http://disqus.com/by/bcourter/enMon, 09 Dec 2019 12:30:47 -0000https://3dprint.com/260667/duann-scott-on-ntopologys-programming-language-for-geometries/#comment-4718221591<p>Was a pleasure to meet you at FormNext, Joris. Thanks for the lovely write-up. :P</p>bcourterMon, 09 Dec 2019 12:30:47 -0000https://www.fabbaloo.com/blog/2019/2/25/at-last-the-end-of-stl-is-in-sight#comment-4354325513<p>Check out Stratasys' new Advanced FDM which takes multi body CAD parts as input. Also 3DS 3DExpert. Should probably mention our products, Element and nTop, although not straight up slicers, too.</p>bcourterMon, 25 Feb 2019 21:36:50 -0000https://blog.grabcad.com/blog/2015/09/18/standard-tessellation-language/#comment-2286554340<p>Good points. We had similar thoughts when we first joined Stratasys.</p><p>It's not hard to slice a B-rep to produce precise curves. There are some wrinkles, such as handling non-manifold cases, joining up points that don't match due to the small gaps endemic to B-reps with non-analytic geometry, but there are known approaches to solving these issues. The problem is that once you have your nice compact curves, you still have to tessellate them into polylines for the printers. One benefit of the STL files is that they are unambiguous, as the CAD system that created them was responsible for deciding how to deal with issues like its gaps.</p><p>Sure, in CAM we've had arc interpolation and spline interpolation for decades, but interpolation isn't on the critical path to print performance. For extrusion-based printers, the bottleneck is usually the liquifier and extruder.</p><p>There's a lot we can do to improve the experience. Working within CAD systems and removing the need to save with inscrutable tessellation parameters is a good place to start. Here an example of progress: <a href="http://www.ptc.com/about/news-room/press-releases/2015/ptc-and-stratasys-collaborate" rel="nofollow noopener" target="_blank" title="http://www.ptc.com/about/news-room/press-releases/2015/ptc-and-stratasys-collaborate">http://www.ptc.com/about/ne...</a></p><p>(Full disclosure: I work for the software division of Stratasys.)</p>bcourterFri, 02 Oct 2015 15:01:03 -0000https://blog.grabcad.com/blog/2015/08/05/3d-printing-complex-geometry/#comment-2177548967<p>For FDM and other extrusion based technologies, the Z aliasing is something you have to manage. When I have two parts that need to slice next to each other, I print them at orientations so that when they slide, it doesn't sound like we are scratching a record. Some prep software, such as Insight, can help you orient the model to mitigate the coarser finish.</p><p>If you really care about surface finish, there are better printers for the job. PolyJet is my go-to choice, as the surface tension does a very good job with the aliasing, especially with the highest quality settings. If you don't mind a matte finish, SLS also provides decent results.</p>bcourterThu, 06 Aug 2015 09:29:53 -0000https://blog.grabcad.com/blog/2014/08/19/engineering-drawings-are-dead/#comment-1552993424<p>Last I checked, the main problem with PMI is that the standard wasn't 100% model based: some GTOLs were still view dependent. Although I understand that views are an important part of PMI for model checking and inspection, not having the PMI be 100% 3D seems like a lost opportunity. Has the situation improved?</p>bcourterThu, 21 Aug 2014 11:40:23 -0000https://designandmotion.net/new-post/grabcad-workbench-just-got-a-lot-smarter/#comment-1131810008<p>Thanks for the excellent review, Gavin. There's one subtlety of Partner Spaces that becomes more important when you are using Workbench for daily file management: when you upload a new version of a file, you get to control when the files get revved in your partner space. That way you can proceed with a design and not worry about your customers or suppliers getting confused.</p>bcourterWed, 20 Nov 2013 16:57:02 -0000http://www.lifecycleinsights.com/technology-insights/concurrent-product-development/#comment-1101454782<p>The external relationships of feature-based CAD are very useful when managed properly, but they can also create havoc. If the part you are designing has external references to a part that is undergoing changes, every time you get the latest version of that part your part will be out of date. Each one of those changes introduces the possibility of rebuild errors and more work.