Once upon a time we shoved pounds of paper into binders and carted them off to the conference rooms for design reviews, sifting through them under bad lighting and a hazy caffeine buzz from coffee that was clearly past the expiration date.
Things improved slightly when we realized our models could be smashed into .zip files and shot across the internet within a few minutes. This was so common in fact that SolidWorks built the ability to zip files up right into the software, keeping the file structure and references intact. Pack and Go. Most often, I use SolidWorks Pack and Go to duplicate models, but a lot of the time it’s how I ‘collaborate’ with other people and teams on a project. That way of working seems ancient now, so when I used Pack and Go again YESTERDAY, I thought, “I need to pack and go ON A VACATION,” then I thought, “There has got to be a smarter way to share these flippin’ files!”
So, here we are. We mindlessly Pack and Go assemblies. That’s how it’s done, it’s how we’ve always done it (a topic to kick in the face another time), it’s how others expect it — but then what?
Let’s attack the possibilities.
Email
Email’s cockroach-like resilience is how we get so much done. It’s also how we get absolutely nothing done. It’s been there for us when we need to send Pack and Go attachment after Pack and Go attachment. But here’s the problem. It’s not secure–not in the slightest. It has file size limits–email rejected. It has attachment limits–email rejected again. Then if the recipients get it, SolidWorks version incompatibility–”SolidWorks cannot open future version”, yay teamwork. And they don’t even have SolidWorks–Fu…crying out loud. And how do you track changes? Nope, there has got to be a better way.
Dropbox (or equivalent)
I am such a fan of Dropbox, it’s borderline psychosis. This (and Box) are how I’ve kept organized for nearly a decade. The simplicity, the flexibility and the access just add up to a great solution to store and backup life. But for SolidWorks models? That’s where I hit a wall. We’re back to a lot of the same problems seen with email. They need SolidWorks. They need a compatible version of SolidWorks. And suddenly, there are multiple copies of the files–one on your computer, one in the cloud, and another on the computer(s) of the people you sent it to. How do you track changes in that scenario? Outside larger storage limits, it’s the exact same problem.
Onshape
Onshape? How does a cloud-based 3D modeling system fit into this list? How about I show you?
It was dead simple for me to upload and share that with you. You can open and see the model right away in all its 3D glory. As long as the zip file name is the name of the top-level assembly (which is automatically done in SolidWorks), it’s ready to upload and view. File compatibility–not an issue. I now have better control of the data. I can revoke access, change access or include others AT ANY TIME. Then there’s version control and the ability to do direct editing on the model. If I need to collaborate, like really collaborate together with someone, there’s follow mode for real time meetings, even chat if I like. Mobile? Yeah, people can access, view and interact with the model on a mobile phone, an iPad or any computer with an internet connection (PC, Mac, Linux, Chromebook). I can get it to a designer for review, manufacturing for comment or vendor for quote. I can get it to a buyer earlier to lock in agreements and to marketing teams earlier for screenshots and media prep.
Ever since I was a kid, blowing Play-Doh through letter shaped cookie cutters, I’ve had an interest in typefaces. I like to think I’m the one who confronted my second-grade teacher about the font choice and kerning on that awful “Mister Seahorse’ bulletin board. I imagine Loren Kulesus, a Brooklyn-based designer, has had a similar lust for letters, and his Heatwave typeface *ahem* 3D PRINTABLE TYPEFACE is perfect proof.
After seeing the 3D printed typefaces of Thomas Wirtz and the incredible material effects in the video for his Master’s Thesis (see below), I was hoping he would be on of those cool people who shares the models.
He built each letter separately, but also created a series of three letter acronym prints. The channels in each were deep enough to fill with different material. Materials like colored dyes, oil, paint and yes, FIRE. I would have slammed some Jell-O in there and serve up some WTF treats to my co-workers. WINNING IT. The effects are simply stunning.
Unfortunately, he doesn’t share the flippin’ models, but fortunately I are an engineer, so I’m work on my own unique typeface. Looking for inspiration, I came across Loren’s set on Thingiverse. Called Heatwave, the letters are 2″ x 2″ and constructed with a series of 1 mm thick walls and series of channels. The height of each wall changes that, when lit just right, would look totally dope in your cubiclesuburban homestead downtown loft.
Honestly, I think Loren’s typeface is a much cooler design. More examples are below and you can see more of Loren’s work on his Instagram page, which includes a lot of abstract 3D and even more 3D printed typefaces.
You have to hand it to Autodesk, they wiggle their marketing nose and out pops some brand spankin’ new content. I caught their latest What is Fusion 360? video, where they show Fusion 360 doing collaboration, freeform modeling, animation, rendering, simulation, CAD and CAM, claiming, “it’s not any one of those things – It’s ALL of those things”, and so different from what I remember about Fusion 360. They say it’s the “next generation product innovation platform” for Mac, PC, and mobile devices. Catchy name, everything in one place, and the video has a funky beat I can bug out to, so I think it’s time we take a closer look at Fusion 360.
We’re doing a Fusion 360 series. We’ll dig to see if there’s enough depth to warrant the distinction of a fancy new product category. Three years ago, Carl Bass hinted at what Fusion 360 is today in our interview with him. That 1990’s CAD on the cloud isn’t enough. That we need a total solution, including 2D & 3D CAD, CAE, CAM and Data Management (phew!), all in one application. With all that, my assumption was features that are spread thin, basic at best. That’s where I was surprised. Each tool has deeper functionality than modeling apps I’ve used in the past, and there’s one thing in particular that completely blew me away. So, let’s break it down.
Part Modeling
This right here. This is what completely amazes me. It’s what I’ve wanted in 3D modeling. Now the key point here is this: if you use Fusion 360 the same way you used SolidWorks, Inventor, Creo, etc., you’re missing out on the advantages of having parametric, direct, surface, and freeform all in the same tool.
Oh, you didn’t know Fusion had all those types of modeling? Well, I did, and I’m still surprised. The freeform modeling is incredible, but let’s look even deeper. Let’s pick one feature within Fusion’s Sculpt workspace and explore for a moment. Obviously the Edit Form command is how you push / pull your way to form, but what about when you need to transition from one shape to another? The Bridge command creates segments to connect two opposing faces within a body or within two bodies. Once the bridge is created, hit Edit Form and continue defining your shape. When you’re ready to switch back over to parametric features like holes, bosses, or ribs, you simply switch workspaces. Now, here’s where I hit a wall – explaining to you how to do that in SolidWorks. I could do it, but it’s just going to take me much, much longer to do. This ability is an absolute time-saver in Fusion 360.
Top down or multi-body modeling isn’t a new concept, but it’s hardly been adopted as a best practice. Many of us model a part, slam it in an assembly and start vomiting shapes around it. While I enjoy a good purging, you can calm the model vomit and take full advantage of some new features in Fusion 360.
Fusion 360 has a feature called ‘As-Built Joints’ that (WARNING) may lead you to more top down modeling in the future. Don’t fear, that’s how it should be done Image may be NSFW. Clik here to view. When you design in context of an assembly (top down), the parts are already positioned in the assembly. As a result, you don’t need to “move them” into place. As-Built Joints simply enable the relative motion between the two components. This is a huge difference between constraints (or mates) and joints. With constraints you are removing degrees of freedom, with joints, you are enabling motion. Got it? So, with As-Built Joints you don’t have to take apart your assemblies in order to mate them back together, or strip out all the sketch relations. That’s a huge waste of time and so common in CAD today, be it desktop or cloud.
Oh, and you can use it on imported geometry too. Instead of adding mates to an imported models, fixed groups and As-Built Joints cut the time out of applying all those mates.
This one’s slightly outside my comfort level, but I know my first design is rarely the best design. Simulation. While not everyone needs it, it’s handy for sanity checks, but most software makes set-up and getting results a barrier to broader use. There are two really annoying things here; interferences between components and finding the right balance between mesh density, solve times, and simulation accuracy.
In the Simulation Workspace, there are built in mesh adaptation tools for parts and assemblies, which will completely ignore interference. You can set up contact points with a mix of opacity controls and degree-of-freedom view. Mesh refinement to check for convergence is reduced big-time, and as far as I can tell, zooming, rotating and zooming to find that interference is eliminated.
Now, if you do find something that needs to be adjusted in the part, you can switch over to the Model Workspace and use those direct edit tools we talked about earlier.
Some are fortunate enough to work closely with manufacturing, but this is where one design mistake can cost a lot of time and money. So, in Fusion 360 switch to the CAM Workspace. Under the 3D options there’s an Adaptive Clearing feature. Wow to the wow. Adaptive Clearing is a roughing strategy for clearing large quantities of material. Autodesk claims it reduces part roughing time by an average of 40%, reducing tool wear by half and virtually eliminating tool breaking. All I know is, I can set up different passes in exactly three clicks and run the scenarios past machining.
It takes a completely different approach to calculating roughing toolpaths. Instead of using a maximum stepover distance, Adaptive Clearing lets the you or the machinist specify the Optimal Load of the cutter. Then, Fusion 360 creates the roughing toolpaths that maintain a constant tool load when possible. That’s unique because it guarantees a maximum tool load at all stages of the machining cycle and ensure that wear is uniform across the length of the tool.
The fact that Fusion 360 is a cloud application is secondary to what it enables. First you have the tool capabilities, then you have the Mac and PC compatibility, and also the device/OS accessibility for design, view, markup, or comment. You never have to ask somebody which version of Fusion 360 they are on, which viewer they support, or which neutral file type they accept. You won’t have to mess with email, copies of models, or creating an account to just share a link to a file.
