How to 3D Print Consumer Product Prototypes, at different levels of realism
Realism helps the prototyping process of Consumer Product Goods, but not always in the ways you'd expect. We climb through the 4 levels of 3D printing realism, using both single color FDM and full-color Polyjet printing, and discuss the cost and benefits of each.
-
Step 1: What are we printing?
A large chunk of my customers design Consumer Goods, meaning products like headphones, speakers, handheld electronics and tools, things that have to look and FEEL high quality.
And people are always asking me "How realistic can 3D prints REALLY get?"
So I decided to put myself into the shoes of a medium-sized Industrial Design firm and see how 3D printing realism plays into each phase of the ideation process.
To come up with a design I was going to print, I started by asking myself one simple question:
Take off any sort of cover or case you've added to make your cellphone easier to grip and try to hold it. Really TRY to hold and use your cellphone in its as-designed, rectangular state. It's not super comfortable, right?
I mean, it's like it's actively DESIGNED to hurt your hand during operation:
And if you try to use that thin, flat, heavy shape for a long time, gripping it different ways to try and find easier ways to scroll:
So why not make cellphones curved to fit a human hand? Not just the outer case, but the whole screen, all the electronics inside, and the phone itself?
Now, I'm not saying this should be a real product. I'm just saying this is the type of project Industrial Design firms attempt all the time, as sort of a "concept car" to present to their clients.
So without further ado, I give you:
We're going to propose a new, radical cell-phone concept that's curved to better fit your hands, and we're going to use 3D printing to make people in my office understand what we're talking about.
Let's start with the first level of 3D printed realism and go up from there.
-
Step 2: Realism Level 1: SIZE
All renderings look the right size on a computer screen.
But how do we know if a certain design has the right curves, or fits right in your hand? This is the lowest level of realism since it can be done on a cheaper, single color FDM printer, and doesn't even need to be that high a resolution to yield results.
For my curved cellphone idea, I wanted to make sure the shape felt right in my palm, but I also had an idea that some bumps on the back would help me grip the phone as well as my poppers do:
I really do love my cellphone poppers, I can't imagine using a cellphone for a long time without one now. But I wanted to see if those ergonomics could somehow be built INTO the design of the case.
So I made a few simple outlines in SOLIDWORKS and brought those CAD files directly onto my GrabCAD Print tray:
At this stage, I'm testing if the curves and bumps are an improvement over a rectangular design but I'm also testing the SIZE of the phone I want to make. (Who says the size of phones companies have given us is the right one?)
I hit 'print' and by late that afternoon we had this:
We broke those samples off the tray, passed them around the office, and within seconds we had our first data point: NO ONE AT ALL LIKED THE BUMPS ON THE BACK.
This isn't some manufactured drama for a better tutorial, I had really thought that those ergonomic bumps on the back would help me grip the phone better than a slick, flat, thin surface. But within seconds, anyone I handed these samples to agreed: the bumps did NOT simulate cellphone poppers in any way.
People described them as "painful" and "awkward" and "Why did you add these?"
But that's exactly what Level 1 realism prints are for: to quickly and cheaply weed out horrible ideas before you spend too much time developing them.
Whether it's a part that won't fit your assembly or having annoying bumps in places, Level 1 is just about getting the outer shape of your part right. And it's cheap! Within one day and a few hours of work (I did other things while the printer printed) I had a much better idea of what worked (and what didn't) for a curved cellphone.
For rev 2, I decided to replace the bumps with simple straight grip lines and curve the cellphone INWARD as well:
This is what my FDM tray looked like (again, all single material FDM):
And once again, within seconds of finishing the print we had our answer:
When people think of "realism" in 3D printed prototypes, they usually don't think of rough, non-functional, SINGLE-COLOR FDM prints to simulate electronic devices. But after passing this really really simple 20 deg. model around my office, my co-workers finally understood what I was trying to do. I heard:
"Oh, it DOES fit in my hand so much better than my iPhone!"
