XANSYS Message: 40986 [Go back to message list]
Average Rating: 5 (1 votes)
Rate item:

Subject: Re: [STRUCT] Bug with EKILL for PLANE182
Author: Markus Fink
Date: 2002-08-22 02:41:00

Hallo Steven,

thanks for your answer. If I anderstand the statment from ANSYS right, all
is right, if the simulatio don't chrashes?
Best regards,

Markus Fink

Infineon Technologies
CAT AIT TI SIM
Tel: 0941 / 202-7003
Fax: 0941 / 202-2884


Markus,

Here is a statement from Guoyu Lin (ANSYS Developer):

Dated 03/26/02
RE: [xansys] STRUC: help with EKILL: ANSYS Crashes with FATAL ERR OR

Q1. Why the crash?

For element 180,181,182,183,185,186,187,188,189, when they are killed (by
command EKILL), the element stiffness goes to zero. This will may result in
the element nodes get removed from the global assembly, and thus causing
crash. This problem will be fixed in 6.1.

Q2. Any other way to do this type of analysis?
1. Before you kill the elements, add a dummy run, this can be a solv without
any loading. So that all the nodes get stayed.
2. Using CP to couple those nodes which will lost its stiffness to some
other nodes who have stiffness.
3. Of course, switching to other elements that are not in the above list is
always a choice if you can do so.

Guoyu Lin
ANSYS Inc.

Best Regards,
Steve Groothuis
Micron Technology Texas LLC
-----Original Message-----
[mailto:Markus.Fink2@i...]
Sent: Wednesday, August 21, 2002 3:25 AM
To: 'xansys@yahoogroups.com'

Hallo all!

I have found the following Mail from Steven Groothius in which he says there
is a bug in PLANE182 with EKILL option. Can anybody please tell me in which
manner this bug will influence my simulation?
Thanks,

Markus Fink

Infineon Technologies
CAT AIT TI SIM
Tel: 0941 / 202-7003
Fax: 0941 / 202-2884


Sent: Thursday, April 04, 2002 1:43 PM

Haider,

In doing a Vapor Deposition process, you must understand the differences
between it and a bimaterial beam in VM174.

THERMAL STRESS PROBLEM
- Initially, from 25C to 400C, only the substrate exists (VM174 assumes both
materials exist).
- Most coatings have not only thermally-induced strains (due to CTE), but
also process-induced strains.
- Those process-induced strains can be introduced in the 1st load step by
using ISTRESS for coating.
- Initial coating stress value used for ISTRESS command needs to be derived
by any radius of curvature measurements that you can do and through the use
of the Stoney equation.
- I would assume that your stress-free condition (for TREF) is at 25C,
however MP,REFT command takes over anyway. NOTE: The value input on the
TREF command applies to all materials not having a specified material
property definition through MP,REFT.
- I would recommend using BF,ALL,TEMP as your loading condition instead of
TUNIF
- Depending on your substrate size, you may wish to use a smaller section
(as opposed to half-symmetry) with periodic boundary conditions.

ELEMENT SELECTION
Depending on (1) coating & substrate thermal properties and (2) deposition
process time, you probably can solve this problem as a uniform delta
temperature using PLANE42 or PLANE82 (not PLANE182 due to bug with EKILL
option, to be fixed at ANSYS 6.1).

SIMULATION PROCESS
The coating deposition process would be 4 load steps:
(1) EKILL Substrate and use ISTRESS on Coating
(2) EKILL Coating and EALIVE Substrate and take to Coating deposition
temperature
(3) EALIVE Coating & Substrate at Coating deposition temperature
(4) Cool-down Coating & Substrate to room temperature

Of course, you may have additional thermal dwells and/or ramps along the
way, but this process should satisfy the sequential procedure coating
deposition on a substrate. Remember to reapply the symmetry boundary
conditions after each step to avoid artificial DOF problems.

STONEY EQUATION REFERENCES:
Stress Development in Mo/Si and Ru/Si Multilayers (Stanford)
http://raiden.stanford.edu/PDF_docs/LLNL_talk.pdf
Residual Stresses in Thick and Thin Surface Coatings (Cambridge)
http://www.msm.cam.ac.uk/mmc/publications/twc010.pdf

Best Regards,
Steve Groothuis
Micron Technology Texas LLC

Posts possibly associated with message #40986AuthorDateScore
40959[STRUCT] Bug with EKILL for PLANE182Markus Fink2002/08/21 
40967Re: [STRUCT] Bug with EKILL for PLANE182Steve Groothuis2002/08/21 
40986Re: [STRUCT] Bug with EKILL for PLANE182Markus Fink2002/08/225
40990Re: [STRUCT] Bug with EKILL for PLANE182Steve Groothuis2002/08/22