A lot have tried to liken design collaboration to social media, to Facebook or Twitter. You can share this post on either with a click of a button (and you should), but it’s not like that with 3D CAD apps. However, with ‘Share Public Link’ in Fusion 360, it is just as easy. Copy the link, click the option to allow download (f3d, Inventor, IGES, SAT, SMT, STEP, DWG, DXF, STL, or FBX). Copy link. That’s it.
Now we come to that fancy name. Product Innovation Platform. What do you think? Is there any merit to Autodesk suggesting that Fusion 360 warrants a discussion that it is more than 3D CAD in the cloud? That it’s an entirely new product development category? Perhaps, but it’s pointless if the feature set within each area (Part and Assembly Modeling, Visualization and Animation, CAE, CAM, and Data Management / Collaboration) isn’t complete enough to produce a product.
So over the next several weeks we’ll dive into each of these areas in order to better understand the entire toolset. I promise, I’ll resist the urge to try and model like I have in other applications. That’s gonna stretch me I’m sure, but it’s the free-form modeling, the joints, the sim, the cam, the collab and all the rest-all of it together-that makes Fusion 360 so powerful. What Autodesk has revealed is that it’s much more than a just CAD app or just a CAM app. That misses the mark completely.
When we first looked at the upcoming SolidWorks Apps for Kids, we had only a quick glimpse of what SolidWorks intends to shape our children’s minds with. But we just had to know more. Fortunately Chin-Loo Lama, Senior UX Design Engineer at SOLIDWORKS, has been all too kind to answer a few more questions we have about the apps.
At SolidWorks World 2016, SolidWorks revealed 6 apps to introduce kids, not just to 3D modeling, but a 3D modeling workflow, starting from ideation and concept and going through to design and manufacturing. The idea behind the apps is to involve children in every step of the product development process. We love it. Others love it. And we’re sure it’s going to be a hit with quite a few kiddos out there. Here’s what Chin-Loo had to say about features of the apps, how they will be accessed, their core technology and whether or not we’ll see more apps in the future.
Image may be NSFW. Clik here to view.SolidSmack: What age range are the Apps for Kids made for?
Chin-Loo: Apps for Kids are designed for children 4-14 years old or basically Kindergarten to 8th grade (in US school levels). Through our years of community engagement, we have learned that during the middle school years that this is time when kids start to form their expectations of what they feel will be their strengths and weaknesses and thus start to decide what they want to do in the future. So if they don’t have the self-confidence on their science and math skills, then they will likely never go down the path of STEM careers. We are hoping to open their eyes to the skills that they already have and help them see the connection between those skills and the skills needed to be in the STEM fields. In short, we want to remind them that they don’t need to be a math genius or a stereotype to be a right fit for a STEM career.
SS: What kind of features will we see in Apps for Kids? I’ve seen some children and young adults pick up SolidWorks. Will the features be similar?
CL: We are designing Apps for Kids from the ground up with the intention to foster imagination without the intimidation of software. We want to have the experience of learning be through play and exploration. To a grownup familiar with CAD and design workflow, features will look similar to those you find in SOLIDWORKS, like the view orientation cube. However, many of the workflows may take a different form more appropriate to the way a child would think rather than a seasoned professional engineer.
SS: How will people access the software? Is it downloadable or will it be accessed through a browser? Able to use on mobile devices?
CL: Currently, the apps will be accessible through a browser and the creations made by each child will be protected under a private account. It is our intention that these apps will be accessed through various mobile devices via the browser on those devices.
SS: Is there a focus on getting this into schools or any schools taking part in the beta?
CL: Starting this summer, we will involve teachers to work with us in developing lessons with Dassault Systemes’ SOLIDWORKS Apps for Kids. We are fortunate to have an extensive global Value Added Reseller network in SOLIDWORKS education that works with schools, school districts, and departments of education on a K-12 level. Our community will be an integral part of defining how Apps for Kids will be used in the classroom.
SS: Are the Apps built on Dassault technology? What kernel does it use? How does it relate/connect with 3DEXPERIENCE?
CL: Just like some of the kids we are designing these apps for, we are playing in a sandbox and exploring all possibilities to achieve the experience we feel most appropriate. Some will leverage technologies we know and some will be new areas we want to explore. Ultimately, we would love to have these kids grow up with us and mature into our professional products, but for now, that journey is still taking shape and will likely take some time to fully form. Sorry for all the metaphors, but working on this project has really opened my mind to a simpler part of life. It’s like building a sculpture. You kind of know what you want it to ultimately look like, but along the way, you are trying all sorts of ideas to see what works and build upon each piece as you go.
SS: Are there plans for more Apps beyond the six mentioned at SWW?
CL: Absolutely, with the six that we have announced, we wanted to set the stage and create a connected world where we tell the story from a possible beginning to a possible end, where your idea can form and take shape using each app. However, we are already working on other apps ideas that will branch out and around further into this world. There is no shortage of ideas coming from our team.
A big thanks to Chin-Loo for the interview. Beta is still set to start early spring with a release scheduled for later in 2016 with SOLIDWORKS Education Edition. You can find out more and sign up for the beta at appsforkids.solidworks.com
You know, Mother’s Day is coming up, and there’s one thing yo’ mamma (or future mamma) really, really wants… You’re not going to guess are you? … No? Ok, it’s a 3D printed self-watering planter. Don’t act like you knew that.
Joe Carpita sent in this week’s design. His mom has like 328 of these things sitting around her home, growing all sorts of stuff, but mostly marigolds. He and several of his design friends created a brand of 3D printed consumer products called Parallel Goods. The focus is on designs that are beautiful, useful, and designed with consumer 3D printers in mind. Their prices are low, allowing customers to print and assemble products themselves using detailed instructions for guidance. BRILLIANT.
One of our fav products from them is the Sawhorse Bracket. Sawhorses all over my yard? Yes please–the neighbors will LOVE IT. They’ve also released a couple designs for free on Thingiverse. The most popular is… wait for it… a Self-Watering Planter. You heard me right, a SELF-WATERING PLANTER. The same one we were talking about earlier, in fact. It’s been downloaded over 18,000 times (!) and has over 100 posted makes. Not too shabby team Parallel!
They’ve got a new collection that I’m told is centered around the “entryway” – Lamps? Umbrella holders? Puppy storage? WHAT COULD IT BE?? You’ll have to bite your nails for a few weeks until they reveal those products at the end of April.
Two weeks ago we did the unthinkable. We taped back our eyelids, slapped on some CAD spandex (that exists right?) and launched a series of articles on Fusion 360. It’s finally time we took a deeper dive into the toolset and explore: Part Modeling, Assembly Modeling, Simulation, CAM, and Collaboration. This week, we’re going to start smack dab in the comfy confines of our home field – Part Modeling. While Autodesk touts Fusion 360 is in a class of design tools that that go beyond 1990’s CAD in the cloud, it needs to have those futuristic features as a foundation to prove it. Let’s have a look, shall we?
I’ve not done any big, commercial projects in Fusion 360, but have done some small designs and have shown the kiddos how to go about using it. It may sound odd coming from someone who has spent 20 years (?!) using SolidWorks, but I actually find the sketching and modeling quite intuitive. I did however have do a little bit of that thing called LEARNING. I even figured out how to ‘Select Other’ when I was listening to Beastie Boys and held my LMB down for too long. Money makin’, money, money makin’.
But here’s a question for you. What’s your process for going between modeling and sculpting, or sculpting and modeling? With the form editing power in Fusion 360, that’s the one process I want to master. I’m getting ahead of myself, though. Let’s start off with some light mental stretching… Nah, let’s go ahead and blow your mind with the most over the top, simple concept that I’ve not seen in any other CAD system.
Delete
Delete? Say what? What’s so flippin’ special about deleting something in Fusion 360? Well, speed for one thing. This is how Fusion 360 handles deleting faces, features, bodies, or components. First, it magically understands what each of those are. Then, when you nub-punch the Delete-key, it adjusts the geometry accordingly. In order to delete a body or face in SolidWorks, you decide which one you want to do first. Select a face, then decide to delete, patch or fill. Delete a body and… great, the features are merged and you’ve got the pain of redoing references and updating features. I do like the approach in Fusion 360 and have yet to experience any oddities with it.
If you’re an old school boolean modeler this feature might turn your skeptical frown upside down. (I know you’re making that face.) Boundary Fill joins or removes volumes using bounding volumes formed by tool selections. It’s similar to Intersect, Combine and Subtract Features in SolidWorks, but this works across multiple bodies and components, or a combination of the two. Then you get an option as to what to do with the resulting geometry – Join, Cut or Intersect or even New Body, or New Component.
We all know direct modeling has been around for a while, but Fusion 360… wait for it… FUSED parametric, surface, freeform, and direct into the same package. Fused. Clever. I’m pretty sure that was intended though. It’s relevant for this example because of the ability to flip that feature history on and off, you can move geometry around without worrying about getting a little knot in your feature tree.
Oh, and here’s an added bonus. Re-anchor. This would be used in a case where you need to move a feature in relation to a reference point. I can grab a face or body, select ‘Re-anchor’ from the dimension pop-up and redefine that feature from another point. It doesn’t change that any initial dimension or reference point. It just allows me to define an exact dimension from a different reference point. Love it.
Image may be NSFW. Clik here to view.
