Beug constraint references?

Hi all

When we insert a part with a stress reference in SolidWorks 2016 SP3 (here the face of part A coincides with the face of the other part B), part A orients itself on several of the faces of part B except for one where it takes an unwanted tilt, regardless of the shape of part A.

When we change the orientation of part A (via the TAB key), the face that presents the problem varies and another face becomes problematic...

You will find the example attached ;) !

Could you help us understand the origin of this behavior ?

Thank you:)

Aurélien

 


teste_references_de_contraintes.zip
1 Like

Hello

A solution that will most likely work: wait for the SP4 version:-D

That's why it's advisable to avoid SP1 and SP2 and install SP4 directly!

1 Like

Well and after the SP4 I am waiting for the SP5? :)

5 Likes

Hi @ Aurelien

not having been able to open your room via your version or at least the same one (prehistoric SW 2012)

but via the subject described 

I think you have to recreate this part if it's an import to block the sides of the parts

and if instead of taking in reference the faces but rather the planes of these said faces

So create a shot on each side

can this change

Unfortunately, you're too far ahead of me and I can't open the rooms.

 

On the other hand, a stupid question. But before you perform your constraint, if you check the angulation between the two parts, is it good?

Sometimes, I feel like I have a flat face before realizing that it is free and unconstrained (and therefore, neither angled nor horizontal/vertical) and I find myself with this problem

1 Like

Hi José22 :)!

Already tried but without success....

To explain more simply, I'm going to do some screeshot ;)!

Defining the stress references on the test part:

Reaction in the assembly:

See the attached video:)

 


20160504_094022.mp4
1 Like

First of all, special bigup for the video. It explains so much better!

 

Indeed, it's hot as hell. If you update your part, do you keep the same problem?

If you create a new "big cube" on which you put the piece, always the same?

What if you do more "classic" constraints? Like a coincidence side, then middle shot and middle shot? Can't it be pre-constraints that are poorly defined on the big cube?

1 Like

@coin37coin,

Yes we tested with a new cube, other constraints, etc, always a side that messs up!

On our tests, the stress references are only on the add-on part (here the pad) and not on the cube. 

Originally, we were testing the stress references on quite complex assemblies and we saw this bug, following that we created this simplified model to test  thinking that it could come from our assembly but ditto.... :/

Is this bugger referenced somewhere?

But on your piece we can see that you only have two constraint references. If you fill in the 3rd, is it the same?

Edit: it looks a lot like the following bug:

Assembly with multiple mate references does not snap / automatically mate to expected location when inserted into another assembly
 

Number: 681040

https://customerportal.solidworks.com/eservice_enu/start.swe?SWECmd=InvokeMethod&SWEMethod=GotoRecord&SWEService=SWGotoRecord&ViewName=SW+All+Defects+List+customerportal+-+Search&BusObject=Product+Defect&BusComp=Product+Defect&Id=1-1GL6BJZ&SRN=

 

Hello

I'm new here, so hi everyone
I am a colleague of Aurélien

A small clarification in our approach
We looked for a lot of examples on the internet but we only found cases with cylindrical coins while we have prismatic coins

We tested by putting one, two or three references and the result was identical
Once the part was placed in position on the cube, in the list of constraints, even if we had put 3 references, we only found the first constraint (Coincidence of two planes)

We have the impression that parallelisms are not taken into account
It's really strange

@.PL
I tried to open your link but I think we don't have an account to be able to access the position, I'll see with aurélien

1 Like

Thank you PL for the link, we managed to open it.

The SPR was supposed to be solved in the 2016 SP1 version but we are in SP3....

1 Like

Personally, I've been using smartmates since 2006 and it works more or less when it feels like it (for my screws, the coaxiality works pretty well but the constraint of the plane/plane under the head is when SW feels like it). In the end you save more time putting your last constraint back in hand than trying to make smartmates work properly.

The same goes for creating library functions (you can see surprises when inserting if you don't insert the functions on the same planes or worse if you insert your library function on a real 3d part. At the time I had created this room to test my functions as a library)

2 Likes

Isn't he trying to make the front and right planes on your cube coincide?

So on the 3rd side, it will be perpendicular and goes loose.

@froussel: Yes, I also have the impression that there is a malfunction with the constraint references. But it's still a shame that it works well for some sides and not for others randomly....

 

@opiep27: Honestly it's total blurring, I set the constraint reference only of the face coincidences and not of the plane, so normally it should not take into account the planes...

@ aurelien bonjours

And if for example you create...................... (test test)

  1. a tertiary reference in + for  the 2 components

what's going on

  1. Face / Face = Coincidence
  2. Width large rectangle / width small rectangle = parallel / distance
  3. Length large rectangle / length small rectangle  = parallel / distance

@+ 

Hello GT22!

Well, from what I understand, there's no point in defining secondary and tertiary constraints if  they are defined on a single part.

We would therefore have to create a secondary reference on the other part (here our cube). Now, our cube here is a simplification for the sake of example, we don't always have that shape there.

Our goal is, for example, to position a manhole (as a smart part to create the right opening on the receiving plate) on a container and for this manhole to be easily positioned in order to speed up the process.

[edit]: Actually, your proposal can work if we always have a standard container that already has the constraint references applied, but we often start from scratch to create this container

 

yes you have to create the same constraints on each diff part at least for the creation