Measure tool inaccurate and inconsistent?

[CAD_CAM_MAN]

Using Inventor Professional 2022.2 

The measure tool seems to be giving inaccurate / inconsistent results in some cases. I have not noticed this behavior in previous versions. See below example...

On a new .ipt file a work plane was created with an input of exactly 43.7 degrees from the XY plane through the Y axis. If you measure from the XY plane to the newly created plane the measurement is 43.700007251 deg. or 43.699984915 deg. or 43.700004875 deg. ? Every few times I measure plane to plane I get a different result. Nothing was modified between the measurements. There is no other geometry in the file. It is a new file with only 1 work plane added (File attached). 

We often use  measured angles for setting component angles in assembly fixtures and am concerned this may cause issues. For example if I measure 43.700004875 deg. and build components around that dimension then try to constrain the part that actually has an angle of 43.700000000 constraint may fail etc...

Am I doing something wrong? Is this a bug, or is this expected behavior in the new version?

[mcgyvr]

I believe Inventor internally uses radians and the conversion between the 2 is resulting in this inconsistency.. 

No way you can measure to that precision nor machine parts/jigs to it..

[WHolzwarth]

I've seen similar in the German forum. Here's a file of a triangle, done in 2021.3.3. Angle measurements are exact 60° in 2021.

No more exact 60° values, if opened and measured in 2022.2.

[CAD_CAM_MAN]

It seems if it was a conversion issue the dimension would be the same every time. Maybe inaccurate but at least consistent.

I tested this same scenario in 2020 and the result is correct and the same every time...

I realize of coarse we cannot measure to that extreme in reality but constraints may fail in the Inventor assembly.

[nathan_wostrel]

This also is not limited to angular measurements.  Linear dimensions also suffer from this BUG.  Yes, it is a BUG.

[johnsonshiue]

Hi! Yes, this is a bug. We are aware of the issue. The measurement readout was rounded prematurely. The geometry is precise but the measurement looks shaky. At the moment, you may change the decimal places to 6, so that the noise does not appear.

The issue has been fixed on our internal build targeting 2022.3 update (April).

Many thanks!

[matthew]

Is there any update as to when the 2022.3 will be released? This issue is causing real problems as the accurate decimals is a critical thing when you are making complex sheet metal assemblies where fixing holes need to be aligned. When it worked this was perfect to ensure everything was bang on before adding oval slots for some adjustment.

[johnsonshiue]

Hi Matthew,

All I can say is "soon." I am sorry I cannot commit to a date.

Many thanks!

[matthew]

No worries thank you for the reply much appreciated.

[clive.w66LR9]

Does this same bug effect 2023? 

if i measure a part in the part environment it reads this

if I measure it in a assembly environment AND rotate it slightly i get measurements like.

and this varys depending on what free rotation angle i put it on?????????????????

[johnsonshiue]

Hi! Yes, the behavior can be seen on 2023.0 and 2023.0.1. The fix to 2022.3 did not make it to 2023.0 due to the time constraint. It should be fixed in the coming 2023.1 update. A testable patch will be available soon.

Many thanks!

[bltrubey87]

Autodesk clearly doesn't care about fixing this issue. If they did, it wouldn't still be a problem in Inventor 2025.

[johnsonshiue]

Hi! I am sorry I lost track. I am not 100% sure your issue is the same as the original thread. If possible, please share an example part file here. I would like to understand the issue better.

Many thanks!

[CAD_CAM_MAN]

The issue does not appear to be the same as my original post, but I have also occasionally noticed some peculiar behavior in the current Inventor version I am running (2026.2.1). This appears to be an issue at the assembly level where the original post affected parts as well. I have been working more with API development projects with 2026 and not using it so much for design. So, I may not be up to speed on all the more recent measurement tool quirks / bugs. But here is an example of what clive.w66LR9 and bltrubey87 may be seeing...

I have attached a .zip file with an example assembly showing the issue. TEST_PART-3 is grounded and TEST_PART-4 has a mate constraint to TEST_PART-3. At the part level the constrained face to the measured face is at 90 degrees. when constrained and measured in the assembly the measurement is 89.999999992 degrees.

It appears to be an inconsistent issue. I first brought the components in, grounded TEST_PART-3, added the mate constraint to TEST_PART-4. At that point everything measured fine. I then dragged (rotate) the TEST_PART_4 component (hot key G) and after dragging, the measurement was no longer 90 even after clicking the update button etc. 

[bltrubey87]

I am unable to share the data on this forum because it is proprietary. I'd gladly set up a Teams meeting to show you the issues I am experiencing though. It's wild that I can measure from a component surface to a plane 4 times and get 4 different numbers, and none of them match the dimension that I used to offset the plane. 

[johnsonshiue]

Hi @CAD_CAM_MAN,

Many thanks for sharing the example! Indeed, these imprecise behaviors may look similar but the source of the imprecision could be different. I suspect your case is more of an angular constraint tolerance issue by the assembly constraint solver. The Measure tool does seem to report the accurate angle. My proof is that if you constrain the parts differently (more along the orthogonal planes), the angle will be 90 deg precise. In some cases, as long as the angle is within the internal solver tolerance, the result could be legitimate.

Hi @bltrubey87,

If possible, please share the files with me directly at [email protected]. I would like to understand the Measure imprecision behavior better.

Many thanks!