Precision of COORdinates

[Jfrisch]

Dear FEAP-Users,

what is the precision a coordinate of a node can be given? I have meshed a problem using gmsh and scripted a little python-script to convert the mesh to FEAP. But wether I write the coordinates as a floating-point number or in scientific notation FEAP meshes every node with the second coordinate vanishing (and also the first coordinate of some nodes).
I have attached the four versions I tested and all of them have the same above described behaviour. E.g. node 2 is generated at (0,0) although it is defined at (-1,0) or node 3 is defined at (-1,1) but FEAP creates it at (1,0). What is the problem in my coordinate definition?

[Jfrisch]

Attached the output-files for comparison of node creation.

[Prof. R.L. Taylor]

FEAP only allows 15 digits, including the sign, to define any variable.  The blank or comma does not count, so field widths are set to 16 digits.   You have more than this.

[Prof. S. Govindjee]

Also, please see section 5.2.1 of the manual for the correct syntax for using COORdinates; you are missing a field.

[Prof. R.L. Taylor]

There is an option that will avoid adding all the missing field:  If you are doing an input for ALL the coordinates at once you can use:

COOR ALL
NOGEN
1,,x,y,z
2,x,y,z
.....
to the last node

[Jfrisch]

Thanks Prof. Govindjee, I missed the generator-parameter in my script to convert the node coordinates!

And thanks Prof. Taylor for the information and suggestion of "NOGEN". I haven't found the 16 digits precision printed in the manual - is it there somewhere and I missed it? But I don't get the coordinate I have defined with more than 16 digits ... the lowest ones are of order 10^(-9) and four digits after the comma plus one digit before the comma plus one digit of sign makes at maximum 15 digits?

[Prof. R.L. Taylor]

See Chapter 4 of the user manual for input restrictions.

[Prof. R.L. Taylor]

Actually, if you use the COOR ALL then the reads are read(ior,* ...) which means the inputs are not field width limited but may be resticted by the specific compiler on whether E+000 is accepted or only E+00.  Otherwise the restrictions of Ch 4 are applied to the other data input items, such as forces, displacements, etc.