Synchronous Throwdown

I’m writing this because I think the CAD community at large, including users, media, and vendors, have sold Synchronous Technology short. I’m just taking this on as a personal project, it’s not inspired, encouraged or sponsored by anyone, although I am using a free temporary license of Solid Edge provided for another purpose.

If you are a Synchronous Skeptic or even an out-right unbeliever, I want to hear from you. Leave your comments below. (I’ve changed comments so you don’t have to have an account, just post with your email address – email not used for marketing). Tell me why you don’t think Synchronous is worth your time, and show me examples of stuff you don’t think it can handle. Where possible, send me a model. I’ll be honest with you about what I think is good and bad about this stuff. I love examples, and will do my best to document nice examples of various features and techniques.

Here’s my pitch for synchronous:

1.You already know it

You already know how to create things in synchronous – it’s the same method as history-based modeling, plus you can use primitive bodies. This puts you more than half way there. Creating new geometry in synchronous is just like history – sketch and extrude. Select edges and apply fillets. Place holes… This is very familiar to all of you.

There is a big lie going around that synchronous is not parametric. It is. You can drive geometry with dimensions and relations. Synchronous is not history-based. The order of the features doesn’t make any difference.

What matters about your CAD model? The final shape, right? That’s exactly all that synchronous is worried about. Get your model to the correct finished shape, and make a drawing of it.

It doesn’t matter how you made it. It doesn’t matter how you edit it. It ONLY matters that it’s the right shape. Honestly. I’ve spent a lot of time writing best practice lists that only deal with how you make a part. And then I realize that how you make it doesn’t matter AT ALL.

2. Real tools for imported data

The people who call imported data “dumb” aren’t using synchronous. Imported data is just as editable as native. Move faces or put dimensions on them and make changes. You can even convert existing patterns on an imported part into an intelligent feature.

Or did you ever get locked out of a “future version” part? Somebody upgraded the software and saved the changes, but you are stuck on the old version? Never get left out again. Import the Parasolid, and party on, my friend. Parasolid, IGES, STEP data all work like native parts.

3. Parent/child dependencies are the big difference

When you think of what makes things hard to edit in your current software, one of the big culprits is parent/child dependencies. This means things have to be built and then edited in a certain order that isn’t obvious unless you’re the one that built the parts and you remember how you did it. The ironic thing is that the simpler your history is, the less you need it. The more complex it is, the more it is likely to hinder your model editing. Either way, you’re better off with synchronous.

Even in an assembly, you can change parts using other parts as references, but without the troublesome in-context relations. You can change multiple parts at once in the assembly. This is often kind of mind blowing stuff to history-based users, but leaving parent/child dependencies behind is a big thing.

4. Understand the BREP

Do you remember when you first learned how to do history-based modeling? You remember being confused for a couple of days? You eventually got it. Learning Synchronous is going to be a little like that. There is one big concept you need to understand, and then the light will come on for you. BREP means “boundary representation”. You need to understand that certain types of faces can be extended or trimmed. If you move one model face, other faces have to extend, or get trimmed back, or untrimmed. Once you understand this, its easy to visualize how edits work, and when an edit fails, why its failing and what to do to make it work.

Arguments against

The big arguments against synchronous that I hear are “I’m too old to change now”, or “It’s not parametric”, or it doesn’t use sketches”, or “There’s no way to control anything”.

To those, I say: I made the change. I’m someone who has a lot invested in the history-based way of doing things. I spent time figuring out best practices, solving problems, writing books, doing consulting work. I didn’t believe it at first, but after a while it started to make sense.

Synchronous IS PARAMETRIC. Parametric means it’s driven by parameters – dimensions. You can also use geometric relations like the sketch relations you’re used to – parallel, coincident, perpendicular, concentric, etc… You can write equations, you can make tables. It’s parametric. It’s not driven by sketches. Not driven by dialog boxes. It’s driven by dimensions and relations.

The “no way to control things” argument I believe comes from two things. First is the desire for sketches. Why you need something 2D to drive a 3D shape isn’t exactly clear to me. The second is the need for a lot of users to see things as “fully defined”, starting with the sketches and applying that to 3D. Like with sketches, I think people assume that your model is going to fall apart unless you have a dimension or a relation on every face. Imported models don’t have any dimensions or relations, and they don’t fall apart, why would a synchronous model? It only changes when you tell it to.

This is the one place where I think people demoing the software have gone wrong. I think they started the story of how to make changes in the middle, without talking about how to do the simple stuff. The big advantages of synchronous definitely come in the area of making changes – editing. Demos of editing technique really should start simple, and ironically, the first thing everyone wants to know is how to prevent things from moving. I’ll put together a demo for this post where I start very simply. With just a cube. Show how to lock down 3 faces to X, Y and Z to keep it from moving (although even without that, it’s not going to move). Then I’ll move on to show how to edit any face on the part. I think people get carried away with the visual arrow pulling in synchronous, and then people trying to learn it just figure it’s a shallow demo system that only allows visual edits. But really it’s just very flexible. It allows dimensions AND visual editing. You can have some areas take both or either one. Really, I think as you get to know this, you’ll understand why I like it so much.

Best of both worlds

If you really hate synchronous, Solid Edge still lets you model with either system. Even in a single part you can use synchronous or ordered or both. In fact there are some techniques that I recommend you use synch, and some where I recommend ordered. There are definite strengths to both, but let me just say this:

History-based modeling has outlived its original purpose. The purpose was to break down computation of mechanical models into bite-sized pieces, so the hardware of the ’80s could compute it step-by-step. By ~2008 or so, commercially available hardware was getting to the point that it could handle live manipulation of analytic NURBS Brep models. So why not do that? In very complex models, the rebuilding is a huge liability. As parts get periodically rebuilt/recalculated for various reasons, the performance of large models gets very slow. Only editing what needed to be edited really simplifies things a lot.

There’s a lot more to synchronous than just simple direct edit, but we’ll get into that as we go along.

More to come

I’ll put together more posts that start to show the real basics of synchronous, without starting in the middle. And I’ll try to answer questions as I go along. I’ll also show questions I had when I was starting to learn synch. I just wanted to get this introductory post up first.