Why Don't You Use Synchronous?

by MLombard ‎12-16-2013 04:00 AM - edited ‎06-11-2015 02:53 PM (5,201 Views)

Usually I write articles trying to answer a question, or trying to advocate for a certain point of view. But this time is different. The main thrust of this entree is to ask you, meaning Solid Edge users, a question. I'm a curious fellow, generally with more questions than answers, so asking questions is actually easier than trying to answer them.

 

Sometimes I hear about people who have been using Solid Edge for a while, since way before Synchronous Technology. Some of these people have clung to history-based methods as their primary tool, and have not adopted Synchronous techniques. I admit that I myself was not crazy about the first two versions of SE wST, before they allowed you to combine techniques into a single part model. But even as a guy who had definitely bought into the history-based way of doing things, I was able to see the benefits of mixing methods - not replacing one method with another, but using the two methods in combination to really get the most power possible.

 

Here's my question. For those of you who have been Solid Edge customers through the change from straight history to mixed mode Synchronous Technology, why have you not changed? I don't know how many or what percentage folks I'm talking to here, but I'm just curious if you think the new stuff is just too hard to learn, if you think there is some threat to your old data, if you just haven't taken the time to learn the new methods, or what.

 

I didn't live through the switch. I was watching from outside as Solid Edge implemented the changes. It's my personal opinion that Solid Edge handled the transitions very well. Your pre-ST data is compatible with ST1 & ST2 as well as the post-mixed-mode data from ST3+. Straight through there were no compatibility issues that I'm aware of.

 

We all know Bob Mileti's inspiring conversion story, of going from a big foe of ST to a big supporter of ST. For those of you who have not yet moved forward, I'm interested in knowing "why?"

 

Take as much space in the comments to respond as you like. I'm just interested in why, I don't really have any great personal need to change your mind, and I certainly don't want to make this into a religious argument, I'm just curious about why.

Comments
by Esteemed Contributor
on ‎12-16-2013 11:41 AM

We build Window and Door products which mainly consist of linear peices of wood, aluminum, vinyl, and fiberglass.  As you can imagine, we have features that are used across many of those parts as we have several linear pieces on one product and those same features are also used across product lines.  To drive consistency, provide a "where used" path, and generally make editing easier for ongoing ECO maintenance, we create a lot of "parts" of those features and then use Insert Part Copy as a sort of associative "tool body" to cut those features from a new part.  Our structure replicates a wood moulding process where:

  • individual features are modeled (decorative shape on inside edge of a wood sash)
  • features combined to create a knife (top, bottom, left, right) of a moulded profile
  • knives inserted into a part profile
  • Profiles inserted into actual part

We may also have tenon/mortise features for the ends of the linear parts as well as hardware routes and such.

 

Now to why we do not use Synchronous... Since Synchronous does not support a associative part copy, we cannot use it in this process as there is abolutely no advantage of a synchronous rectangular block, and in the case of a complete profile, it must be the first feature. Now it is possible to do this somewhat through an assembly because you can create an Inter Part Copy in Synch and use Face Relations to map the faces automatically, but this construct has failed to migrate to the Insert Part Copy and of course lacks some of the benefits of using the Include with Face option.  If we could get an associative tool body solution in Synch, then we could reap the benefits of placing Synch features on the linear part after it was extruded...

by MLombard
on ‎12-16-2013 04:47 PM

Yeah, Ken, I can see that with live links. Part of the beauty of ST is that it doesn't use live links, but I guess that can also be seen as a problem.

 

I can't speak for how Dan Staples and the guys feel about this philosophically, but live links require rebuilds, and rebuilds are part of the history-based scheme. Not sure if it really makes sense to add live/associative links to the synchronous side of things.

 

How about with sketch blocks? Those feature parts could really be represented by sketch blocks, couldn't they? But those won't update in ST either. And it doesn't make sense to have the history-based section before the synchronous geometry.

 

Yeah, I can definitely see why you have had to stick with history-based functionality. Unless you used a big multi-body part where everything is native instead of inserting external parts.

by Phenom
on ‎12-17-2013 11:14 AM

We do primarily mass amounts of sheet metal details and flat patterns, and Synchronous does not easily allow us to make the type of edits that we do every day to allow for more weld shrinkage in a corner, tip a face for trim after bend, etc... There is no benefit to using synchronous for us, and you lose some of the sketch history of part changes that you'd have in Ordered. On the rare occasion that I have a part model, I blaze through it on the Synchronous side of things, but Synchronous tools for our type of sheet metal work are weak and would require revamping of steering wheel edit functions to make this process faster.

by MLombard
on ‎12-17-2013 02:15 PM

Chally,

I just assumed that the sheet metal stuff was pretty solid in Synch. I don't really understand the kinds of edits you're talking about. If you'd be willing to show a part, I'd love to learn exactly what's up here, and especially about the steering wheel improvements you'd like to suggest.

 

All this goes to demonstrate that there's a very good reason why Solid Edge has chosen to combine history and Sync rather than replace one with the other. I firmly believe that direct edit on its own does not constitute a full-powered CAD tool.

 

Thanks for the comment!

by Esteemed Contributor
on ‎12-17-2013 02:37 PM

Matt, what I don't understand is they essentially allow it already but requires an assembly file as a go between.

  1. Create a protruded part and add it to an assembly.
  2. Now create another part (Synchronous) in the assembly derived from the first such that it is an extension of the protrusion.
  3. Now create Inter-part Relationships between the two parts and specify the second synch part as the one to be driven.
  4. Now edit the first part and watch the magic.

