What is an OS Design Workflow Supposed to Look Like

joshua_pohlman

Hello,

I have been using onshape for awhile now and I have finally decided to start making the move from SW which I have been using for about 5 years. And I have started a project to kind of force me to figure out this workflow, and I will say I have been having an odd time adjusting. I think my  biggest things is I don't feel like I understand the design philosophy of OS. I am used to making a part and pulling that into and assembly, and if I feel so compelled I may create and in context part in that assembly. But I feel like I am misusing these part studios when I make one part and try to add them together in an assembly. And I see how I would design in context but what about parts I have already designed how am I to pull those in and mate them up in a part studio. I guess what I am saying is does anyone have some good resources for someone who has a good grasp of the buttons and controls but just need to get a better understanding of onshape itself.

Thanks
Joshua

Jake_Delano

Hi Joshua, we've actually created a webinar for moving from SW to Onshape just for this purpose! You can check out the link here https://www.onshape.com/en/resource-center/videos/moving-from-solidworks-to-onshape-best-practices 

Evan_Reese

I think this depends a lot on what kinds of things you work on. Is it big machines with tons of subassemblies, or a few plastic parts that fit together, or something completely different? If you can give an example, I could at least describe how I'd approach it.

joshua_pohlman

Mostly I work on small robots consisting of no more than 200 parts, some cots some custom.

billy2

I think it depends on a lot of details also. It's good that you're asking the question.

I typically think in terms of an assembly and the datums that'll control the design intent for whatever I'm designing. OS is a parametric modeler and should be able to handle change. But remember you have to set it up correctly.

I'm not sure I'd care about the robot's datums. Robots I use are just an assembly with 5 rotational joints and geometry I've downloaded from the manufacturer. I'll add it to my library so I can quickly use it in a design.

I've found, designing automated equipment, my datums are based on what I'm building vs. the robot.

Also, what are your deliverables? What's it suppose to produce? ie.. documentation, BOM, prototypes or pretty pictures. What's the legacy of the design and who are you trying to control ie.. machine shops, marketing, injection molding, a government or the fda? You do know that OS will compute the robots position for you so you can download the joint angles into the controller saving time vs. teaching the robot  with a pendant. 

For a robot, you'll implement reverse kinematics to control positions and how are you going to control the positions? At this point you could design fixturing based on position 1, position 2 or position 3. It'd be nice to design an end-effector based on the differing things it needs to do at the various positions.

If you hang something on the outside of the robot, cables, tubes or hoses, you need to verify that their lengths are long enough so that over time you don't rip your cables apart. You can model this in OS and validate your design vs. tearing things apart over time.

And the #1 thing to think about, make it so I can understand what you're doing. Eventually, someone's going to end up working on your top assembly.

shawn_crocker

@joshua_pohlman
I wouldn't worry too much about it.  As you start getting the system more and regular operations start to become more of a reflex, you will end up having your view of the system as a whole, expanded.  I would say the main thing to always remember is anything can just be moved anytime if needed.  If a tab needs to move you can simply move it to a new document or to an existing document and nothing breaks.  Another key practice I think is worth attempting to adhere to right if is, if things need to reference other things in order to be sure they fit and continue to fit as changes happen, strive as much as possible to design them in a single part studio.  If you end up wanting the parts in individual part studios, you can use the derive feature to derive any part into its own studio and then use that part in your assemblies.  I use this workflow everyday and it is very powerful.  Similar to solidworks save bodies command but even better once you start implementing other things in onshape like configurations.  I usually reserve editing in context as a last resort.  Like if I have a product assembly and I want to design an additional product that could be purchased and installed as a kit or something.  I would then usually create an assembly somewhere, drop in the product assembly the new product needs to fit and then create a part studio in context.  In this scenario, I would again attempt to design everything in the one part studio and then derive everything out after into separate part studios.  Something to keep in mind when considering to edit in context is you can copy the context parts right into the part studio.  If you use the move copy tool in the part studio and select the copy part in place option, you can actually select parts from the ghosted context geometry and you will end up with a new part in your part studio.  I have used this sometimes when I needed to created another part exactly the same as one already in an assembly but with some minor differences and I did not want to create a new branch or configuration in the parent part.
As a summary, onshape is so awesome and allows for great flexibility in whatever workflow you may choose to use.  Everything can just be moved around with very limited consequences.  Being able to move things without pain was a big one for me when starting out.  I always found it constricting when using solidworks that there was fear associated with attempting to move something to a different folder if down the road I discovered it made no sense that the part file was were it was.  I like my stuff to be organized following a structure and I often just could not keep myself from moving something only to find that I was unable to account for all the things that were depending on its location.  Many tears have been shed over missing parts.