Mating best practice?

traveler_hauptman

I feel like Onshape inverts the way I normally think of mates. A mate connector starts with 6 locked degrees of freedom (DoF) and I choose how many to unlock. In SW a mate gives me no locked DoF except for those defined by the mate. 

Normally I try to match mates to how an assembly physically fits together with as few locked degrees of freedom as possibly so that the solver will make obvious things that cause assembly or stackup problems.

I'm slowly figuring out what mates to use for different situations but I'm curious what other people are doing.

Has anyone gotten to the point where they are using mates to really capture design intent and catch design problems?

brucebartlett

My most common mate is the revolute mate. I find this great for mechanical assy with pin joints. I find I you have to put the connectors in at the part studio, I struggle with them in the assembly. The move and between function are good on the connectors, once you know about them. I also like the rotate to quadrants in the mate, this also resets the part if sitting on an angle. I also have added connectors to sketches. Very quick to get it up and going just a differnet way of thinking.  

I have not worked out how to do welded frames (rigid parts lock in position) yet, I have been doing them fully in the part studio then dropping in the assembly, grouping and fixing 1 part. Not ideal. I am using the part studio mirror and array but end up with multiple instances of the same part, this will not work with the BOM down the track. I am sure there is a work around but have not got it yet, kind of waiting to see how the drawings and document management(fastener library/sharing files between doc's) will work.

traveler_hauptman

babart77 said:

... I am using the part studio mirror and array but end up with multiple instances of the same part, this will not work with the BOM down the track....

This is a really interesting point... how to differentiate between instances that are repeats of the same manufactured part and thus are consolidated in the BOM, and instances that have the same geometry but are different items in the BOM (different materials, different surface treatments).

I'm guessing that there will be patterns in assemblies to handle the first case and that the pattern feature is what one would use for the second case. 

traveler_hauptman

I think I have a mate combination I'm happy with for screws in plates with simple holes. I can detect errors in screw length, plate thickness, and hole location with the mate solver and errors in hole and screw diameter with the interference checker (section views now, the interference inspection tool later when it gets implemented).

I have a mate connector in the receiving plate at the depth of the end of the screw, a fasten mate between that and the end of the screw, a planar mate between the held-plate and the screw head, and a revolute mate between the holes in the receiving and held plates. 

This doesn't handle valid but non-concentric holes. 

billy2

I would have placed the mate connector under the head of the screw. At the end of the screw? That's interesting.

Like babart77, most my assemblies are static and don't move. I'm using the fixed mate connector a lot. I guess my reasoning for using an assy is to manage the feature list. Breaking up small assemblies into various parts helps keep my feature list shorter for each part studio.

I'm finding it painful having to scroll up & down the feature list. Breaking the design into many part studios seems to help manage this but I do lose association between features which sucks.

I'm still not sure I've perfected the part studio/assembly logic. I do know that the feature list are too lengthy and that I like putting mate connectors in part studios only to bind them together in an assembly. I'm not crazy about this approach and waiting to see someone come up with a great approach.

traveler_hauptman

@bill There are two connectors, one under the head and one at the tip of the screw. If you change the screw length or the plate thickness, the mate solver won't be able to find a solution and complain to you. Likewise each hole in a plate get's a mate to the companion plate. The mate solver will complain if the holes move relative to each other. It helps a lot when making changes. When you come back to a model 8 months later and make a change, you don't have to spend 3 hours pouring over the model to make sure you're not breaking something.

(The screws example is a bit contrived. In practice, if the screw is not the right length, you just pull another one out of the cabinet. However not all peg-in-hole fasteners are so easy to change.)

The part-studio/assembly logic happens to match my solidworks workflow (where I use a layout assembly to create interdependent parts, and a as-built assembly to capture construction). I have faith that the feature tree will get cleaned up to be more efficient to navigate before too long. 

brucebartlett

I have added a doc, Fastners Doc, to the public folder. In this I have been playing with bolts and nuts and the best way to set them up. In the end I have just used a fasten mate very quick and easy to fully define. Have a play and let me know your thoughts. Bruce

traveler_hauptman

babart77 said:

I have added a doc, Fastners Doc, to the public folder. In this I have been playing with bolts and nuts and the best way to set them up. In the end I have just used a fasten mate very quick and easy to fully define. Have a play and let me know your thoughts. Bruce

Really nice detail on those fasteners. The fasten mate is great for placing the screw in a position relative to the other parts... I use it often.  But the point of the overconstraining in to get the mate solver to warn you when you make mistakes during changes. It's a cost<->benefit thing though. It takes extra time to place the extra mates. On the other hand, it takes extra time to get parts reworked....

brucebartlett

Thanks traveler, I got sick of waiting for a standard toolbox of bolts and nuts which I would have to modify anyhow so decided start my own. I have been trying to work out the best set of mate connectors to place on the bolts, I plan to get 1 series right them duplicate tab and edit to crete the a full set.

So you would place a mate connector on the end of the bolt and offset mate connector at the bottom of the hole. Bolt should not change but if the plate thickness changes you see and error?

I have also though about minimum thread lengths and placing MC's on the start of the thread, but have decided the group the thread features at the start of the part studio so when suppressed it will blow through to all parts, checking thread length can be done with the section tool, looks really cool when the interfering threads stand out in red. 

bill said:

most my assemblies are static and don't move. I'm using the fixed mate connector a lot. I guess my reasoning for using an assy is to manage the feature list. Breaking up small assemblies into various parts helps keep my feature list shorter for each part studio.

I'm finding it painful having to scroll up & down the feature list. Breaking the design into many part studios seems to help manage this but I do lose association between features which sucks.

I'm still not sure I've perfected the part studio/assembly logic. I do know that the feature list are too lengthy and that I like putting mate connectors in part studios only to bind them together in an assembly. I'm not crazy about this approach and waiting to see someone come up with a great approach.

I finding i use roll forward and rollback a lot in the part studio, this seem to be the only way to order the multi part list. I agree tree management needs improvement and sure that is high on the list. I have started naming features which i would rare do in SW, I also found I have to leave the feature name in place so I know what type of feature it is.

Look forward to planes, patterns, arrays  and mirrors in the assembly mode this will limit work in building assemblies