Do You Use a Crystal Ball for Design Intent?
Design intent can be a tricky topic. I’ve heard some people argue that there is no such thing. I myself have tried to rename it “Design For Change”, because I think that makes more sense. In any case, it’s pretty clear that with history-based modeling systems, the best you can say for design intent is that you need to see into the future to establish it, and into the past to use it.
It turns out that Design Intent could be the most powerful aspect of Synchronous Technology. With your typical history-based model, you have to understand how the model was built to understand how it’s gonna react to changes. With Synchronous Technology, you determine the design intent at the time you make the change.
Here’s a simple example. I want to widen this clevis part. I can do that in several ways (listed left to right):
- move the lower vertical face sideways
- move the vertical and angled faces sideways
- move the vertical face and tilt the angled face
- move the vertical face by adding a step
These 4 different types of changes would require changes to the design intent scheme in a history-based system, but in SE w/ST, you can change between the options on the fly by either selecting an additional face or by changing an option in the active Command Bar.
When you started this project, how did you know which of these methods you were going to use? You probably had no idea. So when it came time to make the change, you had to revisit the sketch, and either re-dimension it, or you may have had to delete and recreate features to get it to change the way you need.
Here’s another way to ask the question. Have you ever limited your options for design change to fit your model’s design intent? This is the cheap way out, but don’t deny it. I know I’ve done it. Rather than go back and rework something, I’ve accepted a result that wasn’t ideal.
You don’t really need a crystal ball to look forward or backward in time. All you need is Synchronous Technology to enable you to determine design intent at the time when you make the change.
Hi Matt,
First off, I’m not a SE user. I’m a SW “power user” by SW standards and I would like to think i’m fairly seasoned, however I learn something new every other day. That being said, I find the feature tree very useful in SW. Admittedly i sometime stare blankly trying figure out my design intent until the little, and sometimes very dim, light goes off in my head on how I’m going to approach a design. But when i do have it figured, that design tree and top level sketches used to drive an entire assembly really rock! Sure it takes a lot, and I mean a lot, of computing power, but the benefits outweigh the drawbacks…
I know you know all this, so then my question: How can not having a feature history tree be a good thing? So what if you can do something on the fly, but what if you need to undo a few critical features to restore and “save” a model? I recognize that I am in a SE forum and I’m not trying to wave the SW flag. But I just don’t see the advantages of SE that you’re promoting. Again, I don’t know or use SE, so please educate me.
“How can not having a feature history tree be a good thing?”
Solid Edge, even in Synchronous mode still has a list of features, but they are really just collected sets of faces made by those features, not a set of variables. And with the list, the order doesn’t matter. Sometimes you might wish that a face or a set of faces had been made with a different feature, so instead of editing a single feature and watching it rebuild, you have to edit multiple features and watch them rebuild. With Synchronous, you can redefine that sort of stuff as you go – you’re not tied to how you did it the first time.
Have you ever done something like put a fillet on a part, then added another feature and dimensioned to an edge the fillet created? That kind of thing is a serious no-no in history-based models, but it doesn’t matter in SE. How many times have you edited a feature way back in the tree, and the whole thing turns red and your features all fail? Why do you mess around with sketches? Why don’t you just manipulate your part’s geometry directly?
When you get to the topic of in-context design, feature trees become downright confusing. History-based stuff only applies in the part, not in the assembly. And then the relationships and file management gets pretty ugly.
When you first learned history-based modeling, the concept was not intuitive. It took a while to understand until the light finally went off. Synchronous is easier for people who don’t have to un-learn history.
As far as undo goes, Solid Edge has a very nice undo. Undo in Works used to get all confused when you went back through changes in sketches.
Stick around and keep asking questions. I made the leap, so I think others can do it too.
“But when i do have it figured, that design tree and top level sketches used to drive an entire assembly really rock! Sure it takes a lot, and I mean a lot, of computing power, but the benefits outweigh the drawbacks…”
Jeremy,
Your History Based approach is still applicable in Solid Edge irrespective of Synch since you can do everything you need in the Ordered Environment. I like many other SE “power users” used the same Master Assembly Sketch approach that you do, so wrapping my mind around how Sync changes that does make it seem that you’re losing control. And in some ways that’s correct. But what you loos in control from the parent Sketches that drives numerous parts, is direct editing of those parts all at once, as Matt is showing. It took me awhile to let go, but after experiencing Synch’s efficient workflow, I almost always avoid Ordered/History based modeling. Whatever confidence I once had with my old History based Sketch approach to designing an Assembly has been more then satisfied with Synch’s more intuitive approach. But it did take time to totally let go.
But one of the techniques that has been promoted for those wanting to “walk before you run” with Synch, is to build all your Sketches in Synch then build your features in Ordered/History. This turns out to give the best of both worlds on many edits. And once an Assembly gets close to finalized, you can transition all those Ordered/History based features into the Synch environment with one command. You now have for the most part a very robust History derived Assembly with the ease of editing of Synch.
Anyway my point is you don’t have to give up your History approach with Solid Edge. In fact you can stay totally History based and only wander into to Hybrid History/Synch or full Synch editing as your comfort level increases.
Bob
I guess for me design intent is making a part to fulfill a need and all this esoteric planning stuff I have never felt I have had to do. My mindset with ST has been that if requirements change on down the road I will just do them and not worry about it. How can I anticipate much of this anyway? I don’t have to plan for these things because change is just to easy and the closest I would come to planned “intent” would be to lock features/dimensions down in such a fashion as to create predictable behavior. Locking holes to different ends of a part where a family of parts will be created by changing the length with the holes relationships preserved relative to critical ends would be an example. So planned intent for me is does the part fit, can I manufacture it and how many can I talk you into ordering? The geometry creation is to easy to worry about more than this.
Matt,
As you know my initial impression of Solid Edge using Synchronous Technology was very positive, but then after ST1 shipped I became one of it’s biggest and most vocal critics. I was an SE user since V7 and ST1 was the first version of SE since V20 that had the Sync option. I say Option because you needed to decide from the beginning of the part creation as to how you’d proceed. Well like many users I gave it a try, but the Part Edit you illustrated above was possible with ST1, it was a nightmare trying to get it to do these edits with confidence. So like many old time users I waited on the sidelines for the technology to mature. But still by ST4 it still had this contempt for the user Design Intent. Again not that it couldn’t do it, but making the user feel they were in control of these type of edits. And granted, most users also needed to STOP fighting the different workflow to let Sync just do it’s thing. But Dan Staples and his Development team at Siemens listened to the gripes of the users, and by ST5 Synchronous Technology had matured into a very productive tool. Now with ST6 shipping, Dan and company have made Synch so much more polished, and have now stepped up this productivity with even more efficient editing of parts. Take your Part above and now place that in a complicated Assembly…. all those edits can be done from the Assembly Environment while using mating and other parts as reference and help drive the Designer’s intent.
I’m sorry but anyone who has used ST6 in this fashion will NEVER want to model in a History based CAD tool ever again. Forget the hours of saved time… it’s too much fun to worry about the time.
Solid Edge is so far ahead of the curve on this, that the folks still using SW and other CAD packages will feel like they’re stuck on AutoCAD 12 in another few years. Every release seems to get more and more powerful while the other CAD packages are chasing themselves in the cloud! Solid Edge is now in a class by itself.
Bob