3D printing/additive manufacturing can help designers simplify a part, reducing assembly steps in many cases, and creating a better end product. Is it possible, though, to over engineer a part? 3D printing engineers like Eric Utley of Protolabs see situations where a part was over engineered. Here are some tips to help reduce excessive engineering and get parts out to market faster.
One of the most common design areas where engineers can go too far is with tolerances. A companion area is “over defining” areas in a CAD file, notes Utley.
“I often see either over defining the drawing or having too tight of tolerances on the drawing. Engineering school, even CAD design schools, tend to teach you to define everything on the drawing because otherwise you’re leaving things for interpretation down the road and it can lead to a bad part.
“However, once it’s passed to a manufacturer, calling out design aspects like surface roughness, for example indicating a 200 micro-inch RA surface finish, will add to the costs of the part. Often, what the designer really wants is a matte finish, but that’s not what it says on the drawing.”
The inclusion of such details means the manufacturer must confirm that level of surface finish is what you want.
“You don’t really need to succinctly define just how matte is matte, especially if it’s really just an aesthetic finish.”
Utley advises letting the manufacturer know in real terms what you want out of the part.
“Usually there’s a lot of insights that happen in those little short meetings between customer and manufacturer.”
Over dimensioning a part can also result in conflicting measurements.
“Typically, what I see on drawings would be like a tight tolerance, or something called out on what is ultimately a non-critical feature. And that can drive secondary processes such as secondary machining or inspections to validate that feature. And those measurements are not critical to the design or to the functionality of the part.”
Utley suggests understanding what really drives the functionality of the part and the critical aesthetics of it. And then leave loose tolerances on the non-critical features.
Another tip is to talk with the manufacturer. Even if the design is largely constrained, conversations can result in a less expensive solution.
As mentioned earlier, the surface finish can be over engineered. Utley considers this area more of a communication issue than a design issue. For example, if a customer wants a part to be black, a manufacturer will need to know if that’s a matte black, a glossy black, or a soft touch black.
“For the surface finish, what you need to communicate is a color and a finish. This ensures customer and manufacturer are on the same page as to what that part looks like. Move away from calling out surface roughness unless it’s a functional issue.
“I’ve seen some drawings go as deep as defining what the specific grit blast media is on a part, and that may be overkill. Unless you’re really concerned with parts matching on the shelf, but the difference would be extremely minor in the end result.”
For 3D printed parts, the CAD file defines the geometry. Beyond that, designers work with material choices and the resolution. Note that different service bureaus may define resolutions in different ways. What one calls high resolution may not be the same as what another bureau calls high resolution. A good approach is to define the layer thickness instead of defining high resolution or normal resolution. The layer thickness speaks more directly to the resolution of the print.
“But even then, I encourage people to have a little flexibility there, says Utley. “If your drawing says three thousandth inch layers and someone offers four thousandth inch, or two and a half thousandth layers, you’re still going to have pretty close part quality if there’s a slight difference there.”
The more you know
In the last five years or so, more designers are taking advantage of the full capabilities of 3D printing technology, primarily because they have been exposed to this technology at nearly all levels of education.
“I see it on both ends of the spectrum,” says Utley. “Some people are trying to get everything they can out of 3D printing and make a fully functioning assembly, print it all at one time, with spinning gears and moving rods and so on. And that is possible. But you really have to be familiar with the technology and design that very specifically within the limitations of even a specific 3D printing technology. It’s difficult to design something to be generally used in different 3D printing processes to do that.
“But on the other end, we see some newer engineers now developing for 3D printing and they’re able to effectively take something that was 15, 20, 30 parts in an assembly, and then print it all as a single piece. And that does require a redesign. You can’t just unify all those pieces and then go with it. You have to kind of take a step back and look at the assembly as a unique part and rethink the build direction and the implications it has on the print to develop that. Such parts are designed and they grow or build very organically, which is kind of the telltale of a well-designed 3D printed part.
“Each manufacturing process requires a different design method. For example, if you’re designing for CNC machining, it’s like you start with a block and you start whittling away at it to form parts. And that’s a great way to design CNC machining.
