找回密码
 注册
Simdroid-非首页
查看: 258|回复: 9

[其他] OFELI-- 一个中规中矩的面向目标有限元软件

[复制链接]
发表于 2012-7-9 16:16:55 | 显示全部楼层 |阅读模式 来自 日本
主页: www.ofeli.net
开发者:Rachid Touzani
Laboratoire de Mathématiques
Université Blaise Pascal (Clermont-Ferrand II)
License:  GNU Lesser General Public License
使用语言;c++
依存外部库;无
基本功能; 热传导,固体变形,流体变形,电磁场
文件格式; 自定义xml格式,附属工具可将输出文件转换成数种格式的可视文件
并行计算;不可

一个完全独立的有限元软件,没有出乎意料的独特功能,其软件设计可按照常用的方式由网格,矢量,矩阵,材料,形状函数,方程组成。
不依存外部库,极好编译。尤其对windows,提供了VC,公开源码开发环境CodeBlocks(http://www.codeblocks.org/)的开发文件,虽有一些小的错误,但基本可以运行。但不用外部库的缺点是——运行速度绝对不会太快。
所附文档可算详细。
发表于 2012-7-10 02:15:34 | 显示全部楼层 来自 美国
Simdroid开发平台
I don't know if you would agree with me. I find that all the open FEM source codes did a very poor job in providing the theory manual. In one sense, it is understandable: No one would like their many years of work to be picked up in just a couple of months by providing detailed theory manual.
回复 不支持

使用道具 举报

 楼主| 发表于 2012-7-10 09:07:32 | 显示全部楼层 来自 日本
tonnyw 发表于 2012-7-10 02:15
I don't know if you would agree with me. I find that all the open FEM source codes did a very poor j ...

Well, I agree. But I don't agree to your second argument. If no one would like their many years of work to be picked up, just don't open their source code, right?
回复 不支持

使用道具 举报

发表于 2012-7-10 10:42:07 | 显示全部楼层 来自 美国
hillyuan 发表于 2012-7-10 09:07
Well, I agree. But I don't agree to your second argument. If no one would like their many years of ...

They would like you to pick up the source codes. But they don't let you pick up in a very short time. Without theory manual, it is harder to understand the source codes.
回复 不支持

使用道具 举报

发表于 2012-7-10 11:55:17 | 显示全部楼层 来自 浙江杭州
tonnyw 发表于 2012-7-10 02:15
I don't know if you would agree with me. I find that all the open FEM source codes did a very poor j ...

个人觉得开发者默认下载并阅读代码的人了解有限元基础理论和编程知识,所以详细的理论手册不是必须的
此外,有些开源代码的使用范围很狭窄,不对这个领域有深入了解的话,不会去下载并阅读这类代码
回复 不支持

使用道具 举报

 楼主| 发表于 2012-7-10 12:38:17 | 显示全部楼层 来自 日本
本帖最后由 hillyuan 于 2012-7-10 12:40 编辑
tonnyw 发表于 2012-7-10 10:42
They would like you to pick up the source codes. But they don't let you pick up in a very short ti ...


Maybe following pasuka's comments say something about it. You may consider the volume of ABAQUS's theory manual, whcih is still not that detailed. It's a hard work for developer of open source and maybe with little meaning considering their reader may know the detail themself allready. I myself, e.g., has found little difficult to read those source code, of course in the field I am quite familar to, finely orginized.

On the other hand, even if you know the relevent theory quit well. It is still hard to really understand a software. Just like, let me see, even if you know how a car run, it doesn't mean you know how to manufacture a car.
回复 不支持

使用道具 举报

发表于 2012-7-11 01:01:45 | 显示全部楼层 来自 美国
I don't mean understanding the theory in general which is not hard. I guess what I mean by theory manual is toward the implementation part. Namely, it should include all the important theoretical details which helps user read the code and then modifies it for his own purpose.

For instance, in the case of FEAPv open source code, you cannot find theory description about shell element. How could the user read the source code? Most likely, the user has to read all the relevant papers and figure out all the theoretical details before he can read the code.
回复 不支持

使用道具 举报

 楼主| 发表于 2012-7-11 09:06:46 | 显示全部楼层 来自 日本
tonnyw 发表于 2012-7-11 01:01
I don't mean understanding the theory in general which is not hard. I guess what I mean by theory ma ...

It is impossible to explain the implementation detail in any theory manual. Even in books specilized in FEM provide no details, expect a few like "computational method of plasticity" of Owen. Maybe some commercial softwares have such documents, which shouldn't be published. And even much commerial softwares doesn't have those as much as I know. As to me, I write down implementation details such as relevent papers, meaning of each variables,  and sometime equation used by tex, in internal documents. It is a pratical approach.

Yes, when you read the code, try to understand its theory at first.
回复 不支持

使用道具 举报

发表于 2012-7-19 12:22:31 | 显示全部楼层 来自 美国
本帖最后由 nkliuyueming 于 2012-7-19 12:24 编辑

了解过一些OFELI的设计,各个模块都很清晰,这是我做过的一个OOP基本概念的对比:

本帖子中包含更多资源

您需要 登录 才可以下载或查看,没有账号?注册

×
回复 不支持

使用道具 举报

发表于 2012-7-19 14:46:46 | 显示全部楼层 来自 上海闵行区
nkliuyueming 发表于 2012-7-19 12:22
了解过一些OFELI的设计,各个模块都很清晰,这是我做过的一个OOP基本概念的对比: ...

要是有数据结构的对比就更好了
回复 不支持

使用道具 举报

您需要登录后才可以回帖 登录 | 注册

本版积分规则

Archiver|小黑屋|联系我们|仿真互动网 ( 京ICP备15048925号-7 )

GMT+8, 2024-3-28 19:50 , Processed in 0.049556 second(s), 11 queries , Gzip On, MemCache On.

Powered by Discuz! X3.5 Licensed

© 2001-2024 Discuz! Team.

快速回复 返回顶部 返回列表