Symmetry
Now for the really fun stuff. This is what I love so much about Fusion 360. In the Sculpt Workspace, the Symmetry feature says, “make the other side of the part look like this side.” This is a must use feature, but the way it was incorporated is super slick. You don’t have to enable symmetry prior to selecting the Edit Form command. You can enable it as needed. For example, you’ve spent time modeling a yak saddle, using symmetry in some spots and not in others. You realize the front of the yak saddle needs to be symmetrical, because you’ve decide to part your yak’s mane evenly on both sides. Select Symmetry, pick your faces, watch it update. No need to undo a bunch of work you’ve already done. And, happy yak.
Cool ‘lil quick tip with Edit Form. Hold down the Alt-key, it adds the faces as you see in the GIF so that you don’t have to go back and split faces. Loooooovely. Yes, I could model that part in SolidWorks. You and I both know it would just take longer.
Summary
Ok, so I really hate to stop there, but we’ve got this ‘5 features’ thing going on, and these are the ones that show a bit of the difference in Fusion 360 part modeling. I know some of these features and concepts are not necessarily ground-breaking, but what I keep coming back to is how they’re all together in Fusion 360–the parametric, the direct, and the sculpting. That’s what makes Fusion 360 unique in this case. And this is where I would like to circle back to how you take on a workflow between the three. With the power that combines, how are you going between modeling and sculpting, or sculpting and modeling?
By the way, when I first tried Fusion 360, I snagged it for free and at $300 annually, the price still amazes me for everything it can do. Snag Fusion 360 for yourself here.
Cloudalize is a company you’ve probably never heard, but you’ve certainly heard of desktop virtualization. You know, accessing all your apps (CAD, CAE, BIM, etc.) from the cloud via any browser or mobile device? Yes? Well, up until now Frame has been the headlinershowing the possibilities, with quick creation of a virtual workspace, and SolidWorks, Siemens and others using them as a cloud solution.
It’s a wide open space and Cloudalize is ready to give Frame some competition. The Belgian-based IT startup has developed their own online workspace solution called GPU Desktop as a Service, or GDaaS. With most CAD, CAM, CAE programs built for utilizing all the CPU/GPU power available, desktop workstations have been the only solution, but with more people going mobile (and the lack of CAD mobile apps) leaves virtualization via a browser a viable option, not only for CAD users, but for software developers who would just as soon not develop mobile or browser-based apps.
GDaaS is client-configurable, browser-based platform pre-configured with an 8-core CPU, 32GB of RAM, 2GB GPU and a 200GB SSD virtual workstation. They have many applications already pre-installed including AutoCAD, LuciadLightspeed, SolidWorks, Revit, ArcGIS Pro, Inventor, and many others. Of course, it’s scaleable from there.
As shown in the video, their cloud-based platform is built on Nvidia’s GRID technology, which is sort of a hybrid solution that provides virtual GPU (vGPU) power, powered by Nvidia’s K1 or K2 video cards, so that commands from virtual machines are transferred directly to the server-based GPU without the need for translation protocols and therefore don’t sacrifice server performance.
Cloudalize offers virtualization via HTML5 compliant browsers as well as through Citrix Receiver. traditional subscription package for those that rely on an annual service and a pay-as-you-go option for those needing a single project solution. Users can even rent GDaaS for a few hours if the need arises. A free trial is available here.
If there’s one threat I’ve always given when being kicked repeatedly in the lung sacks, it’s “If you kick me one more time! I will open my steampunk spectacle! After which you will incur the wrath of 5000 hungry steampunk ladybugs! That willst shoot from my steampunk face orb! Engulf your body! And commence dismantling your putrescent flesh!” Or something along those lines.
Steampunk ladybugs are way scary. How do I know? Well, GrabCAD user Dape (like ‘Dope’ with an ‘a’) gives us a perfect visual of a steampunk ladybug (or LADYBIRDS for those who, I’m told, speak ‘proper’ English–sorry, this thing looks way more like a BUG than a BIRD). Though but a wee creature, a ladybug with metal gears, twice the talons and some hardcore circuitry looks like it could mess you up six ways from Sunday. Now just imagine a SWARM OF THEM *shiver-twitch*. Yeah, try not having nightmares about that.
You know what that mop of yours is missing? Nope, not a 3D printed hat. A 3D printed LOW POLY hat, sucka! No funky floral patterns or flare on this icon of 3D printed fashion. It’s pure, triagular, ear chafing discomfort, ready for a night out on the town.
Thanks to a one ‘Jwall’ who runs ‘Print That Thing’ you can have your very first 3D printed hat. Is it functional? Hellz no. Will it make you look like a 3D print geek? HELLZ YEAH. And that’s what gets you high fives from your bros… and those cute little babies who just learned to high five–isn’t that the cutest… high fivin’ babies, Ha! Let’s see the timelapse that Adafruit created of this cap.
The Onshape guys think Part Studios set them apart from other multibody CAD systems. I’ve been skeptical. After a few weeks of testing, I have some thoughts.
It’s the summer of 2014. I’m on the phone with Scott Harris, who just finished walking me through his Part Studio demo with the telephone-equivalent of jazz hands. An awkward pause follows as I scrunch my brow and look off into the distance. I hear what I think may be a lone cricket chirping somewhere far away. (For a longer demo: Onshape multipart webinar.)
You can now design multiple parts in a single document! It’s called a Part Studio. Cue the confetti cannons! Wait, what?
A: How is this different from any other multibody part file?
S: Because you can add them to an assembly.
A: Uhm. I can do that in SolidWorks.
S: No you can’t.
A: Yes I can.
S: Nuh uh.
A: Nuh huh.
S: Nuh uh uh.
A: Nuh huh huh huh.
This went on for a long time.
Multi-wha?
MCAD geeks can skip this section. For everyone else, a little context.
Mechanical CAD (MCAD) systems have traditionally divided the design process into two main tasks: part design and assembly design. Think of a car: it’s an assembly made up of thousands of individual parts. CAD systems were designed with this paradigm as the primary data management metaphor: an assembly file organizing the car as a whole, and thousands of part files for each and every part in the assembly.
Old-school CAD systems were single-body part systems. One part file could only contain a single contiguous chunk of solid material. One part = one file = one lump of material in the physical world. Simple.
In the mid 90’s we started seeing multibody systems. These allow the user to create more than one hunk of material in a single part file. At first that may seem counter-intuitive, as the one-file-per-part paradigm was simple and (at the time) intuitive. In practice, however, there are lots of reasons why an engineer might want to create more than one solid body in a single part file.
The reasons multibody became important are esoteric and irrelevant. If phrases like “complex sequential boolean operations” or “deeply interdependent parametric relationships” are not your thing, what matters is that multibody part design was a big improvement for many mechanical CAD design situations. Multibody isn’t the answer to every problem, but it’s a powerful tool in the hands of a skilled modeler.
So yeah, on second thought the above is really more like a multi-body part (cut up pig) becoming a single body part via boolean union (whole pig). An assembly, then, would be a duck riding a pig riding a horse. I said it was a bad analogy. Stop asking stupid questions.
Part Studio vs. Multibody: Part Modeling
Back to Onshape.
When Scott first demoed Part Studios to me, my reaction was confusion. Isn’t a Part Studio just a multibody part system? Well, let’s break it down.
Let’s say we’re modeling a hinge.
In a single body part system, you would have to model each half of the hinge in a separate part file. Working this way is robust but an inherently technical process, often involving spreadsheets, variables, or surface body imports.
By contrast, in a multibody system you can design both halves of the hinge in a single parametric environment. Any change you make to the barrel updates both halves simultaneously, and you can see the change’s effect on both of them at once. (There are reasons why I tend to only do this in very simple cases, but let’s leave those aside for the moment.)
When it comes to basic solid modeling, a Part Studio is ostensibly the same as any other multibody system. Mostly.
Since Part Studios were designed with multipart design in mind from the beginning, the feature creation tools in Onshape are all inherently multipart. In SolidWorks, they are not. For example, a single Fillet feature in SolidWorks can only affect edges on a single solid body, even in the context of a multibody part. In Onshape you can do them all at once.
This might seem trivial, but nearly all features in Onshape are inherently multipart. Want to extrude multiple profile sketches at the same time to create multiple bodies in one feature? Do it. Even the Move Face command allows you to select as many faces on as many different bodies as you want in one go. When you can apply features to multiple bodies at once, you quickly find yourself thinking differently about the whole modeling process. It’s freeing.
So yes, there are some nice conveniences for part modeling in Part Studios that aren’t always available in traditional multibody systems. But part modeling is only the beginning, and this is where a Part Studio starts to pull ahead.
Part Studio vs. Multibody: Assembly Basics
Now we want to do a motion study to show how our hinge moves in the context of a larger assembly, and for that we’ll need to bring our hinge halves into an assembly file. This is where a Part Studio diverges from humdrum multibody systems, and where I think real workflow improvements are possible.
In a multibody system, we would have to export our hinge halves to separate part files before importing them into the assembly. This means that we now have four files to manage: the master part file containing both halves,the two split part files containing each individual half, and an assembly file articulating them.
In a Part Studio you can insert the individual halves of the hinge directly into an assembly. No fuss, no muss, just insert and done. In multibody we needed four separate files to make a simple hinge assembly.
How many files do we need to do the same thing in Onshape? If we consider a Part Studio the functional equivalent of a part file in desktop CAD, then Part Studios have just done with two “files” what multibody desktop CAD needed four files to accomplish, and that’s just for a single hinge.
