Intel® Fortran Compiler
Build applications that can scale for the future with optimized code designed for Intel® Xeon® and compatible processors.

About error #6362

Mohammad_T_2
Beginner
2,343 Views

Hello all

I have received this error in an ABAQUS Umat written in FORTRAN:

"error #6362: The data types of the argument(s) are invalid.   [DSQRT]"

The part referred to in the Umat is as follows:

Jacob = 
     &   DFGRD0(1,1)*(DFGRD0(2,2)*DFGRD0(3,3)-DFGRD0(3,2)*DFGRD0(2,3))  
     & - DFGRD0(1,2)*(DFGRD0(2,1)*DFGRD0(3,3)-DFGRD0(3,1)*DFGRD0(2,3)) 
     & + DFGRD0(1,3)*(DFGRD0(2,1)*DFGRD0(3,2)-DFGRD0(3,1)*DFGRD0(2,2))
C or using the defined subroutine FindDet at the end:     CALL FindDet(DFGRD0,3,Jacob)
      MATA33 = MATMUL(DFGRD0,TRANSPOSE(DFGRD0))
     
    DO k1=1,N_sys
       DO k2=1,3
        n_temp(k2,1) = n_loc(k2,k1)
       END DO
       MATA13 = MATMUL(TRANSPOSE(n_temp),MATA33)
       SCALARA = MATMUL(MATA13,n_temp)
     lambda_loc(k1,1) = 1.d0/DSQRT(Jacob)*(SCALARA(1,1))**(3.d0/4.d0)

     ...

    END DO

Could any one help me how to resolve this issue?

Thanks in advance!

0 Kudos
2 Replies
Steven_L_Intel1
Employee
2,343 Views

How is Jacob declared? If it is not REAL(8) or DOUBLE PRECISION, you will get this error. Generally I recommend against using the "specific" names such as DSQRT and suggest using the generic name SQRT, but in this case that might hide an unintentional error since it seems pretty clear you intend to be using double precision throughout.

0 Kudos
Mohammad_T_2
Beginner
2,343 Views

Thank you Steve

The problem was resolved. I had not defined Jacob at the beginning of the file. Now I have defined it and the problem has been fixed.

Sincerely,

Mtorki

0 Kudos
Reply