在结构设计的时候需要保持两个结构之间有一定的间隙，如电子结构和管道之间要保持间隙，防止相邻结构会碰到。该间隙可以通过测量两个节点之间的相对位移得到，传统的方法用MPC，但是比较麻烦，现在介绍一种新的方法通过CBUSH单元的位移方便的测出两个结构之间的间隙，并且可以用于结构设计优化的约束条件。
如何用CBUSH测量相对间隙
CBUSH单元的应变等于位移乘以回复系数，缺省情况下回复系数（EA，ET）=1.0，这样应变就等于位移。
对于有长度的CBUSH单元，A点和B点不在同一个位置：
用GO或
新浪博客
在结构设计的时候需要保持两个结构之间有一定的间隙，如电子结构和管道之间要保持间隙，防止相邻结构会碰到。该间隙可以通过测量两个节点之间的相对位移得到，传统的方法用MPC，但是比较麻烦，现在介绍一种新的方法通过CBUSH单元的位移方便的测出两个结构之间的间隙，并且可以用于结构设计优化的约束条件。
如何用CBUSH测量相对间隙
CBUSH单元的应变等于位移乘以回复系数，缺省情况下回复系数（EA，ET）=1.0，这样应变就等于位移。
对于有长度的CBUSH单元，A点和B点不在同一个位置：
用GO或
Xi定义方向
力：拉伸(+)或压缩(-)
应变和力的方向定义相同
如右图，X方向的应变就是沿着长度方向的相对位移（EA=1.0）
对于0长度的CBUSH
用坐标系定义单元方向
力F= Ke*(Ub-Ua)
如果Ub有正向位移，就是分离，也就是正向应变是分离
悬臂板结构，测量两块板之间的相对位移（间隙）
结构受惯性载荷，方向和Z轴负方向一致的重力加速度1g，两块板的两端都为固定约束
在两块板之间创建如图所示的CBUSH
CBUSH有长度，所以连接CBUSH单元的节点顺序不重要
PBUSH设置6个方向的k=0.0，无刚度
请求输出STRAIN=ALL或STRAIN=set id
CBUSH X方向的应变可以直接测到
正=分离
负=闭合
优化目标：重量最小 设计变量：板厚T1,
T2
约束条件：两块板的节点之间至少保持0.9”的间隙
方法一：约束相对位移
DESOBJ(min)=101 $ minimize weight
DESSUB=201 $ keep rel disp >=-0.1
DESVAR
DVPREL1 11
DESVAR
DVPREL1 12
$ pbush strain-x = item code 2
$ response strain-x for all PBUSH with ID=500
DRESP1
DRESP1
DCONSTR 201
方法二：约束间隙
DESOBJ(min)=101 $ minimize weight
DESSUB=201 $ clearance >=0.9
$ response strain-x for all PBUSH with ID=500
DRESP1
DRESP1
DRESP2
DRESP1
$ keep total distance greater than 0.9
$ positive rel_disp => separating
$ negative rel_disp => closing
$ actual_distance = initial_distance + rel_disp
DEQATN
DCONSTR 201
优化结果
以上两种方法结果相同
重量
初始：
优化后： 3.198E-4
T1, T2
初始：
优化后： 0.063588, 0.6000
相对位移 = PBUSH X方向应变 [间隙]
初始：
优化后：–0.1 [0.9]
附有示例中使用两种不同方法的Nastran模型文件如下：
two_boards_200a.dat和two_boards_200a2.dat
具体信息如下：
two_boards_200a.dat
INIT MASTER(S)
$
ID MSC, SWRO
TIME 100
DIAG 5,6,8,56
SOL 200
CEND
$
TITLE = CANTILEVERED BEAM MADE OF PLATES
SUBTITLE = CBUSH RELATIVE DISPLACMENT EXAMPLE
$
$
$ CBUSH STRAIN OUTPUT = (CBUSH_DISP)*EA for translations
$ FOR NONZERO LENGTH CBUSH USING GO or Xi for ORIENTATION:
$
$
$
$
$ OBJECTIVE
$
$ VARIABLES
$
$ CONSTRAINTS
$
$
$
$
$
DISPL=ALL
STRAIN=ALL
$
SPC=1
$
DESOBJ(min)=101 $ minimize weight
DESSUB=201 $ keep separation >=-0.1
$
SUBCASE
BEGIN BULK
DESVAR
DVPREL1 11
DESVAR
DVPREL1 12
$ pbush strain-x = item code 2
$ response strain-x for all PBUSH with ID=500
DRESP1
DRESP1
DCONSTR 201
doptprm p1
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
param,post,-1
$ create dummy springs to grids between the two plates to measure rel disp
$ this reported as PBUSH strains
$ reverse grids to test to see if that makes difference on forces
$ QRG says compression is negative, tension positive
$ when GO or Xi for orientation
CBUSH
CBUSH
CBUSH
CBUSH
PBUSH
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
spc1,1,123456,101,201,311,411
GRID
=
=9
GRID
=
=9
$
CQUAD4
