Should I switch to 3D CAD?

rose_barfield
edited June 2020 in Other

Hi, I'm currently researching the topic of 2D versus 3D CAD. I'd be interested in hearing people's thoughts on this topic. Do you believe there is a case for switching to 3D drafting or do you think that 2D CAD will remain the predominant form of CAD for the foreseeable future?

«1

Comments

  • There's no such thing as 3D drafting - what you do in 3D is construct a model.

    A 3D model is near useless for conveying construction or manufacturing instructions, except as a backup illustration - you can't rely on the builder or toolmaker/engineer to deduce from the 3D model all that you hope he will, unless supplemented with written spec/notes which draw attention to all the things and relationships in the model that the designer intends as important. And if you have a written spec, why not embellish it with visuals created/annotated to clarify the point? In other words, a 2D (drafted) drawing.

  • Michael Mayer
    edited June 2020

    Hmmh,
    I never did much 2D.
    I started CAD with Microstation, 3D only, from the beginning many years ago.

    I think today the majority of worldwide buildings are planned in Revit or Archicad,
    modeled as 3D BIM models and 2D Plans are generated.

    2D predominant ?
    I don't see this here for middle sized commercial projects.
    There may be situations where 2D CAD may be still more effective but I don't see
    much future for 2D only CAD.

  • Anthony Apostolaros
    edited June 2020

    I've always assumed that eventually we'll all work exclusively in 3D -- first with 2D plans generated from a fully-detailed and accurate 3D model, and later by just giving that 3D model to the construction crew. I assumed they'd have computers at the construction site instead of giant stacks of paper, and could zoom in on whatever area they're working on at the moment, to see the detail, to take whatever measurements they need, and to right-click on any material to see its specification and quantity.

    Twenty years ago, when I started modelling all my projects in 3D in Sketchup, I predicted that the transition would be complete within 20 years. It is for some people, at least the first phase, but not for me. I'm still working exactly the same way I was back then. I still model everything in Sketchup, but then create 2D drawings in Bricscad, aided by 2D exports from the 3D model. I still think 3D-only is the future, but now I don't expect it to happen soon enough for me to be part of it.

    I suspect it's architects and engineers who are dragging their feet on that second phase, and maybe building officials are too. I'm pretty sure construction crews would go for it -- they're already working from PDF files of my 2D drawings, viewing them on laptops and tablets carried around the site, rather than working from the old giant stacks of paper. They have to keep a stack in the trailer, of course, because that's where the building official's approval stamp is.

  • That 3D will replace 'drawings' is more developers' ideology than practitioners' reality; that 'singularity' will never arrive because the 3D model alone, even with added data structure, remains impractically insufficient. Sure, 'drawing' will evolve massively, is already, and will merge into ever more bi-directional integration with the 3D model/data structure.

    https://tangerinefocus.com/about , by the author of Microstation's pioneering Hypermodeling (infusion of 2D into the 3D model), Is on the cutting edge of this question.

  • to "switch to 3d cad" is kind of confusing, as both autocad and bricscad are 3d by default.
    If you mean "switch to drawing everything in 3d", that is confusing too, as some things like an electrical schematic would not be useful if one simply drew the physical transistors the chip will have. Many things have diagrams you just would not do in 3d.
    In fact, you can argue the whole civil engineering industry is 2d diagrams, as we separate our parameters for plan and profile view.
    You can display the things we design in 3d, but fundamentally the plan and profile geometry are separate.
    Then you get into "switch to doing an xIM model" where the X is B for building, and so on, but now the fun part.
    To switch to some kind of encapsulated model, like you do for walls and doors in BIM, there must be software that allows you to do that.
    The architecture and mechanical worlds seem to have done pretty well on that. The civil world is sorely behind, as we store what I would call backbone info (alignments), and then shell info (surfaces), and can even build that into full 3d things that look like our design.
    Its not encapsulated though, into objects. Also, the data sharing mechanisms are fragile and heavy for things like civil3d.
    So some industries do not have the option to just jump to doing objects.
    Assuming you do though, such a switch is very dependent on what a model would solve. In civil, 3d things help tremendously, and I have managed the production of our site utilities for several construction jobs where they conflict detect with navisworks.
    We do all our modeling in basic acad or bcad, our tools work in both, and are company written. No civil3d needed.
    It only took 20 years to evolve, and it all started from just wanting better labeling tools for Land Desktop.
    So for industries where there is no obvious choice for true modeling, you are simply adding the things drawn in 3d to your 2d diagrams.
    It would not be a switch to 3d cad, but the addition of displaying your items in 3d.

