Last week, Siemens launched what is very well the most feature-busting update to the Solid Edge with Synchronous Technology craze thus far. It’s Solid Edge ST3. I’ve not been able to get inside it yet, so I’m holding off on a proper review till I do. In the meantime, I went to Siemens with some questions about the ST3 release. After hearing back from Kris Kasprzak of Siemens Global Marketing, I thought it would be interesting to get some more perspective form a Solid Edge user/blogger I know, Scott Wertel. The results? Interesting to say the least.
Solid Edge ST3
To provide some foundation, Solid Edge ST3 lets you create both history-based features and direct features. They keep the features organized as Ordered features and Synchronous features in the sidebar, so you’re free to work in either one, in the same part or assembly, with the ability to even create relations between the two. So, Scott and Kris commented on this. I asked them questions. They responded with their perspective. I couldn’t post just one. Both have great points and the contrast is wonderful, particularly on topic of the thought process in design and engineering.
I’ll stop gabbing now so you can read, but I’m curious about your perspective on ST3 and on the perspectives given below? Agree? Disagree?
|How have the ST features changed from ST1 to ST3?|
|Scott Wertel – Considering that the features of ST are direct modeling features, the term itself is a bit misconstrued. But, the tools used to make ST what it is have been made more user friendly. I’m specifically referring to Live Rules and the Steering Wheel. Live rules has some behind the scenes improvements that make it less intrusive to the user. In ST1, it always seemed like Live Rules just got in the way of your design intent. Now, not so much. Of course, changing the location of Live Rules in the UI and the feedback based on subtle highlighting also helps. The Steering Wheel still looks the same at first glance, but now has a distinct flavor depending on the type of push/pull modifications you can do based on the geometry selected. Use of the steering wheel is easier, but the full power of it still requires extensive proficiency in SE to know what grip to grab, what geometry to attach to, and what hotkeys to hold while interacting with the Wheel.||Kris Kasprzak, Siemens – The features have not changed (except as we have added more) but we have made synchronous features more accessible to existing users. We have integrated both feature types (ordered and synchronous) in the same model file so users (new and existing) can use the feature most appropriate given the operation. For example synchronous features can be used to create models faster and provide greater flexibility and speed when making edits, and be able to use ordered features to capture a process such as cast/machining steps.|
|The capabilities of ST, and Direct modeling in general, seem to be increasing. Will this tech find it’s way into rendering and simulation?|
|Scott – The power of CAD has never been in the creation of geometry. Moving from paper drawings to 2D CAD, it took most engineers/designers/draftsmen the same amount of time for the initial creation of the print. Most certainly there were tools (like blocks) to make it quicker in some regards, but in general the time savings in one area of CAD was taken up by time expenditures in others – like file management. The real power in CAD comes with how quickly an original design can be modified. Changes in 2D CAD was much quicker, and cleaner, then hand drawings. Changes to 3D models are much quicker than 2D CAD. Since direct modeling makes making changes easier, I expect its use to increase. With ST3, I think Siemens hit the nail on the head on how to implement direct modeling. Start with parametric features that allow a designer or engineer to impart some, do I dare say, design intent. Then, make it easy to change for that inevitable change that wasn’t foreseen. Most importantly, don’t make the engineer/designer have to choose which method to use because no one has a crystal ball. Expand that concept into rendering and simulation. Anyone in television knows, what you see on tv is nothing like real life. Making something LOOK lifelike is not the same thing as BEING lifelike. There are plenty of geometry tricks that need to be done to manufacturing models to make them look real in a rendering. Considering rendering just makes a “dumb” picture, there is no reason to maintain complex feature history or design intent. Analysts also know that a finite element model looks little like the manufactured part in order to get accurate results that solve in a reasonable amount of time. Analysts need tools that allow for quick modification of geometry without worrying about design intent, feature history, OR accidentally changing the manufacturing model.||Kris Kasprzak, Siemens – We first need to clearly define what these technologies are (see below), but in Solid Edge, the capabilities of synchronous technology are already available in Simulation. Most mature 3D modeling systems include some level of FEA, but the real need is in refining designs. If and when an issue is encountered, the history-based approach forces change a certain way which limits the creative process. However as synchronous modeling is available during FEA studies in Solid Edge, designers can refine designs much faster and with greater flexibility. For example, should a part fail due to lack of material, users can push or pull the geometry into shape. Editable 3D driving dimensions can be added to a completed model allow dimensional control, and editable cross sections can be inserted to allow 2D edits to 3D models. With these tools, users should be able to make more refinements in less time—ultimately improving product quality.|
