Question on void command Log Out | Topics | Search
Moderators | Register | Edit Profile

FlexPDE User's Forum » User Postings » Question on void command « Previous Next »

Author Message
Top of pagePrevious messageNext messageBottom of page Link to this message

Inge Buanes (iiiinge)
New member
Username: iiiinge

Post Number: 1
Registered: 08-2008
Posted on Saturday, August 30, 2008 - 03:00 am:   

Hi,

im currently making a 3D model as seen in the picture, which I have split into 3 regions, left, middle and right, with 3 layers, bottom, middle and top.

My problem is that I want the top layer to only cover parts of the cube, and I am able to do so using the layer 'xx' void command in the middle region.

The question is then, is there any way to remove parts of the top layer without making a cumbersome third region? I want to be able to use only 2 regions so I don't have to mess around with boundary conditions across the middle region.

If not, what happens if I dont specify any boundary conditions in the limit between the left/middle and middle/right regions? Does FlexPDE automatically couple the differentials?

Thanks in advance

Inge3D figure
Top of pagePrevious messageNext messageBottom of page Link to this message

Robert G. Nelson (rgnelson)
Moderator
Username: rgnelson

Post Number: 1163
Registered: 06-2003
Posted on Monday, September 01, 2008 - 08:50 pm:   

There are several ways to handle the configuration you have shown, depending on the material properties in the various regions.

1) the way you have done it is acceptable, and allows you to put different materials in any of the colored boxes. This is the most general form. You don't need to do anything about boundary conditions on the interior interfaces. By default they are treated the same as any other cell boundary. The surface integrands corresponding to the Natural BC (see NATURAL in the help index) will be assumed continuous across the interior interfaces. Usually, this means flux conservation.

2) You could define only two regions, if the material properties stacks on the two outside columns are the same. Then region 2 is merely the void part.

3) you could define two regions (as above), but declare the void region as a LIMITED REGION, resident only in the top layer. In this case, the first two layers would be the same throughout the figure. (See LIMITED REGIONS in the Help Index).

Top of pagePrevious messageNext messageBottom of page Link to this message

Inge B (iiiinge)
New member
Username: iiiinge

Post Number: 2
Registered: 08-2008
Posted on Wednesday, September 03, 2008 - 04:25 am:   

Seems like it is option number 1 I have to use.

In the figure the bottom layer including the red, yellow and orange regions are of the same material type. However, layer 2 is made up of two different material types, divided as the following; The left darker green plus half of the light blue layer are of one type, and vica versa for the right side of the model.

Any suggestions on how to approach this problem?
Top of pagePrevious messageNext messageBottom of page Link to this message

Robert G. Nelson (rgnelson)
Moderator
Username: rgnelson

Post Number: 1164
Registered: 06-2003
Posted on Wednesday, September 03, 2008 - 01:52 pm:   

REGIONS delimit areas of uniform material properties. If half of the light blue "region" is one material and half another, then the light blue is not a REGION. If layer two is divided in half then you need two regions in layer two. Each of these REGIONS must be made up of the same material.

The best way to do this is with LIMITED REGIONS.
REGION 1 is the entire figure.
REGIONs 2 and 3 are the halves of layer 2, and are LIMITED to layer 2.
REGION 4 is the two top patches (if they are the same material - otherwise REGIONS 4 & 5) and is limited to layer 3. This leaves the gap on layer 3 defaulted to REGION 1, so you can declare REGION 1 VOID on layer 3.

application/octet-stream
three_layers.pde (0.7 k)

Top of pagePrevious messageNext messageBottom of page Link to this message

Inge B (iiiinge)
Junior Member
Username: iiiinge

Post Number: 3
Registered: 08-2008
Posted on Wednesday, September 03, 2008 - 03:43 pm:   

I thank you so much for your help.

The reason I didn't try out the limited region as from your first reply is that I use a licensed version of FlexPDE 3, and I don't get it to work with the limited command. I'll see what I can do about that.
Top of pagePrevious messageNext messageBottom of page Link to this message

Marek Nelson (mgnelson)
Moderator
Username: mgnelson

Post Number: 71
Registered: 07-2007
Posted on Wednesday, September 03, 2008 - 06:04 pm:   

Version 3 does not support LIMITED regions.

