After I gave it a try...
I started down this path bacause of instructions gregmilliken gave in an earlier post…
gregmilliken said:
In the meantime, or after that you can indeed move back to V8.2, but designs created in 9.0 are not backward compatible, although you can export STEP and get the geometry back without features. Not such a great option I know, but it is one path back.
Then in a later post in response to my comment about errors using STEP…
gregmilliken said:
All you need to do is select the "Make Tolerant" option presented on STEP import and it should work fine, no errors on import or with Check Part, and 3D section views work as expected. "Make Tolerant" is one of the standard options when importing STEP.
Please give this a try and let us all know if it works.
Well I did try and it didn’t [fix my problems anyway]
These findings below bring up some interesting questions...
How can anyone expect other applications to accurately import our data if AD cannot accurately export and import it [even by its own perspective of the standards]?
Is our good data being translated to be " inaccurate or leaky" in the export or the import [or both]?
1. In order to transfer data between 2 different applications, export/import functions are supported in one of the applications...as they are implemented in AD.
2. In an export to one of these file formats [standards] AD takes its model data and translates it so it can be read by the other application and vice versa.
When I do a part export to one of these standards and then turn right around and import the data back in...sometimes the part comes in with [what AD considers] bad geometry.
I for one would be interested in knowing why? And the knowledge may be helpful to others as well...because many people think they can export/import to fix percieved corrupted parts.
I started with a part constructed entirely in AD 9.0 that passes "check part" with no errors.
Then I exported the part to each one of the export file types...then immediately imported the part back in, and it comes in with "check part" and "Healing Analysis reports" of these errors...
This first one is grossly unacceptable because this is Alibre’s own stp format.
[these are check part "low accuracy" errors]...
Alibre file (stp)... Edges = Total 139, Failed 40 ...Vertices = Total 94, Failed 24
STEP AP 203... Coedges Total 278, Bad 94 ...Edges T 139, B 47 ...Vertices T 94, B 24 ...percentage good geometry: 87%
STEP AP 214...Coedges T 278, B 94 ...Edges T 139, B 47 ...Vertices T 94, B 24 ...percentage good geometry: 87%
IGES...Coedges T 278, B 119 ...Edges T 139, B 59 ...Vertices T 94, B 24 ...percentage good geometry: 84%
Even at best...is 13% of the geometry broken acceptable?
ACIS R13...Coedges T 278, B 0 ...Edges T139, B 0 ...Vertices T 94, B 0 ...percentage good geometry: 100%
ACIS R12...ACIS R11...ACIS R10...ACIS 8...ACIS 7...ACIS 6...ACIS 5...ACIS 4...ACIS 3 [same as ACIS R13]
I understand that WE can choose to "make tolerant" or to "heal" the data on import...because this makes data exported by other applications work in AD.
What I can't understand is why we should have to...when the data was just exported by AD.
The import advisor says..."Because the geometry of a tolerant model is allowed to be less precise, inaccurate or leaky data can often be imported using the Made Tolerant option. Making a model tolerant leaves its underlying geometry unchanged. By contrast, Healing attempts to fix problems detected with the model by changing it."
Why should my model change because AD had to heal the geometry to fix problems that were introduced by AD in the export translation,
And why [in order to make it work] it needs to be less precise?...because that makes errors surface later that bite us in the behind...thru failed processes.
One of these failed processes [inserting a 3d section view] is addressed in the next post.
Steve
