刚度矩阵不正定
做练习时经常出现类似如下的错误:*** STIFFNESS MATRIX NOT POSITIVE DEFINITE ***
NODE= 940 EQUATION= 1486DOF= Z-TRANSLATIONPIVOT= -1.88657158E-06
NODE= 979 EQUATION= 1561DOF= Z-TRANSLATIONPIVOT= -7.74860382E-07
NODE= 1044 EQUATION= 1689DOF= Z-TRANSLATIONPIVOT= -3.28445435E-03
NODE= 1050 EQUATION= 1700DOF= Y-TRANSLATIONPIVOT= -7.56978989E-05
NODE= 1094 EQUATION= 1789DOF= Z-TRANSLATIONPIVOT= -4.95910645E-03
NODE= 1100 EQUATION= 1801DOF= Z-TRANSLATIONPIVOT= -1.14443427E-05
NODE= 1196 EQUATION= 1992DOF= Y-TRANSLATIONPIVOT= -2.14589420E-07
NODE= 1196 EQUATION= 1993DOF= Z-TRANSLATIONPIVOT= -3.80032454E-07
*** IN SMALL DISPLACEMENT LINEAR ELASTIC ANALYSIS ***
CHECK THE FOLLOWING --
(A) BOUNDARY CONDITIONS -
THE B.C. MUST NOT ADMIT A RIGID BODY
DISPLACEMENT OR ROTATION OF THE TOTAL
STRUCTURE.
(B) DELETION OF DEGREES-OF-FREEDOM -
ALL D.O.F. AT NODAL POINTS WITHOUT
STIFFNESS CONTRIBUTIONS FROM ELEMENTS
MUST HAVE BEEN DELETED.
(C) ELEMENT GEOMETRY AND CONNECTIVITY -
ALL ELEMENTS MUST HAVE BEEN INPUT WITH
PROPER NODAL NUMBERS, AS DESCRIBED IN
THE USERS MANUAL.
(D) ELEMENT INTEGRATION ORDERS -
THE ORDERS OF NUMERICAL INTEGRATIONS FOR
EVALUATION OF ELEMENT MATRICES MUST BE
SUFFICIENTLY HIGH.
(E) MATERIAL DATA -
ALL MATERIAL DATA MUST BE PHYSICALLY
REASONABLE (E.G. YOUNG*S MODULUS MUST
BE GREATER THAN ZERO).
*** IN MATERIALLY NONLINEAR AND/OR LARGE DISPLACEMENTANALYSIS ***
IF THE MODEL HAS BEEN INPUT CORRECTLY (WITH ALL
THE ABOVE CONSIDERATIONS TAKEN INTO ACCOUNT), THEN
THE COLLAPSE LOAD OF THE MODEL HAS BEEN REACHED.
其中(A)和(E) 能看懂,并且没有出错,其他三项就看不大懂了,查阅帮助文件也是看不出个头绪,晕晕乎乎,不知所云。
尝试过重新划分网格,但是问题依然解决不了。
麻烦各位大侠指教!
多谢! 不一定是网格的问题
材料,边界都有可能
也有可能是实际结构已经破坏了 B大概是指模型存在刚体位移吧,就是说系统约束不够。 3# bian222
Choose an arbitrary point in the model and fix all the dofs to see if you can solve it. If you can, it is definitely because of lack of restraint.
It is even better if you can do the modal analysis first. Just calculate three or four modes to see if any part of the model falls apart in the mode shape. This is to make sure the elements in the model are connected to each other. 我的也出现过类似的问题,之后我把刚度矩阵非正检查选项给关了,然后运算成功,这样做是不是会得出不正确的结果 对后处理中的各种云图一直不是很放心
页:
[1]