The Importance of Material Properties in Analysis with SolidWorks Simulation

Have you ever considered the importance of Material Properties to your Finite Element solution? What about the accuracy of the data provided by material vendors? As Designers and Engineers, we are used to dealing with tolerances. We usually provide default tolerances on our drawing title block. We may add tolerances to some of the model dimensions. For the really bold and daring – or wise and experienced – you might even add Geometric Dimensioning and Tolerancing to your designs. Do you ever see tolerances on material property data sheets? In my experience, the answer is somewhere between rarely and never, with never in the lead.

Let’s consider a material from the Simulation material database. Alloy Steel has the following properties (numbers rounded):
Young’s Modulus – 30 x 10^6 psi; Poisson’s Ratio – 0.28; Mass Density – 0.278 lb / in^3; Yield Strength – 90 ksi

What will happen to the Finite Element solution if one material property varies? I’m going to start with a simple model in tension and apply Alloy Steel as the material using the default property values. Then I will change Poisson’s Ratio and re-run the study several times in order to compare the displacement and stress results. Recall that Poisson’s Ratio is a measure of the lateral strain to longitudinal strain for a material, or E_lat / E_long (pretend the E’s are Greek epsilon’s). Poisson’s Ratio is relevant to the linear elastic portion of the stress-strain curve and is unitless. One thing to note, if you do not define Poisson’s Ratio for a material, Simulation will assume that Poisson’s Ratio is equal to zero. There is a pop-up warning, too, just in case you forget to enter a value. Note that if you do not have Poisson’s Ratio for a material, 0.3 is a good initial estimate. But definitely exercise all of your options to find out the correct value for your design materials.

2011-1012-Material-Properties-Blog-OctoberBlog-Results-Displacement1.analysis-620x444

For a ½” square bar, 4” long, I have fixed one end and applied a 10ksi force at the opposite end, putting the bar in tension. After running the analysis with default material properties, I set a baseline with Trend Tracker. After creating several custom Alloy Steel materials, varying Poisson’s Ratio from 0.0 to 0.5, I re-run the analysis with each custom material. Trend Tracker will record the details for maximum displacement and stress in the model.

2011-1012-Excel-Chart-620x409

As you can see from the chart, the Von Mises Stress results vary approximately 16 ksi and the displacement results vary 0.00006 inches. As percentages, this is a 28% variation in stress and a 1.2% variation in displacement. I don’t think most of us would be concerned with 1.2% variation in our models, but 28% is an entirely different matter! I did, however, choose the model with this purpose in mind. The high stresses are at the fixed end at the sharp corner – something most Engineers would avoid in their designs. Now that a 28% variation has your attention, let’s look at a more practical model.

2011-1012-b-Material-Properties-Blog-OctoberBlog-2-Results-Displacement1.analysis-620x431

This is a simple bracket, a modification of a part in the SolidWorks Essentials manual. I’ve applied a fixed boundary condition to the bolt holes in the base and a normal force to the counter bore face. I’ve repeated the rest of the analysis, just like the square tensile bar, including using Trend Tracker and varying Poisson’s ration from 0 to 0.5.

2011-1012-b-Excel-Chart-620x409

For this “practical” model, the Von Mises Stress results vary approximately 4,300 psi and the displacement results vary 0.0006 inches. The percent variation in this model is 5.8% for stress and 6.6% for displacement. If you’re designing for a large Factor of Safety, less than 6% variation in your stress results is not significant. Using SolidWorks Simulation, however, most of us are designing for the lowest acceptable Factor of Safety in order to save the maximum amount of money possible on material costs. In this scenario, a 6% variation can be significant!

So the next time you’re analyzing that awesome design, consider reviewing the sensitivity of your analysis by varying a material property or two. Now you’re armed with powerful information you can share in your next design review. Material property variations from your vendors are no longer an issue! Now go make your products better with SolidWorks Simulation!

***

Bill Reuss is a CAE Specialist at 3DVision Technologies, a SolidWorks Value Added Reseller with locations across Ohio, Indiana, and Kentucky.  He is a regular contributor to 3DVision Technologies’ Blog where you will find new ideas to improve your productivity with SolidWorks Simulation.