|There’s a thought process when modeling a part with features. It may be molding how it’s machined or modeling it in a way to reduce the amount of features. How is this thought process different with ST?|
|Scott Wertel – This is actually the one area in ST, or any direct modeler, that is one more ripple in the pond to dumb down engineers. OK, maybe that’s too harsh, but it does separate the engineer from the actual manufacturing practice. For example, when I first started 3D modeling using prismatic features with no means of editing short of deleting the “feature” and recreating it, I took great care in following manufacturing methods. Just because I could model it doesn’t mean it could be machined. Forcing myself to visualize all the manufacturing steps, setups, and operations a machinist, tool maker, or mold maker would have to do made my designs better. Stretching, pushing, and pulling geometry is not something that can be done on real parts. It will be too easy to model something that cannot be manufactured and the designer/engineer will be so far removed from the thought process that it will cost thousands of dollars in ECO changes after the machinist calls the engineer department, yet again, to complain about a crappy design. Looking to the future, additive manufacturing methods won’t care so this argument may be moot. And business will have to develop better design (review) processes to catch these types of mistakes before they make it to the shop floor.||Kris Kasprzak, Siemens – Interesting opening sentence—a thought process. That’s actually part of the issue with 3D systems—especially for users coming from a 2D system. It’s the concept of pre-planning that causes issues such as reducing features to improve update performance, or to accommodate future unexpected changes (how do I model my part so it can be reused elsewhere?). These are some of the primary issues synchronous technology solves. This happens because 1) edits don’t need to follow creation steps, so users don’t have to think about how to model a part but designs can easily adapt existing designs for newer uses without having to understand and “re-program” the construction history.|
|What are the major differences between Direct modeling and Direct Editing? How does Solid Edge ST3 approach each?|
Scott Wertel – To me, there is no difference between direct modeling and direct editing – 2 different terms for the same thing. I draw a solid block and add geometry to it. Am I not just editing the block to include a boss? Or am I modeling a boss? Same goes for cutting out a pocket. Am I modeling a pocket or editing the block? If I were to put a fine line on the definition, direct modeling is history-less and direct editing maintains a history. In other words, with direct modeling, I push/pull a face and the geometry updates but the feature list doesn’t change. With direct editing, I push/pull a face and the feature list shows a push/pull feature at the bottom of the history tree. But that is a poor definition because it is based not on technology or design methodology, but user interface. Picture a REAL feature tree that shows all features and their parent/child relationships like a flow chart (aka I-deas). A direct edit wouldn’t necessarily show up at the “bottom” of the list and therefore history has no meaning. As an example, I think VX goes overboard on recording edits as features (like recording “delete” features). But, when in the hands of a proficient user, it makes it a very powerful means of design.
SE ST3 approaches each like my initial definition. When in synchronous mode, a direct edit only edits the geometry and if a PMI happens to be attached, edits the dimension. When not in synchronous mode, a direct edit feature is added to the feature list.
Kris Kasprzak, Siemens – Great question that often gets confused. First let’s define the technologies, then I’ll address synchronous technology.
History based: features are captured as created and must be edited in the same exact way. Generally features build upon each other where this feature is defined using the geometry of that feature. Constraints are used to manage how features interact with each other so some “programming” is needed to get features to react a certain way when edits are made. During a change, edits are destructive meaning downstream features the “tree” are deleted and re-created upon an edit. The delete and re-create process is masked by shown the existing model then the updated without showing the rebuild process. This is why these systems are so fragile—if features aren’t built carefully, they will not regenerate. PS that’s one of the reasons PTC forced complete sketch constraining in the early days. Users can create highly automated designs at the cost of intense pre-planning.
Direct editing: A technology where an operation added to the history tree where faces can be moved outside the feature definition. While helpful in moving simple and small amounts of data, edit issues are compounded as edits to the original feature are over-ridden and additional features must update. The big difference is that direct edits only effect the selected geometry. You can’t maintain intent across geometry unselected and operate only at the part level. For example, say the concentricity between mounting holes and an outer round is to me maintained, should just the holes be selected it can be moved but the concentricity will be broken. Direct editing is only helpful in making small localized edits.
