With EM Simulation Quickly To New Drive Concepts

Conventional motor solutions in electrical drive technology (EM = Electrical Machines) usually struggle with two fundamental problems even with constant load requirements: firstly, their basic efficiency is often too low, secondly, they usually only achieve their maximum efficiency in a relatively narrow speed range, so that additional gears and / or regulator are required. For many applications, the greatest energy-saving options are specific adaptation of the respective working point to the current load situation.

Picture gallery

Picture gallery with 6 pictures

What are the challenges?

Cross-domain simulation with electro-magnetic-mechanical parameters is anything but trivial. The manufacturer's many years of experience in development, application and a sensible combination of powerful algorithms and the ability to display them in a user-friendly user interface (“GUI”) count here. Extensive, practical and, above all, high-quality libraries are necessary to support the developer in his specific task and not to have the wheel reinvented every time. "Downstream" optimization tools must make it possible to quickly and easily optimize and compare different solutions from different points of view.Ansys, as the long-time market leader in electromechanical simulation, not only offers a variety of simulation packages for practically all tasks, but also appropriate training and on-site support. In the special application in the example, the Electrical Machines Design Toolkit is mainly used.

Additional information on the topic Why EM simulation?

In many modern applications such as robotics, complex dynamic requirements arise that can hardly be solved with conventional technology. Here it is almost inevitable to be able to simulate all possible operating states in advance - also to be able to make well-founded statements about the expected service life and reliability. Modern simulation tools make it possible to simulate on several levels or in different domains ("Multiphysics": electrical, magnetic, mechanical, etc.) and also to include possible loads and to obtain information about dynamic behavior. In this way, complete systems can be viewed in “virtual” operation and their interactions can be optimized in their environment.

Easier working with templates

The example shows the replacement of a three-phase asynchronous motor by a brushless DC motor in a typical drive technology application. Thanks to the optimized design of the motor for the specified speed range, the new solution eliminates the need to use a gearbox, thereby saving weight and costs. For a quick and easy description of electrical machines (simply put: motors and generators), the Ansys Maxwell Simulator contains a user-friendly interface (RMxprt), with which the desired machine type is first selected and parameterized. For this purpose, the basic performance data depending on the physics such as structure (number of poles) and size (dimensions) are first determined.Templates are available for all common synchronous and asynchronous machines, whether single or multi-phase and whether with sliding contacts or in a brushless version (Figure 1).

Simulation in Simplorer

After complete parameterization with material and geometry data from the data sheets, an analytical system model is created from the selected template. This has interfaces to the well-known Simplorer simulator as well as to the 2D and 3D FEM simulators contained in Maxwell and other simulation packages. With Simplorer, complete systems can be described on an analytical (“behavioral”) level, thus simulating effects such as when a load is applied or the effects of various controls. Existing solutions can be mapped quickly and easily and dimensioning and type variants can be tried out. Meaningful areas of application and work of the machine can also be evaluated and verified.A developer will want or need to optimize his solution for more complex tasks. Depending on the level of detail required, the Maxwell 2D and 3D algorithms can be used, whereby the accuracy of a 2D approach is sufficient for most tasks. Highly efficient, optimized algorithms that support modern processes such as multiprocessing ensure the shortest possible simulation times.

Flexible optimization

With the extensions UDO (User Defined Output) and UDD (User Defined Documents) predefined for engine applications, the simulation results can not only be presented in a variety of forms, but can also be used as input for further simulation series for optimization. Not only simple sensitivity analyzes and parameter optimizations are possible, but also complex optimizations with several, also interdependent goals (multi-objective optimization) can be carried out and displayed. Figure 2 shows the efficiency diagram of the DC brushless motor as a function of speed and torque. Darker red illustrates the area of ​​highest efficiency. The large deep red area shows the successful optimization:Even without an upstream gearbox, the engine works very efficiently in a wide speed and torque range. Figure 3 shows the selection menu for further reports of the same simulation. In the picture, the dependencies of input current, total losses, core losses (core losses) and the power factor, as with the efficiency card, are shown depending on the speed and torque. In this way, individual parameters can be specifically optimized or the effects of optimizations on these parameters can be viewed in a targeted manner.Core losses and the power factor as shown on the efficiency map, depending on the speed and torque. In this way, individual parameters can be specifically optimized or the effects of optimizations on these parameters can be viewed in a targeted manner.Core losses and the power factor as shown on the efficiency map, depending on the speed and torque. In this way, individual parameters can be specifically optimized or the effects of optimizations on these parameters can be viewed in a targeted manner.

Content of the article:

• Page 1: Using EM simulation to quickly find new drive concepts
• Page 2: Automated optimization loops
• Page 3: Example for Pareto optimization

> Next page