XANSYS Message: 37969 [Go back to message list]
[bookmark on del.icio.us]
No rating yet
Rate item:

Subject: Re: STRUC-Substructuring for Transient
Author: Tom Davis
Date: 2002-05-22 15:39:00

Thanks Sheldon! I think Case 2 is the one I want.

Would you happen to have the answer to my other question? i.e., Could I
use IC to establish an initial velocity? (Using "VALUE2" field?) And if
so, could I eliminate my first load step? Using IC would be an appreciable
simplification to my problem (I think).

Thanks.

TWD

Thomas W. Davis, P. E.
FSI International (http://www.fsi-intl.com/)
3455 Lyman Boulevard
Mail Station 4-1035
Chaska, Minnesota 55318
PHONE 952-361-7359, FAX 952-361-7393
tom.davis@f...

|---------+---------------------------->
| | "sheldonimaoka" |
| | | | nsys.com> |

| | 05/22/02 02:08 PM|
| | Please respond to|
| | xansys |

|---------+---------------------------->

| To: xansys@yahoogroups.com

| cc:

| Subject: [xansys] Re: STRUC-Substructuring for Transient

Hi Tom,

The note you read is related to the generation pass. You can apply
constraints in the generation pass or the use pass.

1) In the generation pass, since you are creating the superelement
once (although with different load vectors), D=0 or M cannot change
since, otherwise, the [K] of the superelement will change, and you
will have a different superelement. This is for the case with D of
UX/Y/Z=0.

2) For case of nonzero displacements, that actually becomes part of
the load vector, if I recall correctly, so you should be able to
change it between load steps in the generation pass.

3) In the use pass, D can change at master DOF in each load step.
The way you do this is as follows:
A) In generation pass, define M at DOF where you want to apply D
B) In use pass, apply/delete D at master DOF (M), as needed.
This works because in the generation pass, M is defined and does not
change. The superelement is then created. In the use pass, we can
then apply/delete constraints or forces at the master DOF, as needed.

The three cases 1-3 are mutually exclusive, so I was just pointing
out the different options. The example you cited from the docs was
assuming case 1. For your situation, case 3 (or maybe even case 2)
should work.

Hope it helps,
Sheldon
ANSYS, Inc.

--- In xansys@y..., tom.davis@f... wrote:
> I am trying to do a transient analysis with substructuring. There
are two
> load steps. The first load step is strictly for applying an initial
> velocity which I do with D and TIME commands. As per the manual, I
have to
> DDELE the displacements from the first load step when I start the
second
> load step. HOWEVER, there is a small note in an example problem in
section
> 5.3 of the manual that says D and M may not change between load
steps.

> So, how can I apply an initial velocity and then move to my second
load
> step without changing D? The manual also talks about using IC to
specify
> initial velocity. I'm not sure I understand how to do this. Do I
use the
> "value2" field to specify velocity? If so, could I then use IC and
skip
> the first load step?

> Any and all suggestions appreciated.

> Thanks.

> TWD

> Thomas W. Davis, P. E.
> FSI International (http://www.fsi-intl.com/)
> 3455 Lyman Boulevard
> Mail Station 4-1035
> Chaska, Minnesota 55318
> PHONE 952-361-7359, FAX 952-361-7393
> tom.davis@f...

Posts possibly associated with message #37969AuthorDateScore
37948STRUC-Substructuring for TransientTom Davis2002/05/22 
37959Re: STRUC-Substructuring for TransientStan Schwantes2002/05/22 
37963Re: STRUC-Substructuring for TransientTom Davis2002/05/22 
37968Re: STRUC-Substructuring for TransientSheldon Imaoka2002/05/22 
37969Re: STRUC-Substructuring for TransientTom Davis2002/05/22