"Offset free" goes to wrong side, no way to change that?

[JST]

When using "offset free" to get a part parallel to a plane, such as the XY or YZ plane, if the part goes to the wrong side, persistently, no matter what position you put it close to, there seems to be no way to reverse it.

The support needs to face the other way, but using "offset free" with the base plane does not allow any choice as to where it goes.

Should that be true?

I tried order of selection and other things. Usually it will go the shortest move, but not this time, regardless of MM, and positioning, etc, it always went as shown.

 

[JST]

I devised a work around for this problem.

The problem still exists, and I have no idea why. It makes the "offset free" orientation constraint pretty useless half the time in certain situations. Still needs to be fixed somehow.

I just was forced to assemble the parts in a different way, to avoid needing that capability altogether. That was the second time in one day that the entire assembly had to be done differently because of a mysterious refusal of Alibre to do what it was supposed to.

 

[DavidJ]

Planes and faces have 'direction'. A 'free offset' is the equivalent of the old 'orient' constraint, which also respected the direction.

Personally I agree it would be nice to have the 'flip' button available with free offset (as it is with a specified offset). You can use an angle constraint instead (when the flip button is available).

I believe that Alibre is doing what it is 'supposed' to do, but that may not be what you want it to do.

 

[JST]

..................

I believe that Alibre is doing what it is 'supposed' to do, but that may not be what you want it to do.

I absolutely KNOW that is the case! No question, this is an "intended behavior"! Problem is it is very inconvenient, and it really makes the function far less useful.

Yes, the old version did have a fixed response, BUT, for some reason, it did identify and respect the face direction. You could use the other side and it would flip. (Not sure about what it did for planes)

Now I am aware that , for instance, a point has NO direction, and only one "side" (the "outside" all around it). So the flip is not applicable to a point. A line, likewise, has only length, but no "sides".

A PLANE, however, has TWO sides, even though it is infinitesimal in thickness. It is possible to select that plane from one side, or from the other. So it is possible to imagine the "offset free" responding to the "selecting side". That would be perfectly acceptable, if it were possible.

I do not know if Alibre receives the information as to which side of the plane it was selected from. The selection is always of a line that is part of an "indicating rectangle", and it may be that no "selection direction" information is provided, and the idea won't work in consequence.

But, it is food for thought on this.

 

[NateLiquidGravity]

I've run into this exact situation. I ended up switching to an angle constraint. Seems "free offset" constraint should just be coded exactly like an angle constraint with the flip button just flipping between 0° and 180°.