Laser Engraving Glass Cups: I Tried It. 240 Cups Cracked.

The Order That Seemed Too Easy

September 2023. A regional brewery asked me to engrave their logo on 250 pint glasses. Their regular supplier couldn't deliver before the release event, and they needed everything in three weeks. I had a fiber laser marking system sitting in the shop. I said yes within the hour.

Should have been straightforward. Put the glass under the beam, mark the logo, ship the boxes. I'd been running custom marking orders since 2018, and I'd learned to be careful with metals and plastics.

Glass felt like the easy one.

It wasn't.

It Wasn't the Settings. But I Treated It Like It Was.

I tested a cup, fired the laser, and watched the logo appear. Clean. Crisp. I texted the brewery that we were good to go.

First production cup cracked at the bottom edge of the logo. Second cup: same crack. By the fifth, I had a pattern on my bench and a knot in my stomach.

For the next two weeks, I ran through the usual troubleshooting playbook. Power down. Speed up. Lower the frequency. Adjust the focus. Swap a 110mm lens for a 160mm lens. Read every forum thread I could find on fiber laser glass marking.

Nothing worked. Some cups looked perfect out of the laser and cracked overnight. Others never made it out of the fixture. By the end of week two, 240 of the 250 cups had failed in one way or another.

It looked like a settings problem. It felt like a settings problem. Settings weren't the problem.

The Real Problem: 1,064nm vs. Glass

A fiber laser module produces light at roughly 1,064 nanometers. That wavelength gets absorbed well by metals, certain plastics, and dark anodized surfaces, leaving a clean permanent mark.

Clear glass doesn't absorb 1,064nm well. The beam mostly passes straight through. Instead of a clean mark, you get uneven heating, thermal stress, and hairline cracks. Sometimes immediately. Sometimes hours later.

A CO2 laser, by contrast, operates around 10.6μm, and bare glass absorbs that wavelength much more readily. That's why so many frosted-glass tutorials feature CO2 systems. My machine was never going to reproduce that result on bare glass.

Why does this matter to you? Because the laser wasn't actually marking the glass. It was heating its way through a material that didn't want to be heated. No amount of power tuning fixes a wavelength mismatch.

But here's the deeper reason this happened, and it's not about glass.

I bought my first fiber laser system the way I buy most tools: I compared three spec sheets, ran the numbers, and picked the best price. I asked about power, price, and lead time. I never asked what the system was actually good at.

I ran the numbers on three systems before buying. The spreadsheet said the budget option was 15% cheaper, with the same rated power. My gut said to talk to someone who'd actually used one first. I went with the spreadsheet. That gut instinct was worth about $5,000.

The supplier sent me a brochure and a YouTube link when I asked about glass marking. That should have told me everything I needed to know.

The Question That Keeps Coming Up

Around the same time, a customer asked me a question I still hear often: “Can you laser engrave paper?”

I almost said yes without testing. I'm glad I didn't. A fiber laser will scorch or ignite plain paper before it makes a clean mark. For paper and other organic materials, a CO2 laser is the right tool—not a fiber laser module.

The paper question and the glass question were the same question in different costumes: do you know what your laser is actually good at?

I thought I did. I had power ratings, lens specs, and a delivery date. What I didn't have was an understanding of how different materials absorb the beam.

That's the gap that cost me the job.

What It Actually Cost

Let's do the math, because I promised myself I'd never hand-wave this.

The contract was worth $4,800 in revenue, roughly $1,700 of which would have been margin. After the failure, I refunded the $1,500 deposit and paid $960 to replace the ruined glasses. I also lost two weeks of shop time—call it $1,200 to $1,500 in billable hours. Don't hold me to the exact labor figure; I'm reconstructing from memory. Direct hit: about $5,400.

The worst part was the cups that passed inspection. They sat in a box overnight and cracked anyway—quiet pops you could hear from across the shop. It wasn't settling. It was failing.

The bigger cost isn't a line item. The brewery won't consider us for future work. Their reorder would have been annual, and they'd probably have referred us to other local breweries. I'm not going to invent revenue projections for something that's gone. But the referrals were worth more than the contract itself.

Looking back, I should have spent $80 on sample cups and run a 48-hour test before quoting anything. At the time, the demo video looked so clean that I thought I could skip the boring step. That $80 would have changed the outcome.

And the part I'm least proud of: I nearly marketed my way out of the problem anyway. A handful of test cups looked great before they cracked. I was about to add “glass engraving now available” to our website when I stopped myself. Per FTC guidelines, advertising claims need to be truthful, not misleading, and substantiated. I had a few lucky test prints and a growing pile of broken glass. That's not evidence.

Can You Engrave Glass Cups? Yes—If You Do This

Direct answer: yes, you can engrave glass cups with a fiber laser. But not the way I tried to. The fix has four parts, and none of them involve endlessly adjusting power settings.

• Use a marking solution. CerMark, or any spray formulated for glass marking, creates a layer that absorbs 1,064nm. The mark develops cleanly, and the thermal cracking risk drops dramatically.
• Run a thermal test before production. Mark ten sample cups, then leave them on the bench for 48 hours. If they survive that, they'll probably survive delivery.
• Get the application engineers involved. Reading through IPG Photonics' news and application notes became a habit before I even bought anything from them. They publish material-specific guidance that rarely makes it into spec sheets. When I moved to the IPG Photonics IX-200, the hardware wasn't the only difference: the application engineers asked what I was marking before they talked about power. That one question was more useful than any spec sheet I'd ever read.
• Ask what's NOT included. Not just “what's the price?” but “what's not included?” Marking solutions, application support, training, beam delivery accessories. In my experience, the suppliers who answer that question honestly are the ones whose quotes end up being the real price.

This approach works for us, but we're a small contract shop running mostly 10-to-200-piece batches. If you're doing high-volume glassware, your setup—conveyor speed, positioning, cooling, scrap tolerance—will be different. The principles stay the same.

I keep the pre-flight checklist I built in Q1 2024 on my bench. We've logged 47 catches since then—materials that would have scorched, coatings that would have failed, and one batch of acrylic that would have melted into an expensive puddle. The first line of the checklist is still:

What is this material, and what does it do at 1,064 nanometers?

If I'd asked that before quoting the brewery, I'd have saved $5,400 and kept a client. Now it's the first question I ask on every new job. If you're about to start laser engraving glass cups, I'd recommend you ask it too.