Anyone who has ever built assemblies with dozens, hundreds, or thousands of components will know that trying to manage multibody part files in a complex assembly is almost exactly as appealing as death-by-sticking-your-head-in-a-microwave-oven. (RIP DFW)
Part Studio vs. Multibody: Mechanical Changes
So now we’ve created our assembly, and we decide after reviewing the motion study that we really need a two-point hinge with three separate pieces–oh, and last time we forgot to include the pin in the hinge, so this time we’ll include those too. So now we have a total of five bodies in our multibody part file, but only two of them are being spat out to part files for use in the assembly. Grr.
Now we have to edit the feature that exports those part files, making sure that the correct bodies are being exported to the existing files while also creating new ones as necessary. Then we go to our assembly and add in the new part files. We now need seven separate files to make our multibody part file into a workable assembly.
Imagine a giant hairball you just pulled out of the clogged shower drain. You now have a perfect visual for how multibody-to-assembly workflows play out in practice.
In Onshape it’s easy. Just add the parts you want to the Part Studio, then drop them into the assembly. Done. No separate “split” features or master model imports, no piles of redundant part files with complex interrelated cross-references. Clean and simple.
Have you ever heard of the “Layout” toolbar in SolidWorks? (In a room of a hundred SolidWorks users, three raise their hands.) How many of you actually use the Layout tools in SolidWorks to design mechanisms? (All hands go down.)
The Layout tools in SolidWorks are meant to facilitate what’s known as a “top-down” design approach for mechanical systems. Without getting too technical, CAD geekery lumps design strategies into two overarching categories: top-down, and bottom-up.
Top-down strategies begin with a big idea and work down into progressively finer detail, adding individual components as needed to facilitate the broader vision. It works well when designing things from scratch.
Bottom-up strategies are just the opposite, usually beginning with a bag of off-the-shelf parts and assembling them gradually into a larger mechanism. Top-down and bottom-up methodologies are both useful, and most designs require a bit of each.
Top-down design in multibody CAD is tricksy, especially in a team environment. CAD systems have tried for decades to work around this inherent limitation of file-based CAD, and yet I still have to teach week-long seminars on the topic. The Layout tools in SolidWorks are an example of this effort. It tries to allow you to divorce the mechanical design process from the geometry design process. It may seem simple in a canned demo, but it’s a jungle out there, folks. Shit gets weird.
The Part Studio paradigm dispenses with the need for kludgy workarounds like the Layout tools in SolidWorks. In Onshape you have the option of working with traditional multibody CAD strategies, but it opens up a powerful new possibility: a Part Studio can effectively separate the mechanical design process from the geometry design process.
Here’s an example. First you lay out your mechanisms using a simple 2D sketch. No details are needed, just a few lines to indicate where the moving parts lie and how they connect with one another. Next, use that sketch to extrude a few simple place-holder blocks, one for each component. These can be simple rectangular extrusions. Now–and here’s where we diverge from traditional multibody CAD–we add those blocks directly into an assembly. No separate files needed, and no fear of downstream consequences.
But how do we connect the blocks together for motion studies? We don’t have any geometry to mate together. This is where the power of mate connectors comes into play. Back in the Part Studio we can add mate connectors to the layout and assign them to the output bodies. In the assembly we can use those mate connectors to mate our parts together into a working assembly–all without ever building a single mechanical detail in the geometry itself.
Onshape’s Part Studio allows for true top-down thinking, letting you explore mechanical options without the need to build detailed geometry just to test an idea.
This is a huge advantage, and one that will only get more powerful as mechanical mate connectors of various kinds are added into Onshape. Once you like the mechanism, you can add geometric detail gradually without breaking the mechanical assembly.
Top-down design was an afterthought in multibody CAD. In Onshape, it’s built in from the beginning.
Does it Really Matter?
There’s no question that a Part Studio has some really nice convenience features over traditional desktop-bound multibody part files. Once you get used to being able to apply features across multiple bodies simultaneously it’s really hard to go back.
Like multibody modeling before it, it’s a big step forward in the history of parametric CAD.
As with most things, the proof is in the pudding. All hand-waving, beard-scratching, and belly-thumping aside, I’ve been using Onshape in test projects for quite a while now, and going back to old school multibody part systems is painful.
In the end, I have to admit defeat: the Part Studio is better by far.
The more I work with them, the more I appreciate the many subtle differences that add up to a better overall experience.
Some projects are best approached top-down, bottom-up, with a master model or interface control document, or a combination of all of these. The Onshape Part Studio vastly improves the top-down workflow, and with the recent addition of linked documents, we can now do true bottom-up design as well, but in a much more robust and straightforward way than a file-based part library could ever offer.
The Part Studio is indeed different from traditional multibody CAD, and so much for the better. It’s not revolutionary, but it’s certainly a welcome evolution with very real advantages.
Scott, I’m sorry. I was wrong. (Grumble grumble grumble…)
It’s time to talk about everyone’s favorite topic. Mating. (You two in the back stop snickering.) We’re in part three of our under-the-covers *snicker* peek at Fusion 360 and this time, we look at Assemblies. Over the years we’ve passed along tips and tricks, best practices, and overall bad-assery when it comes to parametric modeling and assemblies. It burns the nostrils to say this a bit, but I want you to forget about mates for a few minutes. Or rather, let’s think about how we would build assemblies without mates. Ready?
“Where we’re going, we don’t need mates.” – A nerdy cad nerd somewhere
If you’re a SolidWorks or Solid Edge user, you’re comfortable with building assemblies by mating components together. But even before that, you consider how to build and organize the parts and assemblies? Bottom-up? Top-down? Sketch-driven? Parametric modeling is great for many, but not so friendly for conceptual modeling. I faced this frustration many times when just wanting to model up an idea–thinking about assembly structure when I just wanted to think about assembly design. Sound familiar at all? Well, this is where Fusion 360 approaches assemblies a little differently.
In our first article, I mentioned five features in Fusion 360 and one of those features was As-Built Joints. I’ll cover that feature here in a moment, but it will make more sense if we have a better understanding of top-down design (or layout, skeleton, or multi-body modeling). A smarty-pants, scholarly person on Wikipedia described top-down design as “breaking down of a system to gain insight into its compositional sub-systems in a reverse engineering fashion.” What-evs. Here’s my definition of top-down design: Creating parts in context of an assembly. You know, in contrast to bottom-up modeling: Creating parts and adding them to an assembly. Now, let’s break these down and see how Fusion 360 approaches it differently.
Top-Down Design
I’m going to pretend for a moment I haven’t spent 20+ years making sure my assemblies properly reference my parts. Yet, just this week I opened a model and somehow all references were lost. It wouldn’t have been so bad, except that there were over 100 unique parts in the model.
GIFImage may be NSFW. Clik here to view.
Image may be NSFW. Clik here to view.
There are other things we just accept as normal. Things we write manuals and methodologies around. Things like: renaming a file while keeping the reference, breaking up an in-context assembly parts to reassemble them, editing each part separately to apply a simple fillet to both of them, or going even deeper and editing sketches that drive multiple features to fix something that went wonky. Normal, everyday assembling modeling life, right?
Well, if you were to wad those processes up, fire them into the sun and re-write a CAD system, how would you do it differently? Let’s start with sketches. My favorite top-down approach is to use sketches to drive multiple parts. In Fusion 360, a sketch is available to any feature, body, or component in a design without the need to copy or convert it. There’s no need to edit each part separately to change your sketch or add features. Features can be applied across multiple bodies or components, all during the same active command. Joining a feature to an existing body and adding a new body (or component) is all done within the same command. Here’s just a small example:
GIFImage may be NSFW. Clik here to view.
Image may be NSFW. Clik here to view.
What I’m getting here is that, this multi-body approach within Fusion 360 doesn’t interfere with your ability to model assemblies quickly. Instead of moving in and out of parts, you work on the design as a whole. You create bodies, turn them into components, and edit everything as if it were one. The shock and joy to me in all of this? You start thinking, primarily, how features affect the entire design, not just the individual part.
Mates vs. Joints
Ok, but what about Mates? After all, that’s what we started off talking about. How does Fusion 360 handle bringing individual parts into an assembly? Instead of mates, you roll joints. (I promise that won’t be the last ‘joint’ joke.) Are Joints just another name for Mates? Well, yes and no. Let’s take a look at the difference.
In every CAD application I’ve ever used, Mates remove degrees of freedom and, in most cases, require multiple mates applied to each part to properly position and restrict movement. On the other hand, Joints define movement between two components. Now, since Fusion 360 starts by assuming zero degrees of freedom, in most cases, only one Joint is required to define the relationship (or motion) between components. I struggled with this at first, but, wow, have Joints made assembly time more laid back.
So, when you apply a Joints in Fusion 360, watch for the graphical feedback. After you select each part, the Joint will animate. The context menu that appears under the cursor is handy to quickly select a different Joint Type once you use it a couple of times, and spend a little time getting to know which Joint to apply to which situation.
Image may be NSFW. Clik here to view.
Something that bugged me was setting up the equivalent of a Width (or Symmetry) Mate. It’s there, you just have to right-click during the placement of a Joint and select “Between Two Faces”. And, as in the example below, the ‘Between Two Faces” option understands the context, applying the equivalent of a Width and Concentric in one operation. Disco.
GIFImage may be NSFW. Clik here to view.
Image may be NSFW. Clik here to view.
As-Built Joints
Let’s recap quickly. If top-down modeling is the prefered methodology for assembly modeling in Fusion 360, and Joints add a degree of freedom as opposed to multiple mates removing degrees of freedom, what would be the benefit of As-Built Joints? Well, I’m glad to pretend you asked, so I can ask you a question.
