SOS! how do I change the shape of indenter in the punch-example

[Jesse]

hello! my name is jesse fang, I'd like to ask how do I modify the fortran code, if I want to change the shape of indenter  in the punch-example.I still haven't make it clear,please help me,I will appreciate very much!

[Ellad Tadmor]

In the example, the shape of the indenter is specified by the displacement
boundary conditions which are set in routine user_bcon() in user_punch.f.
To change to a different shape you would need to change the code to fix
the displacement of the atoms based on the current value of "propfact"
(which would normally be the indentation depth) and the shape of the
indenter. For example, this (untested) code fragment, should make a wedge
indenter with a 90 degree opening:

propfact = GetLoadPropFact()
do i=1,numnp
xx=x(1,i)
yy=x(2,i)
if (abs(yy-ycind).le.tole.and.abs(xx-xcind).le.propfact) then
id(1,i)=1
id(2,i)=1
id(3,i)=1
f(2,i)=-1.0_dp
b(2,i)=f(2,i)*propfact+abs(xx-xcind)
endif
enddo

Note that this would be a "perfect stick" indenter, i.e. any atom that
comes in contact with the indenter sticks to it. Sliding boundary
conditions are much harder and would require more significant changes to
the code.

-Ellad Tadmor

On Sun, 28 Oct 2012, Jesse wrote:

> hello! my name is jesse fang, I'd like to ask how do I modify the fortran code, if I

> want to change the shape of indenter ï¿œin the punch-example.I still haven't make it

[Jesse]

Dear professor Tadmor:

  Thank you for your reply. I'm using QC to simulate a  nanoindentation experiment which the shape of indenter is sphere.I changed the boundary condition array

 f(2,i)=-1.0_dp*sqrt(contact**2-(xx-xcind)**2); 

 It's obvious that it can only simulate the state that the sphere indenter have been pressed in.I've found it's difficult to control the time that nodes begin to move down if I change the shape to sphere and simulate the process of press in. Could you give me some suggestions, appreciate it very much!

with the photo base on the modified code above:

在 2012年10月30日星期二UTC+8上午5时47分13秒，Ellad Tadmor写道：

[Ellad Tadmor]

Hello-

You cannot model a spherical indenter with a 2D code, so you probably mean
a cylindrical indenter. A good way to do this is with the force field
approach suggested by Kelchner et al. (1998). In this case the indenter is
represented as a repulsive force field that pushes atoms away as it is
pressed into the model surface. This mimics a frictionless indenter. This
is supported in QC by the user_potential routine. See Section 2.11.4 in
the QC Reference Manual version 1.4 for a detailed explanation.

-Ellad Tadmor

On Mon, 5 Nov 2012, Jesse wrote:

> Dear professor Tadmor:  Thank you for your reply. I'm using QC to simulate a
>  nanoindentation experiment which the shape of indenter is sphere.I changed
> the boundary condition array
>  f(2,i)=-1.0_dp*sqrt(contact**2-(xx-xcind)**2); 
>  It's obvious that it can only simulate the state that the sphere indenter
> have been pressed in.I've found it's difficult to control the time that
> nodes begin to move down if I change the shape to sphere and simulate the
> process of press in. Could you give me some suggestions, appreciate it very
> much!
>

> [export2.jpg]
>
> [export.jpg]

>
> with the photo base on the modified code above:
>

> ? 2012?10?30????UTC+8??5?47?13??Ellad Tadmor???

> > want to change the shape of indenter ???in the punch-example.I still