Hanwha polymer resin support for EVA, compound, and industrial conversion teams. Request documentation
Hanwha material note

Why Our SMC Plastic Parts Kept Failing — and It Wasn't the Price Sheet

2026-11-17

I blamed the molder for two years of failing SMC plastic parts. The real problem was that my RFQ said 'SMC, gray, 600 pcs' and nothing else. Here's what I got wrong.

The PO was for 600 compression-molded SMC covers — about $13,800 at the unit price we'd agreed on. It was 2022, and I approved it on a Thursday. Three weeks later, 38 of them wouldn't seat on the housing they were designed for, and the ones that did were visibly lighter than the sample sitting on my desk.

I've been handling purchasing for a 240-person company since 2020 — facilities, office, MRO, and whatever the plant manager needs that doesn't fit anywhere else. Roughly $2.1M a year across 14 vendors. I'm good at that job. But I didn't know what I was buying when I bought that first batch of SMC, and it took me about 18 months and one stopped assembly line to figure that out.

What I thought the problem was

Ask me in 2022 and I'd have told you the problem was the molder. Small shop, second-generation ownership, cheapest of the three quotes I collected. My conclusion was embarrassingly simple: you get what you pay for, we should've paid the 18% more the other guy wanted, lesson learned.

That conclusion was wrong in a way that cost us again the following year.

Here's what I actually thought I was buying: a plastic part. Same mental category as the polypropylene brackets we buy for cable trays, same as the nylon bushings on the conveyor. Material plus dimensions plus quantity — pick the lowest price that meets spec.

If that sounds basic, consider what happened the first time I searched for it. Typing 'SMC' into a search engine in 2022 returned a kennel in Ohio called SMC Mastiffs, a community college, a medical center, and a sheet-molding compound supplier — in roughly that order. It took me a week to understand that 'SMC plastic' wasn't a plastic at all in the sense I was used to.

SMC isn't a material. It's a recipe with a process attached.

Sheet molding compound is a composite. A typical formulation is unsaturated polyester resin, 40–50% calcium carbonate filler by weight, 25–30% chopped glass fiber, magnesium oxide as a thickener, an internal mold release agent, a peroxide initiator, and — this is the one nobody puts on a quote — sometimes a low-shrink additive.

Change any of those and you've changed the part. Drop the glass content from 28% to 22% and the part still looks right coming out of the mold. It still passes a visual check. It fails six weeks later when somebody over-torques a bolt into a boss.

You can measure it. Burn-off testing per ISO 1172 separates glass content from mineral filler; ASTM D2584 gives you total ignition loss. Takes about an hour and a muffle furnace. Nobody runs it on incoming shipments unless somebody asks them to.

Then there's the process half. Compression molding at 140–150°C for a cycle time long enough to actually cure the part — or a cycle time short enough to hit a quoted piece price. If the part isn't fully cured in the tool, it keeps curing on your shelf. It shrinks after you've measured it. It cracks at the inserts three weeks into service.

Here's the causation reversal that took me longest to see. I assumed cheap molders produce bad parts. What actually happens is that molders who can't control their process have to compete on price, and the way they hit a lower price is by removing the things you can't see — glass content, the low-shrink additive, cure time, a proper post-bake. Vendors who deliver consistent quality can charge more. The causation runs the other way from what I assumed.

What the mismatch actually cost

41 scrapped parts at $23 each — $943, not the end of the world. The mold rework to move the boss was $3,800. The assembly line sat idle for a day and a half because we couldn't build finished units without covers.

The invoice was the part that got me in trouble. The molder shipped the replacement batch without a PO revision — verbal go-ahead from me, which was my mistake — and finance bounced the $2,400 invoice back to my department because the amount didn't match the open PO. I spent two weeks explaining to my VP why a purchase I'd made to fix a problem had created a second problem.

The other thing I was getting wrong at the same time

While I was chasing SMC quotes that spring, I was also sourcing workwear for the shop floor — the summer order included nylon briefs and undershirts, about $9 a unit, 300 units. Same purchasing system, same three-bid process, same spreadsheet.

That's a mistake I'd do differently now. A $2,700 uniform order and a $14,000 tooling-dependent molded part are not the same purchase, and running both through the same three-bid workflow is how you end up treating a molding process like a commodity.

The material questions had the same shape. Somebody on the plant side asked me to price out UV resistant plastics for an outdoor equipment cover. I got quotes on polypropylene and nylon and put them side by side with a note that PP was cheaper. That comparison was useless.

Polypropylene vs. nylon isn't a question with an answer outside of an application. PP melts around 160°C, nylon 6 closer to 220°C, so nylon holds up better near heat. But PP absorbs almost no water — well under 0.1% in a 24-hour immersion per ASTM D570 — while unfilled nylon 6 picks up somewhere around 1.5%, and that water changes its dimensions. On UV, neither one is naturally good at it. PP needs a stabilization package (HALS plus a UV absorber) or it chalks and embrittles outdoors within a couple of years. Nylon usually gets carbon black. 'UV resistant' isn't a material property. It's an additive package plus a surface.

And I'd been treating polypropylene resin suppliers as if they were interchangeable with molders. They aren't. One sells pellets by the truckload and the other sells a finished part and takes on the process risk. Both were on the same vendor comparison sheet, which tells you how much I understood at the time.

Where the process actually broke — and the three things we changed

None of this was the molder being dishonest. The RFQ I sent said 'SMC, color gray, 600 pcs, per drawing.' That's not a specification. That's a wish with a quantity attached.

1. The spec sheet names materials and process

The drawing package now carries a one-page material spec: glass content by weight with the test method named (ISO 1172), filler range, low-shrink additive required, mold temperature range, minimum cure time, post-bake if applicable, and the color as a Pantone number with a Delta E tolerance. Industry practice puts Delta E under 2 as a match for brand-critical colors; 2 to 4 is visible to a trained eye; above 4 most people notice without being told. For a gray equipment cover, we spec'd Delta E 2.0 max. Nobody's argued about it since.

2. First-article sign-off, in writing, with samples kept

Three retained samples in labeled bags — one in the conference room, one at the molder, one in the plant — plus the parameters the molder ran at. We skipped this on the first order because we'd worked with the shop before and I thought, what are the odds? The odds caught up with me on 38 parts.

3. Three questions before any quote gets compared

  • How do you verify glass content, and can I see a burn-off report from the last six months?
  • Is there a low-shrink additive in this formulation, and what shrink rate did you design the tool around?
  • What's your cure time, and do you post-bake?

A molder who answers each of those in a sentence is controlling their process. A molder who gets vague is hoping you won't ask.

The part about specialists

Two of the three shops we shortlisted said some version of 'yeah, we can do that' to every line on the RFQ, including the flatness callout on a 24-inch panel. The third said: 'We don't do large flat panels with tight flatness — that's a different tooling approach. Here's who does it well.'

We moved the panel to the shop they named and gave the third one everything else.

The vendor who said 'this isn't our strength — here's who does it better' earned my trust for everything else they quoted. I'd rather work with a specialist who knows their limits than a generalist who overpromises. When a supplier tells you what they're not set up for, they're telling you they know what they are set up for — and that's the only thing that makes a quote mean anything.

I still don't know much about polymer chemistry. What I know is that 'SMC plastic' on a requisition is the opening of a conversation, not the end of one. And that 18% price difference I was so proud of avoiding in 2022 was never really a price difference. It was a difference in what was inside the part.

Ask Hanwha about this topic

Send a related resin question