Tuesday, June 17, 2008

Synchronous Technology from Siemens - History-Free modelling - Myth or Reality?

Of late the CAD Market has been active with discussions on the 'new' technology marketed by Siemens as Synchronous Technology - a new paradigm in Mechanical CAD. Some CAD gurus term it revolutionary, path-breaking and a panacea for the ills of Parametric Technology ( not PTC !), while other seasoned users and I-have-seen-this-often CAD specialists want to be cautious and adopt a wait-and-watch approach.

One of the common thread underlying the discussions on Pro's and Con's has been that Parametric approach has been beset with two issues , namely errors during regeneration and time taken to edit a design having a long history-tree. Hence history-free modelling approach seems to offer a solution to these issues. Instead of taking extreme stand with respect to either History-based or History-free approach, a middle path trying to capitalize on the benefits of both approaches seems to be emerging. This is definitely a positive trend for the 3D CAD Industry and the user as well.

From the perspective of a Design Engineer using 3D CAD, following points remain in focus for any tool to retain its popularity and acceptance:

1. Ease of Use
2. Bi-directional Parametricity with easy-to-generate-drawings having driving dimensions
3. Tolerance Analysis for stack-up calculations
4. Knowledge Management for re-use of design procedures
5. Design Drawing Automation
6. Error-free 2D Drawing for downstream usage in Manufacturing, Inspection, Assembly, Service among other functions.
7. Dimensional Management for Fit, Form and Function

Most of these requirements would need parametricity, constraints, relations between Features of Sizes that goes well with Parametric History-based approach.

One of the points that needs to be addressed critically, is Freedom to Edit Designs developed on any CAD system. Let us look at a scenario wherein an organization develops a design and transfers the same to another manfuacturing facility for prototype development. If the Design data can be changed inadvertantly without leaving a trace of the change (read history-free) it sounds scary with obviously undesirable implications! Freedom can result in abuse !!

Every technology has its advantages and dis-advantages.

Development of 2D Drawings that enables parametric updates and changes to dimensions need to have higher focus. This gets complicated with History-free approach on account of the bi-directional facility enjoyed by the Engineering community.

Simply said, History-based technology provides a facility to Undo, Modify and Redo a design concept at any stage in design process. History-free approach necessitates correctional approach that may involve re-work. Things done in a fraction of a second may require more steps with History-free approach. Parametric dimensions on a drawing would necessarily have to be re-visited by the Designer when a model undergoes a change in History-free approach while it would not be so with History-based approach.

I guess a consensus would emerge with more players such as SolidWorks (that already has Instant3D and SWIFT to address direct editing and regeneration/re-ordering issues) working on ways to address this requirement without sacrificing on the benefits currently enjoyed by the Design community.

Convergence of methods is sure to be in focus for time to come. At the end of the day, Users would be the winners !!


Anonymous said...

Ya. I agree with you. Parametric modeling looks safer. One of the simple example: I can suppress and unsppress a feature depending on design variations or during design iterations. But how it is possible in direct modeling?

We have appreciate the new technology. But in reality how it can help a designer to do his job better?

It is still a question mark for me. Let us wait and see .....


Matt Lombard said...

This is not really a convergence, this is one approach or the other. You can use the direct editing technology or the parametric technology. But you can't use them both at the same time on the same data.

It's realy like having two separate software packages, which is the part that Siemens is not communicating very well.

You are also completely stuck if you have any geometry other than dead simple prismatic shapes with round fillets. No options whatsoever with general NURBS shapes.

I think for very simple parts, this will be useful, but only marginally more useful than any of the existing tools that do the same thing or nearly the same thing. The only really unparalleled thing here is the hype.

Nat said...

You are right. Though it sounds attractive, Synchronous Technology does not overcome inherent limitations in Direct Editing. For example, a Designer develops a Feature such as a rib (I am using rib as a simple example). During the process of design he/ she deletes it and then goes about progressing on other aspects of design. Now, when the rib is required to be back, a history-free approach would necessitate a re-design of the rib while a 'suppress' option in History-based Parametric CAD would be a simple solution. This is one of many cases wherein Parametric approach would be a better choice.

How about a case where a geometry becomes non-manifold from a manifold topology?

Too bad, the user has to choose between Synchronous or Parametric. As you said, this is not explicitly stated by Siemens' marketing !!

Anonymous said...

Dears, i have been a user of both works and Edge for the last 7 years. In my opinion there was no difference between the two except a few bells here and a few whistles there. While Solid Works would always keep point no.1 "The bi directional associativity between model and 2d views" as a must have, i am yet to see anyone using that in real production environment.
Same goes true for many a things at Solid Edge level.
However, synchronous technology from Siemens surely seems to be a diffrentiator, let's give it to Siemens for their real breakthrough. I am sure if Solid Works would have come out of this first, they would have blown even bigger bells and Drums.
I also saw the video from Solid Works ridculing Synchronous Technology by making a few changes to a engine casing, the video showed Works took 13 seconds with Traditional modelling approach to change the size of the hole, while the same change with ST took a second.
So once again guys, let's give it to Siemens PLM and let the user be the real judge.
Any technology that get's launched has room for development. This is the first release and i can see it is production ready for prismatic components atleast.

Nat said...

Hello A2Z:

I agree with you that ST from Siemens is new in their product. Revolutionary? Maybe not. The user has to choose between ST and conventional SolidEdge parametric interface. Why? History-free editing has been there for awhile. It would be ideal to leverage on the ability to 'recognize' and edit a 3D model while retaining parametricity for its obvious advantages. Spaceclaim has also had similar capabilities. When history-based technology from PTC came-in first, history-free modelling was the other obvious choice. It was evaluated and the merits of history-based approach was taken after weighing in the pros and cons. My point is, history-based approach helps convey design spec, how the design evolved, ability to store design knowledge and re-use model data in part or a whole. Something that would be interesting to evaluate with ST. A common approach with parametrics and direct editing would help gain best of both worlds. The way ST has been projected, it looks like the panacea for all ills of parametric technology that has suddenly become obvious and commonplace!! As you said, and I agree, USERS would be the judge and the result would be popularity and acceptance.

Anonymous said...

I keep hearing "parametrics" and "history" being interchanged as if it is the same thing. It is not! Solid Edge with Synchronous Technology still has parametrics in the ST model, what they do not have is the multiple layers of associative dependency (history). This means I can still drive features with dimensions and I can have one parameter drive another without having to worry about a face being deleted, which orphans a sketch plane, which orphans a sketch, which fails a feature.

Nat said...

Parametric with history-associativity has made 3D CAD modellers a functional design tool as it is today.

Let us take an example of a part with ribs. A designer goes about editing the part (in history-based approach) and removes a rib (by suppressing the feature). Somewhere down the line the feature needs to be added back. In a history-free approach the feature has to be re-created while in a history based approach the feature just needs to be unsuppressed.

For a designer the design intent has to be captured during the modelling process and design specification enumerated on a GD&T drawing. For both these functional requirements, history-based approach (with parametric modelling) facilitates the user, in more ways than one, to achieve the necessary goals. With history-free approach, it would be a cat-and-mouse game.

This would be a serious concern when tolerance analysis needs to be performed at assembly level to study stacks.