  • Steven_g
    edited June 2020

    Even so called 3D models are constructed from 2D geometry, how many 3D modellers will directly place a 3D cube using just two sets of 3D coordinates. Not many, usually a 3D cube is created by first drawing a 2D rectangle and then extruding that rectangle to a thickness (again a linear treatment). That is an oversimplified view but that is how 3D is approached, 3D is still just a way to visualize 2D data and a 2D thought process. When programs start designing buildings using actual bricks then we might start seeing improvements. But when I am 'taking off' quantities from a 3D model I need to reduce the 3D walls to 2D surface area and then work out how many bricks there are in a square meter to arrive at a usable quantity. Imagine making a drawing enquiry and having it list how many bricks are needed for a project along with the number and sizes of 'cut' bricks around each opening, now that would be 3D CAD.

  • @Steven_g
    That same thing happens in civil, where you don't count individual bricks, you get the wall area and use that.
    What I am thinking of, is how pipeline designers lay things out essentially the same as a road.
    You choose the horizontal location, which is a bunch of lines and arcs.
    Then you do the profile, which will say follow the ground surface but 4' below it. Then you put bends in to go above or below crossing utilities to miss them. The end result is the bends that go up or down, do not line up with the horizontal bends.
    You don't "break up" the plan segments by the profile bends, and vice-versa.
    Just like you don't modify a wall by moving 10 bricks, you move or resize as if its one big panel with two params - width and height.
    What programs like civil3d and others currently do with pipes is break them up into "parts" wherever a change occurs, horizontal or vertical.
    They wanted to get parts that were only straight pipes, or arcs. It destroyed the design creation and editing abilities.
    Luckily, you can use the road design parts of the program to design your pipe invert.
    I point all this out as the wrong "3d object" design, can make it useless to real professionals if done even slightly wrong.
    The crazy thing is its easier to do it right, but big companies do not have the money nor confidence to redo things once done wrong.
    They instead try to cover it with marketing but only the people that don't understand the issues fall for it.
    Funny how that tends to be the other people at the same CAD company that made the tools. They all cheer for each other as the boat sinks.

  • I imagine a small detail, eg a flashing or even an invisible detail such as the application of a waterproofing paint. Even if it is somehow included in the 3D model, the builder may never see it unless it is called up in a 2D detail drawing, or in the specification. The 2D drawing is an instruction. Drafting conventions like linetypes and hatch patterns are meaningful. The 3D model is very helpful, but 2D drawings can be held up in court. When the owner sues the builder for water damage, you're gonna need that drawing.

    The 3D model can be used by the architect to create valid, consistent 2D drawings. Users working purely in 2D have to keep a 3D model in their head and that can lead to mistakes. The 3D model is fantastic for takeoffs, coordination between disciplines (trades), sun studies, client liaison - lots of things! But 2D drawings won't go away.

  • I shall confess I made this a deliberately "loaded" questions based on my experience on the forums. It's nice to see you guys did not disappoint! :)

    I had a similar conversation with a colleague and wanted to see if other people reache the same conclusion as us.

  • Tom Foster
    edited June 2020

    Nicely put, Damian.

    3D models would be even more useful, for building purposes anyway, if equiped to easily reveal what's happening in muliti-layer situations, often including layers of 'near zero' thickness. This is something that a 3D model can do far better than 2D drawings.

    I'm thinking of the top corner of a window opening, maybe complicated by long things like beams. There are multiple layers of stuff one on top of another, and the most crucial of all is the way the membrane waterproofing the horizontal top of the opening, laps down over the membrane of the vertical side of the opening. This can be complicated by other layers of structure, cladding, insulation, lining, which differ considerably between the horizontal and the vertical case. All have to be butted/lapped/jointed together just right, or else is a prime trouble spot.

    2D drawings really struggle to show this 3D intersection; the industry relies on siteworkers' custom and experience to get this undocumented situation right. A 3D model can faithfully illustrate it, but only if it's easy to digitally 'peel back' layers to see how what lies hidden 'under the (multiple) skin(s)'. 3D CAD has no such tools to make this easy, instead relying on the siteworker's diligence to guess that there's something vital to be viewed and understood somewhere unknown beneath the glossy surface.

