Mr. Yeqing Wang. Thanks for your answer. For FE model orientation in standard axis, for e.g Weld start point (0,0,0) and weld end point (0,0,160), Flux application works fine. However, if i orient my FE Model in to some random position, for e.g weld start point (5,15,-35) to weld end point (10,18,-78), the flux application is not working properly. The temperature results are non-physical. Bottomeline is, how can i generalize my Fortran codes so that it can work on any orientation of FE model ?

I have successfully completed this excersice.

I have developed universal general Fortran codes which can be used for any kind of model with any kind of orientation in global space. Refer animation file. Goldak-thr-DFLUX.avi

Dear Krunal

Congratulations for your results

could you also help me

i sent you a message

thanks for your assistance

Dear Krunal Thakar

I have problem with DFLux like u.

Can u help me.

Problem during compilation:

*USER SUBROUTINE

SUBROUTINE DFLUX(FLUX,SOL,KSTEP,KINC,TIME,NOEL,NPT,COORDS,

1 JLTYP,TEMP,PRESS,SNAME)

include 'ABA_PARAM.INC'

DIMENSION FLUX(2),TIME(2),COORDS(3)

CHARACTER*80 SNAME

*code

! REAL SOL,TEMP,PRESS

! INTEGER JSTEP,JINC,NOEL,NPT,JLTYP

! REAL t,dt,x,y,z

! REAL FLUX1,FLUX2,FLUX3

INTEGER flag,kstep

REAL t,x,y,z

Real Q,a,b,cf,cr,pi,m,n

*times

t=time(1)

a= 6

b = 2.5

cf = 3

cr= 1

pi = 3.14159265359

*start point parameters [m]

FLUX(1) = 0

FLUX(2) = 0

x = coords(1)

y = coords(2)

z = coords(3)

C FORM HEAT FLUX FROM SAMPLING POINTS

if(kstep.EQ.3) then

Q = 0.5*10*90

x1 = x

y1 = y

z1 = z + 5*t

m=exp(-3*((x1)**2/(cf)**2+(y1)**2/(a)**2+(z1)**2/(b)** 2))

n=exp(-3*((x1)**2/(cr)**2+(y1)**2/(a)**2+(z1)**2/(b)** 2))

FLUX(1)=(6*1.732*Q)*((1.5/cf)*m+(.4/cr)*n)/(pi*sqrt(pi)*(a*b))

end if

return

end

Someone can help me?

Hi Dai Le

Did you solve your problem ?

I have problem using dflux subroutine. No results were shown after analysis. Help Please