Fixing Scratched Acrylic Surfaces

Had a guy call me last Tuesday absolutely convinced his $2,800 display case was totaled. Three months old, deep scratch across the front, wanted a rush replacement. I asked him to send a photo.

 

It was a fingernail scratch. Maybe 30 minutes of work with NOVUS #2.

 

I could've sold him a new panel. Would've been an easy $800 after markup. But here's the thing-I've been doing this for five years now and honestly? The replacement game is exhausting. You quote it, fabricate it, ship it, install it, and then six months later they scratch the new one too because nobody taught their cleaning crew what Windex does to acrylic.

 

So I told him to fix it himself. Probably lost money on that call.

Okay look, before I get into the actual repair stuff, I need to rant about something that drove me absolutely insane before.

COVID ruined everything.

Not in the obvious ways-I mean, yeah, supply chain, whatever, we all dealt with that. I'm talking about cleaning protocols. Every single client we have switched to aggressive alcohol sanitization in 2020 and half of them never stopped. Do you know what alcohol does to PMMA over time? It penetrates the surface layer, creates internal stress, and then one day you hit the panel with a heat gun for some unrelated work and boom-crazing everywhere.

We've had maybe fifteen cases this year alone where clients called about "scratches" that turned out to be chemical stress fractures. Completely different problem. Can't really fix crazing the same way. Most of those panels needed full replacement.

The frustrating part? ACRYLITE's own maintenance guide specifically says don't use alcohol. Don't use ammonia. Don't use acetone. Don't use any aromatic solvents. But try getting a facilities manager to read a maintenance guide. They just grab whatever's under the sink.

Anyway. Scratches.

The fingernail thing actually works

I didn't invent this. Picked it up from Jerry, who's been in the trade since before I was born. Run your fingernail perpendicular to the scratch-not along it, across it. Like you're trying to stop at the scratch line.

There's nuance here that the fingernail test doesn't capture, though. Jerry can look at a scratch and tell you within 10 minutes what the final result will be. I'm not there yet. Some scratches that feel medium turn out to be deeper than expected. Some that catch your nail hard are actually just sharp-edged but shallow. You get a feel for it after a few hundred repairs, but honestly I still misdiagnose maybe 10-15% of the time on first assessment.

The actual repair process, more or less

I say "more or less" because every tech has their own quirks. This is how I do it:

Light scratches: NOVUS #2 Fine Scratch Remover. Eight-ounce bottle runs about eleven bucks retail. Microfiber cloth, circular motions, don't use paper towels. Paper towels will create new scratches and you'll be chasing your tail all day.

Medium scratches: Wet sand with 1000-grit (I use 3M Wetordry because that's what we've always stocked, honestly couldn't tell you if there's something better), then 1500, then 2000. After sanding, NOVUS #3 first-that's the heavy scratch remover with the larger abrasive particles-then #2, then #1 for final cleanup.

Also-keep everything wet. Dry sanding generates heat, acrylic softens around 105°C, and then you're dragging material around instead of cutting cleanly. It makes a mess and the finish is garbage.

Deep scratches: Same process but start at 600-grit. Sometimes 400 if it's really nasty. Add 800 between 600 and 1000. This is 5-8 hours of labor for a large panel. Tedious, but it works.

That museum job I'd rather forget

We had a display case last year, museum client, medium-sized panel with a 4-inch scratch. One of our techs sanded just the scratch area-maybe a 6-inch circle. Looked great when we delivered it. Scratch was completely gone, panel was clear.

Two days later the client calls. There's "a weird distortion."

What happened: localized sanding created a subtle depression in the panel surface. Not visible when you look at the panel directly. But when you looked through it at the displayed artifact? Lens effect. The object behind it had this very slight warping that was driving the curator crazy.

We redid the whole panel. Six hours of work we didn't get paid for. Client was happy in the end but I still cringe when I think about that invoice.

On the cast vs. extruded debate

I know I'm going to sound like a salesman here but I genuinely believe this: cast acrylic is worth the upcharge in high-touch applications.

The technical reason is molecular weight and polymerization completeness. Cast has higher molecular weight, better scratch resistance, better chemical resistance, lower internal stress. Extruded costs 15-20% less but in retail environments we see extruded panels need attention 2-3 times more often.

Had a cosmetics display client a couple years back who insisted on extruded to save budget. Six months in, their panels looked terrible just from routine cleaning. They spent more on repairs in year one than they would've saved over three years going with cast from the start. I don't even quote extruded for retail applications anymore unless the client really pushes.

The safety thing I probably should've mentioned earlier

Panel thickness matters a lot for repair decisions. Anything under 3mm, I usually recommend replacement over repair. The material removal from sanding can compromise the remaining thickness too much.

For safety-critical applications-aquarium panels, protective barriers, anything load-bearing-I don't even quote repair. Straight to replacement with proper engineering review. The liability isn't worth it.

There was a shopping mall aquarium failure in Shanghai back in 2010 that stuck with me. 15cm thick panel, 33 tons of water, failed catastrophically. Investigation pointed to installation-induced scratches that became stress concentration points. Fifteen people injured. That panel was way more robust than any display case I work on, and it still failed because deep scratches weakened it.

Large aquarium projects now spec safety factors of 11-14x because of incidents like that. I'm not an engineer but I know enough to stay out of structural applications.

The numbers, since I know someone's going to ask

So you're saving 40-55% on repair vs. replacement, depending on severity.

We have a retail chain client with 47 locations who averages 2-3 repairs per store per year. At roughly $500 savings per repair, that's $47K-$70K annually they're not spending on replacement panels. Makes sense for them to have a repair-first policy.

Whether it makes sense for you depends on your situation. If you have one display case and it's scratched, the math is simpler-get quotes for both and pick the cheaper option that meets your quality standards.

A few things I tell every client after installation

• Specify cast acrylic for high-touch areas. Yes, it costs more. No, the 15% savings on extruded isn't worth it if you're replacing panels three times as often.
   
• Get the cleaning protocol in writing. Actually write it down and post it somewhere the cleaning crew will see it. Mild soap and water is fine. Anything else creates risk.
   
• Keep the protective film on as long as possible. During installation, transit, any adjacent construction work. That film is cheap insurance.

I should probably wrap this up. If you've read this far, the short version is: most scratches are fixable, the fingernail test works for initial assessment, wet sanding and polishing is tedious but effective, don't repair safety-critical applications, and for the love of god stop cleaning acrylic with alcohol.