SW Move Face vs Synch Tech
It sounds like some of you are still confused. Some people are asking me to compare SolidWorks Move Face to the Solid Edge Synchronous Technology method for making edits. I know that the interfaces for the methods look similar, but these are not the same tools. I suppose if you’re a salesman, and you have to say that SW has equivalent functionality, that would be one place where you might expect to hear someone say that these tools are essentially the same. But they’re not, and here’s why.
Let’s keep looking at the punch holder model because it’s not a bad one.
You would think that any comparison that allowed one software to operate with native data while another is forced to use imported data would be patently unfair. And it is. In this case, it’s very unfair to make SW use its own native data and to allow SE to use the SW file imported directly into SE. Why is that? Because for geometry like the punch holder, history serves no purpose. Well, you could take the existential route and argue that material must exist before it can be cut, but there is nothing that says you have to have separate features, you could just extrude a single sketch, and the sketch doesn’t care about history. Or you could extrude the holes as positive, and then subtract them from a block created later. So order doesn’t matter.
There are times when you need to make one feature first, and then a feature that has some dependency to the first feature. Yes, I do believe that history modeling is not dead. It is just grossly misapplied in some cases, including the case of making a prismatic punch holder part as a history-based part. But this is exactly the kind of part that a lot of machine design is involved with.
So because this is not a great part to make a poster child for history-based advantages, some people are arguing that you should use a method that looks like direct editing. Use Move Face instead of editing the original sketch. Would you do this in real life? Some time ago I wrote about using sleazy CAD tricks to just get the job done sometimes. Move Face is clearly a sleazy CAD trick. Sometimes you do it. But it’s sleazy.
The problem with trying to make a history modeler work like a direct modeler is that you’re still making history-based features. So you are actually adding to your original problem. You’re putting an infected band-aid over a wound. Sure it stops the immediate bleeding, but there is a price to pay later on. Let’s take a look.
In this image, you can see that the Move Face feature is added at the bottom of the tree, and that it adds another 20 seconds of rebuild time. This was using Move Face to edit the length of the plate by 1″. And this is in addition to the time it takes to make the feature, which you have to sit and wait for the preview to show, and all.
But what if you wanted to also edit the width by 1″? Or what if you wanted that change to be symmetrical? The answer is that you would have to add another feature, which would add another 20 seconds or so to the rebuild. This isn’t any good. In fact, it’s pretty bad. You’re just making matters worse than they already were.
I’d make a video of this, but it would be long and boring. You can simulate it by staring at your monitor for a couple of minutes doing nothing.
What happens in SE? It takes about 20 seconds to do the edit. And it doesn’t add a feature to the tree. The next time you have to make the same or different edit, it takes about the same amount of time. Edits are not remembered, so the times to create them don’t get compounded.
Plus, you have some additional options, like the Tip, Lift and Extend options we looked at a couple of posts ago. SolidWorks Move Face just works with the existing BREP faces. It may extend or trim or even delete them, but it will not change the faces adjacent to the selection set. Here’s a movie of a simplified case:
And then there’s live rules. Live Rules is really a method for selecting faces according to certain geometrical rules. Using different rules for selecting faces in different situations enables you to change the design intent without editing features. If you’re still struggling with the whole concept of Synchronous Technology, this is one of the key concepts. Features, faces, selections, and moving faces according to rules. And it works with imported geometry, although there are some advantages to native features. Best practice is a thing of the past. It doesn’t matter how you make something, as long as the geometry is correct, that’s all that matters.
Alright? So I can forgive a bunch of you for wanting to compare SW Move Face to SE Synch Tech, some of you even going as far as to say that the comparison was “apples to apples”. Apples to apples to me is to use the preferred method for each modeler. There is no question that continuing to use Move Face in SW to cover up problems with history is really making the problem much worse. I’ll remodel the part in SE ordered just for comparison in the next day or two.
If in SW problem is in rebuild time then deal with rebuild time and enable feature freeze.
Here’s another difference between SW and SE that I didn’t mention, but someone reminded me of. In SW Move Face you can only make a move using a relative dimension, so if you want to move a face 10 mm, that’s how you do it. But in Solid Edge, you can move a face so an overall dimension is a given value, or you can move a face up to a selection. It seems like a small distinction until you have to work with only offset amounts.
Matt if we want to compare “apples vs apples” make sure you switch to linear mode (ordered) inside Solid Edge. Then from the modify group use the move face command. (same for sheet metal environment that will expose specific sheet metal modify commands)
Also we need to understand that move face add an incremental value to the model. If the initial part was 10 inch. When the move face is apply the part will be 10+something and not =10+something. That will make a difference when it come times to edit the part.
If one of you colleague measure the final part change the initial feature, he will need to go back and edit the move face value to compensate. In some more complicate situations it could be confusing. I remember some cases when this was introduce. Where people did not pay attention to the result when using the move face and the y complain about it.
Other situation is when the variables are exposed as a property and the property is use in a BOM. Your BOM no longer match the part.
A quick tips, instead of holding the CTRL key, place the left hand on the lower left corner of the keyboard and extend the thumb. It will be over the spacebar. press the spacebar once to switch the select tool mode to add/remove mode, press a second time for the add mode only, press a third time for remove mode only, press again to go back to initial state.
http://www.solideadn.com/SEHelp/ST4/EN/selectb3d.htm
To access the Selection Manager, use the SHIFT+spacebar then click a surface and pick an option in the contextual menu
Matt I think most us do appreciate the difference between move face as a history operation and a direct edit move face. It still does not get around the point I made about there still being underlying issues with kernel level operations. If direct editing is indeed direct face editing why isn’t the response instant? Like it is in any polygonal modeller.