    '2D drawing' can include quasi-3D isometrics, and to be able to set up such diagrams from the 3D model, by easily 'peeling back' layers and adjusting same for graphic legibility, would be invaluable.

  • @rose_barfield
    what was the conclusion you and friend came to?

  • @James Maeding I'm trying not to lead the discusion here. ;)

  • @rose_barfield
    Ha ha, don't worry, you won't sway things. Each of us are specific to our experiences and industry.
    The good people have realized the more they know, the taller and funner the mountain gets. So much to automate, so little time...

  • My friend suggested that regardless of how good the 3D model was, you still needed to be able to create high-quality 2D drawings. He also thought 3D was more useful for people that had trouble visualizing 3D from 2D drawings.

    Personally, I think that with the rise in digital purchasing, 3D modeling will become more important, not for production, but for sales. I also agreed that it can help sell an idea to someone who is not used to visualising 3D objects from 2D drawings.

    I also suggested that, in some instances, for example, technical drawing, it can be more time effective to draw in 3D initially and the export 2D drawings than spending time creating 2D drawings, especially if it is likely that many changes will need to be made to the drawings at a later date. I myself have spent many an hour drawing something from one angle, only to go back to it months later with a request from the author, to draw it from a top view instead. In these instances, it is certainly more time effective to draw something in 3D once and then quickly export and clean up the drawing in 2D.

  • ^ I agree with all above.

  • I'm in the pressure vessel/steel fabricating field and we're for the most part 2D.
    It's what our shop is used to but our capabilities are "different" than a lot of outfits
    in that we have no modern CNC, nesting equipment and etc. 3D comes into play
    when we get something "weird" then I'll do a 3D layout for the shop to get them
    past the sticking point. Brics has made this so much easier with each new release.

  • For our architecture and interior design practice, it is indispensable to have both 3D models and 2D drawings, including plans, sections, elevations, etc., of course, but also axonometric and perspective projections of a 3D model, for all the reasons already stated. We shift back and forth non-linearly between 2D and 3D CAD. The more fluidly we can translate between the two the better, so I appreciate how BricsCAD facilitates having both 2D and 3D (plus BIM!) in one .dwg format.

    As an example, just this morning, I extruded the elements relating to a custom stair railing and casework into 3D from our 2D construction drawings. Most of this project had been drafted in 2D only, although we had done an earlier 3D model in blender for rendering. Having that model in a separate format made it inconvenient to keep it up-to-date as the project advanced. It had been sufficient to develop the design in 2D, up to a point, until it became important to study certain details further in 3D.

    In other projects, we have taken advantage of BricsCAD BIM to generate 2D drawings from 3D models. So far, however, we always reach a point where we break the associativity to the 3D model due to limitations (whether of our own or of the software) in getting the output to have the appearance we want and to convey all the information necessary. As our skills and tools improve, we are able to postpone little by little when that break point occurs.

  • @rose_barfield
    This whole 3d thing is just starting too. Seeing 3d on a flat screen is one thing, and a few do the stereo viewers, but when the wearable visor hardware is worked out, that will be a whole new way of communicating.
    For an interesting 2d/3d use example, look up https://revizto.com/en/
    These guys have a 3d model viewer, that lets you orient 2d drawings in space so you can see both.
    Its like a 2d/3d catalog you can use to discuss and comment on to coordinate construction.
    Navisworks is one thing, but revitzo seems to bring it to the next level, and is VR ready. I have no connection to them.
    thx

  • Hans Lammerts
    edited June 2020

    @rose_barfield said:
    Hi, I'm currently researching the topic of 2D versus 3D CAD. I'd be interested in hearing people's thoughts on this topic. Do you believe there is a case for switching to 3D drafting or do you think that 2D CAD will remain the predominant form of CAD for the foreseeable future?

    Hi Rose, without going to through all of the reactions sofar .. i just wanted to share some very intresting thoughts and writings of a friend here. You should look at drawings and models as different 'media' serving different purposes. No, 2D geometry will not go away! Hell no!

    One simple math equivalent... Adding Z to X&Y will not make X&Y go away!

    Any drafter or modeller, using [ fill in your 2D 3D software of choice] will tell you you need to get the base geometry in 1 plane right first. Call it XY, XZ or YZ. All without Z are 2D focussing techniques.

