Design Optimization and Fatigue Studies in SolidWorks Simulation Professional by Brian Zias
Intro Every product is subject to real-world conditions that ultimately cause failure in one way or another. SolidWorks Simulation is a tool to computationally quantify the performance of our assemblies and components inside SolidWorks. Using the extended capabilities of SolidWorks Simulation Professional, a company's development team can model more real-world challenges and scenarios.
Typically designs are sensitive to cost, safety, durability, and quality. Making use of today's software leads toward higher product quality, lower costs, longer durability, and far less physical testing. This will propel company innovation.
How does software, such as Simulation, add value to a company? In other words: As engineers and designers, what is our value? Our value is our production of technical work. If we have tools with which we can create, publish, and validate our technical work more quickly, with fewer errors, higher quality, and greater consistency, how will this affect our top and bottom lines? These topics are exactly what we refer to when we say "SolidWorks helps you Design Better Products." We are ultimately going to increase the effectiveness and efficiency of your workflow by adding key capabilities such as SolidWorks Simulation Professional.
The example product is a medium-duty paint sprayer. This is a product I am sure you have seen in the home improvement store.
Design Optimization Study
The nature of design work is iterative. The preliminary design is subjected to scrutiny, maybe physical testing, and then modification after the evaluation. This cycle is repeated causing project time and costs to increase. The more complex the product, the more design options exist. A Design Optimization study is a tool to rapidly evaluate many design options and pick the best.
One benefit of the design study is the ability to take the designer out of the loop for awhile as SolidWorks rapidly iterates and evaluates design options. This shortens design time dramatically. As we try and get our products to market faster, the proper use of simulation and design studies becomes invaluable. We not only save design time, but ultimately we can trim down material usage, which also leads to decreased cost.
It is also very simple to use. Inside SolidWorks, we specify variables, constraints, and goals. Simply, variables are SolidWorks parameters, constraints are product specifications, and goals are quantitative measurements of the design. An example would be the design of a sheet metal bracket or shelf: we vary the sheet metal thickness, achieving the lightest possible weight while keeping factor of safety under maximum load above three.
Parameters, or variables, are usually model dimensions. In reality almost anything that has a number in SolidWorks can be linked to a variable. If you can configure it, you can vary it. I took a portion of the paint sprayer handle and specified for variables the wall thickness (anywhere from 0.025" to 0.060"), the number of ribs (2, 3, or 4), and the spacing of the ribs (0.25" to 1.0"). Rib thickness is linked to shell thickness to keep the mold designer happy. The goal was to obtain a model with the least plastic volume while still keeping stress below 3500 psi during a worst-case squeezing load (100 lbf each side). While these parameter changes are easy enough to do manually, there are literally an infinite number of combinations of possible values. Design studies leverage design of experiment routines to figure out the sensitivity of each parameter to the goals and constraints. After 5 minutes and 15 studies, I get the output of optimum design, and we can take a look at some of the versions attempted by the software:
Ultimately I was able to rapidly evaluate different design options and pick the best one that resulted in 10% decrease in plastic volume from the initial design.
Analyzing Fatigue in this design
What components in our design will be subjected to thousands or millions of loading cycles? We need to consider whether or not material fatigue will be an issue. Fatigue analysis will let us balance the optimum design with the life of the product. This could help to estimate working life and more importantly show me the change in life after I make a design change.
Stress-life fatigue behavior is quantified by S-N curves which are developed experimentally. Typically, plastics are not characterized by S-N curves, so we don't apply Fatigue in a quantitative manner to these types of materials. Instead, what I do is compare my original (non-optimized design) to the final design. Given the same S-N curve, how much life am I sacrificing by optimizing my volume (and thereby decreasing yield FOS). Let's take my initial design and final design from above, and assume a typical logarithmic decaying S-N curve. Applying the previous static study in a zero-based fashion results in a "life" plot, showing the number of cycles until ultimate fatigue failure:
By saving that 10% of volume, the peak stress increased. While the increase was still below my maximum design stress of 2500 psi, this increase caused a decrease in stress life from 1,756 cycles to 251 cycles. Granted this is a repeated application of the worst-case squeezing, but it goes to show that there are trade-offs to every design. More information is typically better. As we push the limit with material optimization, we utilize the fatigue to further bound our design space to ensure a quality product.
Look for our other Study examples elsewhere in our Blog on Thermal and Frequency Simulation.