Cool(FEM) Analysis

[nishit78]

I finally completed my first cool(FEM) analysis using AMI-2012. I have some very general questions i wanted to ask with respect to modeling and results.

Modeling:

I created my mold surface (region) using the mold surface wizard and then converted it to a 3D mesh with my parts and curves layers turned ON. The insides of the mold block were properly captured. However, i had to assign the entire mold block just one steel material.

However if there are multiple mold components (all different materials) i experimented with importing those as seperate files into my study (simplified to just blocks, no internal stuff). Then i assign the mold surface property, 3D mesh it and assign the mold block property. However the insides of the blocks were not automatically captured as they were before.

WHY??? is it because the "region" created by the wizard is treated differently from an imported surface.

What is the best way to deal with multiple steel components for accuracy with Cool(FEM).

Results:

There is one group of results called "transient" and one group called "transient from start up". The latter is obviously from the start of the analysis i.e. cycle 1 @ 0.0 sec till steady state is achieved, but which particular cycle does the "transient" refer to. My guess is the last cycle in the Cool(FEM) i.e. "transient" is just a subset of the "transient from start up".

Please help clarify these points. Are there any best-practices for doing this efficiently.

Thanks

Nish

[nordhb]

Hi,

Modeling:
Yes,- " the "region" created by the wizard is treated differently from an imported surface"
In Moldflow Synergy 2012 one have to import mold block and insert geometries in same study as cad bodies, for all geometries to be recognized when meshing to Cool FEM.
The mold cad bodies should represent mold block, inserts, cavity, and holes for cooling channels, runners (need center lines of channels/runners to mesh to beams).
In 2012, using Synergy mold surface regions, and insert regions created within Synergy, those cannot be recognized when meshing to Cool FEM.
There is a Synergy 2013 beta, where you can test to create mold surface regions and insert regions within Synergy, and mesh to Cool FEM. ( http://forums.autodesk.com/t5/Autodesk-Moldflow-Insight/Moldflow-Beta/td-p/3242606 )
In Synergy 2012, to model mold with inserts, use a solid cad geometry, with the internal features modeled.

The Transient you refer to is Transient within cycle.
The assumption of Transient within cycle, is that it is at steady state.
The Transient within cycle analysis option can simulate the cyclical nature of the mold temperature within a cycle and enables you to see the cyclical temperature variation and deviations from the cycle-averaged mean temperature.
The Transient within cycle is faster than Transient from production start-up, as Transient from production start-up runs many Cool analysis, with input from previous Cool cycle to see how many cycles are needed until stable operating conditions, a steady state, is reached.

This to be compared to Averaged within cycle
While not a transient analysis, the Averaged within cycle analysis option uses the Cool (FEM) solver to determine the average mold temperature during the molding cycle for each location in the mold. This solution is the default selection, and it is equivalent to the existing boundary element method (BEM) Cool analysis sequence. 

Hope this clarifies.

Regards,
Berndt

[nishit78]

Yes Thanks,

I think I understand. So just to summarize the steps:

If i have a design with Cavity H-13, Stripper S-7 and Core 420SS all of which have cooling circuits in them. I should import all of the following:

1. The 3 steel comps (with internal feed and cooling circuit details)

2. The plastic part

3. Centerline representation of feed and coolant channels

Then i should assign the steel comps a "mold surface" or "mold insert surface" property.

Finally i can begin meshing the part

Then i select Cool(FEM) as the analysis type

Then mesh the mold blocks/inserts using the Cool(FEM) mold mesher

Ultimately assing the mold tetras the correct properties.

If i did not have MDL or if i had to import these files as IGS then i'd get into a little bit of a trouble trying to get everything to match perfectly, right?

Another question is that if i import the mold comps, mesh them and assign them the relevant properties for mold steel then i would not need the Cool(FEM) mold mesher?

I guess i could experiment with various combination but i wanted to see if there is a quicker answer.

[nordhb · Accepted solution]

Hi,
the workflow I use is as below:
Import the model/assembly.
Mold, inserts and cavity/part can be in same study. When importing an assembly each cad body is assigned to a layer.
Set Mesh type 3D and
Set analysis sequence to Cool FEM

Turn off Mold block layer.
Mesh cavity/part and inserts. The property import has Part (3D), and keep that and mesh.
(there is an option when meshing to "Stop after surface mesh generation" to be able to inspect the mesh, and fix if needed.)
Change insert part properties of mesh to mold insert properties.
Mesh cooling channels and runners.

Turn on Mold block layer, and set properties to Mold Block.
Do Cool FEM mesh, with the cavity and insert mesh layers on.
(there is an option here as well when meshing, to "Stop after surface mesh generation" to be able to inspect the mesh, and fix if needed.)
So one use the Cool FEM mesher to create volume mesh of the mold block.

Note: Do not forget to change cooling channel beams properties that goes to have same material properties as the material they go through.

Regarding Autodesk Moldflow Design Link:
From the release Autodesk Moldflow Design Link 2012, you can install the default version that
will operate with *.ipt, *.iam, and *.sat files, with no license required.
You can use Autodesk Inventor Fusion 2012, to import various cad formats, and save in .sat-format.

Regarding contact, of course it is ideal to ensure precise match on contact interfaces, but not always achievable.
One might have to put a lot of effort to make a precise match depending on cad model quality.
Each node on interfaces will have a corresponding node on the other side, and it makes data communication between adjacent bodies easier during analysis.
However, this is not the only way to transfer data from one body to its neighbors. The Moldflow solver development has proved that this can be done without this limitation of precise matching, so the solvers of Moldflow Insight 2012 are more capable to deal with not precise match.
So, it is good if nodes are matched on interfaces. If CAD model is clean, "precise match" is a good option, but not compulsory.

Regards,
Berndt

[sadashiv]

Hallo 

Thanks for the information. I have created mold surface as well as mold insert surface region inside moldflow in beta 2013. The mesh is still running will let you guys know if it is working or not. Moldflow creates many layers e.g. some ami triangles, internal triangles, exteral triangles n so on.Selecting each surface one must give the properties. i.e. for mold surface one and mold insert surface other steel and so on.. and have also added cooling using moldflow cooling so to cut the story short a complete mold concept for thermal analysis is created in moldflow which will help to find accurate cycle time.

The rapid heating channel can be used on hot runner for externally heated hot runner the heating coil can be placed on the hot runner block and the heat transfer between hot runner and mold for rapid heating with cool FEM can be hopefully simulated will have to test it.

The stitch surface is a gr8 help to add inserts and have to test the Rapid heating but the development is really impressive and hopefully it will work. Nice to see that we can import the complete Mold in Moldflow at least cold runner mold and Cooling analysis is also developing fast for complext cooling hopefully the beat elements will go away soon and everything will be in 3D for complex parts.

The 3D large deflection is taking really long it is showing little more deflection tats all.