Direct Modeling In Creo

In “What Exactly Is Creo? – Part 5” I wrote:

For me the key breakthrough (if any) will be in the way Creo does direct modeling. If Creo’s direct modeling is simply automating the hack and whack approach now being used by history based parametric modeling users, then I would hardly consider that as a breakthrough. It will be a time saver, that’s all. Autodesk is trying to achieve something unprecedented by making the software so intelligent that it can pick apart the history tree and incorporate the direct modeling changes made to it, all the time. From the limited understanding that I have on Creo (because I don’t have the software yet) I gather that the software can edit the feature tree if the direct modeling changes made are simple. But if they are complicated, Creo Elements/Direct simply adds a move face feature to the feature tree (which is invisible to the user) which then shows up in the Creo Elements/Pro. If that is indeed the case, then this whole thing is simply reduced to mere automation of an existing task.

Today in a post titled “Smoke, Mirrors, Creo, Ninjas: All good things“, Al Dean of DEVELOP3D wrote:

At its very core, Creo is still a history and feature based modelling system – whether you’re using the Creo Parametric or the Creo Direct app. What PTC has done is an excellent job of hiding the fact away from the user of the Creo Direct app. That’s the one fundamental thing to realise. Creo Direct looks, feels and works like a direct modelling system. All of the user experience flags are there. Grabbing faces, deleting data, re-applicable of rounds, maintaining of geometric relationships where possible. It looks and feels like CoCreate does, like any of the other systems do. But underneath the hood, its storing a history of every edit you make, every feature you create. You just don’t see it.

So as I guessed eight months ago and as Al understands it today, PTC has implemented Direct Modeling in Creo by simply automating the process of adding features to the bottom of the feature tree. If that is indeed the case then, then in my opinion, Creo is hardly a breakthrough in 3D modeling technology. But as Al points out, who gives a shit, as long as users can get their job done and this method of direct modeling solves more problems than it creates.

  • Cmon

    Dave,thanks for showing SE. i left a comment yesterday, but somehow it didn’t appear here. this one is done on an imported IGES from solidworks 2011. which can be seen in the feature tree. again, it won’t work without selecting the bottom surface.
    http://bit.ly/kZ7Yre

  • Tas_georg

    I really think that you guys have not really replicated what is Cmon showed us about his problem in SC. Even thougn Solid Edge is able to solve the situation easily, it is all a matter of the initial selection set.

    In my video you see three possible ways of sychronous edit  using Solid Edge ST3. You will notice that based on the intial selection set prior our edit (move of the cylinder) we can face the same kind of problem our not.

    This is happening because the boundary conditions of our possible edits are based on the topology of the selection set and the rest of the geometry.

    Is it only me that thinks there is some kind of pre-planning enen in the synchronous (direct) modeling?

    http://www.youtube.com/watch?v=NTPZQp-GEsc

  • http://twitter.com/bcourter Blake Courter

    Well, most history-based CAD systems also have a move face feature.  If that works for you, use it.  

    That is a simple case that should work in any system, direct or history-based.  It probably will soon in most CAD systems after this thread.  Many others work immediately.  The important point is that it’s still very easy to do.  I think that most direct modeling vendors, including us, show a disproportionate amount of drag and drop when slice and dice is more a more dominant workflow.  The main benefit is the simplicity of the data model.  You can make 100 different tweaks and have a model as good as if you did it in one.  It’s very freeing, once you get used to it.  And, as 3D print says, SpaceClaim is probably the fastest at doing these edits.

  • Paul Hamilton

    Tim, I finally got to try it on v17, and sure enough it does not work as well as 18. I guess that is progress :).  I can make the change easy enough, but not with a simple drag. I also played with 18 more and there are some positions in the part where I try to place the cylinder that won’t solve – so it is not perfect, but better than 17. But even if the simple drag doesn’t provide the needed results, there are many other simple ways to accomplish the task. In direct you should never have to start over. Any geometry is of value.

  • Paul Hamilton

    Neil,
    It
    is true that a simple drag or push in direct modeling does not
    always give you the results you want – but that is not the key advantage
    of direct modeling. With direct modeling
    there are always simple, and relatively fast ways to work
    through the geometry to get the results you need. There is never a need
    to throw away geometry and start over – something I see history-based  users doing frequently. It’s not that
    their geometry is wrong; it’s usually that
    the tree structure won’t support
    the needed edits/design intent. I wonder how many that are reading this have
    had to do that.

     

    In
    CoCreate besides dragging the feature, I
    could rather move the edges and patch in new surfaces, I could
    cut the surfaces and paste them back into the solid, I could cut the solid
    like Blake did in SC and move and unite. Whatever works best for you depending
    on the geometry and topology. In the end the result is the same – a valid B-Rep solid.

