Partitioning help

[limtom77]

I am creating a part which needs to be partitioned into about 6 separate volumes (but still be considered ONE part).

I notice that if I don't create the features in a certain order, when I start to partition, I get errors.

Is there a certain convention about the order in which I create the features (extrusions)?

 

[fsi]

Partition should be done last after all construction operations. Partitioning part can not include any tangent edges or curves.

 

[limtom77]

cannot "contain curves", meaning, for example, i cannot partition a cylindrical volume within a rectangular block?

thanks

 

[fsi]

I meant it can not contain a line and curve that are tangent to each other.

 

[RCDLtd]

Partition part should work with any profile that any other operation will work with, including lines and curves that are tangent.

With regards to ordering operations then once partitioned, a part will become non-manifold. Some operations after this point may not react the way you expect, such as offset and shell. Operation order is therefore important, but no more important than say for fillets or drafts.

It's a little odd that you want a part to have 6 separate volumes within the same part, this may be better represented as 6 parts within an assembly. Partition part is more a means to an end, ie. to have ribs on the B surface of a part once shelled without having to put in additional operations.

Perhaps more explanantion is required for me to understand fully.

 

[limtom77]

my intent for the partition is to model an FEA with separate volumes that i can specify different material and mesh characteristics.

i seem to have trouble doing FEA on assemblies, and this method has seemed to work well, up until now, now that the parts are becoming more complex.

any advice?

 

[limtom77]

a follow up question...

if i create a partition and it tells me that the "partition is not fully defined", and it also turn yellow on my History Tree, is there a way to query what exactly is "not defined" ???

thanks

 

[RCDLtd]

I did a little bit of thinking after my last post and the new "volume" entity that came in with V9 has had an effect on what you can do with partition part.

So I'll revise my last statement, it's no longer suitable for ribs on a B Surface, although it can create pockets. - Not sure I see the point now within the Design Application.

However, if your modelling specifically for meshing then I do see the point.

Your error message is also related to the Volume entity introduction. The software feels the partition is "not defined" unless the partition itself creates a separate volume.

And I understand it's sentiment if you are using the Partition for the creation of further solid features. ie. if you shell a partitioned part that is "not defined" then the resultant now contains free edges - unlike it used to pre-V9.

As that's not your intention you can probably overlook the yellow nodes and carry on.

Perhaps I need to review the V9 update training, but I don't see this change as a move forward. I'd be interested to hear others' views to perhaps shed some light on what I'm missing.

 

[fsi]

RCDLtd
U are incorrect in ur statement:
"Partition part should work with any profile that any other operation will work with, including lines and curves that are tangent."
The following statement is from:
D:\SDRC\I-DEAS9\SDRCHelp\LANG\English\qascr02h.htm
"Partitioning Restrictions
You can't partition solid parts when the edges of the dividing surfaces are tangent. Although in the figure below, (A) and (B) look like they could divide the solid into two volumes, partitioning won't work. You also can't use the surfaces of (C) and (D )to partition the second solid part into two volumes."

limtom77
"partition is not fully defined"
This means your partitioning part or section did not go across the full x-section of the part being partitioned. It probably stopped at a surface location not at an edge or vortex. Make your partitioning part bigger or section larger.
I hope this helps.
fsi

 

[hinterneder]

limtom,

Getting back to your original question on the 'order or operations' being important:

2 quetions:

1) Are you fully constraining the geometry you're using to build the partitions with?

If not, issues might (will likely) arise when changes to the geometry (either partitions or original model) occur.

2) Are you only building 2-D sections, then using the resulting features to try and define partitions that really need to be defined seperately and then 'cut' into the original part (ie, partition with a boolean operation)?

If so, that might be another way to go.

Regarding the other posts, the key with most solid modelers (and I-DEAS in particular) is to avoid forcing it to decide what is Zero when it comes to 3-D feature creation/modification (ie, the tangent and coincident items discussed above). If you are sure to over/under extend your geometry (depending on the goal), you will generally have more succcess in building a particular feature or part.

If this doesn't help you may want to try a more in-depth description of you problem.

-ed

 

[RCDLtd]

fsi,

On my machine, partition part works fine with a profile that includes tangent lines and curves.

Reading your quote from I-DEAS Help, then I thought that it may be ambiguous and was refering to a profile that met the main solid tangentially when extruded. However, just for research, I tried that also and it worked fine.

With regards to "partition not fully defined", on my machine, a closed section that is fully within the "x-section" of the part, not crossing any of edges or vertexes does NOT give this error. Only when an open section is used that does not complete divide the base part into two separate volumes do I get this message.

 

[limtom77]

i finally succesfully partitioned the model.

you guys were correct in saying to build the entire solid before paritioning it.

i was building each feature and partitioning it immediately after. i believe this is what caused the problems.

thanks for all your help!

 

[limtom77]

I've come across another problem.

I'll try to explain this a best as I can.

Let's say I have one large cylinder (5" diam, 4" height).

Inside this cylinder, I want to stack two smaller cylinders (like a wedding cake).

I want a 4" diam, 2" height cylinder on the bottom, and stacked on top of it, I want a 3" diam, 2" height cylinder. Both of these cylinders are within the large cylinder.

I can succesfully create the large cylinder, then create a partition for one of the inner cylinders, but when I try to partition/create the 2nd cylinder, I have a partition that is not "fully defined".

Any advice?

 

[fsi]

Remember:
"Partitioning Restrictions
You can't partition solid parts when the edges of the dividing surfaces are tangent. Although in the figure below, (A) and (B) look like they could divide the solid into two volumes, partitioning won't work. You also can't use the surfaces of (C) and (D )to partition the second solid part into two volumes."
Try using "Joinpartition" command. May be it will work.Try it in several different orders.
fsi

 

[RCDLtd]

limtom77,

I had no problem performing the Partitions as long as the smaller diameter cylinder fitted within the top of the larger diameter internal cylinder.

If the smaller cylinder profile "crossed over" the top of the larger cylinder then I got the "not fully defined" error message and it left a surface out on the base of the smaller cylinder.

 

[limtom77]

ok, i got it to work.

i found out that when defining the section to extrude, i have to select both the outer edge of the large cylinder, as well as the edge of the inner circle.

thanks