脚手架的屈曲分析（请高手帮我看看）

zyq309

最近我做了一个脚手门架的屈曲分析。但做非线形屈曲分析时做不下去了，我看到论坛里关于屈曲分析的帖子一般回贴的很多。恳求大家指点我一下。
我的命令流如下：
   
/filname,JIAOSHOUJIA
/title,BUCKLING
/prep7
et,1,beam189
mp,ex,1,2.1e11
mp,nuxy,1,0.3
mp,dens,1,7850

k,1,0,0,0
k,2,1.018,0,0
k,3,0,0.4144,0
k,4,1.018,0.4144,0
k,5,0.204,0.6184,0
k,6,0.814,0.6184,0
k,7,0,0.9895,0
k,8,0.204,0.9895,0
k,9,0.814,0.9895,0
k,10,1.018,0.9895,0

kgen,2,7,10,1,0,0.3379,0,4
kgen,2,7,10,1,0,0.7225,0,8

k,19,0.2215,1.712,0
k,20,0.7965,1.712,0
k,21,0.1865,1.712,0
k,22,0.8315,1.712,0
k,23,0.2215,1.7345,0
k,24,0.509,1.7345,0
k,25,0.7965,1.7345,0
k,26,0.509,2.022,0
k,27,0.509,2.059,0
k,28,0,2.059,0
k,29,1.018,2.059,0
k,30,0,2.11,0
k,31,1.018,2.11,0
k,32,0.1865,2.059,0
k,33,0.8315,2.059,0

!直径为42的钢管
l,1,3
l,3,7
l,7,11
l,11,15
l,15,28
l,28,30
l,2,4
l,4,10
l,10,14
l,14,18
l,18,29
l,29,31
l,28,32
l,32,27
l,27,33
l,33,29
!直径为32的钢管
l,15,21
l,21,16
l,16,19
l,19,20
l,20,17
l,17,22
l,22,18
l,19,23
l,20,25
larc,23,26,24,0.2875
larc,26,25,24,0.2875


!直径为26的钢管
l,3,5
l,5,8
l,7,8
l,8,12
l,11,12
l,12,16
l,21,32
l,4,6
l,6,9
l,10,9
l,9,13
l,14,13
l,13,17
l,22,33
lfillt,28,29,0.126,,
lfillt,36,35,0.126,,

sectype,1,beam,ctube
secoffset,cent
secdata,0.0188,0.021,8
sectype,2,beam,ctube
secoffset,cent
secdata,0.0138,0.016,8
sectype,3,beam,ctube
secoffset,cent
secdata,0.0118,0.013,8

lsel,all
lsel,s,,,1,16,1,
latt,1,,1,,,,1
lesize,all,0.05
lmesh,all

lsel,all
lsel,s,,,17,27,1,
latt,1,,1,,,,2
lesize,all,0.05
lmesh,all

lsel,all
lsel,s,,,28,43,1,
latt,1,,1,,,,3
lesize,all,0.05
lmesh,all
lsel,all
cp,1,all,277,191
finish

/solu
!dk,1,all,0
!dk,2,all,0
dk,1,ux,0,,,uy,uz,
dk,2,ux,0,,,uy,uz,
dk,30,uz,0
dk,31,uz,0
fk,30,fy,-2.874e4！为了让屈曲荷载系数接近1，我调了多次，调到这个荷载值，这时的屈曲荷载系数为1.0004
fk,31,fy,-2.874e4
acel,,9.8
ftran
antype,0
eqslv,spar
pstres,on
solve
finish
/solu
antype,1
bucopt,lanb,1,0,0
mxpand,1,0,0,1,0.001
solve
finish

/solu
upcoord,0.01,off  ！修改单元节点坐标
antype,0
nlgeom,1
sstif,on
dofsel,s,fy
fscale,10
dofsel,all
time,1
autots,on
nsubst,500,,,1
kbc,0   !逐级递增荷载
outres,all,all
solve
finish
/post26
nsol,2,9,u,z
varnam,1,load-uz
xvar,2
plvar,1
finish

但是最后得到的荷载位移曲线太不可思议了，然后得出了特征值屈曲分析临界荷载和试验值（100多KN）有很大出入。
请大家帮我啊！我这里得出的特征值临界荷载只有57.48KN。

后来我把非线形屈曲分析部分的命令流换成如下
/solu
upcoord,0.01,off
antype,0
nlgeom,on
sstif,on
nsubst,100
arclen,on,4
outres,all,all
fk,30,fy,-6e6
fk,31,fy,-6e6
solve
提示出现错误，算不了。现在我正郁闷！

[ 本帖最后由 无水1324 于 2008-8-4 21:32 编辑 ]

sxman

曲线之所以怪异，是因为你把横纵坐标颠倒了，你把输出变量y改为2，x变量改为time，就是正常的一条曲线了

zyq309

谢谢sxman兄给我的指点，我现在就改。

zyq309

还是不行啊！郁闷！谁还帮我看看啊！

sxman

我映像中求解特征值，模型上只加载一个荷载，你的例题中是两个，不知道这种加载怎么求法，起码我在结构力学中没有学到过这种稳定求解的模型。我建议你只需要建立一半对称模型，然后求解。我算下来是-522.89N

[ 本帖最后由 sxman 于 2008-8-5 16:52 编辑 ]

zyq309

好的，我试过，我算出来的没你结果大，没有522.89N。你是用我的这个命令流算了吗？sxman兄。

qlp529

终于看到同期的帖子，学习了，回去试一下
多多交流啊

muqiang

:lol 全是命令流啊