Laser Cutting Paper, Plexiglass, and Welding Metal: Why One Setting Doesn't Fit All

I've reviewed enough laser process sheets to know one thing for certain: there's no such thing as a universal setting. When I first moved into quality management for laser systems, I assumed a machine that cut paper could handle acrylic with a few parameter tweaks. My first 40 acrylic parts came out with fuzzy, melted edges. That batch cost us roughly $600 in material and a full day of rework. It also taught me to stop treating "laser cutting" as a single process.

Here's the thing—laser processing splits into different scenarios, and the settings that work for one will wreck another. If you've ever pulled a process sheet off the internet and wondered why it didn't transfer to your material, that's probably why.

The Three Variables That Decide Everything

Before I get into specific materials, it helps to understand what's actually driving the differences. Three variables matter most:

  • Wavelength absorption. How well your material absorbs the laser's energy at a given wavelength. This is the big one—it decides which laser type you can even use.
  • Thermal conductivity. How fast heat spreads away from the cut. High conductivity means heat bleeds into surrounding material (bad for precision); low conductivity means heat stays put (bad for melting).
  • Edge tolerance. Whether the cut edge is a cosmetic surface or a functional one. This sets how much you're willing to slow down or add gas assist.

Get these three right and the rest is tuning. Get them wrong and no amount of parameter tweaking will save you.

Scenario A: Paper, Cardstock, and Thin Organics

This is the scenario most people start with, and it's also where overconfidence does the most damage. When people ask me about laser cutting paper settings, they usually want a single recipe. There isn't one—but the failure modes are at least predictable.

The main issue is edge charring. Run too hot or too slow and you get a brown, brittle edge that crumbles when handled. The counterintuitive part: don't max out speed with minimum power. That combination gives you a rough, stringy edge because the beam doesn't spend enough time to cleanly vaporize the material. A moderate power setting with a slightly defocused beam and a moderate speed usually produces a cleaner edge than the "fast and light" approach most tutorials recommend.

For paper under about 0.5 mm, air assist is optional but helps. Above that, use it—the smoke has nowhere to go and will deposit on the lens.

One more thing: paper is hygroscopic. A ream that's been sitting in a humid shop for a week will cut differently than one straight from a sealed box. If you're running production, store your stock in the same environment as the laser.

Scenario B: Acrylic (Plexiglass) — Yes, But the Wavelength Matters

"Can you laser cut plexiglass?" is one of those questions where the honest answer is "yes, with the right laser." Not every laser. This is the scenario where wavelength stops being a detail and becomes the whole decision.

Acrylic absorbs CO2 laser energy (10.6 μm) extremely well—over 90% at the surface. That means a CO2 laser cuts it cleanly with a flame-polished edge. A fiber laser at roughly 1.06 μm, on the other hand, passes through acrylic almost like glass. You'll get melting, scorching, and a mess—not a cut. This isn't a knock on fiber lasers; they're just optimized for metals, where that wavelength gets absorbed efficiently. Match the tool to the material.

Edge quality depends on the acrylic type. Cast acrylic cuts cleanly with a polished edge. Extruded acrylic tends to bubble and leave a rougher finish. If the edge is visible in the final product, cast is usually the no-brainer choice, even at higher cost.

Thickness matters too. Up to about 6 mm you can single-pass with a decent CO2 system. Beyond 12 mm you're looking at multiple passes or a different process entirely. Don't let anyone tell you a 150W CO2 will cleanly cut 25 mm acrylic in one pass—it'll cut, but the edge will look like it was chewed.

Scenario C: Metal Welding — Different Physics, Different Checks

Metal laser welding is not "cutting with more power." The physics are genuinely different, and the quality checks have to be too. This is where I've seen the most expensive mistakes—and where prevention pays off the most.

The basics: you're running a high-power fiber laser (typically 1–6 kW for production welding), using shielding gas to prevent oxidation, and controlling a melt pool that's moving at millimeters per second. The variables that wreck a weld are heat input, focal position, gas coverage, and fit-up. Get fit-up wrong and no amount of laser power will save you—you'll just blow a bigger hole faster.

Here's where I learned my most expensive lesson. We had a batch of stainless brackets where the weld looked fine to the naked eye. Six weeks later, three of them failed in the field—lack-of-fusion defects that visual inspection never caught. The rework and field service cost us about $22,000. Looking back, I should have insisted on a pre-production weld qualification with cross-sectioning. At the time, the schedule felt tight and the coupons looked good. They weren't representative.

Now we monitor the process, not just the output. And that's where camera specs start to matter. When I was speccing a weld monitoring setup last year, I pulled the Edmund Optics #68-576 camera specs first—it's the natural starting point when you're comparing options. The lesson I took from that exercise: for post-weld inspection, resolution and shutter type dominate. But for catching the weld pool oscillating, frame rate is the spec that decides whether you can see the defect at all. A camera with a frame rate in the class of the Edmund Optics 33-163 might suit that job—but match the frame rate to your actual weld speed, don't pick a model number and hope.

The bottom line on metal welding: put your money into pre-production qualification and in-process monitoring. Post-weld inspection catches some defects, but by then you've already paid for the material.

How to Figure Out Which Scenario You're In

Most people try to categorize by material name. That's a mistake. Categorize by these four questions instead:

  1. What wavelength does your material absorb? If it's transparent at 1.06 μm (acrylic, glass, some plastics), you need CO2. If it's metallic, fiber is likely your tool.
  2. How much heat spread can you tolerate? Thin, heat-sensitive materials need lower power and faster travel. Thick or conductive materials need more power and slower travel, but with better gas coverage.
  3. Is the edge cosmetic or functional? Cosmetic edges justify slower speeds and gas assist. Functional edges (where you'll weld or bond later) care more about dimensional accuracy than surface finish.
  4. Do you need in-process monitoring? If a defect is invisible until the part fails, you're in Scenario C—and you need a camera, not just a checklist.

If you answer those four, the scenario usually picks itself. And if you're on the fence between two, default to the one with the stricter quality check. It's cheaper to over-verify than to rework.

The Checklist I Wish I'd Had Earlier

Here's the 6-point check I now run before any new laser process goes into production. It's saved us an estimated $8,000 in potential rework over the past two years:

  1. Confirm the wavelength actually matches the material (absorption curve, not guesswork).
  2. Run a parameter ladder on scrap—three power levels, three speeds, three focal positions.
  3. Inspect the edge under magnification, not just by eye.
  4. For welds, cross-section at least three samples from different positions in the batch.
  5. Verify your monitoring setup can actually see the defect you care about (frame rate, resolution, lighting).
  6. Write down the winning parameters and the material batch they came from. Material properties drift.

Five minutes of verification beats five days of correction. That's not a slogan—it's the math I keep coming back to every time a "quick job" turns into a rework ticket.

One caveat: wavelength and absorption figures here reflect standard material properties as of 2024. Specialty coatings and composite materials can behave differently—verify absorption curves with your material supplier before committing to a process.

Leave a Reply