"It's really comfortable!"
"Wow, this will even fit in my jeans pocket better than a flat phone- it hugs my body!"
This print was realistic enough to get the rough concept across.
So I decided to move up to the next level.
-
Step 3: Realism Level 2: FEEL
I debated long and hard about whether 'Color' or 'Feel' should be the next level of 3D printed realism, since we have customers doing both at this point in the design.
But for hand-held consumer products, I decided that letting a client FEEL a realistic shape in their hands and turn it over, squeeze and bend it, would be more important than getting the colors right but having your new headphones or ergonomic kitchen tool feel like an unforgiving rock.
So now we're going to add a little 'soft-touch' to our model:
Moving past FDM, integrated soft touch is possible in Stratasys polyjet printing because those are printers that can put down 3 or 6 different materials on the same layer, meaning you can mix hard and soft materials in the same print. And we have materials at different Shore values (hardness) to allow that:
Shore D and Shore A are different hardness scales so sometimes it's difficult to compare them, but I found this webpage from Teknor Apex that does a great job of showing them all together:
So what I'm going to do is print the body and screen of my cellphone in Vero White for that hard PVC pipe stiffness, but print my outer grip and back ridges in Agilus Black to get that soft mouse pad/bottle nipple-like comfort:
But this raises a question.
In normal production processes, when you have a rubber overmold around something, usually you want a way to 'capture' the overmold to your main body so it doesn't pull off. On this Protomold page, they call it an 'interlock':
So do we need to do the same for 3D printing?
I didn't know, I had never printed Vero and Agilus together before.
So I designed these 3 test pieces in SOLIDWORKS to see:
In theory, the No Interlock sample should be the weakest bond of the three, and the dovetailed High Interlock should be the most strongly captured.
(Side note: because we're 3D printing this interlock layer by layer, we could have made that dovetail capture look like ANYTHING inside our model: shapes that twist and turn, hook around features, start and stop ten times, shapes that could never be inserted by traditional means. It was pretty startling to realize the amount of freedom we had, bonding these multi-material parts together.)
And then we printed them:
The surprising result from this experiment was that the "No Interlock" sample, the one that had the Agilus just sitting on top of the Vero base, was strong enough for our purposes!
I really pushed on them, as shown in the picture above, and the edge of the rubber wouldn't peel off under normal use. It seems like the Agilus and Vero bond well enough at the interface to take a lot of stress, meaning we don't need any special design changes here at all!
Then we printed our 'real-feel' phone on a J750 with Agilus Black:
You can see that we printed multiple screens, and had to print the screen and body as separate pieces (more on that in a second) but from the final result, it looks like a phone at first glance!
It's definitely good enough for a quick photo-shoot to show off your new product:
And it's probably on par with renders a lot of design houses currently use to explain their concepts in these early stages:
If you're trying to explain a wholly new product concept to a client, wouldn't having a printed part in hand to touch and turn over be better than an equivalent render? Why not have both?
And this only took me a day of work to CAD up, and then I printed it overnight.
You can see a big part of looking like a realistic phone is the colored apps on my screen, but before we move onto realism level 3 (color), I wanted to make a quick diversion about how to get even MORE out of realistic feeling parts.
-
Step 4: Realism Level 2B: TEXTURE
You may remember from my previous tutorial I discussed how SOLIDWORKS 2019 has a new feature that automatically displacement-maps (textures) solid bodies based on a repeating image:
So I thought that would be great to do for this polyjet print too, to further enhance the feel of our rubber molded outside. So I started down that path...
And hit problems right away.
The easiest way to explain this is that, to select multiple materials in GrabCAD Print, you need to have multiple BODIES in your assembly, to give you different volumes to select:
That works for imported CAD assemblies with more than one part, STEP assemblies with more than one part (shown above) and imported STL assemblies with more than one part. It also works for multi-body SOLIDWORKS part files.