“When designing for injection molding, you almost start with the midline of the part or the cross section of the part, and then you extrude it out from either side and add draft, just visualizing like a waffle, essentially, in the design.
“With 3D printing, you want the design process to mimic the manufacturing process. Often you can just drop the critical features in space, just unattached. And then visualize the best approach for the build direction. And then decide upon a build direction and kind of commit at that point.
“And it’s good to visualize the print as an animation, or like a flip book. Imagine it like you’re printing a cone and you’d have a flip book you’d flip through and have consecutively smaller circles. And it’s like, you just remove the paper, you’re left with that cone. And so, when you’re designing for 3D printing, you want to visualize it like a flip book and go through it.
“And then think of how clean is that animation? Are things blinking in and out sporadically, or are things jumping from one end of the paper to the other? That tells you it’s poorly designed for 3D printing. You want it to transition organically from layer to layer. The less you have of aspects of a part blinking in and out, and the less you have of things drastically changing in shape and size layer to layer, that’s ultimately going to be a better 3D printed part.”
When to use 3D printing for low-volume production
Experts used to advise that production runs of 10,000 units was the cut-off number to switch from 3D printing to more traditional manufacturing methods.
“But the number has been trending up over time as the materials improved and the technology got faster. But 3D printing is still in that low volume, high mix space where you need a wide mix of different parts at low volumes.
“Currently, 3D printing is frequently used to make first products that end up in people’s hands, such as electronic devices, or a VR headset. And you want to make it on a scale of a few hundred to get the product out into the hands of people to determine what they think of it before you invest in molding. It’s a good way to get their feedback before investing and taking that product to the next step.”
Be careful about designing for 3D printing when you expect to finally produce a part using other manufacturing technologies.
“You want to have some sight lines further out of what will be your eventual production levels of this part,” notes Utley. “I think a good rule of thumb to go by would be like, say 500 copies a year of your product. If you’re looking at like 500 copies a year of something, and it is plastic, you really want to look at injection molding and try and find a way to injection mold that part.
I think many people have this concept that you’re looking at tens of thousands of parts a year for injection molding to be cost effective. And, really, that bar is much lower than that. If you’re leveraging like an aluminum tooling for injection molding, I’ve seen injection molding be more cost effective as low as eight units. You definitely want to design for the manufacturing method you’d be using tomorrow, not necessarily the one using today.
The best thing to do is to be technology agnostic, to use all the tools in your toolkit.”
Key design choices
Many design choices come down to material, quantity, and complexity. Those are three key features of a part that drive what manufacturing process is going to be best for that part.
“If I need one of a part,” says Utley, I’m going to 3D print it. If I need a thousand units of that part, I’m going to use injection molding. It really takes all three of those factors to decide what manufacturing process is going to be best.”
Another consideration is reducing weight. In many instances, weight reduction works better with 3D printing. If a part will be more expensive at a lower weight and manufactured through conventional means, look into 3D printing.
“Light weighting is very big,” says Utley. “Hyper cars like Bugatti and Porsche use 3D printed metal parts because they’re in that space. They’re happy to spend three X for that component to shave a little bit of weight off, but it’s also personal customization. Customers are willing to pay a little extra for custom braces made for their teeth versus some generic ABC that may not be as comfortable a fit. High-value components are also an area for the use of 3D printing.
“Don’t try to pigeonhole yourself into a particular manufacturing method. Be open to even using a manufacturing process you may not be familiar with and educate yourself on it.
“Think longer term, don’t design yourself into a corner. Have a sight line. Do you think you’re going to be making a million of these in five years? If so, have a plan to get there, and the manufacturing method that’s going to enable you to get that, drive towards that. Also, as the product succeeds in the marketplace, consider whether high production manufacturing methods will be needed and will the design need alteration to accommodate those methods.”
Designers are often faced with coming out with a model refresh every few months. Often, there’s a “half step” between the next version of a product. The expectation is to be coming out with a new model every two years.
“These half-step iterations of a product, these model refreshes are where 3D printing can be very strong. As long as the part design allows it, you can switch from one manufacturing process to another.
As the quantities go up, or if something changes, like the material changes, or the demand changes, you’re going to have to change that design potentially for a new manufacturing process.”