Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Assembly (Identity Pairs + Continuity) and Compose Geometry (/Union): Do they represent the same problem?
Posted 07.04.2013, 13:25 GMT-4 Geometry, Structural Mechanics Version 4.3 1 Reply
Please login with a confirmed email address before reporting spam
Dear all,
I am trying a simple example to better understand the Assembly mode. The example is a multilayer (3 layers), where only the layer 2 is at high temperature (so it expands and "pull" the other layers, 1 and 3). The constraints are assimilated to a cantilever beam (the left extremity fixed and the right one free)
I have simulated these configuration:
1. Union mode
2. Compose (Geometry > Boolean operations) applied to the 3 layers, thus independently of the use of assembly or union
3. Compose (Geometry > Boolean operations) applied to the 2 layers (2nd and 3rd), and Assembly mode + Identity pair (1) + Continuity (in Solid Mechanics > Pairs), to couple the 1st layer to the 2nd+3rd
4. Assembly + Identity Pairs (2) + Continuity applied to the 3 layers
Only the results with 1. and 2. are identical, the others are different (as in the pictures that I attached). Why don't all the 4 models give the same result? They should represent the same problem but implemented in different ways in Comsol, or I'm doing wrong?
Thank you in advance!
I am trying a simple example to better understand the Assembly mode. The example is a multilayer (3 layers), where only the layer 2 is at high temperature (so it expands and "pull" the other layers, 1 and 3). The constraints are assimilated to a cantilever beam (the left extremity fixed and the right one free)
I have simulated these configuration:
1. Union mode
2. Compose (Geometry > Boolean operations) applied to the 3 layers, thus independently of the use of assembly or union
3. Compose (Geometry > Boolean operations) applied to the 2 layers (2nd and 3rd), and Assembly mode + Identity pair (1) + Continuity (in Solid Mechanics > Pairs), to couple the 1st layer to the 2nd+3rd
4. Assembly + Identity Pairs (2) + Continuity applied to the 3 layers
Only the results with 1. and 2. are identical, the others are different (as in the pictures that I attached). Why don't all the 4 models give the same result? They should represent the same problem but implemented in different ways in Comsol, or I'm doing wrong?
Thank you in advance!
Attachments:
1 Reply Last Post 07.04.2013, 13:39 GMT-4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
the easiest way to understand where "assembly" applies (that is boundaries that will be duplicated) are to take a look at the Selection view when selecting the geometry Finish node. Then any common boundary of objects in the list will be dedoubles, while any internal boundaries of an object appearing with 1 name will be as in union mode, have one single common boundary.
Therefore if you make a block, and add "layers" (see layers tab), in assembly mode it should normally appear an an "union entity, as these layers belong to the same geometry object. While if you make 3 blocks adjacent and apply assembly mode then you get several domains with dedoubled overlapping boundaries.
Another point to consider (will not be enhanced in your model as all blocks have the same shape) is to turn on the create imprints, then any overlapping boundary of different shape will be cut into parts such that there is one common dedoubled boundary and one or more isolated external boundaries
There are several treads about union and assembly mode.
For the physics 2 thing to consider: assembly boundaries require you to define the continuity or specific physiscs linking the up and down boundaries, the mesh does not need to be identical on both sides, as the domain mesh attached to each respective boundary. A too different mesh density might give poorer results as the precision of the projection might be reduced (all very model dependent)
--
Good luck
Ivar
the easiest way to understand where "assembly" applies (that is boundaries that will be duplicated) are to take a look at the Selection view when selecting the geometry Finish node. Then any common boundary of objects in the list will be dedoubles, while any internal boundaries of an object appearing with 1 name will be as in union mode, have one single common boundary.
Therefore if you make a block, and add "layers" (see layers tab), in assembly mode it should normally appear an an "union entity, as these layers belong to the same geometry object. While if you make 3 blocks adjacent and apply assembly mode then you get several domains with dedoubled overlapping boundaries.
Another point to consider (will not be enhanced in your model as all blocks have the same shape) is to turn on the create imprints, then any overlapping boundary of different shape will be cut into parts such that there is one common dedoubled boundary and one or more isolated external boundaries
There are several treads about union and assembly mode.
For the physics 2 thing to consider: assembly boundaries require you to define the continuity or specific physiscs linking the up and down boundaries, the mesh does not need to be identical on both sides, as the domain mesh attached to each respective boundary. A too different mesh density might give poorer results as the precision of the projection might be reduced (all very model dependent)
--
Good luck
Ivar