Getting Started With SolidWorks Simulation Basics

SolidWorks Simulation 2011 is a built-in finite element analysis module that comes bundled with the standard SolidWorks suite. It sits under the Tools menu when you open a part or assembly file. The interface looks like every other SolidWorks feature — PropertyManager on the left, graphics area in the middle, standard icons along the top. There is nothing fancy about it visually. The software handles meshing, applies boundary conditions, runs a linear static analysis, and presents results in color contour plots. That is the full workflow on the surface. I learned FEA through this tool about ten years ago because it was already on our computers. I did not have ANSYS or Abaqus available at the time. It turns out SW Simulation was sufficient for a lot of the structural checks our shop needed — bracket stress, shaft deflection, basic bolted joint pretension. It is not going to handle nonlinear contact problems, plasticity beyond basic yield, or transient dynamic loading without workarounds. But for straight static analysis on solid geometry, it works fine.

Introduction To Finite Element Analysis Using Solidworks Simulation 2011

Setting up a basic analysis requires four sequential steps. First, you create or open a solid part file. Make sure the geometry is clean — no missing faces, sliver bodies, or tiny features that will eat your mesh. Second, you apply a fixed geometry or imposed displacement to constrain all rigid body motion. If your model has any unconstrained degrees of freedom, the solver will either fail or produce garbage results that look plausible at a glance. Third, you apply loads and connections — force, pressure, thermal load, or mates acting as physical connections. Fourth, you run the study and review the results.

The mesh is where most people go wrong. By default, SolidWorks applies a global mesh size of L/20 where L is the smallest feature dimension. That often produces far too few elements for accurate stress prediction. I usually set my initial mesh to a manual element size based on the critical dimension — something like 5mm for a bracket, 2mm for a small gear. You can add local mesh refinement around holes, fillets, and load application zones without remeshing the entire model. Right-click the mesh folder in the FeatureManager and select Apply Mesh Controls. This cut my setup time from about 30 minutes per model down to maybe 8 minutes once I stopped relying on defaults. Mesh convergence is something the software will not tell you automatically. You need to run the analysis at two or three different mesh densities and compare the maximum von Mises stress. If the stress changes by more than 5-10 percent between your coarsest and finest mesh, your mesh is still too coarse in the critical region. I keep a table in a separate spreadsheet tracking mesh size versus peak stress for each model. It takes a few extra minutes upfront but saves you from submitting a report with a number that is off by 20 percent.

Common Pitfalls I Learned the Hard Way

The biggest issue beginners face is ignoring stress concentrations at geometric discontinuities. The software will give you a peak stress value at a fillet or hole edge, and that number is almost always artificially high because the mesh cannot resolve the true singularity. A sharp re-entrant corner in the mesh will produce a stress that keeps climbing as you refine — mathematically, it goes to infinity. In practice, you just need to refine until the number stabilizes within a reasonable band, then move on. Never report the absolute maximum node result as your design stress without checking what happened at the node.

Another problem that caught me early on is misinterpreting the unit system. SolidWorks Simulation does not do automatic unit conversion. If your model is in millimeters but you input a force in Newtons and a material property in Pascals, the solver treats them as written and the results are meaningless. I spent an afternoon debugging a study that produced von Mises stresses in the gigaPascal range for a part made of aluminum before realizing I had entered Young's modulus in kPa instead of MPa. Always check your units in the Properties panel before running anything. I now keep a conversion cheat sheet open on my second monitor. Material selection matters more than you would think for this version of the software. The default library contains a few hundred materials, but they are not all accurate. Some entries have incorrect Poisson ratios or temperature-dependent properties that are simply wrong. I found this out when a thermal stress study on a steel bracket produced expansion values that were half of what the handbook says. I ended up entering custom material properties for anything beyond the standard structural steels and aluminum alloys. It takes about 10 minutes per material to look up the correct values and enter them manually, but it prevents bad data from propagating through your analysis.

A Specific Edge Case I Encountered

I was analyzing a thin-walled aluminum housing with press-fit bearing seats using SW Simulation 2011. The assembly had contact pairs between the bearing outer diameter and the housing bore. The software's default contact setting was "no penetration" with a default bond option. The initial run showed zero deformation at the interface — it was treating the contact as a bonded joint rather than a frictional press fit. I had to explicitly set the contact type to "frictional" with a coefficient of 0.15 for aluminum-on-aluminum dry contact and define the normal pressure from the interference fit calculation. Even then, the solver struggled with convergence because the thin walls were flexible and the contact detected them separating locally. I resolved it by adding mesh refinement around the contact zone, switching the solver to use the "Newton-Raphson with substeps" option, and reducing the initial contact tolerance from the default 0.01mm to 0.001mm. The full analysis that previously failed in 45 minutes after 200 iterations now ran to completion in about 12 minutes on a standard workstation.

What This Tool Cannot Do Well

SolidWorks Simulation 2011 is strictly a linear static solver at its core. It can do basic modal analysis and a limited thermal study, but it cannot handle large deformation, material nonlinearities past yield, time-dependent contact, or explicit dynamics. If your design involves impact loading, buckling past the critical load, or any scenario where the geometry changes significantly during loading, this tool will give you answers that look professional but are structurally incorrect. I have seen engineers use it for drop tests and get results that looked convincing until someone compared them against actual test data. For more complex problems, you need a dedicated solver. ANSYS Mechanical or Abaqus Standard will handle nonlinear contact, hyperelastic materials, and explicit dynamics. The tradeoff is that those programs have a much steeper learning curve and require separate licensing. SW Simulation occupies a useful middle ground — faster than starting from scratch in a dedicated FEA package, but not infinitely capable. Know where that line is for your particular problem before you invest time building a model.

Practical Workflow Tips

Use the Design Study feature if you need to evaluate multiple design variations. You can define design variables (like thickness or hole diameter), set up responses (like maximum stress or displacement), and let the software run a parametric study. It automates what would otherwise be eight or nine manual runs. I used it to optimize a support bracket by varying wall thickness across 12 iterations. It took about 20 minutes total versus what would have been several hours doing it by hand. Save your studies as separate .sldstudy files within the same part document. Do not create new part files for each load case — you will end up with version control problems and will waste time rebuilding constraints. The software allows multiple studies in a single file. I typically keep a base study with mesh and material properties, then duplicate it for each load condition. Changes to the geometry automatically update all studies, though you may need to rebuild the mesh if you modify anything that affects element topology. Documentation of your assumptions is something I started doing seriously after a design review pointed out that I had applied a safety factor of 2.0 on one study but 1.5 on another without explaining why. Now I include a short notes section in every study — what loads were applied, how constraints were modeled, what mesh size was used, and what assumptions were made. It takes 5 minutes per study and has saved me from having to redo work when questions came up months later.