One question often facing engineers performing any type of simulation analysis is, “How can I simplify my model and still get accurate results?”
A case in point for FEA analyses: When engineers have a large assembly with multiple parts but are interested in the simulation results of only one of the parts, only the part of interest needs to be fully modeled and meshed. The remaining parts can be reduced to a point mass where their effects are accounted for, but they are not fully meshed. As a result, this approach significantly reduces the size of the overall model as well as the solution time. In other words, it is a much smaller model to solve.
As electric machines become smaller and faster, designers face the growing challenge of delivering machines that meet all the electrical, mechanical, and thermal design goals in a very limited timeframe.
Electric Machine AnalysesFor example, in the automotive industry, electrical machines commonly need to be:
Compact, given the limited space available
Lightweight and Efficient, with increased driving range and extended battery life
Quiet, for maximum passenger comfort
With multiple types of analyses required, designers need to seamlessly move from one analysis to another efficiently. And a problem that presents itself downstream may require that upstream analyses be revisited. For instance, if the thermal analysis reveals an overheating issue, then perhaps the upstream electrical analyses needs to be adjusted and the process started again. Furthermore, once an acceptable design is completed, the design may not yet be optimal. It just means that the design is in the ballpark.
In manufacturing settings, a mixing tank is often used to prepare materials for production. Large amounts of materials are put into an enormous tank in order to blend components together through refined mixing.
CFD Simulation of Mixing Tank Process Mixing tanks are an important piece of equipment in industries such as:
Oil & Petroleum
Flue Gas Desulpherization
Paints & Coatings
Adhesives & Sealants
Polymer & Plastics
The $64,000 question for the manufacturer: Do the quality and consistency of the final mixture meet the intended goals? If not, what needs to change? Obviously, the manufacturer can adjust the material inputs, the speed of the mixer, and other factors through trial and error. However, this approach provides information only on the inputs and outputs. If I add so much of product A and so much of product B, and mix at a certain speed for a certain time and at a certain temperature, then I get product C. This method doesn’t offer much information on the actual mixing process itself.
Computational Fluid Dynamic (CFD) simulation software can provide much more insight into the mixing process, showing the manufacturer a detailed picture of what is happening inside the tank and how the materials are reacting over time. For example, what if the mixing process could be shortened by 30 seconds? What kind of economic impact would that make?
Stress linearization is a technique used to decompose a through-thickness elastic stress field into equivalent membrane, bending, and peak stresses for comparison with appropriate allowable limits.
Membrane stress is the average stress through the thickness (P/A)
Bending stress is the linearly varying stress through the thickness (Mc/I)
Peak stress is the total stress minus the membrane plus bending stress
Stress linearization was developed in the pressure vessel industry by Kroenke and Gordon in the 1970s. It was first implemented in Section III (nuclear vessels) and Section VIII, Division 2 (commercial vessels) of the ASME Boiler and Pressure Code. Although widely used in industry thereafter, specific guidance was not included in ASME Boiler & Pressure Vessel Code (BPVC) Section VIII Div. 2 until 2007.
While ANSYS Workbench can be used to validate designs, care must be taken in how the stresses and their locations are handled.
Ron Gagnon Sr FEA Consulting Engineer Fatigue & Fracture ExpertMechanical fatigue is a VERY popular topic with the ANSYS user community, with many engineers and designers attending our fatigue-related webinars over the past few years.
Luckily, all our fatigue-related webinars—as well as all recent webinars—are recorded and available anytime. Learn more about mechanical fatigue and how to use ANSYS and nCode DesignLife to solve fatigue models in these webinars: