← New Improved SolidWorks Knowledge Base

SolidWorks and Animations →

Direct Modeling: The future?

April 23rd, 2008 | CAD Industry News

I’ve put it off writing about this subject matter for quite some time now but with the recent announcement of yet another CAD company touting Direct Modeling, CAD 2.0, Non-History based modeling, or whatever other industry term that is out there, I thought the time was right.

Siemens announcement yesterday no doubt shook some foundations around the CAD Industry.  I’ve read articles or posts that range from “This is THE next big thing” to “It’s old news”.  So which one is it?  Or does the answer lie between the lines somewhere.  Here is my take on it.  Is Direct Modeling the next big thing?  Simply put…No.  It HAS been around for some time.  Is it an important feature set that should be looked at by all CAD companies?  YES.

Matt was correct yesterday when he said that SolidWorks has had some direct modeling functionality for some time now.  A combination of the Move Face command and Instant3D does cover many of the topics we all saw in the video yesterday.  Are those features developed to the level they should be?  No.  The Move Face command is very capable but there is no access to it via the Instant3D tools.  Its success can be very dependent on what faces you select (In particular when there are fillets present).  It is still a great tool for making tweaks and adjustments down the tree without having to understand the design intent of the sketches or worry about child dependencies.  As Josh pointed out, Instant3D does require the feature history to be there in order to make changes to a model quickly.

Even with these existing features, I would still like to see additional capabilities and improved UI functions built into SolidWorks that allows more direct editing of imported models (which of course have NO history).  You can do quite a bit with the tools they have now, but they need to be more powerful, easier to use, (especially for less experienced users) & should require LESS steps.

Will History Based Modeling be a thing of the past?

I do not agree with those who say that Direct Modeling will replace history based modeling.  I believe that combining the two methods in the correct manner IS the best solution I have seen.  I can’t tell you exactly what the correct manner will be but I’m sure the next 12 months will give us all much more insight.  We all saw some pretty impressive things yesterday but there really wasn’t much in the way of an explanation on how the direct edits were shown in the tree itself.  (Actually there was a quick view of a feature tree that shows a “Direct Edit9″ feature about 7:22 into the video.)  With that said, I really like the approach that SolidWorks and Siemens is taking with this.

As with many things in the industry, time will tell.  I do think there is something to the fact that 2 of the 4 major MCAD players have now quickly adopted some sort of marketing stance in direct modeling.  The coming months will be interesting to say the least.

Those are my thoughts….I welcome yours.

Share and Enjoy:
  • StumbleUpon
  • Digg
  • Technorati
  • del.icio.us
  • Facebook
  • LinkedIn
  • E-mail this story to a friend!
  • Google

Viewing 9 Comments

    • ^
    • v
    "believe that combining the two methods in the correct manner IS the best solution I have seen."

    I totally agree. I think the direct modeling will add ease of use but the history will provide the power. The focus is going to be on the DM because it's flashy and looks good in video.

    thanks for the link Ricky!
    • ^
    • v
    Hi Ricky, great blog. As for DM I agree, that it will not replace parametric modeling, like stated, power in parameters, ease and quick fixes in DM. However, I have had both abilities already for the past 10 years and its been a godsend to have both. I can't tell you how DM got me out of some binds and last minute deadlines. Also when time allowed to have parametric features help out for simple updates to standard productlines when properly implemented.
    Cheers
    Tom
    • ^
    • v
    ProE will have functionality to create parametric features out of "dumb" imported models in WF4... I think this is the best of both worlds.
    • ^
    • v
    Mike,

    I haven't seen this advertised anywhere. (Even checked the PTC site tonight). From the sounds of your description it sounds a bit like FeatureWorks which does work fairly well for prismatic parts.

    While creating parametric features can be a handy feature, direct editing if implemented correctly can be fast and easy.

    Like any software out there, the details of how these new features work will reveal how useful they can be. I appreciate you bringing it up as I will be on the lookout now to check it out.

    Best Regards,

    Ricky Jordan
    • ^
    • v
    anybody seen DS Catia V6 video demo that was out in may...same kind of thing, except they didn't make as much of a fuss about it.

    combination of the two ways of doing things, looks like the future.
    • ^
    • v
    Imagine a parametric model that you inherit from someone else where the first 100 features are all parametric and the last 20 are hack-and-slash DM features.

    One of the problems with mixing DM with history based modeling is that the history based modeling always wins. the DM edits become history based features, as we see in SolidWorks with Move Face, Delete Face.
    • ^
    • v
    Matt,

    I agree that 20 "hack-and-slash DM features" is a bit much at the bottom of a tree and is not a practice I recommend.

    Here is a good example of useage: An injection molded Master Model (3 solid bodies) with 500 plus features. After SEVERAL weeks on the project and three prototype sets, you need to move an interior rib wall by 0.020" without effecting any other feature. The move face worked well for this...especially when the change was last minute and literally had to be done in 10 minutes so that the files could be shipped off for more prototypes.

    Yeah...I agree...it is kinda funny that the DM edits are history based features. As long as the regeneration time of the DM features are low I don't think this is too much of a problem. If we get regeneration times like the Deform tool...then we have problems.

    The bottom line for me is that we (SolidWorks users) do need more tools that will allow us to make changes to imported parts. It seems to me some useful enhancements to the Direct Editing features is that best place for it without having a completely separate package (like SpaceClaim).

    Thanks for commenting...and congrats on the Surfacing Book.

    Best Regards,

    Ricky Jordan
    • ^
    • v
    I have used both and there are positives and negatives about both systems. Without a doubt DM is easiest to use when u inherit a part from another system or even another user of the same system. Its a very "hands-on" approach to modeling, more intuitive, more tolerant of different modeling methods and faster to hit the ground running. In a team based environment when you throw models around to be modified by other members it wins hands down. Not to say parametric is dead, just horses for courses. Unfortunately, one would have to look at the entire design cycle to truly evaluate which is the better method, something noone has ever or will ever do.
    • ^
    • v
    Editing imported files like iges & step, vene corrupted files with missing surfaces, has been around for years in VX. Thsi does create history after importing these dumb parts (solids or surfaces). That makes life much simpler, being able to just get on with the job.
blog comments powered by Disqus