Cut Costs by 99%: How 3D Printing Replaced a $110 CNC Coupling with a $1.05 Composite Part

From $110 to $1.05: What Happens When Your Machine Shop Gets a 3D Printer

A production manager with zero additive manufacturing experience replaced a CNC-machined stainless steel coupling with a 3D printed composite part and stress-tested it for a week straight. It never went back to the Bridgeport. Here’s exactly what happened, and which parts in your facility could follow suit. 

Key results at a glance 

• 99% part cost reduction: $110 → $1.05 per coupling
• Zero signs of damage after one week of continuous actuation (~5 years simulated service life)
• Improved part geometry: star-profile impossible to mill, trivial to print
• Lower field failure risk vs. the machined original

3D Printing a Culture Change at Air & Liquid Systems

The Part That Started Everything

Ryan Wenzlick had a problem most production managers know well: a small, complex valve coupling that was a nightmare to make. The star-shaped cross-section was impossible to cut on the Bridgeport. His team machined a squared-off approximation instead, broke tooling regularly, and paid $110 per part for the privilege. 

Ryan is the Production Manager at Air & Liquid Systems, an OEM equipment manufacturer in Michigan that builds heat exchangers, sludge removal systems, and liquid filtration equipment for customers like GM, Tesla, and Ford. His team had no additive manufacturing experience when they brought in their first machine. Within weeks, that $110 coupling became a $1.05 printed composite part, one that fit the valve better than the original ever did.

It took our part cost from $110 down to $1.05. We were like, holy cow.

– Ryan Wenzlick, Production Manager, Air & Liquid Systems

Ryan’s team didn’t just measure the cost and call it a win. They put two printed couplings into valves and ran them under continuous actuation for one week, the equivalent of roughly five years of normal service life. When they pulled the parts apart afterward, there was no measurable wear, no deformation, no failure of any kind. 

That test converted a skeptical team. When you don’t know additive manufacturing, the default assumption is that printed parts are prototypes, something you’d use for fit-checking, not production. A week of continuous mechanical stress in a working valve killed that assumption. 

ENGINEERING NOTE  When evaluating 3D-printed parts for production use, accelerated life testing—continuous actuation, cyclic load, or elevated temperature exposure—is more informative than static strength measurements. The geometry freedom of additive manufacturing often enables designs that outperform their machined equivalents under real operating conditions, not just in tensile tests.

Why This Part Was the Right First Target

Not every machined part is a good candidate for printing. Ryan’s coupling hit several criteria that made it nearly ideal for a first production application:

Factor Machined stainless version Printed composite version Unit cost $110 $1.05 (99% reduction) Geometry fidelity Squared approximation (tool-limited) True star profile — functional improvement Tooling wear High — frequent end mill breakage None — eliminated Operator time Skilled machinist required Unattended print — labor freed Risk of field failure Higher — geometry compromise Lower — optimal geometry achieved Setup and changeover Fixturing, workholding required File re-open, print — minutes vs. hours

The geometry constraint was the decisive factor. A part that’s expensive to machine because of complex geometry is almost always a strong candidate for printing, because the geometry constraint that makes it hard to mill often makes it easy to print. Internal channels, star profiles, organic forms, and thin-wall geometries that defeat end mills are native territory for additive.

What Parts in Your Shop Should You Target First?

Based on the Air & Liquid Systems experience, and patterns across similar OEM and production environments, these part categories consistently offer the strongest ROI case for early additive manufacturing adoption: 

• Complex small machined components: Parts with internal profiles, non-standard cross-sections, or geometry that requires multiple machine setups to produce. The coupling in this case study is the archetype.
• Fabricated plate assemblies: Anything that involves cutting plate stock, drilling, and then welding or fastening into a bracket or structural mount.
• Retention clips and hardware: Clips, clamps, and retention features that are bought in bulk or machined as one-offs.
• Low-pressure fluid fittings: Valve fittings, couplings, and connectors in systems that don't require full metallic sealing. Composite materials handle a wide range of fluid environments.
• Tooling, jigs, and fixtures: Assembly fixtures, drill guides, and inspection aids. No validation testing required for non-product tooling.
• Low-volume custom parts: Any part ordered in quantities of 1–50 where setup cost, MOQs, or lead time make traditional procurement inefficient.

Learning Curve: The “Zero Knowledge” Problem

One thing worth addressing directly: Ryan’s team had no additive manufacturing experience before their first machine arrived. No one on staff had used CAD for 3D printing. No one understood composite print parameters or how to design parts for FFF deposition. 

This is common. Most production managers evaluating industrial 3D printing carry the same assumption: we’ll need to hire someone, or train for months, or bring in a consultant before we print anything useful. That assumption is worth challenging, and Ryan's team tested it by going straight to Markforged University, a free additive manufacturing online training program, before ever unboxing the machine.

The tutorials were perfect. Everything was on point. For somebody that has zero knowledge of any of it—we sat through that and it was like, all right, let’s unbox this machine.

– Ryan Wenzlick, Production Manager, Air & Liquid Systems

What Ryan is describing is a real design philosophy in industrial additive: the system needs to be operable by people who run production, not just people who design products. The machine and software workflow are built around manufacturing applications—scan a part, upload a file, set support structure, print. The expertise accumulates through use, not classroom time. 