The problem is, with the new '3D Textures' command in SOLIDWORKS 2019, you can texture ONE body of your part file, but when you try to export it as an STL, you can't get a multi-body STL file that we need.
So if I applied that hexagon texture in SOLIDWORKS, I'd have to export the outer rubber grip as a separate body and glue it on later, which we did:
Now, that result did feel a LOT more realistic than my original smooth outer grip (those bumps can be used to simulate leather, or injection molding, or any other kind of texture you normally might have), but the super glue actually held LESS tightly than bonding the Vero to Agilus during a print would have!
I could pull these apart after gluing, while Agilus and Vero printed together wouldn't! That's sort of a bad look, to have your textured grip come off while your client is holding it.
And the superglue left some residue...
And it was a pain to line it up correctly...
Better to have your textured parts bonded to your Vero parts during printing, which meant they would have to come out of SOLIDWORKS as a shared, multi-body thing.
So I went 'brute-force' and cut every divot out of SOLIDWORKS myself:
It's hard to convey over a web tutorial, but that textured print definitely FELT a lot more realistic than my first smooth Agilus version. The air gaps made it even softer than before and the textured contour felt more like something I'd find on a high-end device, like a speaker or kitchen tool.
So if you're printing consumer goods and you want to increase the realism in your soft-touch sections even more, consider adding a TEXTURE, even if you have to brute-force it a little in your CAD package.
And hopefully, future advances in the SOLIDWORKS' texture tool, and GrabCAD Print, and other programs which can do texturing (like Rhino) will allow mated, multi-body parts to easily be exported and printed in multi-materials.
So let's look at how we can improve the realism of that phone even more.
-
Step 5: Realism Level 3: COLOR
I've already written an entire tutorial on how to soft-proof and hard-proof your J750 colors, so we won't duplicate that here. If you want the colors coming off your printer to get as close as possible to a physical sample you're trying to match, definitely read that tutorial first.
(And if you're trying to match a specific PANTONE color with your J750, contact your local reseller, because we're running a beta on that feature RIGHT NOW.)
What I'm going to cover here is how your specific Level 2 FEEL choices affect your Level 3 COLOR potential.
We all know that digital displays use 3 colors (Red, Green, Blue) to show their RGB images. And we all know that physical paper printers use 4 colors (Cyan, Magenta, Yellow, Black) to show their CMYK images.
(And if you don't, you really should go back and read that color-proofing tutorial.)
So with 6 heads available to us on a J750 (not counting the support head), if we wanted to set up our printer to have the widest color gamut, we'd load something like these materials:
With 6 addressable heads, we can take 4 to match the CMYK of paper printing. Paper printing is done on bright white paper, so they don't need to add that color, but we are printing in air, so we need one more head for Vero White, to give our colors some consistent 'backing' for brightness and color accuracy.
That leaves one more head open, and customers often fill that with VeroClear, so they can print textured models with clear around them, or clear window sections inside their electronics prototypes.
But we've just gone through Realism Level 2 to add soft touch to our parts, and you'll notice there is no Agilus loaded in the 6 head diagram above.
The actual loadout on the J750 I used to print my textured phones looked like this:
So this was a problem.
I could print a combination of Agilus Black and Vero White to get my feels-like case with overmold, but I didn't have Vero Black to make the depths of my screen images stand out:
The picture on the right looks much better, because of the dark blacks. But GrabCAD Print won't substitute Agilus Black in when the model is looking for Vero Black, because that would make a rigid model oddly squishy in places (not what I want for my cell-phone screen).
(Side note- GrabCAD Print WILL do this with Agilus White, look for THAT story in a later tutorial!)
But isn't there some other combination of Cyan and Magenta and Yellow inks the machine could have used, to make the black sections?
It turns out no.
