Sketched Dimension and Measure Tool Inconsistency

[kaori8996]

Hi,

Please check the below pictures, there are very small differences between sketch dimension and measurement.

I swept a profile along curve.

  

[kaori8996]

[TheCADWhisperer]

Can you File>Export your *.f3d file to your local drive and then Attach it here to a Reply?

What measuring instrument will you use in the real world to detect that “difference”?

[kaori8996]

Sorry I cant export and post it here, cause it's the company's property.

There should be no difference in software, it's not a real world. And it cause a problem and was not able to make fluid volume.

[kaori8996]

[TheCADWhisperer]

---

@kaori8996 wrote:

Sorry I cant export and post it here, cause it's the company's property.

---

Can you make up a dummy file that exhibits the same behavior without revealing any proprietary data.

I can then check in other CAD software programs.

[kaori8996]

I will try to make and post it here. But I designed in Autodesk Inventor without any issue. I am not caring about the difference, I was not able to make fluid volume so I increased precision and found the difference between measurement and sketched dimension. I thought that could be the source of problem.

[kaori8996]

Please check the taper angle parameter and measurement result.  This small issue is making a big problem for me.

@TheCADWhisperer 

@Phil.E 

Thanks

Ali

[Phil.E]

Can you describe how this impacts you? I totally understand that you see this as a problem, but I'm wondering how or where you see the problem arise, other than these measurements. In order to assess the issue, and prioritize the impact, I need to hear how this affects you.

Does it affect CAM operations, or how the part fits into an assembly, or something like that? Let me know so I can include it in the report I need to write. Thanks for taking the time to explain.

[kaori8996]

Thanks for replying.

The parts that I am creating is brazed plate heat exchanger. I should do CFD simulation therefore I need to prepare the volume between the plates. I created the plates and used fluid volume command to make the volume in between but I got this error "There was a problem combining geometry together ..... (problems can occur where faces and edges are nearly coincident), even though there is a contact between parts in cross section. I check the dimension with higher precision and found out there is a very small difference between what I input as dimension and measurement. I speculated maybe that would be the issue. 

Beside the above problem, I am afraid if that small difference will raise an other issue in CFD simulation. 

[Phil.E]

Thanks for the details.

""There was a problem combining geometry together ..... (problems can occur where faces and edges are nearly coincident), even though there is a contact between parts in cross section."

Did that warning also state that you should ensure a clear overlap or gap between parts to avoid this? When parts are in contact they can give that error. The volume you are creating is having trouble doing the math along the boundary of contact between it and your parts.

Does the same warning appear when you manually build the fluid volume, using Boundary Fill? It also might help for me to see the entire design in this case.

In any case, it's not clear that this error would appear because of the results you are finding.

Also, the results you are finding are not new. Most people do not use precision that is orders of magnitude beyond what can be produced in manufacturing, and many orders of magnitude over the manufacturing tolerances involved to produce physical parts. So most people use 8 place precision or less in drawings, or measurements and do not see this result.

I'm no CFD expert, but I would suspect that a difference in angle or length at 14 decimal places is not measured by CFD solvers. The mesh on your CFD model will be millions of times larger than that difference. Mesh is generally in the millimeter size range, and it is very much an approximation made of triangle faces that are orders of magnitude less accurate than what you are measuring.

14 place precision is far beyond the horizon of solvers and CNC

8 place precision is recommended

The drawing precision is 8 places

So the only issue I think you need to overcome to complete your task is the volume body to use in CFD. As I said, try using a combination of patch and boundary fill to define the volume. Let me know if that works. If not, we can try something else.

I'll file a bug report using this post and your model as an example where these numbers can be confusing to users. Thanks for taking the time to explain it.

Regards,

[kaori8996]

Dear Phil,

Thanks for replying.

I have not tried to create fluid volume using Boundary Fill. I will do and let you know the result. The length tolerance in our design is +/- 0.01 mm and I don't need to go beyond, even though I feel not good when I see this issue. Attached are some picture which I drew a model by other software and created volume in Fusion 360, using fluid volume. Thanks for taking the time to explain.

  

Best

Ali

[Phil.E]

Ali,

Thanks for reporting this and your patience with the issue. We think there is definitely room for improvement and this case is a good example. Thanks for adding your voice to the community!

[kaori8996]

Dear Phil,

Thanks for your great support and Fusion 360 team.

Best

Ali