</p><p>The best approach for large design teams working concurrently is to order the engineers' decisions so the interfaces between the teams become fixed early. The most sophisticated companies organize their teams according to the order in which then need to make such decisions.</p><p>To put it another way, history-based CAD automates the process of making a change. That change still has a cost attached. If the rebuild caused by a change succeed, the cost is minimal, but it is higher than if no change were necessary. If there are rebuild errors, the cost can be very high. Therefore the engineering process should strive to minimize the number of required changes, and to keep them as early in the process as possible.</p>bcourterTue, 29 Oct 2013 17:53:48 -0000http://www.lifecycleinsights.com/technology-insights/concurrent-product-development/#comment-1101299450<p>I think there was a period of time where the industry hoped to make associativity good enough that it could be used for concurrency for downstream deliverables, like drawings and tool paths. However, I don't know of any product that does it well enough today to be workable for the initial design. Associativity does, however, do a reasonably good job of automating changes after an ECO in a serial process.</p><p>That said, concurrency makes a ton of sense for a large engineering team during the design phase. The team will need to use the tools correctly, and PDM is likely to play an important role in coordinating between the users. Ironically, most users who succeed in introducing concurrency in the design phase minimize their use of associativity across users and teams.</p>bcourterTue, 29 Oct 2013 15:55:40 -0000http://www.lifecycleinsights.com/technology-insights/concurrent-product-development/#comment-1101143885<p>The main difficulty I regularly hear about from GrabCAD customers is that it doesn't make sense to start the drawing until you are done with the part. Associativity only succeeds with unrealistically minor changes, and even if all of the dimensions' and annotations' references remain intact, it is likely they would need to be re-positioned after a change. The drawings are also rarely needed until it becomes time for a quote, so there is no reason to touch them.</p><p>In lean parlance, the concept of the last responsible moment prevails over unnecessary concurrency. I suspect that engineers learned that by measuring results and focused on ways of working in parallel that don't introduce waste.</p>bcourterTue, 29 Oct 2013 14:04:07 -0000http://deelip.com/grabcad-workbench-first-impressions#comment-868459172<p>Hey Deelip,</p><p>Measure tools are in! Please keep the great feedback coming.</p><p>Best,<br>-Bake</p>bcourterFri, 19 Apr 2013 14:09:20 -0000http://deelip.com/grabcad-workbench-first-impressions#comment-866204142<p>Hey Deelip,</p><p>Thanks for the great coverage. We're pretty excited too. Your observations are astute, as always. Measuring, for example, will magically appear in the very near future.</p><p>-Blake</p><p>(A product manager at GrabCAD)</p>bcourterWed, 17 Apr 2013 13:42:44 -0000http://www.engineering.com/Videos/VideoPlayer/tabid/4627/VideoId/3044/Space-Claim.aspx#comment-706416676<p>Cary, I would agree that all CAD users should also have access to direct modeling, especially for working on imported models. The problem is that to use feature-based modeling well, you really have to apply yourself to it. There are many people who want to use 3D without all of the overhead and complexity, and only pure direct modeling can accomplish that. At SpaceClaim, we serve advanced engineering groups, simulation users, machinists, sales engineers and other folks who want to get results in 3D without the overhead of features, constraints, parent-child relationships, and rebuild errors.</p><p>-Blake (a SpaceClaim employee)</p>bcourterFri, 09 Nov 2012 14:15:46 -0000http://deelip.com/the-world-according-to-spaceclaim#comment-679179597<p>I love it when I catch my typos once they've been picked up in blog posts. Apologies, everybody.