     

    The
    key benefit of direct modeling is that it eliminates the dependence
    on proprietary data (the history tree), modeling standards,
    and planning ahead.

  • Paul Hamilton

    Looks great Roger. I like the interaction and response. But I do have to pick on you a little. In direct modeling should there ever be a dependency on “workflow”?  I know that is a bit of topic, so we don’t need to start another discussion. At least I think that should be one of our goals for direct modeling.

  • Roger

    Paul, in an ideal world I guess not, but in reality you will never match step for step from one CAD package to another.  In terms of “direct modelling” I wasn’t trying to show Solid Edge’s “direct” credentials but merely its modelling capability.  Actually, the example video I did was done using a hybrid approach which you may see as a compromise.  Personlly I don’t because SE is not a pure direct modeller – it is what it is and for me this is a strength not a weakness.  I have a fairly pragmatic view on the direct argument  and don’t really care how the software does it, I just want a modelling tool that helps me design quicker.  PTC seem to be of a simila mind. 

    The only “pure” direct modeller I have trialled is SC and I really like the implementation of some of the tools – very clever.  But tying down the geometry is clunky and feels like an afterthought.  I think SE has done a better job of combining the two. 

  • Neil

    Paul, my initial comment was prompted by what I saw as less than candid claims from the direct modeling boosters. Now both you and Blake are saying that we have misunderstood what direct modeling is all about. So it is not about “directly” pushing and pulling and twisting and moving of faces, but rather about… what exactly?

    Techniques like copying bodies, faces, and edges, creating patch surfaces, stitching surfaces, booleans, etc. are legitimate tools of history-based modeling especially with surface modeling. So you should not count them among what is exclusive to direct modeling.

    You also claim that one should NEVER start fresh. That starting fresh in history-based modeling is ALWAYS bad. Is that really true?

    You have often tried distinguish CoCreate as a “mature” product, so it was amusing to see you explain to Tim that the 17th version of your product doesn’t work well enough relative to a seemingly simple change!

    I think direct modeling failed to catch on because it “over promised”. I am trying to understand when and where direct modeling is appropriate. Now you have added what direct is modeling, to the questions I am looking for answers to.

  • Cmon

    Well, maybe after you found out that SpaceClaim fails “trivial” cases, because you initially said “it worked for me”, the development team should put some efforts and make the move tool more  capable.After all, this is “the easy to use” CAD tool.  I’ll check that with your next release. BTW, why Direct editing is so dependent on the modeling workflow? i thought its just the final geometry that matters. the edges and faces. but it turned out to be false.

  • d3print

    Should we call all these direct editing movements  “Easy Edit Tools”. If you have EET in your MCad, you might survive with your problems, if not, do it as you can ;).
    Next big question is how we can modifye the coming AMF format or can`t we.

    Ps. Anyone checked this out with Kubotek or Ironcad stuff?

    Thanks,
    d3print  

  • d3print

    As I asked earlier from Roger does it works when you move cylinder first to corner and try then move it again to middle. It failed inST1.
    Question is do I have to know how I`m going to move the object?
    Does it matter if the cylinder is done by trad mode and block done with sync?

    Thanks,
    d3print

  • Tas_georg

    Roger.. how did you make this with Solid Edge ? I am trying to replicate  this but live rules block me. I set the same option in live rules as is shown but still the problem exists. Is it native geometry or imported?

    Thanks, Tas_Geo

  • Tas_georg

    No, if you first move the cylinder to the corner nothing more can be done. As I understand it, this is happening because “no proper geometry set can be computed because geometry is missing from two sides and as so the combinations of the missing geometry are numerous”. I tried this is Solid Edge ST3.

    I demonstrate something like that in my previous post. All is based on the initial selection set. I cannot really know how the modeler solves the moving geometry but these are all I have experienced using Solid Edge.In ST3 you can model both trad and sync features. If you model the block in sync and the cylinder in trad, then you must move the cylinder in the sync environment in order to manipulate it directly.I could upload  a video if you like.

  • Cmon

    Very well said Nail.

  • Tas_georg

    Roger, don’t bother.. I was finally able to find the solution. The fillet is done in the ordered env so every time the cylinder (face set) is moving then the fillet feature is applied to the new geometry. 

    If you try to put the fillet in sync and move the cylinder, you get the errors I was talking about because there is no fillet feature but a set of faces.

    I really love the technique.

    Thanks Tas_Geo.

  • Roger

    d3print,
    Every modelling method has areas of strength and weakness, the real question is can it take you where you want to go without too much fuss.  Have a look at the video and decide for yourself:
    http://www.youtube.com/watch?v=Gv8ySVGNhMY

  • d3print

    Roger, I agree with strenght and weakness, and that`s why I personally like SpaceClaim more than SE in
    most cases, it gives you more flexibility and speed.

