有限元分析：究竟什么是好网格？

Good mesh need resolve physics, not follow the CADmodel.

好网格应该很好地描述物理本质,而不是符合CAD模型。

A good mesh is always associated with the physics. Whenyou change your boundary conditions, or change your loadconditions, or change analysis types, or change flow models, goodmesh (with the identical geometry) may turns to be badmesh.

好网格必须始终反应物理本质。当你改变边界条件、载荷、分析类型或者流体模型时,质量好的网格可能会变成差网格。

The accuracy advantage of hex meshis no longer existing anymore, for most engineering problems.

对于很多工程问题,六面体在精确度上的优势已经不存在了。

When the software vendor needjustify that his meshing software is in the high-end (of courseprice is usually in the high end as well), he will tell you thesoftware allows you to all sorts of manual controls. The hiddeninformation is only manual controls can generate better mesh.

当软件供应商向你推销网格划分功能时,他会告诉你这款软件支持各种各样的手动划分网格的方法。隐含意思就是手动控制网格才能得到好网格。

Yes, to salesman, good mesh need manualcontrol. But to engineer, you need understand this is misleading. Agood meshing software should have the intelligence to analyze theunderlying geometry: calculating curvatures, finding gaps; findingsmall features; finding hard edges; finding sharpangles, applying sensible default settings….

是的,对于销售者来说,好网格需要手工控制。但对于工程师而言,你应该知道这其实是在误导。一个好的网格划分软件应该智能地去分析它将要划分的几何模型:计算曲率、识别间隙、识别小特征、识别锐边、识别尖角,然后进行合理的默认处理(无需人工干预)。

<译者按:这也许是前处理软件的发展趋势>

Fine mesh does not mean good mesh.The purpose of meshing is toget solutions in discrete locations. Good mesh is the mesh thatserves your project objectives. So, as long as your results are: 1)physical; and 2) accurate enough for yourproject, your mesh is sufficiently good.

精致的网格不等于好网格。画网格的目的是得到离散点上的值。只要能满足工程需要就是好网格。保证两点:1)表达物理本质;2)达到工程所需的精度。

Another example of thismisunderstanding is that most users always use full 3D model. Intheir eyes, 3D full model is “realistic”. However, if the problemis symmetry, using portion of your geometry (therefore lessnodes/elements) will produce better results because the symmetriccondition is exactly enforced. If the problem is axis symmetric, 2Dresults is far more accurate than most 3D results.

这一误区的另一个表现是:总是使用完整的3D模型。在他们眼里,完整的3D模型才是最逼真的。然而,如果问题是对称的,使用模型的一部分(这样可以减少节点和单元数量)可以获得更好的结果(因为这样会精确反应对称性)。如果问题是轴对称的,使用2D模型比3D模型更精确。

Currently, CAE still replieson meshing. Good mesh should:

1. be able to resolve the physics to bestudied;

2. have reasonable quality so that the solver will notfail;

3. simplify the geometry based onphysics;

4. be problem specific;

5. meet your project requirements

目前,CAE仍然依赖于网格划分。好网格应该做到以下几点:

1、能够解决所研究的物理问题;

2、网格质量合理,避免计算失败;

3、基于物理本质简化几何模型;

4、具体问题具体对待(对同一个几何模型做不同的分析,可能需要划分不同的网格)

5、满足工程需求

内容转自:

http://caewatch.com/top-5-misunderstandings-on-good-mesh/