If you want to show mechanical motion, what sense does it make to model parts exactly where they go in the final assembly, if you have to disassemble them, only to assemble them again? Yet, that’s how you do it to maintain the references. Most CAD software doesn’t leave you with any other option but to do this. In Fusion 360 however, As-Built Joints assume the component is already defined by the sketch, so all that’s left to do is to select the two components and determine the working behavior between the two by defining the type of motion.
GIFImage may be NSFW. Clik here to view.
Image may be NSFW. Clik here to view.
Hopefully you see why I saved the best assembly feature for second to last. Joints and As-built Joints in particular save loads of time. It goes to the foundation of simplifying the product development process, starting with a top-down approach and working more efficiently with components. But we’re missing one crucial aspect of completing our assembly. Bottom-up design.
Bottom-Up Design
So what you’re saying here is that you have to model every part, every time as part of a top down design? No sir. And if you’re seeing a theme in these posts, the folks over at Autodesk re-thought this little nugget, too. Not only can you create a reference to another component that will update when design changes are made, but let’s imagine this situation: My colleague, let’s call him Mr. Pita, is working on a part that is referenced in my assembly. He makes some horrible change that lights up my assembly tree brighter than the Griswold’s cat at Christmas. Instead of having to resolve these issues, I can choose a version of Mr. Pita’s file that was working previously.
GIFImage may be NSFW. Clik here to view.
Image may be NSFW. Clik here to view.
Not sure exactly when they added this, it hasn’t always been in Fusion 360, but sometime last year (June going by this feature overview) they added the ability to work collaboratively through Distributed Design with the option to Choose Version for a part that is already being used in your assembly.
On top of this, since the Fusion 360 data lives in the cloud, independent of file locations, and a team would be working from the same data, renaming files doesn’t cause issues in Fusion 360. The references will automatically be updated.
Summary
These are the assembly features in Fusion 360 that are making me think differently about the design process. There are still some things I’m working through, like updating Joints after a geometry change, but maybe you have some tips for that. And yeah, of course you’re going to consider the structure of the assembly up front, but with Fusion 360’s approach to top-down design, you’re not constrained by working within one part or another. You can work in context of the assembly across parts, you define movement between parts, and you build assemblies with less worry about references and a faster ability to choose versions. I don’t think the time that all of this saves can be emphasized enough. If there are other assembly features I may have missed that have saved you time, definitely point them out. And if you haven’t yet, you can try Fusion 360 here.
We all know you can’t first become a Jedi Master until you get all whiny about droid maintenance and zapped in the ass a few thousand times by a remote training droid. And yeah, we all know good ol’ uncle Ben gets his chuckles from seeing young Padawans zapped repeatedly by a floating, laser-shooting orb — YOU’LL GET WHAT’S COMING TO YOU UNCLE BEN… NOOOO!! RESIST THE DARK SIDE!!
Raymond Gaustadnes wants to relive his Jedi training days and has provided the foundation to do so with a life-size model of a remote training droid–the Marksman-H training remote, manufactured by Industrial Automaton. It’s approximately 15cm (5.9 inches) in diameter and has all the little details to make it look extra real. He modeled it in Fusion 360 and has provided a STEP version for download, laid out and almost ready to 3D print. Now all that is left to do is actually print it, hollow it out, add some blaster bolt components and whatever anti-gravity drive you can get your hands on. *Chewbacca roar!* Hours of fun!!
You can download the model on GrabCAD. (Note: to download a STEP format, load the model in the 3D Viewer and hit Download! Or download direct here.)
Well, this snuck right past us. In mid-March, a Monday no less, IntegrityWare/nPower Software published a press release announcing the release of something new. You might be familiar with their ‘Power Surfacing for SolidWorks’ or ‘Power SubD-NURBS for modo’ plugins. Well, this announcement wasn’t about a new plugin, or an updated plugin. It was about an entirely new, standalone 3D modeling software. Yet, not only software, but a platform that combines all the best of their plugins across three new products.
From Integrityware: “For many years, IntegrityWare Inc. / nPower Software has helped 3D engineers and designers improve productivity through its popular, award winning plugins (Power Surfacing for SolidWorks, Power Translators for 3ds Max / Maya / MODO, Power SubD-NURBS for 3ds Max / MODO, Power NURBS for 3ds Max, etc.) for 3ds Max, Maya, MODO, and SolidWorks. With its new standalone architecture, Cyborg3D brings these popular plugins, with their advanced features, to virtually the entire CAD / 3D modeling / 3D printing community.”
Cyborg3D. Yes, standalone architecture. As in, standalone software. IntegrityWare goes on to say the new 3D modeling software “uniquely combines organic surface modeling (subdivision surfaces) with precise, parametric CAD (NURBS solids & surfaces) modeling.”
Mmm-hmmmm. Let that soak in. How many years have we been waiting for a CAD software developer to do this? And it drifts by like a spring rain. Well, here’s what it has. Altogether, Cyborg3D is three new products.
Cyborg3D SubD2CAD Cyborg3D SubD2CAD ($995), based on Power Surfacing for SolidWorks (and Power SubD-NURBS plugins)ss will convert typical Subdivision surface (SubD) models into standard CAD formats–convert 3ds Max, MODO, ZBrush into STEP, IGES, SAT. It also provides free-sform, organic modeling using subdivision surface modeling tools.
Cyborg3D CAD2Print Cyborg3D CAD2Print ($495) can import standard CAD formats (IGES, STEP, SAT) and format them for 3D printing producing high quality, water tight STL meshes.
Cyborg3D CAD2Poly Cyborg3D CAD2Poly ($495) tessellates NURBS (CAD) geometry into water tight, high quality polygonal models (including quad meshes) in STL or OBJ format applicable for game engines, rendering, interference detection, 3D printing, and capable of handling large assemblies through as multi-threaded architecture.
Cyborg 3D is built on Integrityware’s own modeling kernel, the SOLIDS++ geometric modeling kernel, a kernel that runs the gamut of 3D modeling capabilities, from curves and surfaces to solids and polygons. It’s breadth of applications is enough for IntegrityWare to secure licensing agreements with Autodesk, PTC, McNeel, MoI and others.
When IntegrityWare first announced Power Surfacing for SolidWorks in January 2013, I was sure it was destined to be gobbled up by Autodesk, as with what happened with T-Splines and who previously acquired the nPower SAT translator, or Dassault, or any of the players who could use that tech, from 3D Systems and Ansys to PTC or Bentley, heck, even Ford who also licenses their kernel.
But do the CAD companies need it now? Last I checked, surfacing still leaves much to be desired in the mid-range 3D modeling space. Outside Autodesk’s implementation of T-Splines into Fusion 360, what is there that gives you solid and sculpting together? Yet, nPower still offers the plugins for SolidWorks, modo and others. And now they’re making a stand in the space with their own standalone software that brings it all together.
If you’ve given Cyborg3D a spin, let me know what you think.
Here we are. Part 4 of this epic exploration of Fusion 360. If you remember our first post, we are working our way through five aspects of Fusion 360 that, which Autodesk suggests, makes it more than CAD. To recap in a sentence, it all revolves around the idea of Fusion 360 as a ‘Product Innovation Platform’ that brings all the features of 3D CAD together with CAM, Collaboration, Simulation and a whole lot more. Now it’s time to look at Simulation. As mentioned, this gets slightly outside my comfort zone, so I’ll leave it to you to you to help me fill in the gaps while I attempt a quick sanity check on a design. With that, here are the reasons I would use Simulation in Fusion 360.
The Integration
The first fab thing to note is the way in which the simulation capabilities are integrated into Fusion 360. It’s one workspace that lives right alongside the other workspaces: Sculpt, Model, Patch, Render, Animation, and CAM. So, as you would suspect, you can move right from one to the other. It’s the way I’m use to doing quick sim studies, and with the commands, context menus, and viewing the same across each workspace, it’s a process that compliments the back and forth between CAD and Simulation.
It allows you to quickly react, adjust the model and validate that the changes had the results you need. If you use SOLIDWORKS Simulation, this sounds familiar right? Build model, run sims, update geometry, and all that. No big deal. Well, there’s purchasing SOLIDWORKS Simulation Professional for the price of a brand new Nissan Versa or gaining the ability to perform the same simulation studies (as far as I know) in Fusion 360 for $300/year. Yep.
Static Stress – I think were most familiar with this one. Add loads. Add constraints. Analyze the deformation and stress on the model. With the results, you can investigate displacement, stresses, and common failure criteria. In other words – Fix a portion of the design, apply a force and use the results to find potential problems.
Modal Frequencies – Here you analyze the natural mode shapes and frequencies of an object during good, good, good vibrations. From the results, you can investigate the shapes in VIBRATION MODE Mode Mode, corresponding frequencies and their mass participation factors. In other words – will this frikin’ thing vibrate itself apart? Find potential problems.
Thermal – With a thermal study, you determine how the model responds to heat loads and thermal boundary conditions under steady state conditions. The results include temperatures and heat flux. In other words – Heat that sucka up and use the results to find potential problems.
Thermal Stress – With a thermal stress study you determine temperature and stress distribution on the model from both, you guessed it, thermal and structural loads. The results are the double-whammy impact of applied heat and stress. In other words – Heat that sucka up, apply a force and use the results to find potential problems.
