Engineering

How We Wasted $3,200 on Our First Markforged FX20 — And What I Learned About Boundaries

Industrial additive manufacturing article feature

If you’re considering a Markforged FX20 or Metal X for production, here’s the short version: they’re incredible machines, but they solve a specific slice of your manufacturing needs — not all of them.

I’m a manufacturing engineer who’s been handling additive manufacturing orders for about six years. In that time I’ve personally made (and documented) seven significant mistakes that collectively wasted roughly $12,000 in scrapped parts, reworked tooling, and delayed deadlines. This article is the checklist I wish I’d had before our first Markforged order.

The core lesson: don’t treat a Markforged printer as a universal replacement for injection molding, CNC machining, or even other 3D printing processes. Know its strengths, respect its limits, and you’ll save both time and money.

Mistake #1: Ignoring the “Shot Capacity” of Injection Molding — And Expecting 3D Printing to Fill the Gap

In my first year (2019), we had a rush order for 500 polycarbonate housings. Our injection molding partner quoted a 6-week lead time because their machines were booked. I thought: “We have the Markforged FX20 — let’s just print them.”

The FX20 can print Onyx (carbon-fiber-filled nylon) beautifully. But the part geometry was designed for injection molding — thick walls, uniform shrinkage, draft angles. 3D printing each one took 8 hours. After the first 50 parts, we realized the surface finish was inconsistent, and the layer adhesion was weaker than a molded part. We ended up scrapping the entire run — $3,200 worth of material and machine time.

What I didn’t understand then: shot capacity of injection molding machines (the volume of material injected per cycle) allows for fast, repeatable production of identical parts. A 3D printer builds layer by layer — it’s slower and the mechanical properties are anisotropic. They serve different purposes. Now, when a vendor tells me their injection molder has a 200-gram shot capacity, I know exactly what that means for cycle time and part consistency. And I know that for low-volume prototyping or complex internal geometries, 3D printing wins. For 500 identical housings, injection molding wins — even if it means waiting.

Put another way: don’t use a Markforged printer to compete with injection molding on volume. Use it for parts that can’t be molded — internal channels, lightweight lattice structures, or short-run custom tooling.

Mistake #2: Using the Metal X to Make a Tool for Cutting Door Knob Holes — Without Understanding the Printer’s Workflow Limits

Around September 2022, we got a request for a custom jig that would guide a hole saw for cutting door knob holes in metal doors. The design called for a hardened steel guide bushing. We had the Markforged Metal X — perfect, right?

I designed the part, sent it to print, and… stopped. The printer needed to stay connected to a computer for the entire print — about 14 hours. Our IT policy locked down the workstation overnight. The print paused when the computer went to sleep. We lost a full day recalibrating. (Should mention: modern Markforged printers can run from an onboard controller or network connection, but at the time we hadn’t set that up.)

The bigger issue: the tool we printed required debinding and sintering. The shrinkage was 20%, and our hole tolerance was ±0.1 mm. We didn’t account for that. The bushing came out undersized. $890 wasted, plus a 1-week delay for a replacement.

Here’s something vendors won’t tell you: do 3D printers need to be connected to a computer? It depends. Most industrial units like the Markforged Metal X have a built-in touchscreen and can operate standalone via USB or network. But if your shop uses locked-down workstations, verify the workflow before starting a long print. Also, metal binder jetting requires post-processing — plan for that dimensional change.

I learned to keep a dedicated laptop for the printer, configure network printing, and always run a test coupon first to measure shrinkage.

Mistake #3: Overpromising “One-Stop-Shop” Capabilities — When We Should Have Said “This Isn’t Our Strength”

After the first two disasters, I became the internal expert on additive manufacturing. Colleagues started assuming that if we had a Markforged FX20 and a Metal X, we could replace all our traditional machining and injection molding. I almost bought into it myself.

Then came a request for a high-volume run of 10,000 nylon clips — the same part we’d prototyped on the Markforged with Onyx. I confidently quoted a 3-week turnaround. Three weeks later we had delivered only 400 parts, and the customer was furious.

The vendor who told me “this isn’t our strength — here’s who does it better” earned my trust for everything else. That’s what I do now. When someone asks for large-scale production, I recommend injection molding and explain why the shot capacity of their machine drives cost efficiency. When they need a one-off repair bracket with carbon fiber reinforcement, I show them the Markforged. When they want a custom tool for cutting door knob holes in field conditions, I suggest a printed jig but caution about post-processing.

To be fair, Markforged printers are exceptional for what they do: continuous carbon fiber reinforcement, metal parts with complex internal geometries, and rapid iteration without tooling. But they cannot (and should not) replace every downstream process. The most honest thing you can say is: “Here’s what we’re amazing at — and here’s what you should go elsewhere for.”

That boundary makes you look more professional, not less.

What I Do Now: A Pre‑Flight Checklist Before Any Markforged Print

  • Confirm the part is a good fit for additive manufacturing — is it low volume (under 100)? Does it have complex internal features? Do you need anisotropic properties in specific directions?
  • Check connectivity — printer plugged into network? Can you monitor remotely? Is there an emergency stop accessible?
  • Understand post‑processing — For Metal X, account for 20% shrinkage. For FX20, consider if continuous fiber reinforcement is needed.
  • Compare with injection molding — If quantity > 500 and geometry is simple, ask your injection partner about shot capacity and cycle time.
  • Quote honestly — If the job fits a specialist outside your printer’s strengths, refer it. It builds long‑term credibility.

We’ve caught 47 potential errors using this checklist in the past 18 months. Saved roughly $8,000 in scrapped parts and avoided three major delays.

Oh, and one more thing: you don’t need to keep the printer connected to a computer 24/7 — we now use a dedicated tablet with the Markforged Digital Forge platform. It’s not rocket science, but it’s the kind of detail that trips you up when you’re focusing on the part design instead of the process.

— A recovering mistake-maker who now runs a checklist.

Discuss this topic Start a prototype sprint

Jane Smith

I’m Jane Smith, a senior content writer with over 15 years of experience in the packaging and printing industry. I specialize in writing about the latest trends, technologies, and best practices in packaging design, sustainability, and printing techniques. My goal is to help businesses understand complex printing processes and design solutions that enhance both product packaging and brand visibility.