    Thanks,
    d3print 

  • Roger

    Yes, Spaceclaim is a great product.  I did a trial a while back but never got as far as the slice and dice techniques.

    Would you like to provide a video to show the different options to achieve this task in SC?

  • Paul Hamilton

    Neil, perhaps we are talking about two slightly different things. For me direct modeling provides a complete different methodology for product design. Direct editing is just a small part of it. Depending on process and product characteristics, direct modeling MAY provide a better methodology. Certainly, depending on process and product characteristics, history-based modeling MAY provide a more suitable methodology. For many companies around the globe, the cost of the structured model does not yield a high enough return to justify the investment. As such they choose direct modeling as their primary design platform.

    What is exclusive to direct modeling? Geometry is the “master”. That is all. Because of this, depending on your situation, it might be easier to learn, there is no need for modeling standards, geometry is geometry regardless of where is was created or how it was created, there is no need to plan ahead before you begin modeling, the model creation process has no impact on the ability to edit, working on the design as a team can be easier, …

    There is no geometry you can create with one methodology that you can’t create in the other. There is no edit that can be accomplished in one that can’t be accomplished in the other (if you don’t care about the structure in the tree). Certainly the process and ease of doing it may vary, but whatever, we just need to get our job done the best we can. If direct modeling/editing doesn’t help you do that, forget about it. Although the number of people taking advantage of it is growing rapidly every day, it is “catching on”. The CoCreate product for example is experiencing record growth – higher than it ever has in its last 18 years of existence. These companies are not purchasing it solely for it’s geometry editing capabilities, but rather for it’s ability to support their “design” process better.

    Most all of today’s CAD tools (history or direct) can now do some level of direct editing. It is very likely that investment in the technology will continue – as long as people are paying money for it – and they are now more than ever before. Use it if it helps you and ignore it if it doesn’t, but it is obviously helping a lot of companies out there be more productive.

  • Dave Ault

    WOW, the thread that has gotten so long that one is lost if you have not stayed here.  It surprises me that this is a problem. When I create a cylinder like Roger did there are problems for sure. But since we are discussing ease of modeling here I am going to make a few comments.

       What we are looking for here is quick and easy part editing right? so far we have moved things around and all that stuff. We have shown some geometry that can’t be moved in the same fashion by all cad aps. But insofar as this goes is it the only answer or the best answer?  Maybe I can’t move the cylinder off or to the corner but I can click on the cylinder on the side and delete,  and then click on the bottom face,circle,extrude and have my cylinder in the corner. My point here is that this was done in the part mode without ever leaving and is done literally in seconds. There is more than one way to skin the cat and lets not forget that how quickly you can get to the desired result counts.  Does fascination with a known tricky problem  over rule speed and ease with which things can be done? In this case I think it is so and looking at the clicks and commands needed to do this and can it do this become irrelevant when there is a different way that is quicker and does so in fewer clicks and tricks. Just a thought here about the validity of what we are doing and discussing here and how it relates to efficiency in cad creation.

       Also, did I miss it or did the editing of imported parts get lost in the shuffle here?

  • Lee

    In my opinion such a behavior of Solid Edge is just a bug. If surfaces can be moved and the result of the move is supposed to be meaningful geometry then you should be able to get it.  Period.

    What about Creo, does it work for all your cases?

  • d3print

    Yes, I try to return to this later, at least I hope so. At this moment I have no opportunity to record screen, operation and system changing going on.

  • d3print

    Roger, here`s my practise with that cylinder movement in imported part (step). Done with few tricks, split and compine.
    http://d3print.blogspot.com/

  • Roger

    Cmon, I couldn’t read all of the options on the local menu you were using in the video, but it seemed pretty comprehensive. It would seem that Creo is the only one can make this edit without a work around. Very impressive!
    Roger

  • Roger

    OK, I’ve had a closer look at Cmon’s video and realize I have been too kind to Creo.  It seems that the Flexmove operation is simply appended to the end of the parametric history tree.  Now that I understand how it works, IMO the Solid Edge solution is much more elegant!

  • Pingback: Creo Explained – Part 4 | Deelip.com

  • Pingback: PTC Creo - Deutsch - Konzeptentwicklung: 3D oder nicht 3D?

  • Pingback: PTC Creo - Français - Pourquoi choisir la 3D lors de l’élaboration du concept

  • Pingback: PTC Creo - 日本語 - 構想設計: 3D で行うべきかどうか

  • Pingback: Anonymous

  • Pingback: PTC Creo - Italiano - Conceptual design: 3D o non 3D?

  • Pingback: PTC Creo - Deutsch - CAD-Tools müssen Flexibilität beim Modellieren bieten

  • Pingback: Anonymous




Archives

© 2014 Deelip.com. All Rights Reserved. Deelip.com is a registered trademark of Deelip Menezes. Log in - Designed by Gabfire Themes