Yeah, there’s a theme here–set it up, run it, analyze the results. Same old process, right? Well, it’s within this process where Fusion 360 separates itself a bit when compared to other simulation tools I’ve used. The way in which you can setup, run, and interrogate the model is what makes the experience… and I love/hate this word… easier. Let’s continue down the rabbit hole.
The Setup
One of the best things about Fusion 360’s sim setup happens when you define a Contact. First of all, Fusion 360 will auto-detect body/part contact and apply a Bonded Contact. This typically addresses ALL contact situations, but can be overridden by editing a contact and changing it to Sliding, Non-sliding, Spring, etc. Even if you build in tolerance and have small gaps in your designs, you can use a contact allowance to adjust the tolerance. Then, you can use a DOF plot to check for missing contacts, which stays activated as you go in to add manual contacts.
GIFImage may be NSFW. Clik here to view.
Image may be NSFW. Clik here to view.
As is common, selecting faces to define manual contacts sets are difficult because, well, by nature, the faces are hidden. I recall exploding, hiding, suppressing or setting up configs in SolidWorks to make that process easier. In Fusion 360, the workflow goes like this. 1) Select a two bodies, at which point all other bodies disappear, and the second goes transparent. 2) Select the first contact face and that part goes transparent while the other body goes opaque for selection of the second face. Adaptive transparency based on selection. Imagine this in a VR environment? Or real life?!
GIFImage may be NSFW. Clik here to view.
Image may be NSFW. Clik here to view.
The Flexibility
There are four different ways to define a load direction in Fusion 360: Normal (the default), Angle, Vectors, and Reference. No need to make sketches or planes to apply loads at odd angles; all of the required abilities are contained in one of these options, aaaand it’s a bit of fun to play with. After the direction has been defined, you can use ‘Limit Target’ to focus the force in a specific area of the selected face. No need to sketch and split face to apply those loads in a small area of a larger face. Have a look.
GIFImage may be NSFW. Clik here to view.
Image may be NSFW. Clik here to view.
The Results
When it comes to analyzing and interpreting the results, Fusion 360 has some capabilities similar to SolidWorks. Things like clipping planes, animations, and ability to use the scale to hide or show values above and below certain values. Where the results differ is how Fusion 360 deals with accuracy with automatic mesh refinement tools and provides cloud based reports and share links.
Adaptive Mesh Refinement – Seasoned FEA users know (way better than I) that the most efficient way to ensure the results have reached their highest accuracy is by checking convergence with adaptable meshes and running the studies over and over again. Fusion 360 allows you to do this quickly with Adaptive Mesh Refinement, a simple toggle and slider that, after run, will will give you a convergence plot of the results to geek out about or scrap your design.
Share Links – Yeah, this is on the general collaboration side of things, but applies beautifully to Simulation. This capability was actually added just last year and allows you to share all details about the design, including the SIM results through the design’s project page on the A360 cloud collab portal. On top of this, the data stays updated with each save.
Oh, and according to their published roadmap, cloud solve is right on the horizon – “Cloud SIM preview allowing you to use grid computing to solve multiple studies in parallel without tying up your local client.”
As studies get more complex, unloading the brunt of the work (the solving) to a remote computer can free up your local computer to either setup the next study, another load case, or to move onto a totally different part of the design in another workspace, all within the same program. For me, all these Simulation capabilities are more than I would ever use, but it’s incredible to see what Autodesk is continually adding to it.
So, if you haven’t already, download Fusion 360 and tell me if these five things are worthy or if there’s something I’ve missed.
My parents often joke that there were five things I was holding in my hand at birth. I have no idea why it was only five, or what they all were, but one of them was a crane cabin joystick. That’s right, a joystick you would see in a flippin’ tower crane.
Well, the trauma of family and friends laughing at my over-sized, crane-grasping baby hands all came back when I saw this incredible joystick modeled by Magnus Skogsfjord in NX. He designed it to demonstrate the capabilities of using NX for ID, taking it though production with 3D printing and how you get mucho respect for modeling cool stuff. He’s provided both .prt and .stp files, and you can bet the surfacing is out of this world awesome–just check out the images.
When Onshape added linked documents a few weeks back, I don’t think I even looked out from my box of Cheetos. On second look, it might just be the most important feature of the year thus far.
Life, the Universe and Everything
As of April 2016, Onshape lets you use parts from one project in another. This means you can have libraries of standard parts, for example, and reuse those parts in all 42 of your current projects. Make a change in the library part, and all 42 update.* (Cue intergalactic cabaret dancers.)
* They don’t update right away. That would be dangerous. What if you don’t want to update yet for a given project? Onshape lets you choose whether to update a given referenced part to the latest version or stick with the old one. More on that below.
If you’ve ever done any of the following, this feature is for you:
Use the same part or assembly in two different projects
Use a supplier’s part in your own assembly
Share a part or assembly for use in a customer’s project
Share one part with a supplier, but keep the rest of the project confidential
We all knew this feature was coming. It’s a prerequisite of any serious manufacturing MCAD system, and the entire Product Data Management (PDM) industry exists, in large part, as a way of working around the problem of data reuse.
PDM’s mythical “Single Source of Truth” is a lie. There is never a single source of Truth. Every project involves multiple truths coming from multiple sources: suppliers, part catalogs, previous projects, competitive products, ergonomic data, industry standards. If it were all about a single source of Truth, it would be easy. As Truth generally is, it’s more complex than a PDM salesman will admit, and it’s in the management of that complexity that there be dragons.
We all need to reuse data across multiple projects, but keeping that data up-to-date (or not) is surprisingly hard.
In file-based data management systems, data linking and revision control is an inherently slippery business. Not so with a central database.
In the cloud, all of your data lives in One Big Database. That means tools like Onshape have the advantage of not requiring special PDM procedures for file management, syncing across data centers for global access, etc. Everybody has unfettered access to the entire catalog at all times from one clean, simple, searchable UI, and all directly integrated into the CAD tool itself.
And lest you think this is easy, note that (as of today) the Other Guys don’t do it yet. In fact, Onshape is the only cloud tool so far to offer linked data across multiple projects, and it’s the only product on the market–cloud or otherwise–to make it look easy.
What’s most interesting about Onshape’s specific implementation is the way they handle engineering changes. Onshape’s built-in PDM system is unique, and easily one of the most important advantages of their ecosystem. The “branching and merging” paradigm they’ve built makes it trivial to try new ideas safely and efficiently, and that same fluid workflow applies to document linking.
A note to the blasé: those of us who’ve worked with enterprise PDM systems for CAD file management, this might seem like par for the course. You undoubtedly also know, then, that ePDM systems have massive IT overhead–usually at least one dedicated server and full-time technician–and that proper use of those systems requires week-long training sessions for system administrators, multi-day training sessions for each end-user, and careful adherence to best practices by everyone. Traditional PDM systems also have significant mental overhead for users, requiring users to think hard about each action to insure compliance with the broader system.
I’ve spoken with hundreds of CAD users, and I can say this with certainty: everyone hates PDM.
Onshape’s most important accomplishment is making PDM a no-brainer. It takes what used to be an expensive, time-consuming distraction, and does it all for free. What used to be advanced PDM features are now so effortless as to seem incidental. I can focus on design ideas without a wasting single thought on data management.
I’m designing a wifi-enabled turret-mounted inter-continental water balloon ballista. We call it the B2, and it’s as epic as it sounds. It’s actually a new iteration of an older ethernet connected version of the same product (the B1). I need to reuse parts from the B1 in the B2, but with a few modifications.
Sounds easy. It’s not.
I actually have a whole defense contracting company built around ballistic turret launchers: the Needlessly Whizbang Mostly Harmless Low-Tech Weaponry Corporation of Northern Virginia (NWMHLTWCNV for short). Traction trebuchets, counterweight trebuchets, bombards, onagers, mangonels, ballistas, springalds… you know. Your basic medieval artillery arsenal retooled for water balloons and desktop computing.
Anyway, we’ve decided to use the ratchet pulls from the B1 in the new B2 model. As it happens, Simon’s been working on improving the ratchet pull system for the entire ballista line. This leaves me with a problem: what if I design my turret around the current ratchet pull mechanism, but then Simon makes an update? Doesn’t that risk breaking my model and setting me back during the prototyping phase?
Not a bit of it.
Milliways and the Golgafrinchans
Revision control ain’t the sexiest topic in the world, but can literally make or break a turret-mounted water balloon ballista, trust you me. When we make changes in the part library, we need to be careful about pushing those to other designs automatically, lest we cause problems for people working on live projects.
The joy of the Onshape linked document system is that you never actually link to a part per se, but rather a version of a part. Since all versions of all documents are saved for all time, this means you can safely reference any version of any document without fear of downstream changes.
Once Simon is done making his changes, my assembly gives me a little icon alerting me to the new version. I know I’ll eventually want to bring it in, but we’re right in the middle of prototyping. I’d rather not introduce the changes until we’ve seen how the first proto behaves. So I wait.
After the 0.1 proto is finished, I choose to update to the latest version, and viola: Simon’s changes are now in the B2 assembly. Looks great.
Just as importantly, the B1 assembly still uses the older version. That’s super important because we’re still shipping the B1. The assembly needs to accurately reflect the product as it exists on the market. With Onshape’s version system, that’s a given.
Thanks for All the Fish
So yeah, linked documents in Onshape. Pretty great in a team setting, but even more so if you’re a parts supplier or a user of parts from a part supplier.
