So, you’re sitting in a design review with a concept model of the product and the topic quickly changes to schedule. “You can just use that model for the engineering, right,” your boss asks confidently in front of the customer. Now, before your sweat glands burst and your eyes glaze over, rewind a few days.
There’s a discussion going on in the SolidWorks forum where good ol’ Mark Biasotti is asking how you use concepts for a final design and if SolidWorks is a good tool for doing so or if it needs improvements.
It’s a great question and reminds me of a method for creating concept designs we discussed a while back which allows you to try out different variations of a concept. But using that for engineering? Where do you stand?
An argument for History-free models?
If you’ve done many concepts models in SolidWorks, you know, the geometry you slap on the screen isn’t always thought out completely. There are stacks upon stacks of features and then there’s the whole parent-child, child-parent thing. So… what if you didn’t have features? Seems an easy solution to move from concept to final model, doesn’t it? However, in the same way that history-free models clear up your side-bar, features can help clear up where the change needs to happen on the concept model to make it usable for engineering. Neither is perfect for swallowing a heavy dose of geometry change though, so the argument quickly switches to that of both being able to accept geometry changes easier.
Here’s what I’d like to see
In my world, the engineering group, we use all of our engineering concepts for the final design. Where we differentiate the process is between Design and Engineering. To explain, Design will do a quick concept model, a cartoon, to bring into a rendering package or to create 2D layouts. This process can go through several iterations and last months. When the Design is finalized, Engineering usually starts from scratch in SolidWorks going off the design elevation drawings approved by the customer. In this case, the Design models may no longer be valid, and honestly, who wants to work on a model someone else threw together for some rough ideas.
Now, what would change this is if Engineering was involved with the design. It’s possible, but unlikely because of the vast difference in schedules between the two. So, some of it is going to come down to the company and what modeling methodology they use. “Do this here in Design, so Engineering can do this.” But that can be translated over programatically. Design can set up ‘markers’ on features or bodies that Engineering will use. It can go back and forth also. Engineering can create models with ‘markers’ for Design. The model can be sent back and forth through the conceptual design process. It’s going to require more clarity and the awareness of each group to actually coordinate, but this would allow Design or Engineering to go through iterations that get closer to the final product.
On top of this, the only other thing I’d like to see is smarter parent-child geometry updates. If I move a feature above it’s parent, why shouldn’t SolidWorks evaluate the options and allow geometry to adjust accordingly without going into the sketches or blow away relations?