between the two runs on your geometry.

Decision factors and trade-offs

Situation	Suggested formulation	Reason
Early design sweep, many contact pairs, runtime critical	Pure Penalty	Fewer inner iterations; acceptable if small penetration does not change the decision.
Bolted joint, gasket, seal, or interference fit	Augmented Lagrange	Lower penetration and more reliable contact pressure for failure criteria.
Frequent contact status changes	Pure Penalty, or test both	Multiplier updates can struggle with rapidly changing contact status.
Safety-critical pressure with code compliance	Augmented Lagrange	Constraint enforcement reduces reliance on penalty stiffness tuning.

Limitations and how to verify

• Mesh sensitivity: Both formulations need interface mesh convergence. Refine the contact boundary and confirm penetration and peak pressure stabilize before accepting a value.
• Stiffness overrides: Manual stiffness too high can ill-condition the matrix; too low permits excessive penetration. Program Controlled is usually the safer starting point.
• Version differences: Default stiffness algorithms and multiplier behavior can differ between Ansys Mechanical releases. Check Help for your installed version.
• Memory and time: Augmented Lagrange stores multiplier data per contact integration point, so large transient models with many contact pairs use more RAM and take longer per iteration.

Practical check: plot contact penetration and contact pressure on the same interface for both formulations. If Augmented Lagrange shows materially lower penetration and a pressure change that matters for your design limit, adopt it for the critical pairs. Formulation is set per Contact object, so a hybrid model is possible.

Before you change the production model

1. List every contact pair where pressure or penetration is a design output.
2. Run the two-formulation benchmark on a representative submodel.
3. Compare penetration against element size; many teams treat under roughly 1% of local element size as acceptable for qualitative checks, but confirm your own acceptance criteria.
4. Document the formulation, stiffness setting, and mesh size in the analysis report.
5. To revert, change Contact Formulation back to the original setting and re-solve; no geometry is altered by this switch.