Imagine, for example, if a catalog like McMaster were added to Onshape as a linkable document. No more downloading STEP files, just link directly to the part in the catalog. Use it in your design. When McMaster needs to update a part in its catalog because of a supplier change, for example, you’ll know your designs won’t be affected unless you explicitly choose to accept the updated version. You can already access the Traceparts library directly from within Onshape. Linked documents would only make integrations like these even better.
In March, Tech Soft 3D finally released Tetra4D Enrich. For those who are working 80 hour weeks and not awares, Enrich is how you crush it with 3D PDFs. It’s a tool to make interactive tech documents quickly, from inside Acrobat.
Isn’t a 3D PDF good enough? Well, if you actually use 3D PDFs, congrats, you’re a step ahead of everyone else. What Enrich does is provide a way to add, yes, a 3D viewer to the PDF… plus parts list, plus viewing options, plus entry fields, plus conditional actions and more. It’s the option you want if you’ve ever wished you could customize that 3D PDF or simply make sharing your 3D data better and safer. They’re simple to make as well. If you’re familiar with placing and editing elements in Acrobat already, this works exactly the same. So, let’s go through the process and show how you can have an interactive PDF in 4 easy steps. Note: The screenshots below are taken in Acrobat Pro DC.
First Things First – Make A Template
Although you can create a 3D PDF using Enrich without a template, it makes the presentation so much slicker if you have one. A template can be any graphical layout, or simply a page with your logo, converted into a PDF. You can use Photoshop or Microsoft Word. Just save it as a PDF and open it in Acrobat. Select Tetra4D Enrich from the Acrobat Tool menu and you’re ready to start adding elements.
Being a product of Tech Soft 3D, developer of the HOOPS Exchange libraries, you can bet Enrich imports nearly all 3D file formats. To add a 3D model, select 3D, Add 3D and draw a box where you want the viewport to appear. Pick your model, select Open and you’re good to go.
Who wants a flippin’ carousel on their 3D PDF? We do. Next select Carousel, Add carousel. Same process as before. Draw a box where you want the carousel to appear. You’ll see some customization options appear in the right-side properties pane, including checkboxes for the views to include.
We could stop there, but we won’t. Let’s add a part table. Select Table, Add part list (flat). Draw your box and set your options, including what columns you want to show and what you want to be mapped to each column. Image may be NSFW. Clik here to view.
Step 4 – Add Some Buttons
There’s one caveat to buttons. You need to create a button first in Acrobat. To do that, select the Acrobat Tools, Add button option. If you have some graphical elements for buttons, draw your box over the graphic and name the button. Then select Actions, On button to make the button do your bidding when clicked. Sadly, buttons are limited, but I expect them to add more (like link to other pages, documents or websites) later on. Image may be NSFW. Clik here to view.
Now, one thing to remember. The elements you add are not part of the Undo stack in Acrobat, so if you delete something, don’t expect to get it back. And if you add something you don’t want, you’ll have to select Edit, then delete–the Trash Can icon in the right-side Properties menu after you select an element.
3D PDFs by themselves are extremely useful. I’m not sure why they haven’t taken off more except that perhaps the 3D software companies haven’t really presented them well as an option for sharing and archiving 3D data. 3D PDFs are still useful in that regard, but Tech Soft 3D has extended that with bringing more, much needed capabilities to 3D PDFs. Capabilities that go beyond viewing to help you make documents that capture and present information in an interactive way. Documents that can be used all along the process from design and engineering to manufacturing.
If you have examples of how you are using Enrich, I would absolutely love to see what you’re doing.
It’s available now starting at $1,748.00 USD with a trial available here and sample PDFs here that even work within Adobe Reader.
Yeah, so here’s the thing. When your kid ask why you have a superhero’s head with knives stuck in it sitting on your countertop, you can say, “Did I ever show you how to make Chimichangas?” But why do… “CHIMICHANGAS!!” It’ll all make sense. And how do you get a superhero’s head on your countertop. It requires two things, a loud grunt and checking out this Instructable.
Britt Michelsen was hankering for a knife block to display her beautiful set of slicers. She came across a Julius Caesar knife block, but that’s so lame compared to what she thought up after seeing Marvel’s Deadpool–A DEADPOOL KNIFE BLOCK. Unfortunately she didn’t have the 3D modeling chops to make such a thing, but found a 3D model of Deadpool, ganked it, bashed out the mealy bits and added the knife slots using a combo of MeshLab and Meshmixer (which probably took more time and skill than just modeling the dang thing).
With her Replicator 2, she printed out five sections. Altogether it took about 40 hours to print. She glued the pieces together, acetoned the sucka, sanded, primed and painted it. And done. A totally cool, but totally not role-model worthy Deadpool knife block for your kitchen countertop.
Well, well, well. We have quite a cocktail coming together here. We’ve poured in some Fusion 360 part modeling and assembly features, and squeezed in a little simulation. Now it’s time to add that dash of collaboration. On the CAD capability excitement meter, this gets about a two. The geekiest of the geeks don’t even get excited about collaboration. (Maybe if we dressed it in jeggings and a Ferrari that would change. No?) Yet, collaboration is likely the aspect that is driving the most profound changes in CAD software today. Features, UI, versions control? That all has to be thought about as any number of simultaneous users viewing and working with the CAD data, not just one.
So in an effort to make this post a bit more exciting, let’s try something interesting. Think of someone in your life who isn’t an engineer. Now, think of them eating a stick of butter. Now, share the project you are working on. Yes, right NOW. You have 5 minutes. Go. No friends? That’s ok. Do what I did and check your parent’s Facebook page.
Like me, I’m sure you’re going through a mental checklist on data preparation, files sizes, suppress/hidden/config states. The comical reality is that this, that process, is not a difficult task for us–We deal with it all. the. damn. time. The fact that we have to “deal with it” at all is the decades old lip rot of the cad software industry. And it’s lookin’ pretty gnarly. So, I’m sitting here, thinking what would be needed to get rid of that mental checklist–Accessibility? Shareability? Securability? What else? Any with fewer than five syllables? Simultaneous Viewing? I guess not. Ahh, Project Info? Version Control? There we go. So, let’s explore what Autodesk is doing in Fusion 360 with regard to collaboration and see if it picks the list of features that get rid of those mental data prep lists.
Data Access Anywhere
Ok, this is at the top of my list. I’ve harped on accessibility for years. Not only data accessibility, but the ability to access the software. (How fast can you get and install [enter traditional CAD software of choice here]?) But let’s focus on accessing the data. Once a project is saved in Fusion 360, the data can be accessed through Fusion 360 of course, but also A360, Fusion 360 Mobile or through a web browser. And to my surprise, and the delight of my network drive that’s bursting at the seams, I don’t have to configure desktop/network access – I can access the data anytime on any device, log in on another laptop and access the same data there. This is something we don’t think about, unless we realize we have it. I’m writing this article in Google Docs. I started on my laptop, switched to my desktop, checked for comments on my tablet. I think were just beginning to see what the same accessibility adds for product development.
As a result of being able to access your data anywhere on any device, you can share that data with others just as quickly with Share Public Link. Share Link is as simple in execution as it is in concept (a rare occurrence in software apparently). Right from within the Fusion 360 Data Panel, right-click on a file and choose Share Public Link, or with a model open click File, Share, Share Public Link. Only a ‘Share’ button on your keyboard would make it faster. A dialog will appear with a link, hit Copy, set a password and make it downloadable if you like, and send it to your co-worker, neighbor, or friend who acts totally interested in what you do.
What does that allow? From that one link, the recipient has the ability to view the file directly from a browser (no login or account creation required), and gives them the ability to download a file type of their choosing just in case they wanted to work it into their design. What? An example?
Live Review is an interactive, real-time design review feature that takes the GoTo out of the Meeting. (Ha! Always wanted to use that.) Updates are streamed live, in real time–I see what you’re doing, you see what I’m doing. You can give control to anyone on the Live Review session. Share markups and move to action more quickly than playing the “imagine if you will” game.
This wasn’t on my list, but I’ll add it here. The power of the red pen. Giddy are those with the authority to deem a person’s entire afternoon of work a steamy pile of wasted effort, and display it in a merciless, rage-induced markup session of passive aggressive delight. Therapy session anyone? *hands go up* Given my disposition on redlines, why then would I even bring it up? Well, it’s not so much that redline and markup tools are available in Fusion 360–that’s hardly ground-breaking–But rather, it’s how they applied it to make it more useful.
Each time a redline is added to a design, a comment is added to the comment stream associated to the project. The view orientation is saved along with it, each of the comments are version-specific and always saved, never lost in a series of email-based decisions with attachments never seen by the rest of the team.
Where Used & Related Data
This is a big one for me. With data being at the center of Fusion 360, Autodesk has done some interesting things to illustrate data that is entirely impossible in Windows Explorer. In Fusion 360 (and tell me if I explain this wrong), several external relationships are created when creating a design. It automatically understands and displays these parent-child relationships in the browser, and maintains them no matter what. I don’t fully understand it, but that’s the beauty of database modeling systems. I rename a file (even while open), it updates, and relationships stay intact. In a desktop, file-based system, you would get all sorts of warnings.
So, when I have a model open, I can create renderings, pumped out a simulation, or create a CAM toolpath. I can share all or any of these with co-workers just by sharing the one model. And if I have a ton of extra image files, sim results, etc., Fusion 360 stores all of those with the design and keeps them organized with the correct version. All of it can be accessed when the model is open and seen clearly on A360 in the browser. And yes, I can upload other documents (images, docs, specs, etc.) within the same project, and while I can’t view these in Fusion 360 yet (a recommendation I’ll make to them), I can organize them appropriately and view them in A360.