</p><p>RE PTC, I'd like to be clear that I have great respect for PTC, and I consider many there to be old friends. It did make me deeply saddened to see them retire the Pro/Engineer brand, only to obfuscate their product line with absurdly similar names. I prefer to see vendors, especially engineering software vendors, communicate with clarity and precision. Typos notwithstanding, that's what we try to do at SpaceClaim.</p><p>-Blake</p>bcourterThu, 11 Oct 2012 14:00:46 -0000http://www.3dcadworld.com/parameterizing-cae-models-with-comsol/#comment-550111303<p>With direct modeling, it is now no longer necessary to put the right parameters in the model when you build it.  SpaceClaim, for example, can add parameters post-facto to dumb, imported geometry from any CAD system.  CAE tools, including COMSOL, can use those parameters to optimize geometry.</p><p>-Blake<br>(A SpaceClaim employee)</p>bcourterThu, 07 Jun 2012 14:44:39 -0000http://prosapr.blogspot.com/2012/06/sketchup-revit.html#comment-548881074<p>We see SpaceClaim getting used to clean up SketchUp files. In our last release, we added some capabilities to do this automatically, which works pretty well on planes, cylinders, cones, and splines surfaces curved in one direction. For models like this, the best way to proceed is to fit splines to the major curves and re-loft.</p><p>-Blake (who works for SpaceClaim)</p>bcourterWed, 06 Jun 2012 12:08:32 -0000https://www.solidsmack.com/3d-cad-technology/wearable-multi-touch-makes-your-shoulder-look-dope/#comment-464620004<p>Everything looks better with EL wire.</p>bcourterTue, 13 Mar 2012 18:49:04 -0000https://www.solidsmack.com/cad-design-news/pirate-bay-introduces-physical-3d-printed-object-download/#comment-420824822<p>There's a place for 3D printing, there's a place for kevlar and epoxy, and there's a place for duct tape.</p><p>I'm trying to figure out whether I want to encourage our AEs to upload models there or not.  In the meantime, I'm going with GrabCAD.</p>bcourterWed, 25 Jan 2012 15:22:05 -0000http://www.evanyares.com/software-toolsets-and-golf-clubs/#comment-382508717<p>Fun article, Evan.</p><p>Although it might be career-limiting to say it, I am not a golfer.</p><p>I assume that in order to get a golf club to sell, it must perform well. I suspect that look, feel, brand, and celebrity endorsements are ultimately more important, although some of those factors are likely derived from performance. My marketing brain suspects that fancy-sounding technology sound bites can greatly increase profitability. Did the Nike Air really cause basketball players to jump higher? Was it a brilliant product/market fit?</p><p>But to answer your question, if I wanted to compete in a high-performance market, I would want to have the best team I could afford using the best tools for the job. CAE would come first. Ideally, CFD would drive the overall geometry shape directly, so the resulting surface geometry would drive the overall shape of the head. Specifically, the hard constraints such as the angle of the striking face, approximate volume, etc., and resulting surface geometry with low drag would result. Some mesh-to-surface workflow might be helped through a reverse engineering tool such as Geomagic. If CFD-generated surfaces aren't practicable, then you're looking and a multivariate design optimization on surface control points. You might be able to get away with that in a history-based modeler, but I would probably turn to an aerodynamic surface heavy-lifter like Multi-Surf.</p><p>Once you have the CFD geometry optimized, you're looking at a three pronged problem. ID, materials, and FEA converge. Let's back off of ID for a moment. Although the outer surfaces are fixed, the internal geometry probably has many degrees of freedom. It seems to me that the next step is to wire the geometry to the simulations using an MDAO tool like ANSYS Workbench, Esteco Modefrontier, Noesis Optimus, or perhaps even manage everything from the top down in Comet. For geometry, I'm partial to direct modeling, so I can dimension and optimize one model many different ways through various simulations without having to create many different constraint regimes to juggle my different simultaneous modeling intentions. To my knowledge, SpaceClaim* is the best product for the geometry part of the equation once the tricky surfaces are defined.