marseilleroulette
Electrical
- Apr 7, 2009
- 1
Hello there, need help with circuit-coupled harmonic analysys... badly (
Got a task to model an 3d induction heating furnace e/m field (educational purposes).
Havent worked with ANSYS before that at all. Decided to start with a 1/4 symmetry model of the inductor coil. Just the inductor and the air. Done some research at help topics. And as a result:
/batch,list
/prep7
emunit,mks
r1=0.054
r2=0.064
l=0.05
n_coil=20
fill=0.8
cs=(r2-r1)*l
pi=2*asin(1)
vol=pi*(r2*r2*l-r1*r1*l)
!***************Element types*****************
csys,0
et,1,97,0 !air
et,2,97,3 !inductor - stranded coil fea DOF - CURR,EMF
et,3,124,5 !coil - DOF VOLT
n,1,1,0,0 !CIRCU124 i
n,2,2,0,0 !CIRCU124 j
n,3,0.054,0,0 !CIRCU124 K
local,11,1,0,0,0,0,0,0,1
type,3
real,2
e,1,2,3
esys,11
r,1,cs,n_coil,vol,0,1,0
rmore,4,fill
r,2,1
csys,0
!***************Material types****************
mp,murx,1,1 !air
mp,murx,2,1 !inductor
mp,rsvx,2,1.8e-8 !inductor
!**************Geometry***********************
wpoffs,0,0,-l/2
cyl4,0,0,r2,0,r1,90,l !1 - Inductor
wpcsys
wpoffs,0,0,-l
cyl4,0,0,0,0,2*l,90,2*l !2 - Air
wpcsys
vovlap,all
nummrg,kp
numcmp,kp
numcmp,line
numcmp,area
numcmp,volu
!**************Meshing***********************
type,2
real,1
mat,2
mshkey,1
mshape,0,3d
vsel,s,,,1 !1 - Inductor
esize,0.005
vatt,2,,2 !mat#=2,elem#=2
vmesh,all
mshkey,0
mshape,1,3d
vsel,s,,,2 !2 - air
smrtsize,8
vatt,1,,1 !mat#=1,elem#=1
vmesh,all
vsel,s,,,1 !curr,emf conductor couple
nslv,s,1
cp,next,curr,all
cp,next,emf,all
nsel,s,,,1
d,all,VOLT,380
nsel,s,,,2
d,all,VOLT,0
asel,s,,,9
asel,a,,,8
asel,a,,,7
NSLA,S,1
d,all,az,0
WSORT,z
finish
/solu
antyp,harm
harfr,50
nsel,all
esel,all
solve
finish
/post1
The problem is that i cant debug that model, but everything seems to be correct - it just doesnt solve at all. No flux lines, no current density, all by zero.
Got the right element types, got the right real constants, geometry, attached the k-node of CIRCU124 element to the FEA, got AZ=0 at model far posts, applied voltage to I and J nodes of the CIRCU123 element.
One moment where i could have been mistaken - is the local coordinate system (required for the r1 consant) definition.
So can you please write down an code example of the right definition?
The second moment im not sure in is - is it enough symmetry definition that is described in the SOLID97 real constant (csym,care)?
While solving ansys shows a note that "COefficient ratio exceeds 1.0e08". How Can this tip me to the solution?
And finally - when doing an circuit coupled analysys like this one - is it enough to apply the loads at the nodes of CIRCU element (ground node and 380 V node) and the loads will be transfered to the SOLID97 fea element?
Please, if you can help with that problem, ill appreciate that a lot, because got noone around here to give me a tip.
Got a task to model an 3d induction heating furnace e/m field (educational purposes).
Havent worked with ANSYS before that at all. Decided to start with a 1/4 symmetry model of the inductor coil. Just the inductor and the air. Done some research at help topics. And as a result:
/batch,list
/prep7
emunit,mks
r1=0.054
r2=0.064
l=0.05
n_coil=20
fill=0.8
cs=(r2-r1)*l
pi=2*asin(1)
vol=pi*(r2*r2*l-r1*r1*l)
!***************Element types*****************
csys,0
et,1,97,0 !air
et,2,97,3 !inductor - stranded coil fea DOF - CURR,EMF
et,3,124,5 !coil - DOF VOLT
n,1,1,0,0 !CIRCU124 i
n,2,2,0,0 !CIRCU124 j
n,3,0.054,0,0 !CIRCU124 K
local,11,1,0,0,0,0,0,0,1
type,3
real,2
e,1,2,3
esys,11
r,1,cs,n_coil,vol,0,1,0
rmore,4,fill
r,2,1
csys,0
!***************Material types****************
mp,murx,1,1 !air
mp,murx,2,1 !inductor
mp,rsvx,2,1.8e-8 !inductor
!**************Geometry***********************
wpoffs,0,0,-l/2
cyl4,0,0,r2,0,r1,90,l !1 - Inductor
wpcsys
wpoffs,0,0,-l
cyl4,0,0,0,0,2*l,90,2*l !2 - Air
wpcsys
vovlap,all
nummrg,kp
numcmp,kp
numcmp,line
numcmp,area
numcmp,volu
!**************Meshing***********************
type,2
real,1
mat,2
mshkey,1
mshape,0,3d
vsel,s,,,1 !1 - Inductor
esize,0.005
vatt,2,,2 !mat#=2,elem#=2
vmesh,all
mshkey,0
mshape,1,3d
vsel,s,,,2 !2 - air
smrtsize,8
vatt,1,,1 !mat#=1,elem#=1
vmesh,all
vsel,s,,,1 !curr,emf conductor couple
nslv,s,1
cp,next,curr,all
cp,next,emf,all
nsel,s,,,1
d,all,VOLT,380
nsel,s,,,2
d,all,VOLT,0
asel,s,,,9
asel,a,,,8
asel,a,,,7
NSLA,S,1
d,all,az,0
WSORT,z
finish
/solu
antyp,harm
harfr,50
nsel,all
esel,all
solve
finish
/post1
The problem is that i cant debug that model, but everything seems to be correct - it just doesnt solve at all. No flux lines, no current density, all by zero.
Got the right element types, got the right real constants, geometry, attached the k-node of CIRCU124 element to the FEA, got AZ=0 at model far posts, applied voltage to I and J nodes of the CIRCU123 element.
One moment where i could have been mistaken - is the local coordinate system (required for the r1 consant) definition.
So can you please write down an code example of the right definition?
The second moment im not sure in is - is it enough symmetry definition that is described in the SOLID97 real constant (csym,care)?
While solving ansys shows a note that "COefficient ratio exceeds 1.0e08". How Can this tip me to the solution?
And finally - when doing an circuit coupled analysys like this one - is it enough to apply the loads at the nodes of CIRCU element (ground node and 380 V node) and the loads will be transfered to the SOLID97 fea element?
Please, if you can help with that problem, ill appreciate that a lot, because got noone around here to give me a tip.