Explicit editing or direct modeling: A technology where there is no history, features or captured steps, but edits are made directly to the geometry. Similar to direct edits, only selected geometry can be moved. Unselected geometry stays “put”—hence maintaining intent is impossible. Useful for making fast flexible changes.
Synchronous technology combines the precision of feature based modeling with the speed and flexibility of explicit modeling. Users can create and edit using a concept of features such as hole parameters can be adjusted as they were created and pattern instances can be altered. However because features are independent of each other, edits can be made to virtually any feature and others updated accordingly—features created last can affect the first operations. For example, take a block with some rounded corners and mounting holes placed at the center of the rounds—a common machined base plate. The most important features are the holes—designers really want to be able to control distance between such they align with other assembly components. In synchronous technology you can easily add a dimension to the hole centers, (if not already present) and alter the distance. The concentric relationship is automatically found and maintained and the blog size adjusts accordingly.
|How would you explain ST to SolidWorks, Pro/E and Inventor users and Spaceclaim, Keycreator, and Cocreate users?|
Scott Wertel – SolidWorks, Pro/E, and Inventor started as a history-based-parametric-feature-based solids modeler. Over time, they added features that could directly edit the geometry instead of the parameters defining the geometry. They did so by recording the direct edit as a feature in the feature list. Therefore, direct editing/modeling is not new, not even for Spaceclaim or CoCreate. To explain the ST to Solidworks, Pro/E, or Inventor, the fact that there are no features is the main difference, unless you want there to be. There are collections, and based on the current UI they look like features, but they are not features in the same sense as a h-b-p-f-b solids modeler. CoCreate and SpaceClaim users would understand that, but have a hard time making the association since they can’t conceptualize features that way (assuming no other experience to CAD). So SW, Pro/E, and IV users can keep their features, but lose the parameters and history unless they CHOOSE to keep them. That’s the difference and happens to be the most powerful aspect because it allows future change without the crystal ball to anticipate the change. Somewhere in there, Live Rules would have to be explained. No one else has anything like Live Rules.
To explain it to SpaceClaim and CoCreate users, I would ask them to create a block with a hole in it. Now what if you want to interrogate or modify the hole, or a pattern of holes, how do you find it/them? SE ST allows you to find the collection containing that hole and you can manipulate the collection, not just the hole geometry. SC and CC users placed a hole, at a specific location, now how do you guarantee it stays there? With SE ST, you can create a parameter and lock the geometry in place, yet still allow for a controlled modification in the future. It’s the best of both worlds in a single user interface (hinting towards Inventor Fusion).
KeyCreator users I would refer to Jon Banquer. Actually, KeyCreator has been a step ahead of the competition for true hybrid modeling since CadKey. So has NX since UG. While on topic, so has VX. That’s what you get when you allow CAM software to create CAD geometry. But hybrid modeling is a completely different animal which is why KeyCreator stays in a class all by itself, unable to compete with the typical midrange MCAD modelers or high-end modelers. There is just nothing to compare that the average user can conceptually grasp.
Kris Kasprzak, Siemens – For any history based system (Pro/E, Inventor, SolidWorks are basically all the same minus some subtle differences in the user interface—sorry but it’s true)
1. Create designs faster because a design doesn’t need to be pre-planned, and features don’t need to be programmed to “talk” to each other”
2. Faster change as model regeneration is eliminated, but control and intent is still maintained
3. Improved reuse of imported 2D or 3D parts into new designs as imported models can be edited with the same flexibility as native designs—the dirty little secret of 3D systems solved.
For any explicit modeling system (co|create, spaceclaim, etc. are basically the same minus subtle differences in the user interface.
1. Faster change as many features can be edited using input parameters (try changing pattern counts in co|create—it’s a delete and re-create process. Also intent is automatically preserved for “strong” or obvious geometric conditions such as tangent, concentric, horizontal/vertical, co-planar, and many others. Basically users can create without pre-planning but get designs that support highly automated change.
2. Improved reuse of imported 2D or 3D parts into new designs as imported models can be edited with the same flexibility as native designs—again still not solved very well with explicit modeling systems.
I’d like to thank Scott Wertel for providing user perspective for this post through an email Q&A that I edited for blogging. If I got anything wrong, he promised to put the smackdown with a Stratego challenge. Find him at @swertel or at Scott’s Harangue.
I’d like to thank Kris Kasprzak, Siemens, for providing the marketing perspective for this post.