Attached is an equivalent script for version 3 that does not use limited regions, only VOID layers in the appropriate regions. This was done by projecting all the figures in the domain down into the XY plane. Then, any one of these areas that has a different stack of materials above it, must be a separate region. Here there are four separate regions.

This script is also valid on version 5.
application/octet-streamFor Version 3
three_layers_v3.pde (0.7 k)
Top of pagePrevious messageNext messageBottom of page Link to this message

Inge B (iiiinge)
Member
Username: iiiinge

Post Number: 4
Registered: 08-2008
Posted on Tuesday, November 18, 2008 - 09:13 am:   

Hi again, thanks for a great tool and help, but I have one more simple question.

In the attached file I have written down the equations I use for solving my system (2D). However, instead of using a relative permeability for some of my materials, I'd like to use a B-H curve.

How can I do this as simple as possible?

Most other finite element programs have a list of integrated non-linear materials in a library that I can just choose from, but not this one.

Thanks in advance
text/plain
Equations.txt (0.7 k)
Top of pagePrevious messageNext messageBottom of page Link to this message

Marek Nelson (mgnelson)
Moderator
Username: mgnelson

Post Number: 83
Registered: 07-2007
Posted on Tuesday, November 18, 2008 - 04:26 pm:   

FlexPDE is applicable to so many problems that we could not tabulate all known material properties. Material properties can be defined using expressions or tables of values. See TABLE and TABLEDEF in the Help index. There is also several sample problems in "Samples | Misc | Import-Export" that shows the use of table input.

Attached is an example based on your problem, where I define 'mu' in the table (the values are just your H values so will need to be changed to B/H).

application/octet-stream
eq.pde (0.4 k)

text/plain
bh.tbl (0.0 k)
Top of pagePrevious messageNext messageBottom of page Link to this message

Inge B (iiiinge)
Member
Username: iiiinge

Post Number: 5
Registered: 08-2008
Posted on Thursday, November 27, 2008 - 11:53 am:   

I am somewhat stuck with this problem. Thing is that the regions where I import the B-H data should be saturated for 2.4 tesla when the H field is about 318.000 A/m. The magnets I use generate an H field much larger than this, and the soft magnetic materials should be saturated very fast.

Question is then, how do I limit that upper boundary for when these materials are saturated.

iiiinge
Top of pagePrevious messageNext messageBottom of page Link to this message

Robert G. Nelson (rgnelson)
Moderator
Username: rgnelson

Post Number: 1190
Registered: 06-2003
Posted on Thursday, November 27, 2008 - 02:45 pm:   

The saturation behavior of the material will be determined by the data in the BH table. If it is not behaving the way you expect, examine the data in your table and the dimensional consistency of all your equations. You can manually plug some values into the table and see what numbers you get. Be sure that the coordinate range of the table is sufficient to cover all the values that occur during the run; if you reference beyond the table span, you will always get the last data value as a result.

Have you looked at our example problem "Samples | Steady_State | Magnetism | Saturation.pde"?
Top of pagePrevious messageNext messageBottom of page Link to this message

Inge B (iiiinge)
Member
Username: iiiinge

Post Number: 6
Registered: 08-2008
Posted on Thursday, November 27, 2008 - 04:11 pm:   

I have look at that example, and if you make an elevation plot of Bx in different parts of the air region, isn't it unreasonably high?

My main problem is that I need to know the H field in the soft magnetic material to determine the induction B and from that finding mu. However, I need mu to know H, because I always end up with: error: recursive defenition of mu.
Top of pagePrevious messageNext messageBottom of page Link to this message

Marek Nelson (mgnelson)
Moderator
Username: mgnelson

Post Number: 90
Registered: 07-2007
Posted on Saturday, November 29, 2008 - 04:46 pm:   

To get this error, you must be indirectly defining something in terms of itself. You should recheck all your definitions for such recursion.

It is actually kind of hard to get a recursive definition, but here is one way:

DEFINITIONS
mu
mu2=mu+1
mu=mu2-4

If you try mu=mu+1 then FlexPDE will complain that mu is not defined yet. If you use mu=1 in the first line above, then FlexPDE will complain that mu is defined twice. So you have to be indirect in order to create this situation.

Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action:

Topics | Last Day | Last Week | Tree View | Search | Help/Instructions | Program Credits Administration