Problem with 3D Dimensions

[PatrickMacMahon]

Hello,

I am having great difficulty applying reliable dimensions in 3D sketches. 

For example, while the dimension states that the length of a line is 2,000.000 mm, it is in fact very slightly less, (1,999.9999995443) when checked with the Measurement tool. 

Of course, this is so small as to be normally irrelevant, but if this is one (1) of eight(8) placement points in a template, (jig) for positioning a "flexible" Assembly, then this tiny difference makes the task impossible.

I attach both a PDF of the problem and part of a PART file for the forum's information & edification and any advice would be most welcome.

Thanks

[kelly.young]

@PatrickMacMahon thanks for attaching the part. I can't really provide any explanation for that.

I forwarded to the development team to inspect. Will post back with any feedback.

[blair]

Possibly a rounding error when the document only calls for 3 decimal places and you are measuring to 7-10 decimals. You can always use the Tolerance option when an 8 decimal place discrepancy occurs.

Looks like a distance was a product of a conversion or mathematical calculation to get that many digits. Not sure of any manufacture working to 1 ten-billionth.

[johnsonshiue]

Hi! I have to say the behavior is a bit surprising to me. Although the measured distance is virtually the same as 2000mm, the input value 2000mm should be honored. This particular case seems to use a slightly bigger tolerance when solving. I will work with the project team to understand the behavior better.

Many thanks!

[PatrickMacMahon]

Thanks,

Just to add that I have attempted used  the constraint +/-  feature from 1-5  mm but to no avail.  

The measurement observation shows a tiny variance but it seems this is significant enough to throw off the entire assembly.  

[Xun.Zhang]

Hi @PatrickMacMahon,

Thanks for reporting it, ticket INVGEN-11624 to track it down.

Thanks again!

[Xun.Zhang · Accepted solution]

Hi @patrickmm,

The ticket has been investigated internally, it related with Inventor application level tolerance setting. However, the tolerance setting is very sensitive and wildly impact every domain of application, so it will take more time to investigate.

The idea more like to have a control-able tolerance setting to help user for large scale modeling.

By the way, even though, it will not impact the accuracy of modeling since the application level tolerance is persist for every function, and the drawing annotation respect the value of 2000 mm for such case. In other words, it should be only a display problem, please continue to leverage parameter driven design for modeling.

Thanks for your understanding! 

[Hochenauer]

2,000.000 mm  = 2000m = 2km or roughly 1.25 miles is a size that is pushing the numerical boundaries of a part file.

There are limits what the modeler can accurately represent in both directions.