That said, there are real design-for-additive concepts that improve outcomes significantly: orientation relative to print layers affects mechanical properties; infill patterns and densities affect weight and strength; composite reinforcement layers are placed where they’ll be loaded. None of these are hard to learn, but they’re worth understanding before designing production parts, not after.

The Culture Change: From “Can We?” to “What’s Next?”

The shift Ryan describes, from cautious validation testing to a team that walks the floor identifying printing candidates, is the outcome that most production environments undervalue when they evaluate additive manufacturing. The ROI calculator captures part cost and labor hours. It doesn’t capture the compounding effect of a workforce that thinks about manufacturing differently. 

At Air & Liquid, the culture shift manifested in a few specific ways: 

• Part review changed. Instead of asking “how do we machine this?” engineers began asking whether the geometry was even worth machining. Parts that had always been made a certain way got reconsidered.
• The machine became load-bearing. Ryan describes the printer as running “almost nonstop” and “always making money.” That’s a different operating posture than a machine that sits idle between R&D projects.
• Small business competitive parity. Ryan frames this explicitly: “We’re taking the small business and we’re doing what all the big guys are doing too.” For OEM manufacturers competing against larger facilities, on-demand custom component production is a meaningful competitive differentiator. 

FOR PRODUCTION MANAGERS  The highest-leverage question to ask when evaluating your first additive application isn’t “what can this replace?” It’s “what do we currently make that we break tooling on, or that requires a skilled machinist to stand there for hours?” That’s your first target. Validate performance, prove the economics, and the culture shift tends to follow on its own.

What Air & Liquid Systems Actually Printed

Beyond the star-profile coupling, Ryan’s team described several other parts that moved to print during their adoption curve. Each one followed the same logic: a part that was expensive, slow, or difficult to make conventionally—and that additive manufacturing handled without those constraints. 

Plate assemblies: Originally produced by ordering individual plates, drilling them out, machining a connecting piece, and welding the assembly together. Ryan’s team replaced it entirely with a single print operation. No fixturing. No welding certification required. No lead time beyond print time. 

Retention clips: Small hardware previously ordered in bulk from a supplier or made one-off on a lathe. The printed version was described as “more robust” because printing can produce gussets, radiused roots, and optimized cross-sections that would add machining operations and cost if done conventionally.

Common Questions from Production Teams

Can 3D-printed parts actually replace CNC-machined metal components in production? In many factory floor applications, yes. The key is application selection. Parts with complex geometry, modest load requirements, and no extreme temperature or chemical exposure are strong candidates. Air & Liquid Systems ran a printed composite coupling under continuous actuation for one week—equivalent to roughly five years of service—and found no damage. The printed version also outperformed the machined original because it achieved the designed star-profile geometry that milling couldn’t produce. 

What does “composite” mean in this context? Is this just standard FDM plastic? 

No. Industrial composite 3D printing embeds continuous strands of reinforcement fiber (carbon fiber, fiberglass, or Kevlar) within a thermoplastic matrix during printing. The result is a part with directional strength properties more like a laminated composite than a thermoplastic. For mechanical applications—brackets, fittings, load-bearing fixtures—this distinction matters significantly. 

How long does it take to learn enough to print useful factory floor parts? 

Ryan Wenzlick had zero additive manufacturing experience before completing Markforged University, and his team was printing production parts shortly after. For straightforward replacement parts—geometry replacements for machined components—the CAD and slicing learning curve is measured in days to weeks, not months. 

What’s the right way to validate a printed part before putting it into production? 

For structural or mechanical applications, accelerated life testing is more informative than static strength tests alone. Simulate the actual loading condition—actuation cycles, vibration, pressure, temperature—at accelerated rates. For the Air & Liquid coupling, one week of continuous actuation (representing ~5 years of service) was the validation gate. 

At what production volume does 3D printing stop making sense economically? 

Additive manufacturing typically becomes less cost-competitive above several hundred to a few thousand parts per year for simple geometries, and remains competitive at higher volumes for complex parts that are expensive to machine. A facility printing a diverse range of low-to-mid volume parts across many SKUs can keep utilization high and economics favorable continuously.

The Bottom Line for Factory Teams

Ryan Wenzlick’s experience at Air & Liquid Systems isn’t unusual, but it is instructive. The path he followed (identify a painful machined part → validate performance rigorously → expand to other candidates → operate continuously) is replicable in most OEM or production environments. 

The economics of the coupling alone—$110 to $1.05—would justify the equipment in most facilities if that part ran in any volume. But the larger value is the operational posture: a machine running continuously, a team that thinks about manufacturing differently, and the ability to produce parts with geometry that conventional machining can’t achieve at all. 

Ryan’s closing line is worth taking literally: “Us without Markforged is tough to think about.” That’s not a marketing sentiment. That’s what it sounds like when additive manufacturing moves from experiment to infrastructure.

All of the blogs and the information contained within those blogs are copyright by Markforged, Inc. and may not be copied, modified, or adopted in any way without our written permission. Our blogs may contain our service marks or trademarks, as well as of those our affiliates. Your use of our blogs does not constitute any right or license for you to use our service marks or trademarks without our prior permission. Markforged Information provided in our blogs should not be considered professional advice. We are under no obligation to update or revise blogs based on new information, subsequent events, or otherwise.

Never miss an article

Subscribe to get new Markforged content in your inbox