=
=8
GRID
=
=9
GRID
=
=9
$
CQUAD4
=
=8
MAT1
PSHELL
PSHELL
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
$ Loads for SOL 101/Linear Statics
GRAV
ENDDATA
two_boards_200a2.dat
INIT MASTER(S)
$
ID MSC, SWRO
TIME 100
DIAG 5,6,8,56
SOL 200
CEND
$
TITLE = CANTILEVERED BEAM MADE OF PLATES
SUBTITLE = CBUSH RELATIVE DISPLACMENT EXAMPLE
$
$
$ CBUSH STRAIN OUTPUT = (CBUSH_DISP)*EA for translations
$ FOR NONZERO LENGTH CBUSH USING GO or Xi for ORIENTATION:
$
$
$
$
$ OBJECTIVE
$
$ VARIABLES
$
$ CONSTRAINTS
$
$
$
$
$
$
DISPL=ALL
STRAIN=ALL
$
SPC=1
$
DESOBJ(min)=101 $ minimize separation over all measurements
DESSUB=201 $ maintain clearance >0.9
$
SUBCASE
BEGIN BULK
DESVAR
DVPREL1 11
DESVAR
DVPREL1 12
$ pbush strain-x = item code 2
$ response strain-x for all PBUSH with ID=500
DRESP1
DRESP1
DRESP2
$ keep total distance greater than 0.9
$ positive rel_disp => separating
$ negative rel_disp => closing
$ actual_distance = initial_distance + rel_disp
DEQATN
DCONSTR 201
doptprm p1
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
param,post,-1
$ create dummy springs to grids between the two plates to measure rel disp
$ this reported as PBUSH strains
$ reverse grids to test to see if that makes difference on forces
$ QRG says compression is negative, tension positive
$ when GO or Xi for orientation
CBUSH
CBUSH
CBUSH
CBUSH
PBUSH
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
spc1,1,123456,101,201,311,411
GRID
=
=9
GRID
=
=9
$
CQUAD4
=
=8
GRID
=
=9
GRID
=
=9
$
CQUAD4
=
=8
MAT1
PSHELL
PSHELL
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
$ Loads for SOL 101/Linear Statics
GRAV
ENDDATA
力：拉伸(+)或压缩(-)
应变和力的方向定义相同
如右图，X方向的应变就是沿着长度方向的相对位移（EA=1.0）
对于0长度的CBUSH
用坐标系定义单元方向
力F= Ke*(Ub-Ua)
如果Ub有正向位移，就是分离，也就是正向应变是分离
悬臂板结构，测量两块板之间的相对位移（间隙）
结构受惯性载荷，方向和Z轴负方向一致的重力加速度1g，两块板的两端都为固定约束
在两块板之间创建如图所示的CBUSH
CBUSH有长度，所以连接CBUSH单元的节点顺序不重要
PBUSH设置6个方向的k=0.0，无刚度
请求输出STRAIN=ALL或STRAIN=set id
CBUSH X方向的应变可以直接测到
正=分离
负=闭合
约束条件：两块板的节点之间至少保持0.9”的间隙
方法一：约束相对位移
DESOBJ(min)=101 $ minimize weight
DESSUB=201 $ keep rel disp >=-0.1
DESVAR
DVPREL1 11
DESVAR
DVPREL1 12
$ pbush strain-x = item code 2
$ response strain-x for all PBUSH with ID=500
DRESP1
DRESP1
DCONSTR 201
方法二：约束间隙
DESOBJ(min)=101 $ minimize weight
DESSUB=201 $ clearance >=0.9
$ response strain-x for all PBUSH with ID=500
DRESP1
DRESP1
DRESP2
DRESP1
$ keep total distance greater than 0.9
$ positive rel_disp => separating
$ negative rel_disp => closing
$ actual_distance = initial_distance + rel_disp
DEQATN
DCONSTR 201
优化结果
以上两种方法结果相同
重量
初始：
优化后： 3.198E-4
T1, T2
初始：
优化后： 0.063588, 0.6000
相对位移 = PBUSH X方向应变 [间隙]
初始：
优化后：–0.1 [0.9]
附有示例中使用两种不同方法的Nastran模型文件如下：
two_boards_200a.dat和two_boards_200a2.dat
具体信息如下：
two_boards_200a.dat
