SolidWorks机械工程师网——最大的SolidWorks学习平台

 找回密码
 立即注册

QQ登录

只需一步,快速开始

查看: 361|回复: 10
打印 上一主题 下一主题

谈谈接触问题

[复制链接]

4

主题

42

帖子

28

金币

天使

Rank: 2Rank: 2

积分
83
QQ
跳转到指定楼层
楼主
发表于 2007-10-17 14:46:59 | 只看该作者 回帖奖励 |倒序浏览 |阅读模式

   经典图书
有用cosmos做接触分析的吗?很想听听大家的心得体会,个人感觉cosmos在这方面功能不是很强大,用户可以选择的范围太窄了,很多东西都是黑匣子一样。
分享到:  QQ好友和群QQ好友和群 QQ空间QQ空间 腾讯微博腾讯微博 腾讯朋友腾讯朋友
收藏收藏 转播转播 分享教程|习题|模型|技巧 点赞点赞 拍砖拍砖
SolidWorks机械工程师网
提示:建议使用谷歌浏览器浏览本网站!如单击这里下载!否则,可能无法下载附件文件!(支持大多数版本的谷歌浏览器,支持360和QQ浏览器的极速模式,即谷歌内核模式,使用IE和Edge浏览器,浏览个别网页以及下载文件时,会误报“***不安全”,此时需要单击“继续访问此不安全站点(不推荐)”才可以继续下载,另外,本网站不含任何不安全的文件,已联系微软公司解决,纯属IE和Edge浏览器误报)
回复

使用道具 举报

7

主题

59

帖子

49

金币

天使

Rank: 2Rank: 2

积分
155
QQ
沙发
发表于 2007-10-18 02:32:01 | 只看该作者
帮你顶。。。

接触----难!!
接触加大变形----难上加难!!!
SolidWorks机械工程师网
提示:建议使用谷歌浏览器浏览本网站!如单击这里下载!否则,可能无法下载附件文件!(支持大多数版本的谷歌浏览器,支持360和QQ浏览器的极速模式,即谷歌内核模式,使用IE和Edge浏览器,浏览个别网页以及下载文件时,会误报“***不安全”,此时需要单击“继续访问此不安全站点(不推荐)”才可以继续下载,另外,本网站不含任何不安全的文件,已联系微软公司解决,纯属IE和Edge浏览器误报)
回复 支持 反对

使用道具 举报

7

主题

38

帖子

41

金币

天使

Rank: 2Rank: 2

积分
135
QQ
板凳
发表于 2007-10-18 03:46:33 | 只看该作者

   经典图书
分析装配体
要求高点
少有一点不合适就分析不下去
当时
总的来说
还是不错的
入门吧
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

5

主题

47

帖子

33

金币

天使

Rank: 2Rank: 2

积分
121
QQ
地板
发表于 2007-10-18 22:14:32 | 只看该作者
说实话,还是一个入门级的软件,不怎么样
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

6

主题

39

帖子

34

金币

天使

Rank: 2Rank: 2

积分
116
QQ
5#
发表于 2007-10-19 08:35:17 | 只看该作者

   经典案例图书
唉,可惜老板叫你分析的都是装配体接触问题,一个算例算n久才动那么几步,通常遇到无名的错误,就停止了,而且任何结果都没有。头都大了。
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

4

主题

33

帖子

24

金币

天使

Rank: 2Rank: 2

积分
88
QQ
6#
发表于 2007-10-20 04:21:29 | 只看该作者
有两种可能:
1)几何模型有问题
2)理解和处理接触有问题
2.1 接触对的处理很有学问,实际和ANSYS一样的,接触类型的选择,而且不需要你傻乎乎的尝试不同的接触刚度
2.2 接触区域的选择,只做关心的区域,尝试采用分割功能来缩小接触区域
2.3  不关心接触应力的位置何必要定义接触对
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

1

主题

54

帖子

19

金币

天使

Rank: 2Rank: 2

积分
104
QQ
7#
发表于 2007-10-20 11:23:41 | 只看该作者

   经典案例图书
呵呵,软件无高下,功力有高低
软件只是工具而已!
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

2

主题

45

帖子

16

金币

天使

Rank: 2Rank: 2

积分
69
QQ
8#
发表于 2007-10-20 13:59:02 | 只看该作者
谢谢斑竹回复。
COSMOS中设置接触问题时供用户可操作的东西我目前看到的很少,接触本来就是个非连续的问题,我们也只能选择有可能接触的区域和接触类型。至于类似罚函数刚度,渗透值,还有为什么COSMOS做接触分析时不能用弧长控制呢?很多时候一个接触问题可能会带来结构失稳,用NR方法的话,正切刚度矩阵可能变成降秩短阵,导致严重的收敛问题。
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

4

主题

50

帖子

29

金币

天使

Rank: 2Rank: 2

积分
79
QQ
9#
发表于 2007-10-20 14:18:41 | 只看该作者
The Hybrid Method
In this method, the displacement and the force methods are combined to solve the
matrix equation. The displacement method is used where external forces are
prescribed while the force method is utilized where the displacements are
prescribed.
In general purpose finite element programs, a displacement-based method is used.
However, in dealing with nonlinearities, such as contact, a hybrid method can be
efficient.