    Enjoy this read. This many other great writings of Rob Snyder,
    Fusion, Hypermodelling, connection worlds.. thats what it is all about using computers in a democratic world.

    https://tangerinefocus.com/blog-tangerine/

  • James Maeding
    edited June 2020

    @Hans Lammerts
    Keep in mind too, that civil engineers use parameters for plan and profile that you could call "2d".
    We separate horizontal and vertical. It should not be done any other way, its the simplest it can get similar to how a rectangle is described by length and width.
    The civil software makers have forgotten this, and are trying to make tools to model pipelines and conduit as "parts", where each end has both a horizontal and vertical "hinge point". Its a mess, as the more it fails the harder it gets marketed. It will never be right though, and lets people like me that write their own solution just keep picking the low fruit with no competition. Its all about understanding parameters, and many times those are best expressed in a 2d plane...

  • Anthony Apostolaros
    edited June 2020

    I just read something that seems to confirm the idea that it's not builders who are resisting the move to 3D. It's about a company called Constructability3D. Working for construction companies, they take all the 2D drawings of a project and make a complete 3D model of the building, including structural members and MEP work. The idea seems to be to avoid the kinds of problems that can arise from designing using only 2D drawings.

  • @Anthony Apostolaros
    That is true for the higher end construction companies. Its a big mix though, on software.
    You have the architects and HVAC/Fire guys that use true BIM software, then the civils are left to attempt whatever they can find.
    Some try civil3d, but quickly learn its a nightmere for pipes. Then they seem to try revit, and hate that too. I watch the pipe/conduit teams delay certain revisions I can do on the spot with our pipe/conduit tools.
    Its because we did not combine the plan and profile params, and it turns out the editing of those is simple if you keep them separate.
    Its not BIM though, we have parametric 3d modeling tools that update as the design (alignments) change.
    Anyway, its quite an experience to see the models that develop, and the games everyone plays now that you can revise things and catch conflicts reliably.

  • Hans Lammerts
    edited June 2020
    What a waist of energy. Turning 2d to 3d and disposing it? What's the point? I will only make things 3d if it gives some more insight of means to accomplish something. I see more people burning and crashcanning valuable 2d information. All in sake of some higher goal named 'BIM'. A big empty mainingless container full of nothing.

    Working hybrid, work with all kinds of media, much smarter..
  • @Hans Lammerts
    fur sure, no point making 3d stuff just for fun. But if its REALLY fun, then... :)
    What happens in this construction coordination is they check for conflicts, revise, check again, and so on until things conflict physically and then you punch holes through pipes and seal with duct tape. The leaks irrigate the plants a bit.
    You can imagine we have got along without 3d conflict detection forever, so people are really good at watching for critical areas.
    If you have decent tools though, you almost get the 3d utils for free on larger stuff (8" dia+).
    Then the smaller is only modeled at construction time.
    Is it a waste? No, they pay me for it, what else could you ask?
    Yah, its way under utilized. Our industry is so behind where it could be, and much of it is due to how Autodesk runs (or ran).
    The big guys have lost steam, it's all up to little places now to make decent pipeline tools.

  • There are many industries for which 2D drafting is the best choice. The main reason is that a drawing is actually an abstraction of reality.

    For example, in a plan view of a refrigeration chamber, I need to indicate where each panel is located. They may be connected to each other using a cam-lock. But, I drawing a cam-lock where they go would not be a good idea. I need to show which side is male and which side is female. So, I use a simple shape looking like a >.

    If I did this in 3D and showed all the features as they are in real life, my cam-lock would show up as a square in dashed lines. If you zoomed-in and looked very close, you could see the gender. But, the chamber is never drawn large enough to make this viable on a piece of paper.

    Much of architecture is like this. You don't normally want to see the all the internal structure of a wall, except for a simplified version of an opening or window. Symbols are used for electrical outlets. Not realistic views of electrical outlets.

    Where 3D makes the most sense to me, is in machine design. There there are physical objects that must move, and clear each other. They all exist in multiple elevations and depths, so a simplified orthographic view may not be sufficient to make the design clear. Though, one conveyor line manufacturer, that had various types of machines on the line, would do its work in 2D. This is because the machinery was already designed separately. The front and side views, with very simplified views of the machinery was easier to understand than a realistic 3D representation would have been.

    One final benefit of 2D, is that it is often faster than a 3D model. Though, if you need an isometric view of an object, then 3D suddenly becomes a major advantage.

    -Joe

  • Very good points, but I disagree a bit regarding 2D being faster than 3D. I think that boils down to project complexity.

    Mechanical design is an application that certainly benefits from 3D models.

