Invetor 2018 - Export translation error?

Anonymous

Invetor 2018 - Export translation error?

Anonymous
Not applicable

I've noticed something very odd and concerning about how inventor is exporting my models. What i see in inventor doesn't get translated properly to any other CAD format, IGES, STEP, X_T, X_B, etc. In this model i have a round sphere tangent to a part surface, and inventor shows zero gaps between the surface of the sphere and the surface of the part. But when i open up the file i export from inventor and open it up in Solidworks i can see a gap between the two. Why is this?? I attached screenshots from both inventor and solidworks looking at the same feature and solidworks shows a gap of about 0.003".

0 Likes
Reply
822 Views
11 Replies
Replies (11)

Anonymous
Not applicable
Can you post the part files I'm sure someone can investigate.
0 Likes

johnsonshiue
Community Manager
Community Manager

Hi! Without seeing the file, I would say INV's solution is probably right, because tighter tolerance was used. But, it can go either way. Please share an example here. There should be a logical explanation to the behavior. Either INV or SWX is right.

Many thanks!



Johnson Shiue (johnson.shiue@autodesk.com)
Software Test Engineer
0 Likes

Anonymous
Not applicable

I'll try recreating the same problem again and upload it. The model I'm having an issue with, i can't release it publicly.

0 Likes

Anonymous
Not applicable

I attached a sample that i was able to recreate the same issue. I included both the IAM and the IGES i exported from the IAM.

0 Likes

Anonymous
Not applicable

Something went funky with the attachments. Its not letting me attach the IGES file. I attached a STEP file and the IPT for the sphere.

0 Likes

Anonymous
Not applicable

You will need to attach the parts as well as the assembly.

0 Likes

Anonymous
Not applicable

Sorry about that. Been a long day.

0 Likes

Anonymous
Not applicable

I opened the .stp assembly, the ball is tangent to the curved surface:

Tangent.jpg

0 Likes

Anonymous
Not applicable
What you may be seeing is a graphics resolution issue.
Try zooming in and out, does the display get any smoother?

Anonymous
Not applicable

Sorry for the late follow up, its been busy.

 

I took your suggestion and it didn't make a difference. It doesn't appear to be a graphical issue. This also doesn't make a whole lot of sense from what i dug into with different software.

 

I took the model and popped it into SolidWorks and did a minimum distance check between the two surfaces. The result was they both intersected according to the software, same as inventor's result. I put the same model into Kubotek Keycreator and did the same dimension check, and this time i got a dimensional value from it of 0.000485", screenshots attached as well.

 

Then i popped the same model into Polyworks and re-created the intersect vector of the two features, since Polyworks doesn't have a direct built in function to check for minimum distance. You can see how i made the constrains in Inventor. I recreated the same axis that i constrained the sphere and created intersect points between the axis and the surface of both features. Then took a distance between the two points and came up with a different number of .00012522". Now i could have a bit of error in Polyworks because i recreated the intersect vector, but its still generated a value.

 

I don't understand where/what the difference is.

 

0 Likes

TheCADWhisperer
Consultant
Consultant

@Anonymous wrote:

I attached screenshots from both inventor and solidworks looking at the same feature and solidworks shows a gap of about 0.003".


1. You are confusing images with geometry.

The image display (not to be confused with geometry) is a function of both the settings of your software and your graphics card and driver.

SolidWorks Image QualitySolidWorks Image Quality2. How are you measuring "about 0.003" in SolidWorks?  By eye?

3. Your Tooling Ball part was originally a SolidWorks 2010 file - could SolidWorks 2010 be exporting incorrectly.

(Not the issue, but a confounding variable.  It is interesting in light of your thread title.)