- 积分
- 33150
- 回帖
- 0
- 西莫币
-
- 贡献
-
- 威望
-
- 存款
-
- 阅读权限
- 100
- 最后登录
- 1970-1-1
签到天数: 230 天 连续签到: 15 天 [LV.7]常住居民III
|
本帖最后由 zengxiaodong 于 2015-3-6 22:20 编辑
以下为引(转帖)用内容:
I am a long-time Ansys user and a casual user of Maxwell 3D. With that in mind, here are my impressions of the two programs.
I find Ansys to be more reliable and robust, however, for a quick answer to simple magnetostatic problems, I choose Maxwell.
Unfortunately, Ansoft has a very "black box" attitude about their software. There's very little information about how and what the program does. They don't have anything like the Ansys theory manual in their documentation set.
Maxwell's default solution mode is based on successive mesh refinement of h-type elements (tets only). I have found that it will sometimes converge to an erroneous solution for magnetic forces. Re-running the analysis usually fixes that. Ansys takes longer to set up and longer to solve (for simple problems), but the solution is generally reliable. Maxwell's built-in geometry creation is close to useless -- you need a separate CAD program to create anything even slightly interesting. Maxwell does not have very good mesh control.
Maxwell has a very nice and powerful postprocessing capability. It also has a powerful macro language, but it is poorly documented and somewhat inconsistent. However, once you figure out how to use it for postprocessing, it makes Ansys EM postprocessing look clumsy.
I have heard that Ansoft will be adding a linear structural solver, but I don't know how well integrated it will be with the EM solvers.
One of the biggest drawbacks of Maxwell is its miserable documentation -- nothing like the very good documentation available for Ansys.
I don't know how Maxwell program support is handled internationally, but in the US it all goes through the two Ansoft offices. My experience has been good with support issues. With Ansys, of course, the quality of support depends on where you are, since it is handled by regional offices.
In summary, I usually use Maxwell for EM solutions, but the problems I need to solve are typically very simple, with structural interactions that can be handled separately. In particular, I would be very cautious of Maxwell's electromagnetic force calculations. For more complex analyses, Ansys is clearly the program of choice.
|
|