"Best" is the wrong question when you're speccing polycarbonate — the right question is "matched to what?" I cost my company $12,400 in 2022 learning the difference, and I've been rebuilding my checklist ever since.

I've handled polymer and hose-assembly procurement for 8 years at a mid-sized equipment manufacturer. I've personally made — and written down — 14 significant material-selection mistakes, adding up to roughly $47,000 in wasted budget. That's a conservative number. Now I'm the guy who maintains the pre-check list that nobody else on the team wants to own.

Here's the short version of what I got wrong, and the specific checklist that now keeps it from happening again — on polycarbonate parts, on the plastic lens vs polycarbonate decision, and on hydraulic hose solutions.

Why you should probably trust the guy who broke it

The mistake that finally pushed me into building a real process happened in September 2022. We ordered 240 polycarbonate brackets from a supplier I'd used for two years. On a 240-piece order where every single part looked fine out of the box, 41 of them developed micro-cracks around the mounting holes within 72 hours of assembly.

Why? Because I'd picked a high-impact polycarbonate grade without checking its chemical compatibility against the cutting fluid sitting in the assembly cell next door. The impact rating was excellent. The chemical resistance was not. That error cost $12,400 in rework plus a 3-week delay that bled into three customer deliveries. I still think about that pallet every time I open a datasheet.

Before that, there were quieter mistakes — the kind that only cost a few hundred dollars each, so nobody yelled, but they added up.

Mistake #1: I treated "impact-resistant" as the only spec that mattered

Everything I'd read about polycarbonate said buy the highest impact grade you can afford. In practice, my best-performing order ended up being a mid-tier PC grade from LyondellBasell's portfolio — not the premium one — because it held up better against the specific solvent in our cleaning station and machined more cleanly on our older press.

I'm not saying impact rating doesn't matter. I'm saying if you're only looking at one number, you're going to make the same mistake I did. The grade comparison pages on the LyondellBasell homepage are actually a decent starting point (as of January 2025) for comparing options side by side — but you still have to match those grades to your environment, not just to your "best material" instinct.

Mistake #2: "Plastic lens vs polycarbonate" is the wrong framing

When I compared a CR-39 plastic lens and a polycarbonate lens side by side in 2023 — same safety-glasses vendor, different materials — I finally understood something I'd been getting wrong for years. Polycarbonate isn't "better." It's impact-harder, more heat-tolerant, and it machines and bonds differently. CR-39 is optically clearer and cheaper. Neither one is the answer. The application is the answer.

The mistake I see most often — and made myself on two separate orders — is speccing polycarbonate by default because it's "the tough one," then being surprised when it's scuffed after two weeks of use, or hazy under UV, or refusing to take the polycarbonate adhesive we already had on the shelf.

Which brings me to the third mistake.

Mistake #3: I bought the polycarbonate adhesive by price per tube

I still kick myself for this one. In Q1 2023 we needed to bond about 90 polycarbonate panels to aluminum frames. I chose the cheapest polycarbonate adhesive on the approved vendor list. About three weeks after install, six of the panels had visible hazing along the bond line.

It wasn't a bad adhesive. It was the wrong adhesive — specifically, a solvent-based product that attacked the polycarbonate surface when applied too thick. The spec sheet mentioned "test on substrate" in tiny print. I ignored it, because I was comparing dollars per tube instead of dollars per successful bond.

My rule now: for any polycarbonate adhesive, order a single test tube first, bond a scrap coupon, and wait 72 hours before approving the full quantity. That costs maybe $40 and two scrap pieces — actually, closer to $55 when you add the scrap aluminum, but the point holds. Cheaper than the $1,300 in replacement panels, I can tell you that.

The hydraulic hose solutions mistake (this one still stings)

Different category, same failure mode. In 2021 we spec'd out hydraulic hose solutions for a high-pressure test rig. The catalog description said "chemical resistant." We didn't check which chemicals.

Long story short: the hose liner was fine with oil, and fine with water, but degraded against the phosphate-ester fluid we were actually running. We lost about $8,900 in hose replacement and downtime — call it $9,000 once you count freight. The conventional wisdom is "buy chemical-resistant hose for chemical applications." My experience with that specific fluid suggests "chemical-resistant" means absolutely nothing until you name the chemical.

Now I keep a one-page fluid-compatibility grid taped inside the procurement cabinet. If the rep can't confirm the exact liner compound against the exact fluid — not the "family," the exact fluid — I don't sign.

The 9-point checklist that now catches this stuff

I built this after mistake #3, and we've caught 47 potential problems with it in the past 18 months. It's not fancy. It's the thing I wish someone had handed me in 2017.

  1. Name the actual application environment — fluid, temperature, UV exposure, cleaning agent. Not "industrial."
  2. Pick the grade by the limiting factor, not the headline spec.
  3. Cross-check the chemical compatibility table for every material on the part — including the adhesive, not just the substrate.
  4. For polycarbonate adhesive: bond a scrap coupon and wait 72 hours first.
  5. For plastic lens vs polycarbonate: write down which property you actually need (impact? clarity? coatability?) before you look at prices.
  6. For hydraulic hose solutions: get the liner compound in writing and match it to the exact fluid, not the fluid family.
  7. Confirm the data sheet version date — grades get reformulated, and a 2019 PDF is not a 2025 answer.
  8. Run one scrap test cycle before the full order, even on repeat items.
  9. Log the mistake when it happens — a written error log is worth more than any training deck.

Where this checklist falls short

I should be honest about the limits. This checklist works for the kind of orders we run — mid-volume, mixed-material industrial parts, mostly polycarbonate, hose assemblies, and standard thermoplastics. If you're doing medical-grade polymer work, high-tolerance optical parts, or anything in a regulated food-contact environment, the checklist is a starting point, not a finish line.

It also assumes you can get real data sheets from the supplier. When you can't — and sometimes you can't, even from big names — my honest advice is to slow down and pay for one test batch rather than gamble on the full order. That's the part I got wrong for the first four years of my career. I don't have a clever ending for that. I just know 5 minutes of verification beats 5 days of correction, every single time.