Mating Problem with Dovetails

I assume that the dovetail side are at different angles. Solidworks doesn't know the term "dovetail", only faces, points, planes, edges, etc. See if you can mate a face of the female dovetail to a point or edge on the male dovetails face that you attempted to mate. If it does mate, then check your sketches ensuring the angles are the same.

Keith

Erica,

As Keith pointed out, check that the angles are the same. To add to this, they must be EXACTLY the same out to 8 decimal places. If they aren't Solidworks (or any other similar program) can't resolve it.

Mark

Erika,

I use a dovetail slide in an optical assembly and found that SWX chokes on the two plane-to-plane mates.

Try this:

Make a plane-to-plane mate between the "60 degree" beveled faces. Then make a coincident mate between the "flat" plane of the housing (female dovetail) and the **EDGE** of the slide (male dovetail) where the "60-degree" and "flat" faces of the slide intersect (assuming a sharp edge).

Now the slide is constrained (I wish SW Corp would call them constraints instead of mates!) in 5 DOF, with only the "sliding" direction free. You can drag the slide with your cursor only along its direction of travel, or you can add a distance mate of a point on the slide to an end face on the housing to finish the job.

HTH, Art

Why bother trying to mate the dovetails at all?

1. In the real world and in your model the male and female parts will never exactly be line to line.
2. If you try to mate the angled portions there is inherent ambiguity in where the dovetails are with respect to each other based on the roundoff error in the angles.

So instead of mating the angled portions, create a centerline plane on the male and female portions and mate that to center the two to each other. Then you will be using three orthogonal mates to locate the feature.

My model is an IGES download from a vendor. After checking, I found the two angles were in error at the 3rd & 4th decimal points (which I rarely have displayed), enough to cause the problem.

Keith and Mark are right about having the angles exactly the same. In that case the two plane-to-plane mates work.

From a kinematic standpoint they shouldn't work, because each one consumes 3 degrees of freedom, leaving a "7th" DOF (which doesn't exist) to control sliding motion. I suppose that SWX ignores the conflict when the angles match exactly, and only constrains 5 DOF.

Art