INIT MASTER(S)
$
ID MSC, SWRO
TIME 100
DIAG 5,6,8,56
SOL 200
CEND
$
TITLE = CANTILEVERED BEAM MADE OF PLATES
SUBTITLE = CBUSH RELATIVE DISPLACMENT EXAMPLE
$
$
$ CBUSH STRAIN OUTPUT = (CBUSH_DISP)*EA for translations
$ FOR NONZERO LENGTH CBUSH USING GO or Xi for ORIENTATION:
$
$
$
$
$ OBJECTIVE
$
$ VARIABLES
$
$ CONSTRAINTS
$
$
$
$
$
DISPL=ALL
STRAIN=ALL
$
SPC=1
$
DESOBJ(min)=101 $ minimize weight
DESSUB=201 $ keep separation >=-0.1
$
SUBCASE
BEGIN BULK
DESVAR
DVPREL1 11
DESVAR
DVPREL1 12
$ pbush strain-x = item code 2
$ response strain-x for all PBUSH with ID=500
DRESP1
DRESP1
DCONSTR 201
doptprm p1
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
param,post,-1
$ create dummy springs to grids between the two plates to measure rel disp
$ this reported as PBUSH strains
$ reverse grids to test to see if that makes difference on forces
$ QRG says compression is negative, tension positive
$ when GO or Xi for orientation
CBUSH
CBUSH
CBUSH
CBUSH
PBUSH
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
spc1,1,123456,101,201,311,411
GRID
=
=9
GRID
=
=9
$
CQUAD4
=
=8
GRID
=
=9
GRID
=
=9
$
CQUAD4
=
=8
MAT1
PSHELL
PSHELL
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
$ Loads for SOL 101/Linear Statics
GRAV
ENDDATA
two_boards_200a2.dat
INIT MASTER(S)
$
ID MSC, SWRO
TIME 100
DIAG 5,6,8,56
SOL 200
CEND
$
TITLE = CANTILEVERED BEAM MADE OF PLATES
SUBTITLE = CBUSH RELATIVE DISPLACMENT EXAMPLE
$
$
$ CBUSH STRAIN OUTPUT = (CBUSH_DISP)*EA for translations
$ FOR NONZERO LENGTH CBUSH USING GO or Xi for ORIENTATION:
$
$
$
$
$ OBJECTIVE
$
$ VARIABLES
$
$ CONSTRAINTS
$
$
$
$
$
$
DISPL=ALL
STRAIN=ALL
$
SPC=1
$
DESOBJ(min)=101 $ minimize separation over all measurements
DESSUB=201 $ maintain clearance >0.9
$
SUBCASE
BEGIN BULK
DESVAR
DVPREL1 11
DESVAR
DVPREL1 12
$ pbush strain-x = item code 2
$ response strain-x for all PBUSH with ID=500
DRESP1
DRESP1
DRESP2
$ keep total distance greater than 0.9
$ positive rel_disp => separating
$ negative rel_disp => closing
$ actual_distance = initial_distance + rel_disp
DEQATN
DCONSTR 201
doptprm p1
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
param,post,-1
$ create dummy springs to grids between the two plates to measure rel disp
$ this reported as PBUSH strains
$ reverse grids to test to see if that makes difference on forces
$ QRG says compression is negative, tension positive
$ when GO or Xi for orientation
CBUSH
CBUSH
CBUSH
CBUSH
PBUSH
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
spc1,1,123456,101,201,311,411
GRID
=
=9
GRID
=
=9
$
CQUAD4
=
=8
GRID
=
=9
GRID
=
=9
$
CQUAD4
=
=8
MAT1
PSHELL
PSHELL
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
$ Loads for SOL 101/Linear Statics
GRAV
ENDDATA