Now you are driving a synchronous part associatively.  What I want is the ability to do the SAME THING sans the assembly file...

by Esteemed Contributor
‎12-17-2013 02:41 PM - edited ‎12-28-2013 05:30 PM

Much like Dylan, we do a large portion of our work through the Sheet Metal environment, and find the Synchronous instance has "some" limitations that make it not worth while to be living there fully, just yet. Sync is perfectly fine for most of our new Part models.

One prime example that comes to mind [due to recent personal blog post from Dave Ault], would be a lofted flange command, when making a rectangle to round transition. [and most of the artistic forms we get to design/manufacture, such as this > http://on.fb.me/1eip6Ql]

by MLombard
on ‎12-17-2013 03:14 PM

Ken,

 

Yeah, but the InterPart stuff comes with a huge overhead of the assembly, both parts, the relations between them, and the file management implications. I guess I see what you're saying. Maybe the associativity itself is the costly item.

 

Sean,

 

Yeah, I can see that. I haven't tried this yet, but I'm assuming you can mix Synch and Ordered methods even in sheet metal, right? Would there be any advantage to that? Getting the ease of change of synch while getting the power of sketch-driven interpolated shapes?

by Valued Contributor
‎12-18-2013 06:15 AM - edited ‎12-18-2013 06:24 AM

At our last user group meeting the same question was asked. I was the only one using ST there and it was a bit shocking. The primary reason was companies were comfortable with what they knew allready and felt they had no time to learn new things. They did not however know ST well enough to make this a valid reason based upon productivity standards. It was also surprising how many were on past versions of SE and apparently the philosophy of some of these companies is to upgrade at some point in time but not every year and certainly not to potentialy disruptive new stuff. I got the feeling that there were no solid reasons here like Ken's or Sean's or at least nothing was mentioned about lack of capabilities on the ST side.

 

  The second largest group were students and it appears that the local college instructor is just to lazy to learn and teach ST.

 

  I find Sean and Ken's replies interesting and truthfully until I had to make a rectangle to round transition duct for the first time in ages I never realized that there were things I could only do in Ordered. I just never use Ordered and have not for years. Last time I needed to make transitions I just used my third party software bought during ZW3D days so until this recent effort I knew nothing of these lacks.

 

  My primary use is for the Baking and Food manufacturing industries. Perhaps our needs are a little different than many as according to Don Cooper we as an identifyable group have a higher ST adoption rate than any other one.

by Experimenter
on ‎12-20-2013 04:46 PM

  The simple answer for me is I like control over my design and as short a feature tree as possible. Synchronous is not very friendly with the former and way too long with the latter. Another issue is neither environments play well with together. In other words you can't freely move from one to the other. Once you put an ordered feature in then you have to continue that way.

  With Sync I end up up with useless sketches that control nothing once the feature has consumed it. Another anoying thing is changing dimensions always causes something else to change that I don't want to change. Granted it is user ignorance, but none-the-less it should be more clear to the user what the change will effect before making it. Better still let me select the face I want the dimension to move before it moves.

  Anyway I use the system like a Bull-in-the-China shop and Sync is just to sensitive for my brutish behavior. As it is I crash SE enough in ordered. I'd hate to think how much more it would happen in Sync.

by Valued Contributor
on ‎12-21-2013 04:53 PM
Krasher, What ST version are you using? There are rules for Ordered and Synchronous and it would be worthwhile to make them your friends no matter which you use.
by PLM World Member Valued Contributor PLM World Member Valued Contributor
on ‎02-21-2014 06:40 PM

We have pockets of users who use Synchronous, but many who don't.

The primary reason most of our designers and engineers don't use it is that we have 14+ years of Solid Edge ordered models and a lot of what we do uses these existing models.  Even when we do have to create a couple of new parts, the major advantages of Sync aren't recognized because 90% of the assembly is Ordered.  So users don't see the huge benefits.

 

Where we have switched over almost completely is in our Tooling group.  Everything they create is new and self-contained within that design, never shared across designs.  It is the perfect situation for using Synchronous.  Their management saw the advantages and mandated that everyone use Synch for new designs.  Some people were open minded and got it quick and LOVE it.  Some people fought it all the time and just never grasped the fundamental differences in the ordered and synchronous paradigms.  

 

Our Product Inovation group is the next area I'd like to see us use this.  This is where new products start and if the new designs are done in Sync this will eventually trickle down since it will one day be a mature product.  However, even in a new design we use a lot of existing models.

 

We also do a lot of automation of our parts and assemblies.  For this we need the ability to drive things parametrically and to suppress features in master templates.  After many homegrown systems we are embarking on our first Rulestream project.  Synchronous models just aren't at the point that we can control them easily and completely the way we need to.  If this improves I'd switch those to,

 

Matt's presentation at SEU 13 helps users understand the low level concepts and helps explain why some of the "issues"/techniques can't be handled in Sync.  This has helped us identify what types of things to do in Ordered as part of a hybrid modeling approach and what types of changes to be careful of (consuming an entire face).

 

Carl Breving

by MLombard
on ‎02-21-2014 10:11 PM

Carl,

Thanks so much for that insight. It's not that often we hear that kind of detailed answer. Thanks also for the mention of the presentation. Coming to the realizations as I prepared the presentation helped me understand what the software is all about.

 

Thanks again.

by Phenom
on ‎04-23-2014 04:32 PM

Hello,

 

I'm working purly with sheet metal and assembly driven parts. Every part is linked to the assembly though assembly planes. Moving a single plane updates flat patterns for many parts automatically.

 

Working this way in ST would force me to manually establish the links rather than the links being infered by how the sketches were created to begin with.

 

I has hoped ST worked with Ordered as the starting point rather than the other way around. No 2 way street makes ST less useful.

Labels