Sure there are more complex issues here but until these kernel level issues are resolved we will still be arguing the toss between direct/ST/history. In many respects I am starting to consider the whole development of ST as a way to circumvent underlying kernel issues that are known to be unsolvable at this stage.
What you say about “limitations” driving the development of Synchronous is true, but the limitations are not in the kernel per se. With Synchronous, there is really very little reason (as you point out) that operations can’t be “near instantaneous”. When we started Synchronous our goal was to be sub-2-seconds on any operation. We have achieved this in most cases, but we continue to work on the outliers. I was actually surprised by Matt’s timing on this part (and have confirmed same) and I consider it to be unacceptable. We are looking at what is going on with this particular model.
That said, you will never achieve sub-2-seconds on this model in a history-based system. Before doing Synchronous we looked very hard at history based alternatives — after all, we were history-based like everyone else and not doing that was petrifyingly scary. So we looked within our existing world first — like doing the direct edit on the end of the tree, and then back figuring it into the tree in a different thread — that kind of thing. It is all dead ends. As long as you are wedded to history from a technology point of view, you can’t get to true interactive performance.
That said, it would be wrong to think about Synchronous as only about performance and therefore it being a “toss” (which it won’t be, Sync is on the upswing and history is maxed out). The key thing that people find about Synchronous when they actually engage over a period of days or weeks is that it changes their DESIGN process for the better. If you are just “modeling” — i.e. you know exactly what you want, then history-based will largely suffice. If you are designing, where you hit blind alleys and don’t know exactly where you are going till you get there, Synchronous will totally change your world. (This can literally happen. We had an early adopter Matritech, who when asked the benefits of Synchronous technology said “I get to spend more time with my family”, because he was doing twice the number of designs in 1/2 the time. To him the benefit was more than just cool. )
Thanks for the feedback Dan. As I said, from an end user’s perspective we just want to create the geometry and edit it efficiently. Sometimes that will be best done with history/associativity other times a direct edit will be best.
But there are underlying issues that the CAD vendors need to address.
I can open this part in Modo ( as a polygonal model) select the faces and move them instantly.
This is one reason why many product designers are starting to bring sub d modelling into our workflows. So take me for example. I conceptualise on paper, in Modo, in Rhino, in SolidWorks, in Shark and others. The format is not important. The outcomes are. The speed is.
As we move around apps we get frustrated by the glaring differences. What is easy in Modo is very very hard to do in any standard CAD app ( apart from maybe Creo now they have sub d or mega pricy CATIA). Similarly what is simple in a CAD app (filleting) is painful in a sub d app.
Similarly as I sit with my 4 core/8 threads workstation watching Solidworks only using one core I get angry. These are kernel issues that slow things down. Geometry is not a done deal until I can mix and match Nurbs approaches with sub d. Fast. There is a lot of legacy inside all the main kernals now. I think the CAD world needs a new more efficient one.
“I’d make a video of this, but it would be long and boring. You can simulate it by staring at your monitor for a couple of minutes doing nothing.”
Good one, LOL
Devon
It is funny to hear Bob say this. I came to SE right as ST1 was released. The idea of direct editing was powerfull and what brought me here along with sheet metal.Warts and all, some pretty big ones in some cases, I started using ST1 for most of my work. By ST3 it was for everything and when I had to go back into ZW for a cam plan history based editing had become such a PITA in my eyes that I would take the part back out, edit it, and bring it back in. I found myself in those moments reflecting on how difficult life used to be for editing and especially on imports.
II have watched these long time straight history guys who are willing to give it a try in SE all get converted. Now they still fuss about some things but I bet you can’t take ST from them without a fight now. I do not know one person who has honestly tried ST who does not like it. Learning live rules is the key here.
Move face is not even in the same building much less the same room as ST.
Matt,
As always you do a great job of exploring the subtleties and power of Solid Edge and Sync’s workflow, and what us users should expect when editing parts. And this comparison of SW’s “Move Face” command versus Solid Edge is a fine example. I especially like you explanation as to Sync’s subtle differences in behavior when using “imported” versus “created” geometry. It took me awhile to appreciate this difference of editing in Sync. As you point out when creating a new part using Sync, it creates “features” within the model. I think since they are made in a linear fashion, just like a History based model, it confuses the user into thinking there is “history” to these features. But again as you point out this is not the case. Even importing History made Solid Edge parts can be a bit of a mixed bag. SE does a great job as you’d expect importing the geometry, however understanding the original design intent, can be tricky and editing these in Sync behaves somewhere between “imported” and “created in” geometry. But that could be just me. But knowing the difference in Sync’s editing behavior as you illustrate here is good to know… irrespective of what CAD program you use.
I’m a 12+ year user of SE and in the early stages of Sync I butted heads with it due in part to these kind of things. I kept applying my “carved in stone” mindset on how this was suppose to work, rather then learn the subtle differences within Sync edits. However with ST4, I took the plunge and although it has taken me 6+ months, I’ve become more adept at understanding how Sync is going to behave. The funny thing is now when I bring up older SE parts, which are History based, I find myself more frustrated then my time spent moving from History to Direct Editing. I miss the inability to edit with just Sync. I guess you could say I’ve turned the corner, and the old ways of SW and SE history based and Face Move modeling seam quaint. That’s not to say Sync can handle everything. History is still required in a few cases, but I think in next few years, Sync will eventually get there too.
So thanks again for your blog and another great tutorial on using Sync. You have an excellent way of explaining how Sync works, and I learn something new every time.
Bob