    My experiences are on the more specialized side of architecture; custom cabinet building. My job is to create a set of drawings that our building team can actually use to construct the cabinets because the original drawings supplied by the architect do not provide the level of detail required to build cabinets.

    However, before drawings get to our building team, they still have to be approved by the architect. Those drawings often require re-design several times before they are approved. In this case, a 3D model of the cabinetry works very well. That model can be quickly re-designed and all layouts are updated immediately. With 2D drawings, I'd have to individually alter each view and detail. Not so bad for a single bathroom vanity, but if your project is cabinetry for an entire vineyard tasting room or total renovation of a hotel lobby, spending time to build a 3D models is well worth it.

    -B

  • Michael Mayer
    edited June 2020

    @Joe Dunfee said:
    There are many industries for which 2D drafting is the best choice. The main reason is that a drawing is actually an abstraction of reality.

    If I did this in 3D and showed all the features as they are in real life, my cam-lock would show up as a square in dashed lines. If you zoomed-in and looked very close, you could see the gender. But, the chamber is never drawn large enough to make this viable on a piece of paper.

    -Joe

    I also think that we still need 2D Plans.
    But that doesn't mean that working exclusively in 2D, manually drawing plans, would be a good idea.

    I think it is a better idea to model in 3D a (B)IM Model and let the Software generate the
    2D drawings. Where all plans automatically get updated with changes.

    But, I drawing a cam-lock where they go would not be a good idea.
    I need to show which side is male and which side is female. So, I use a simple shape looking like a >.

    So it is up to the Software to generate the drawing in a useful way.
    Of course it can't be just a dumb Section through all Objects like it may work for simple Walls.
    There is more complexity or intelligence needed.
    Just like in Architecture where Openings in Walls, above Section Line, should appear in dashed lines
    and Openings below Section Line should show up in dotted lines.
    Also we need different complex appearance according to Scale.

    So maybe certain important 3D Objects may need to include 2D representations that show up in
    Plans - instead of a dumb Section through the geometry only.
    And in different Detail Levels according to Scale.
    A Window Component could include Plan Details that show up at Detail scale levels.
    (We already have an extra 2D top plan representation in Bricscad's Components)

    So the 3D Model's geometry does not need to be too detailed. But it can provide that 2D Information
    when needed.
    In 3D, where you may not be able to recognize all Details, intelligent Components could show an
    Info Button when selected, which could open a Label to show all its specifications and even
    pop up detail views.

    That may not all be there today.
    I still think it is better to work in 3D and instead try to generate as much 2D as possible automatically.
    And as long as currently needed, to fill the missing parts manually in Annotation Space or
    Viewport Overlays.

  • I started with Computervision CADDS 4X many years ago, and the company I worked for rarely used the full implementation of 3D Drawings, until a number of our products were brought in house, and we had to work with Computer Controlled CNC Multi-Axis Systems for working with metal molded products, then 3D point output was necessary. Now that I am in an entirely different business, 3D is used for the Fire Sprinkler business efforts, but the overall layout drawings are still generated for 2D for both Authority and installers use, with minor 3D for some details.

  • Joe Dunfee
    edited June 2020

    @Michael Mayer said:
    I think it is a better idea to model in 3D a (B)IM Model and let the Software generate the
    2D drawings. Where all plans automatically get updated with changes.

    Years ago I tried Vectorworks on Mac. As you inserted things like windows, the blocks had two representations, a 3D and a 2D one. You could toggle between the two display modes. Definitely a nice idea. Though, it caused a lot of confusion when trying to transfer a Vectorworks file to AutoCAD. I forgot how they managed elevation view, vs. plan view.

    I wish there were comparable options, not just for 2D vs. 3D, but even in pure 3D, if you had a choice between a simplified 3D vs a complex 3D model. Things like fasteners could show threads and all the little features of the head of a bolt, but then switch to a very simple model, that wouldn't eat up a lot of processor time and overwhelm your computer.

    -Joe

  • I commonly see people confuse the term 3D with BIM, and especially in the construction business.
    Autodesk encouraged the practice, because they wanted to call civil3d a BIM program, and they are still edging that way.
    People that know Civil3d and other programs do not have the parametric's going on to model what we need have fought it.
    Its interesting because its not just having software that lets you model your industries thing, but then it must also throw back useful 2d drawings since the contract world and my grampa still rely on 2d drawings. That is not easy, but the term BIM is a really high apple on the programming tree so no shame in software that helps but is still not BIM.