The ABC’s of RTAM
3D modeling applications capable of creating watertight STL (Stereolithography Tesselation Language) files are now ubiquitous and cheap. HP (hp.com) will soon OEM a 3D FDM (Fused Deposition Modeling) machine to add dimension to its suite of printers. School kids are typically well-versed in digital literacy. Put it all together, and there’s a bumper crop of newbies eager to cut their teeth on RTAM (Rapid Technology and Additive Manufacturing). So whether you’re one of them or someone a bit more seasoned, here are some things to know. It is distilled from the FAQ’s that I’ve given to my students over the last few years.
1) Think priority. What do you hope to achieve from this process? The question seems simple enough, but there exists a different machine, process, and material for every need. Will the part be used as an appearance model, a casting master, a functional prototype, art, a production simulation part, or a final manufactured artifact? No one machine can meet all such needs at once, so it’s up to the designer to prioritize.
Think of it this way. If I’m creating a motorcycle, I’ll use the SLS (Selective Laser Sintering) process with PA (polyamide) 11 or 12 material for the fairing, since I expect to go over it on occasion, so durability and cost drive that decision. I’d have it sanded, primed and painted, and it would pass for a production part. But if I want to wind-tunnel test it at aircraft speeds (I never said this would be a real-life example), though, I’d likely use SLA (stereolithography) and have it metal plated at Repliform (repliforminc.com) or Metalise It (metaliseit-america.com/index.html), making it rigid enough to withstand such forces. And if it is to find its way to a magazine cover, I’d have it mirror-polished to a beautiful shine. But if I want to race it on the track, I’d go the route of Formula 1 and use SLS with a carbon-loaded PA to make it extra stiff (read: extra expensive). Say I want that fairing to feature that classic 70’s-era sunset scene—complete with horse and cactus—directly from a jpg. That can only be achieved with equipment from Z Corp. (zcorp.com), though it won’t be quite so track-ready. Is the bike electric? May want to go green with a pure, unpainted PA 11 or 12, to keep it recyclable. The clear windscreen? Try SLA, using a water-clear resin (though sunlight may yellow it after a while). And for that stainless-steel skull above the headlight (taken directly from a CT scan of my own skull, of course) I’d use DMLS (Direct Metal Laser Sintering) or EBM (Electron Beam Melting) to fabricate a stainless, titanium, or maybe chromium likeness (no cheap endeavor, but again, this is fiction). In short, decide on the purpose, the effect, the durability requirements first, and then find the material, process and machine to suit the need.
2) Think orientation. If you upload your file to any RP (rapid prototyping) site for quoting, it’s nearly arbitrary as to which side it decides is up. Unfortunately, this means everything when considering a part created additively. Orientation impacts the part’s strength, accuracy, appearance, surface quality, cost, support structure remnants, build time, etc. For example, a part with a concentric pivot will rotate smoothly when fabricated with the axis vertical, though it may not rotate at all if fabricated with the axis horizontal. That said, orienting the axis vertically may increase part cost and increase breakage, so a tradeoff may be necessary (I tend to describe orientation as if the part were made from plywood—it will break on the layer lines first). As a precaution, I always include a screenshot of the part in the preferred orientation with the file, just to clarify. When in doubt, I call the vendor to ask their opinion.
3) Think hollow. A part machined with CNC equipment is often cheaper to create as a solid block, since the machining time to hollow the part adds to the cost. Not so with a process where you pay by the dust particle—in fact, the additive process invites parts that are hollow. Extensive research has gone into processes for generating the complex ribbing that increases the strength-to-material ratio of that hollow part. For further reading, look into Netfabb (netfabb.com) or Paramount’s Conformal Lattice Structures (paramountind.com). The dream of creating bird-bone quality structural optimization is not only possible, it makes economic sense as well when using the additive processes. But be sure to consider where you locate the holes to drain the residual liquid or powder trapped in the fabrication process, because if you don’t, the vendor will, and he may not be as aesthetically inclined when it comes to putting in the holes.
4) Think additive. I’ve seen students design parts that adhere to CNC and injection molding rules, even though they’re intended to be produced with an additive process. This guarantees the worst of both worlds. If a part is best machined or molded, then so be it. But if the additive processes make more sense given the nature of the part, then forget old-school fabrication and enjoy the flexibility that only it offers. Dual-walled parts with conformal ribbing, complex hollow manifolds that would otherwise be unrealistic, impossible undercuts, and internal features—these are at your disposal. Thin-
walls too fragile to be machined? No problem. A complex contour with grossly variable wall thickness that would preclude injection molding? Bring it. An array of ventilation holes that would lead to injection molded knit/weld lines and broken core pins? Simple. Just consider the possibilities first.
5) Think green, think clean. Additive manufacturing technologies are inherently green. The part is often created with nothing more than a laser, a heater and the exact amount of raw material. What could be greener than that? And many materials—polyamide and some metals, for example—are curbside recyclable. I often throw the #7 symbol onto my polyamide parts (recyclelogos.org/recycling-symbols.htm)
But yes, those gritty, white SLS parts do accumulate grime with use. To combat this, I often have them tumbled at a typical industrial tumbler to smooth the surface. Beyond that, SLS parts are also dishwasher safe, leaving them pristine and lemon scented.
6) Think interlocking. You know those cool impossibly interlocked parts that you’re told were never assembled in the first place that they sometimes give away at tradeshows? We all get the urge to create them. Here are a few pointers that may save you from making polyamide paperweights.
The powder-bed machines invite this trick more than the liquid or deposition machines, I’ve found. Just consider that holes will come out slightly smaller than you designed, and shafts slightly larger when you’re designing the parts. You’ll need around 1-mm clearance to ensure that the parts won’t fuse in the process. Remember too that the axis wants to be vertically oriented (see point #2) to guarantee the roundest cylinder. Also consider the STL process will turn your cylindrical mating surfaces into triangles, which may well screw up your tight tolerances. In these instances, it may be worth increasing the tessellation density (adjusting chord height or angle control, or whatever your modeling app allows) to create a rounder approximation of that axis. Finally, be sure to design in a gap or path to allow all trapped cake to escape. Again, when in doubt, call the vendor and ask their opinion before pulling the trigger.
7) Think green (the other one). Yes, students usually face a limited budget. So here are a few tips to help keep your beer money. Orientation, wall thickness and process have the greatest impact. But be sure to shop around. Note that the expensive vendor today may be the lower-cost one tomorrow. This is because their estimate is determined by a complex formula involving part size, part volume, and material amount. That formula changes from one vendor to the next, so one bid is little indica-tion as to their overall pricing. If cost is crucial, ask them if there are orientations or design modifications that will reduce the cost. They may suggest a more optimized orientation, that you split the part, or that you reduce mass in areas to cut your costs.
Another cost save is to have them send it ground mail instead of overnight (typically the default. While you’re at it, request that they skip the Styrofoam peanuts: who doesn’t hate those?). Tell them when time is not an issue, because some vendors may reduce the price since they can fit the part in when they have a run with extra space.
Finally, allow yourself time and budget to explore what these processes can do. This is a very new technology, and everyone in it is a pioneer of sorts. There is much to be learned by trial-and-error, and amazing possibilities ahead. Please share your thoughts and experiences with others, since we’re all on this ride together.
