Shapeways 3D Printing Service vs. CNC Machining vs. Injection Molding: How to Choose
I coordinate rush manufacturing for product teams, which is a polite way of saying I get calls when a part is already late or wrong. I don't ask which process is best, because best doesn't survive contact with a deadline. I ask how many parts, what material, and how many hours are actually left.
Shapeways is worth including in that conversation because it makes process comparison fast. The platform is best known for its Shapeways 3D printing service, but it can quote other manufacturing steps too. That doesn't make the process decision go away. You still have to decide whether 3D printing, CNC machining, or injection molding fits the part.
Here's the mental model I use when I triage these requests. It isn't perfect, but it's practical.
Scenario 1: Complex Geometry, Low Quantity, Short Lead Time → Shapeways 3D Printing Service
If the geometry has undercuts, curved internal features, or a design that changed yesterday, tooling is the last thing you want. A mold or a custom fixture will eat the schedule before it makes the first part. In this scenario, I use a Shapeways 3D printing service mainly because it removes setup as a variable.
People think 3D printing is the fastest process because layer times are measured in hours. The truth is more boring: the speed comes from not waiting for setup. You upload a file, get a quote, and order. There's no toolpath planning call and no mold sampling meeting.
In March 2024, we found a sensor bracket problem 36 hours before a customer demo. One tab interfered with the housing. I changed the model, added a fillet, uploaded it, and paid about $58 for expedited shipping. The part arrived in time. If I'd waited for a conventional machining quote before starting, the demo would have gone out with a broken fixture.
Before you jump to additive, check the material datasheet. 3D printed plastics vary in strength, moisture resistance, and heat tolerance. If the part has to be load-bearing in a safety-relevant way, a printed part might still work, but you need to validate it rather than assume.
Separate Idea: Shapeways Shop for Repeat Low-Volume Sales
There is another use of Shapeways that gets overlooked. A Shapeways shop gives you a storefront for on-demand production, so customers can order units without you holding inventory. This is somewhere between a prototype and a production run.
For the rush-order types who read this, the shop route is useful for replacement parts and small accessories that are stable in design but too niche for a mold. If the design is still changing, don't set up a shop yet. You'll sell yesterday's problem.
Scenario 2: Quantity Climbs and Design Freezes → Plastic Injection Molding vs 3D Printing
The plastic injection molding vs 3D printing debate usually misses the starting point. The real question is whether the design is frozen.
Injection molding has a high upfront cost and low per-unit cost. 3D printing has low upfront cost and higher per-unit cost. If the design is going to change, a mold can become an expensive anchor. If it's locked, molding becomes attractive at high volume.
Last year I went back and forth on a customer clip part for three days. The forecast said 10,000 pieces over two years, which pointed toward molding. My gut said the clip would be redesigned after field testing. We printed the first run, and I paid an extra few hundred dollars to have the geometry reviewed for moldability so the next order could go straight to tooling. That compromise felt slow, but it avoided a tooling change order.
Automation in Injection Molding: What It Actually Changes
Automation in injection molding gets cited as the reason to switch from 3D printing. Automated cells do lower the per-part cost, and they make quality consistent. Robots don't get tired, and they don't round off edges at 2 a.m. That matters for production.
But automation in injection molding adds work on the front end. The mold still needs to be designed and sampled. The robot needs a program, and the cell needs validation. So if you're comparing lead times, don't count cycle time only; count the time from the day you sign the purchase order to the day the first approved part shows up.
People think the main benefit of automation is lower labor. In my experience, the real benefit is consistency. Lower cost follows because consistent parts create fewer downstream failures.
Scenario 3: Metal, Tight Tolerances, or Both → CNC Machining
If a part needs metal, or if the drawing has tolerance callouts that a printed layer line can't meet, CNC machining is likely the answer. CNC isn't automatically slow. For simple geometry, it can be faster than additive because the machine starts with a solid block and can be done in minutes.
People assume that 3D printing is always faster than CNC. That's true for complex geometry, false for simple brackets. Complexity is the deciding factor.
In my workflow, I upload the CAD to Shapeways as a baseline because the platform can quote several processes from the same file. I also talk to a local shop when I need someone to inspect a weird setup or explain tolerance decisions in person. The platform gives me a fast number; the local shop gives me conversational detail. Use whichever matters more.
The CNC Taper Turning Formula I Keep Handy
If the CNC job includes a conical or tapered section, the standard geometry is simple:
K = (D − d) / L
where D is the larger diameter, d is the smaller diameter, and L is the length of the tapered section.
When your CAM system needs an angle rather than a ratio, use the half-angle formula:
θ = tan−1((D − d) / (2L))
I include this because of the factor of two. For example, if D = 0.5000 in, d = 0.4800 in, and L = 0.7500 in, then D − d = 0.0200 in. Divide by 2L, which is 1.5000, and you get about 0.01333. The inverse tangent is about 0.76° for the half-angle. The included angle at the cone is about 1.53°. If you put the wrong one into the CAM program, you'll get a part that looks close but doesn't seat correctly in the mating component.
If you're turning between centers on a conventional lathe, the tailstock setover uses the same taper geometry but is a separate calculation. On a CNC lathe, the machine can move the X and Z axes together to generate the taper, so the half-angle is usually the number that matters.
How to Know Which Scenario You're In
The four questions I ask before every decision:
- Is the material metal, or does the function require molded plastic properties? If metal and tight tolerances are needed, CNC should be on the shortlist.
- What quantity is actually needed over the next 12 months? Low quantity and complex geometry push toward a 3D printing service.
- Is the design frozen? If not, avoid hard tooling and automated molding until the design stops shifting.
- How much time is left after quoting, production, shipping, and any secondary work? I count the longest lead time in the chain, not the manufacturing step itself.
Then I do one last honesty check on the quote. The vendor who shows every cost upfront might look more expensive at first. In my experience, that vendor is usually cheaper by the time the part arrives.
Ask what's NOT included in the price before you ask how the price compares. That question has saved me more money than any discount I've ever negotiated.
Looking back at the rush jobs that went wrong, the problem was rarely the process. It was the assumption that every quote meant the same thing. The goal isn't to find the best process in a brochure. It's to find the process that gets you a good part by the date printed on the purchase order.