请问各位，为何再包一层空气磁场结果就不对称了

文辰天下

模型为两个平行导体，激励为电流，我用http://forum.simwe.com/thread-837950-1-2.html帖子中会员zhouyz写的命令流得到的空气中的磁场分布是对称的。
命令流如下:
/prep7
et,1,117        !air
et,2,117,5        !conductor
mp,murx,1,1    !air
mp,murx,2,1    !conductor
mp,rsvx,2,4e-7 !bus
block,-0.1475,-0.1275,0,0.4015,-0.050,0.050  !conductor1 model
block,0.1275,0.1475,0,0.4015,-0.050,0.050    !conductor2 model
block,-0.6375,0.6375,0,0.4015,-0.400,0.400   !
vovlap,all
numcmp,volu
vsel,s,,,1,2
vatt,2,,2
allsel,all
smart,6
mshape,1,3d
mshkey,0
vmesh,all
asel,s,,,4
asel,a,,,10
DA,all,volt,0
allsel,all
asel,s,,,3
nsla,s,1
cp,1,volt,all
*get,nmax1,node,0,num,max
F,nmax1,amps,24000
asel,s,,,9
nsla,s,1
cp,2,volt,all
*get,nmax2,node,0,num,max
F,nmax2,amps,24000
allsel,all
/solu
asel,s,ext
DA,all,AZ,0,0
CNVTOL,AMPS,1,0.5e-9,2, ,   
CNVTOL,CSG,1,0.5e-7,2, ,
allsel,all
solve

文辰天下

1# 文辰天下
但是如果我在上面这个结构沿y向再对称地包上空气体，

红字是在上个命令流中多加的命令，其它命令流不变。结果算出来的磁场不对称了。

/prep7

et,1,117
!air
et,2,117,5
!conductor
mp,murx,1,1
!air
mp,murx,2,1
!conductor

mp,rsvx,2,4e-7 !bus
block,-0.1475,-0.1275,0,0.4015,-0.050,0.050
!conductor1 model

block,0.1275,0.1475,0,0.4015,-0.050,0.050
!conductor2 model

block,-0.6375,0.6375,-0.4015,0.803,-0.400,0.400
!

block,-0.6375,0.6375,0.4015,0.803,-0.400,0.400

block,-0.6375,0.6375,-0.4015,0.,-0.400,0.400
vovlap,all

numcmp,volu
vsel,s,,,1,2
vatt,2,,2

allsel,all

smart,6
mshape,1,3d
mshkey,0
vmesh,all

asel,s,,,4

asel,a,,,10
DA,all,volt,0

allsel,all

asel,s,,,3

nsla,s,1
cp,1,volt,all

*get,nmax1,node,0,num,max

F,nmax1,amps,24000

asel,s,,,9

nsla,s,1
cp,2,volt,all

*get,nmax2,node,0,num,max

F,nmax2,amps,24000

allsel,all

FINISH

/SOL
asel,s,ext

DA,all,AZ,0,0

CNVTOL,AMPS,1,0.5e-9,2, ,

CNVTOL,CSG,1,0.5e-7,2, ,
allsel,all

/STATUS,SOLU
SOLVE

[localimg=180,130]1[/localimg]

文辰天下

本来想把不对称的结果贴上来，可说我附件大，没法帖。麻烦大家帮我跑一下命令流，就可以得到结果 了。
请大家看看为什么会出现这种情况。难道用空气把两个导体包起来不对吗。导体两端面是不是一定要设置磁通量平行的边界条件呢。

kt266a

感觉是工程电磁场理论那方面的东西，和软件关系不大。

perch88

多了东西以后，体号、面号也多了，就有可能不一样了
所以必须修改命令流里面相应的加载的面号和体号

lieren810

5L正解~~~~