unwanted assembly changes

[neilr]

I have a cnc router I am designing and have run into an odd issue. When you see the pic below you will notice the roller bearings are not in contact with the tracks on one side. The thing is they were until I changed a part.....The part I changed was the sketch for the side plate for the gantry (seen on edge) that holds the bearing mounts and the upper cross track. The change I made was only a small adjustment to the height, but when I saved the change the position of the entire side was altered away from the track and the other parts that are set to extrude to geometry are all to long now, and I am going to have to rebuild the entire gantry it seems. So is this a bug? I am not sure if I can reproduce it....also I am using the latest 62 build.

 

[ASP]

so why don't you constrain this part to the position it should be ?

 

[neilr]

It was I had the bearings constrained to the tracks on both sides and all was well until I made a dimension change that had nothing to do with the location of the part ( I made the side plates 2 inches taller). The constraints disappeared and the side moved out several inches...Now when I try and set constraints to the track so as to reposition the side nothing happens it just does not add the new constraint.

 

[HaroldL]

You mentioned that some of the parts were created with a "to geometry" condition. It sounds like these parts were created in the assembly context. Could there be some inter-design constraint or relationship that is affecting the top level assembly.

What is the measured displacement of the side plate? Is it the same as the change in side height?

 

[RocketNut]

neilr:
I have learned that if you change the face or edge that a is constrained to a part, AD loses that constrain because it's position is different. So when you changed that plate dimension by 2 inches the constrain inside AD became different resulting in a lose constraint. I suggest you go back and edit the constrain to its new position.

 

[neilr]

Rocket nut yeah I was able to go back and fix it, as far as loosing the constraints on the side plates I understand that one but the constraints that I lost that make no sense are the one from the bearings to the tracks (inter design) those parts were not altered.....

Haroldl I am not sure but I do not think so...But I think I have a copy of this at work so I will take a look and see if I can find anything...I will also try and duplicate it...

 

[RocketNut]

Believe me when I say I have chase constrains errors all over the place. It seems that AD constrains are inter locked some how. Meaning I constrain a part to another part with out errors. Then several parts latter the constrain holding the first parts together start throws errors. It's extermly frustrating and confusing. Specially when the error messages do not give you a clue as to whats wrong. :x :x :x

 

[ASP]

i think it isn't AD and the constraining system what fails (most of the time), it is the user -
nobodys brain can handle hundreds of constraints and so the user often tries to constrain things which are already fully constraint in the ASM (or just impossible).
The most common error is overdefined - there are always 2 constraints with this error. If the last set constraint is overdefined it is not set up correct or not necessary.
Witch changing parts constraints often fails because the constraint face/edge does not exists any longer or the part is changed that way that other constraints makes
it impossible to be solved correct.

Most of my ASMs have hundreds of constraints - without any error.
I don't think the constraining system is inconsistent, in 99% it is me doing the mistake.

In some other cases the parts just don't fit because of "misconstruction".