I have a series jobs to analysis a fatgue damage question, and there is some elements deleted between jobs. Now I want to import some STATEV as initial condition for next job by SDVINI subroutine, which is read from field output in last job. But I find some dislocation in STATEV data.
For example, if I delete element label 1514, the STATEV of 1514 will be import to element label 1515 instead of being skipped.
here is my SDVINI subroutine:
SUBROUTINE SDVINI(STATEV,COORDS,NSTATV,NCRDS,NOEL,NPT,
1 LAYER,KSPT)
C
INCLUDE 'ABA_PARAM.INC'
C
DIMENSION STATEV(NSTATV),COORDS(NCRDS)
INTEGER, PARAMETER :: START_LINE = 20, END_LINE = 18335
CHARACTER(LEN=100) :: FILENAME
INTEGER :: CURRENT_LINE, UNIT, IOSTAT
INTEGER :: FIRST, SECOND
DOUBLE PRECISION :: THIRD, FOURTH
FILENAME = 'G:\test_dic\NPTdatafile6.txt'
UNIT = 10
OPEN(UNIT=UNIT, FILE=FILENAME, STATUS='OLD', IOSTAT=IOSTAT)
! Process lines from START_LINE to END_LINE
DO CURRENT_LINE = 1, START_LINE-1
READ(UNIT, *, IOSTAT=IOSTAT)
END DO
DO CURRENT_LINE = START_LINE, END_LINE
READ(UNIT, *) FIRST, SECOND, THIRD, FOURTH
IF (FIRST == NOEL) THEN
STATEV(1) = THIRD
STATEV(2) = FOURTH
END IF
END DO
CLOSE(UNIT)
RETURN
END
and my STATEV date in last job "NPTdatafile6.txt", my import result "result.txt".
THANKS IN ADVANCE!
Hi Gao.
To answer the title of this post, it depends. In theory it should be but when using an user element for explicit analysis (VUEL), it appeared to me that the variable NOEL wasn't equivalent to the element number in the inp / CAE (meaning that the 12th element in the CAE / inp could have the variable NOEL = 49). It may also be the case for implicit analysis (UEL) but I don't remember. If you are not sure, I invite you to check that in your subroutine by comparing the variable NOEL with the elements in your inp (by reconstructing the connectivity table in pre-treatment and then checking the position of the element at the initialization of your computation for example).
Also, I'm not familiar with the SDVINI subroutine but I get that it's used to initialize state variables at each integration point. But since you "delete" some element (even tho I'm not sure what you mean by that), the Abaqus solver shouldn't call these elements at all, right?
Louis Scheidt.
Hi Gao,
Can I please clarify what you are doing?
I've made up the filenames in the following.
- "a.inp" is run and generates "NPTdatafile6.txt"
- "b.inp" uses SDVINI to read state variables from "a.inp"
- "a.inp" had element 1514.
- "b.inp" does not have 1514.
Is this correct?
Regards,
Simon
- Badges
- Science topic