Coaxiality accepted despite the difference in centre distances

3D Design Assembly
1

Hello

I realized a problem during an assembly using coaxiality constraints.

For simplicity, I am sharing with you the problem with the use of two very simple pieces: Two plates each pierced with two holes, named Piece1 and Piece2.

Part1, red in color, has two 50mm center distance holes
(Attachments - A)

Part2, blue in color, has two 50,001mm center distance holes.
(Attachments - B)

When assembling these two plates, I apply a coaxiality constraint between the first hole of Part1 and the first hole of Part2.
The same goes for the second hole.
Despite a delta of 0.001mm, coaxiality is well accepted (no error message):
(Attachments - C)

I now change the center distance on Part2 to 50.0011mm:
Coaxiality is no longer accepted:
(Attachments - D)

So it seems that Solidworks is forcing a coaxiality constraint without triggering an error if the center distance delta is less than at least 0.001mm.

Until now, I very often carried out perfect coaxiality checks by applying a coaxiality constraint between all my holes, i.e. by over-constraining (hyperstatism) my assembly. But at least that way, if a coaxiality was displayed in red, I knew I had a mismatch to correct.
Now that I realize this uncertainty, I have to check my alignments one by one with the measuring tool which gives with precision, the difference in coaxiality between the holes.

Do you have a solution to reduce the coaxiality uncertainty to 0mm? That is, make Solidworks display an error message and a coaxiality in red if the delta is nonzero?

NB: The only info I could find on the Solidworks help is the " Allowing misaligned coaxial stresses ":
(Attachments - E)
https://help.solidworks.com/2021/french/SolidWorks/sldworks/t_allowing_misaligned_concentric_mates.htm?format=P&value=

But it doesn't seem to be my problem because the two " misaligned " coaxial stresses should then appear in this form:
(Attachments - F)

Which, as a reminder, is not my case:
(Attachments - G)

Thank you in advance for your help.

Kind regards

Parts jointes.pdf (494.6 KB)

2

Hello Remi_DOLE,
I notice that you use sketches or edges, that's why you have this defect, it's better to use surfaces/surface or PLANE/PLANE, especially not the rest, (unless otherwise in the case of non-planar surfaces or and put a distance constraint).
There you go. Good luck.
@+.
AR.

2 Likes
3

Accuracy may be related to the accuracy of the units in the document options.
It would seem logical to me:
image

4

Hello A.R,
The coaxiality constraints are each applied to the cylindrical surfaces of the holes, not to their edges or sketches. The sketches that I highlight are just there to show the center distances, no constraints are linked to them.
Here we have a simple example, but in my constructions, I have, for example, a centering between a barrel and a pin, and it is perfectly coherent to constrain the elements by a coaxiality between their cylindrical surfaces, which are the functional surfaces in reality.

5

Hello sbadenis,
Apparently that's not it, I'm 10th-6 accurate on my lengths.

6

Hello again Remi_DOLE,
Ok for your remark, but sorry I don't have any other solution to propose.
Kind regards.
AR.

1 Like
7

Hello;
See in the options " document property" => Constraints:
=> It is possible to allow deviations on the coaxiality constraints...
image
… This is an aberration created by Solidworks since 2018...

But that's not all: (it would have been too simple...)

In the System options, => Assembly:
You can also allow (or not) the creation of misaligned constraints...
image

https://help.solidworks.com/2018/french/SolidWorks/sldworks/t_allowing_misaligned_concentric_mates.htm?verRedirect=1

5 Likes
8

Hello @Maclane,
The " Allow creation of misaligned constraints " option was unchecked.
Once checked, I can apply a coaxiality, even with any delta of center distance, in short, I can align two holes several meters apart...
création contrainte option désalignement

Then the specific logo appears:
lolo coaxialité désalignée

The option that would seem to correct my problem is the following:
options

However, I don't see any difference whatever the selection...

1 Like
9

And if apply the settings as in my two screenshots:
=> Uncheck " Allow misaligned constraint creation"
=> Dimension of Zero in " Misaligned Constraints "

Normally a your constraints should be wrong...

1 Like
10

With your settings, it doesn't change anything.
By now selecting " Never " in " Change constraint alignments when editing ", it finally shows me an error for a delta of 0.001mm...
… But no mistake for a delta of 0.0009mm...
We have still progressed by 1/10 of a micron, which is not nothing :joy:

2 Likes
11

Hello

Under SW2020 Sp5.0 there was already the problem (and no " modify constraint alignments" box).
It's surprising that we have this kind of problem and that it's not related to an adjustable parameter in the options.

I think we should escalate the bug to SW (so that at least the value used by the software is configurable when the box ' allow misaligned constraints ' is not checked).

12

If I understand correctly, Solidworks would have created the checkbox "modify constraint alignments..." rather than trying to solve the problem of coaxiality!!
(Note, it looks like them... :upside_down_face: :upside_down_face:).

That said, there is one point that we did not look at, besides the unit precision in the options (See @sbadenis message)... There are also, perhaps, the settings in the " mass/section properties" options... depending on the settings set, this may distort the reading of the " Measure " tool.
image

@Remi_DOLE ... check that you have the options for " Masses/Sections " on "Document Settings)...

1 Like
13

It's beautiful to dream!! :beach_umbrella:

14

@Maclane
My options are well set as you indicated on your screenshot...

15

And yes, and it's cheap, plus it's free... :rofl: @+.
AR

16

Hello @Remi_DOLE

Perhaps sensors would be more practical for highlighting deviations than measuring them one by one.

17

Hello @Sylk
Sorry for the late reply.
This would be a solution for the verification of some coaxialities.
But in my case I have a few hundred...
I will make a request directly to Visiativ.
Thank you all.