你可以采用非线性分析来分析,另外COSMOS采用混合方法来自动处理接触问题
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

5

主题

48

帖子

27

金币

天使

Rank: 2Rank: 2

积分
124
QQ
10#
发表于 2007-10-20 19:19:01 | 只看该作者

   经典图书
Troubleshooting for Gap/Contact Problems
Following is a list of commonly encountered errors during the execution of
nonlinear gap/contact problems.
1. During the first step, the program stops with the error message: “Stop, the
diagonal term in equation ...., node ...., direction .. is ....." (a zero or negative).
This error usually indicates that the whole model, or portion of it, is externally
statically unstable due to improper constraints. It should always be remembered
that the gap elements do not alter the stiffness. If a portion of the model is
supported only by gap elements, then that portion can be stabilized through the
use of soft trusses (see verification problems NS17 and NS18).
2. The program runs successfully, but the postprocessing module shows that the
one node gap elements go beyond the contact surface which should have
stopped them.
Make sure that you are looking at the structure's iso-scale displaced plot (both
the dimensions of the structure and the deformations have the same scale in the
plot). The default setting of the postprocessing programs show the deformedshape with an exaggerated deformation scale. Therefore, try to use a scale factor
of 1.0 in plotting the deformation (see the DEFPLOT (Results > PLOT >
Deformed Shape) command in COSMOSM Command Reference Manual).
If the gap elements still exceed the contact surface after issuing the above
commands, then the conclusion is that the gap elements are not properly closed.
The following two possibilities then should be considered:
a. The orientation of the gap surfaces might be wrong. The gap elements are
allowed to remain on the positive side of the surface.
b. The original displacement is too large that the displaced location of the gap
elements cannot be compared correctly with respect to the contact surfaces.
This case usually occurs when one of the bodies is an unconstrained structure
supported only by soft trusses. To investigate this possibility, the analyst
should calculate the displacements of the soft trusses under the loads. If the
resultant displacements are excessive and the gap elements are pushed far
beyond the contact surface, it is likely that the gap iterations also will not
converge. This situation can be overcome by decreasing the applied load step
through modifying the “time” curve, or by using stiffer trusses to support the
unattached portion of the model.
3. The program stops with the error message: “Stop, wrong definition for the
contact surfaces.”
Check the target surface connectivity and direction. Make sure that each target
contact surface is represented by continuous sub-surfaces.
4. In nonlinear dynamic problems, the program converges but the structure
behaves erratically after the gaps close.
This condition often occurs due to the assumed perfect rigidity of the closed
two-node gaps. To avoid that situation, some flexibility for the contact can be
introduced through the third real constant of the two-node gap elements.
5. The program completes one or more steps with some gap elements closed, but
finally stops with one of the following error messages:.
a. “Stop, the diagonal term in equation..., node ..., direction ... is zero or
negative."
b. “Stop, convergence not achieved for gap elements.”
c. “*** ERROR: Convergence is not achieved in 200 iterations” or
“Convergence is not achieved in 100 contact iterations.”
These errors basically imply difficulties in problem convergence due to:
– System stiffness has deteriorated and become singular or close to singular
due to other nonlinearities (geometric or material).
– The load increment is too large.
In either case, reducing the load increment is most likely to solve the problem.
However, in case I, if the stiffness has extremely deteriorated, a solution
continuation may not be possible.
✍ If friction forces are present, the analysis is nonconservative (dependent on the
load application sequence). Therefore, the loads must be applied in increments
which resemble the actual load history.
✍ Contact problems which involve large-deflection analysis are likely to require
mesh refinement in the regions where contact is expected.
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

9

主题

46

帖子

54

金币

天使

Rank: 2Rank: 2

积分
148
QQ
11#
发表于 2007-10-21 15:46:29 | 只看该作者
呵呵,谢谢斑竹热心回复我的问题。
关于接触这方面的资料,cosmos可以公开的资料好像很少啊,大侠有什么好资料可以拿出来共享下吗?
SolidWorks机械工程师网
回复 支持 反对

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则

关闭 卷起
关闭 卷起

SOLIDWORKS 2023 机械设计从入门到精通

手机版|小黑屋| GMT+8, 2024-9-29 15:35 , Processed in 0.186141 second(s), 17 queries , Memcache On.

SolidWorks机械工程师网 ( 鲁ICP备14025122号-2 ) 鲁公网安备 37028502190335号

声明:本网言论纯属发表者个人意见,与本网立场无关。
如涉版权,可发邮件: admin@swbbsc.com

快速回复 返回顶部 返回列表