If you go through a few of your texture profiles at the bottom of your 'Tray Materials' screen:
You can see that switching between those profiles tries to better approximate your colors with whatever combination of Cyan, Magenta and Yellow you have loaded, but none of them really capture what my imported screen OBJ was going for:
CMW was the closest, but CMY was my printer's default. So it was pretty surprising when I put my mainly white phone screen into GrabCAD Print and it previewed mostly YELLOW!
(And now if this ever happens to you, you know what to adjust!)
I wanted my colors as realistic as possible, closest to what the artist intended. But I didn't have a deep Vero Black loaded on my machine.
Also, I couldn't print my multi-body CAD Agilus + Vero White base in the same file as my single-body Rhino textured OBJ screen. This is why I had to print the screen and base in separate parts side by side and glue them together.
I mention all this because, if you only have one polyjet machine, you have some decisions to make, regarding realism. Do you want a more realistic color, or feel? Is it okay if you have to print your prototype in pieces and glue it together, or do you want it all printed as one unit?
Your choices in Level 2 affect your options in Level 3.
In my case I wanted both realistic color AND feel, so I went through the process of replacing Agilus Black and Agilus Clear on my machine with Vero Black and Vero Clear (not something you want to do more than once a month!), to give me full color capability again.
And then I textured a lot of files in Rhino and printed them:
So how did everything turn out?
-
Step 6: Results!
If I was an industrial designer and my boss gave me one week to make a product pitch to a big electronics house, I'd feel pretty confident walking in with some renders and those printed samples. They're the right size, feel and color for a client to really understand the concept I was going for.
Does it look exactly like a real cellphone? No. But at Level 3, we might not have the time or effort to go for an indistinguishable, 100% match. What you have to ask yourself is:
And I think the answer is yes.
Some of you might be saying, "One week to get a single Level 3 realism model you can hold in your hand? I can do better than that carving foam!"
But what I haven't told you is that I didn't do all this just to get one model.
In fact, to get to this point, I was able to test THIS many variations:
And now that I've got the workflow down, I can print more at the drop of a hat. Consider this:
That's 17 different screens. That means, every 7 hrs and 40 mins, I can test 17 different variations for my client.
So if your client asks you one morning: "I like that logo on your new phone, but can I see it in a few different colors?", you can answer: "Yes. We can have 17 different versions to you by the end of the day. And every day, for as long as you want."
(And don't even get me started about the 16 hours between when you leave the office at night and show up again in the morning. The printer works while you're asleep.)
So that's how to use 3D printed realism in consumer product prototyping. If you have any questions email me at-
Oh?
What's that?
You remember I started this tutorial talking about FOUR levels of 3D Printing realism, and I've only covered THREE?
Okay, if you insist...
-
Step 7: Realism Level 4: EVERYTHING ELSE
My print may be the SIZE of a cellphone, FEEL like one, and have the approximate COLORING of a cellphone, but there still a lot of ways to improve my realism.
It doesn't have the WEIGHT of a cellphone- I could design my print to let me glue pennies or lead weights inside my model to give it a realistic heft.
It doesn't have the FUNCTION of a cellphone- for larger devices, maybe I just print the case and leave space inside for real electronics?
And METALLICS are something still pretty difficult to print on polyjet. If I really wanted my prototype to have a metallic look and finish, I could do something like this:
That worked with mixed results:
If you want your parts to have a little electronics 'glow', you could also try something like this:
That didn't work. In fact it didn't work so hard, I'm not even going to show you a picture of the result.
But my larger point stands: don't let 3D printing be the end of realism for your prototype. Remember that you can always glue, weigh down, paint or alter your 3D printed parts to achieve even more realism. That's true Level 4.
But for most designers, going through Levels 1-3 should be enough to get your product ready for client pitches, focus group testing or the hand off to manufacturing:
If you'd like more information on how to do the same for your Consumer Product Prototypes, click here, or let me know at shuvom.ghose@stratasys.com!
And if you want to learn more about the affordable, FULL-color printers used to make these models that can sit right next to you in your office and churn out models, check out our new J55 printer at this link.
