12 Replies Latest reply on Sep 26, 2018 12:17 PM by Tom Gagnon

# Advise on Mates (When Tangent doesn't work)

Pictured below are my part and an O-Ring I'm trying to mate.

The cyan surface is a spherical radius, the green surface is a groove cut into the spherical radius, the black part is just a standard o-ring.

I was hoping to select the green surface and the outer surface of the o-ring, apply a tangent mate and move on.

The diameter of the o-ring is sized so that it would be centered in the groove if the tangent mate worked.

However this doesn't work so I'm looking for advise on how how to accomplish this mate.

I can create a plane on the outside of the o-ring that is angled to correspond with the groove and use that with the tangent mate.

I have a multitude of configurations and the groove angle changes with each one, so before I went that route I thought I'd ask if there is a better way to do this.

Thanks in advance for any ideas.

• ###### Re: Advise on Mates (When Tangent doesn't work)

Excellent question. Why is the tangent not working. Time to revisit the ERs.

Workaround:

1. Pick a plane that contains the center of the O-ring and normal to its axis

2. Open a sketch on that plane

3. Run the intersection curve tool. It will result in two circles.

4. If needed repeat steps 1 to 3, to get other circles.

5. In the assembly, make each circle tangent to the spherical surface. 2 should suffice.

6. Add a directional mate (parallel or coincident on planes or axes.

7. Control the position (there are always 2 solutions for the tangency) with a limit mate.

• ###### Re: Advise on Mates (When Tangent doesn't work)

Richard, a tangent mate is not a good way to go in this case.  Unless the OD of the o-ring and of the groove are the same a tangent mate would not leave it centered.  How about using a concentric mate of the axes, using the Temporary Axes would be an easy way to do this.  Then you can use a symmetric mate with the two face of the groove and a plane through the disc of the o-ring to locate the o-ring axially.

• ###### Re: Advise on Mates (When Tangent doesn't work)

I agree. Reference Geometry to the rescue.

• ###### Re: Advise on Mates (When Tangent doesn't work)

I respectfully disagree w/ reference geometry unless it has already been required to form the part to begin with.  Otherwise your design might end up on the HORROR museum thread. https://forum.solidworks.com/thread/224153

• ###### Re: Advise on Mates (When Tangent doesn't work)

I don't know how the o-ring was designed, but logical placement upon primary planes and an origin leads to automatic reference entities included already, at very least a sketch plane or an axis or w/e was used to generate it. I am not saying additional entities need created, but to use what ought to already be present.

• ###### Re: Advise on Mates (When Tangent doesn't work)

I'd suggest you show entities like axis, etc. and use those in the mate.  i.e. offset distance from the edge surface where the o-ring would seat to the c/l of the o-ring cross section -or- use the end point of the rotation axis to be coincident to the planar surface.  Cylindrical/coincident of the o-ring axis to the cylindrical surface of the seat and set it to no-rotation.

You can also use the edge created by the cylindrical cut in the housing.

• ###### Re: Advise on Mates (When Tangent doesn't work)

Then you can turn those entities back off.  The mates will stay.

• ###### Re: Advise on Mates (When Tangent doesn't work)

I use the axis of rotation for the revolve to center it (using concentric mate), then use a plane to position it (using coincidence mate).

Example:

• ###### Re: Advise on Mates (When Tangent doesn't work)

Simplified versions of my files attached.

I've created a plane in the o-ring on the outer surface angled to match the groove.

In the assembly the tangent mate using the groove surface and the o-ring plane seems to work pretty well.

• ###### Re: Advise on Mates (When Tangent doesn't work)

I don't know how the O-ring was modeled, but if I'd done it then it would have been centered on each of the three primary planes, so it would have been very easy to mate the Part using them.

• ###### Re: Advise on Mates (When Tangent doesn't work)

I assumed this might be a good learning opportunity and it looks like I was right.

My O-Ring is centered on all 3 planes as has been suggested.

I've already learned that I don't know how symmetric mates work, so I'll have to look into that.

Adding a reference plane to my O-Ring seems simple enough since the O-Ring will be driven by a design table anyway, adding the needed angle for the plane will be simple enough.

After lunch I'll go thru all the replies a second time and see what else I might learn.

Thanks a ton folks.

• ###### Re: Advise on Mates (When Tangent doesn't work)

Richard Ahlgrim wrote:

I've already learned that I don't know how symmetric mates work, so I'll have to look into that.

Symmetric mates are simple enough once you get into it. I try to remember the order of selection while holding CTRL: that first you select the middle plane, then one far reference, then other far reference. Once the three selections are made, it ought to offer Symmetric mate in the popup window.

Otherwise, it is available from Mate > Advanced Mates > Symmetric, where inputs are more distinct and categorical with labeled boxes for Symmetry plane and Entities to Mate fields.