</p><p>Along the way, I would intimately involve industrial designers. Assuming that the geometry is mostly generated from engineering, I would have them refining the styling and producing deliverables I could use to sell the idea to potential customers and partners. For that, a Rhino or Modo would probably be the best fit, along with a really fast renderer like Keyshot so they can quickly work up different color and style combinations.</p><p>But I am not an expert in precisely what tools are best, which is why I would surround myself with the best talent and let them use the tools they need. 1-2 engineers and a great designer or outsourced design firm should do it. What I would not do is force everybody do use a big box store CAD system that has a long laundry list of features but isn't excellent at anything in particular. In fact, I probably wouldn't bother documenting the design more than necessary to get it mass produced, which is surprisingly little nowadays.</p><p>The process I outlined isn't something I invented to answer your question. I'm extrapolating some workflows my customers have shown me. Although I'm still no expert engineer and designer, this is how I've seen it done by the best.</p><p>Thanks,<br>-Blake</p><p>* I'm an employee of SpaceClaim Corporation<br></p>bcourterThu, 08 Dec 2011 19:15:30 -0000http://thephoenix.com/blogs/817353.aspx#comment-355034271<p>Any chance of a re-post someplace where it can still be downloaded?</p>bcourterThu, 03 Nov 2011 16:09:51 -0000https://www.solidsmack.com/design/cnc-zonahedral-structures-zome-rob-bell/#comment-338331959<p>Those zomes were my favorite structures at Burning Man this year.  I'm kind of biased, though, because I've been on a rant about how standard geodesic domes are not optimal for gravitationally-challenged environments such as aerialist performance. (<a href="http://www.coroflot.com/bcourter/Autosub-Dome/)" rel="nofollow noopener" target="_blank" title="http://www.coroflot.com/bcourter/Autosub-Dome/)">http://www.coroflot.com/bco...</a></p><p>Hope I am fortunate enough to run into Rob someday...</p>bcourterTue, 18 Oct 2011 13:29:21 -0000http://www.evanyares.com/converting-domain-experts-to-cad-users/#comment-324068536<p>We see a lot of customers devise a system that lets organizations maximize the value of CAD for detailed design but still let the rest of the organization communicate in 3D. The model we repeatedly see is for the detailed design team to be focused on CAD, typically using a feature-based modeler, and for the rest of the engineers, including program managers, physical architects, simulation experts, sales engineers, and manufacturing engineers use direct modeling. It would argue that it's detailed design itself, which also uses CAD, that requires specialization, and that it's best if that team agree on stringent rules of how things will be done in CAD.</p><p>But does that mean that no one else in the organization should be allow to scratch out ideas and answer questions in 3D? Of course not!</p><p>Define "CAD" however you like, but get everybody talking in 3D.</p><p>-Blake<br>(a SpaceClaim employee)</p>bcourterFri, 30 Sep 2011 17:11:27 -0000https://www.solidsmack.com/3d-cad-technology/windows-8-product-design-cad-industry/#comment-313197583<p>It's interesting to see how, for any new release of software, some users get excited and others worry.  Feature-based CAD users seem especially concerned about stability.  ("How long must I wait for a reliable service pack?")</p><p>I think I'm going to start my press briefings for new release by saying "You really don't have to worry about this release because..."heh.</p>bcourterFri, 16 Sep 2011 10:05:17 -0000http://www.solidsmack.com/3d-cad-technology/bitcasa-infinite-cloud-storage/#comment-308751032<p>Randall won this time...</p>bcourterTue, 13 Sep 2011 13:47:27 -0000http://www.solidsmack.com/3d-cad-technology/bitcasa-infinite-cloud-storage/#comment-308686835<p>And you didn't even use your special link to get pushed up the the queue. Exercise your blogger privileges!   </p>bcourterTue, 13 Sep 2011 11:58:45 -0000http://deelip.com/spaceclaim-2011-2#comment-308094985<p>Hey there,</p><p>Any chance you could send a copy of that quote to blake (at) spaceclaim (dot) come?  It seems high, although it may be possible to configure it that way with all the options.  </p><p>Also, we now have  mirrored assemblies.  By the way, have you seen the sheet metal lately?</p><p>-Blake</p>bcourterMon, 12 Sep 2011 17:17:36 -0000