I’ve mentioned this before, but the Fusion 360 roadmap is public, out there for everyone to see what they are working on. This is where they publish what’s coming to the product next and such a great way to keep users connected to the development team. I don’t think I’ve seen anything laid out so transparent by another modeling software developer.
One nugget I pulled out of this list was Branching & Merging (pan down to the Data Management section). A hot item apparently. They showed how this will work back in April. Team members can create branches to design in parallel. A branch is like a copy, a version. You can work on your branch/version independent of others in the design team. When a project milestone hits, the branch/version can be merged and Fusion 360 will provide the option for updated parts from any branch/version to be brought into the main design.
So, this is where I come back and look at Autodesk’ claim that Fusion 360 is more than just CAD in the Cloud. Yeah, I’m “new to Fusion 360.” I “haven’t used it” in “ages.” I “don’t let my scabs heal.” But the more I become familiar with it, the more it stands out as a tool I want to have on my product development arsenal. Maybe they would consider changing the fancy “Product Innovation Platform” label for Fusion 360 to “Product Innovation Arsenal”. Has a nice ring to it, imo. What I find interesting about Fusion 360 is how they’ve approached a lot of the traditional problems we face inside (and outside) of CAD software, turned them on their head and are taking the opportunity to approach it all a little differently.
Are these all the features that improve collaboration? There are so many things on their roadmap, I’m sure there’s more to come. You can try Fusion 360 here.
Adam shows you how to use FeatureScript for Onshape, and possibly also how to be Mr. Handy Pants around the house. But mostly FeatureScript.
Quick side note: SolidSmack has a lot of great sponsors, including several of the biggest CAD companies on earth. That’s awesome. One of those sponsors is Onshape.
This post, however, is not sponsored by anyone. This one’s all me.
FeatureScript is, in my opinion, the single most important feature of Onshape. It could very well be the killer app that brings Onshape into the mainstream. It’s a core piece of the Onshape architecture, and as of today it’s available to everyone.
FeatureScript is a scripting language for writing features. (Awkward pause.) If the moniker sounds kinda like “Javascript” or “Actionscript,” that’s appropriate. It’s syntactically similar to Javascript, but it’s ultimately a proprietary language with a powerful built-in IDE, much like Actionscript worked within Flash. These attributes have the advantage of feeling familiar to coders, readable to normal humans, and–because it’s a purpose-built language–much more concise than a general-purpose language like JS or Python could ever be.
The features you write in FeatureScript appear in the Onshape UI just like any other feature, right alongside Extrude and Revolve and Sweep and Loft. The UI for your feature will look and work just like every other feature in Onshape. You can add your features to the toolbar and use them in all of your projects. You can even share your features with others so that they can see how clever you are take advantage of them.
One of the first tools I built in FeatureScript was a one-step tongue and groove feature. It’s designed so that one T&G feature can create grooves in as many boards as you like, and tongues on as many cross members as you need, all in one fell swoop. It also automatically centers the groove and sets the width to 1/3 the board width–but that can be customized, of course.
Things you can do with FeatureScript
The first thing most people will try to do in FeatureScript is essentially a macro-style list of existing Onshape features: run an extrude, then another extrude, then a fillet, then a boolean, then another boolean, etc.
For example, I like the SolidWorks “Cut With Surface” command, but Onshape doesn’t have it. I wrote my own in FeatureScript by implementing a split command, followed by a delete bodies command. Easy peasy.
Taking us one level deeper, every feature in Onshape was written using FeatureScript. That means that by definition, FeatureScript is capable of literally everything Onshape can do. If you’ve seen an Onshape feature do it, you can do it in FeatureScript.
So, for example, instead of using Onshape’s “extrude” command (the one you see in the Onshape UI), you can use the deeper opExtrude function, providing deeper, richer access to the internal workings of an extrude. Using this method I built an extrude command that can extrude multiple profiles at the same time, each normal to its own sketch plane (instead of all going the same direction, as in Onshape’s extrude). This is a trivial example, but there are many more.
I’ve written quite a few FeatureScripts at this point, and I’ve learned a lot in the process. Once you start building features to automate common tasks, you quickly find you want more and more of them. At this point I can do in ten features what used to take a hundred, and I’m only getting started.
Then you can do a lot of stuff that Onshape features can’t currently do, too. For example, there’s no 3D spline tool in Onshape, but FeatureScript gives us access to a spline that we can run through a selection of 3D points, essentially rolling our own 3D spline tool.
In experimenting with FeatureScript, I focused mainly on carpentry tools: lumber, plywood, dado cuts, mortise & tenon, tongue & groove, pegs, rabbets, and lap joints. I made custom features that can perform multiple joinery operations in a single feature, and one that can create an arbitrary number of standard boards around a perimeter.
The beauty of FeatureScript is that things don’t have to be linear: we can add conditional statements like “if distance x is further than 16 inches, add another joist”. You could dynamically generate staircases, railings, ribs, bosses, grilles, perf patterns, or even irregular structures like Voronoi patterns and generative architecture. The possibilities are really exciting.
This feature runs a series of standard boards around the perimeter of a sketch. The sketch can be any shape or size, and the ends can be extended so that they overlap enough to make room for joinery. In this case, one feature is generating 17 boards. Imagine using something like this to frame a house. You could build the basic framing for an entire floor in one feature, then have opening features to generate doors and windows, stairwell features, rafters, etc. What I’m showing here is simple, but the possibilities are huge.
Things you can’t do with FeatureScript
FeatureScript is fantastic, but there are a lot of things it will never be able to do.
First of all, despite having the word “Script” in the name, FeatureScript isn’t really a scripting system, at least not in the traditional desktop sense of the word. It is a script for the creation of features, which have a very limited and specific meaning in Onshape. Specifically, features exist within Part Studios, and help to define the geometry that is generated there.
For example, you cannot write a FeatureScript that creates a Part Studio or an Assembly. Features exist within Part Studios, so they certainly can’t create them. You can’t use FeatureScript to generate a drawing or a rendering. FeatureScript can’t email an administrator via email if Part 12345R6 changes, or run a cost analysis across an entire project. In short, FeatureScript is intended for creating geometric features within a Part Studio, and not much else.
As a proof of concept, this simple feature throws points and tangent planes at user-defined intervals across the U and V parameters of a surface. The reason the points are all bunched up on one side is that, just for fun, instead of a regular grid I’m using an exponential one (see line 28 of the code below, where I square the grid locations). In theory I could use this feature to drive other pattern-oriented features, gradually building up a Grasshopper-like generative modeling tool set.
This is the entirety of the code for the surface grid feature shown above, and it’s more complex than it needs to be. I hacked this together in just a few minutes as a curiosity. Imagine what we could do with a little more time and energy.
There are lots of esoteric programmy things it can’t do, too. For one thing, it can’t create new geometry types. For example, you couldn’t use FeatureScript to create a subdivision surface modeling tool, or to manually generate degree-739 NURBS surfaces. You can’t build a voxel system or interact with the GL meshes rendered on screen.
As with any cloud system, security is extremely important within Onshape. As such, FeatureScript can’t make calls to external web addresses of any kind. So you can’t, for example, have a feature that spells out your most recent tweet in block letters.
Also, as you probably know, Onshape allows multiple users to work on the same document simultaneously, so it’s very important that all feature results be purely deterministic. In other words, a feature should always have the same result with each and every rebuild. This might seem obvious, but there are some important limitations that come up when you realize this.
For one thing, FeatureScript has no date or time functions. If it did, you could make features that rebuild differently based on the day of the year or time of day. This might seem like a humbug, but remember: you might not be the only one seeing it. If you’re looking at an Onshape document with a user in Japan, whose day or time should it use? That’s a superficial problem, but there are deeper ones. Suffice it to say that date functions in FeatureScript would be problematic, so there aren’t any.
Similarly, FeatureScript cannot generate truly random numbers. If it could, each rebuild could result in wildly different things. Your collaborator in Japan would be seeing something entirely different than what you’re seeing, and neither would be “true” to the file. Export would be entirely unpredictable. Instead you can use psuedo-random numbers generated using a seed value. These are random enough for most users, but without the problem of being non-deterministic.
Can other CAD tools do this?
Sort of.
Most CAD systems have some sort of macro system for stringing together sequences of tool operations. Most also have some sort of scripting system for automating tasks, typically using a non-proprietary language like Python or–in the case of SolidWorks–Visual Basic (Adam projectile vomits across the office). Some even have GUI interfaces for storing or recording macros, known variously as Macros, Library Features, User Defined Features, or Power Copy, depending on your tool. Grasshopper lets you create logic-driven geometry using a nodal system that’s essentially a visual programming language. All decent options.
Most CAD tools also have some sort of API for developing richer plugins, usually in c++, but you’re going to put on your Big Kid Pants for those, as they’re typically written and maintained for use by professional developers, not your average user.
All of these options are great, but FeatureScript is a bit different. As with any walled-garden, it comes with both advantages and disadvantages. On the upside, FeatureScript lets you write first-class features in Onshape that look and feel just like the features that ship with the product, and do it with surprisingly little code. Downsides include learning (another) proprietary language, no system-wide automation, and very limited access to the deep inner-workings of the system.
All in all, having written many plugins for other apps over the years, FeatureScript is really exciting. For me, the benefits vastly outweigh the limitations.
More to come
This is going to be fun. I’ll be publishing tutorials in the coming weeks. Stay tuned!
