The 7-Step Checklist I Use to Vet Carbon Fiber Prepreg Suppliers

The 7-Step Checklist I Use to Vet Carbon Fiber Prepreg Suppliers

If you've ever taken delivery of a "meets spec" batch of carbon fiber prepreg, watched it sit in the freezer looking normal, and then watched it fail during layup, you know that sinking feeling. I've been on the receiving end of it twice. Both times, the root cause was a question I should have asked before the PO, not after.

I'm the quality and compliance manager for a composites fabricator. Roughly 200 supplier audits and six years in, this is the checklist I actually use for prepreg sourcing. Seven steps. I'll say upfront: it works for industrial and commercial prepreg. If you're sourcing aerospace or medical-grade material, this is the entry ticket, not the whole building — you'll also need AS9100 and usually NADCAP coverage on top, and a materials engineer signing off. For everyone else, here's the sequence.

One admission before we start. When I got into supplier auditing, I assumed the tensile and flexural numbers on the data sheet were the whole story. Two years and one $18,000 scrapped layup later, I learned that a data sheet tells you what the material can do — not what the supplier will reliably ship.

The Checklist at a Glance

Seven steps, ordered by how often suppliers get vague. If you only do three, do 1, 3, and 6.

1. Pin down the resin system, cure cycle, and working life
2. Ask for tolerances, not typical values
3. Require batch-level COAs on every shipment
4. Confirm volatiles content control
5. Check storage and transport — cold chain is not optional
6. Review batch-to-batch consistency, not just the sample data
7. Match the certification to your end use

Step 1: Pin Down the Resin System, Cure Cycle, and Working Life

This is the first filter, and the one where vague selling starts showing up. When a rep says "epoxy system," what they usually mean is one of several things: a 180°C-cure toughened epoxy with a two-hour hold? A 130°C fast-cure version with a 30-minute ramp? Out-time of 30 days at 23°C, or 10?

This matters because different resin systems are not drop-in substitutes. A press tuned for a two-hour cure will not produce the same part with a 30-minute material, no matter what the data sheet says about mechanicals.

The output of this step is a written spec sheet — resin chemistry, cure temperature, ramp profile, hold time, out-time, storage conditions. Not a data sheet. A spec sheet your supplier signs. If they won't put the cure cycle in writing, that is a red flag, and in my experience it is most common when working through brokers who know less about the actual production line than they claim. When evaluating carbon fiber prepreg manufacturers directly, this is the first thing I ask for.

Step 2: Ask for Tolerances, Not Typical Values

"150 gsm areal weight, 35% resin content" on a data sheet is close to useless without a tolerance. What you need is: is areal weight held to ±3 gsm or ±10 gsm? Is resin content ±1% or ±3%?

I've watched too many suppliers answer "we can hit 35%" using the typical value, then produce batches that actually range from 32% to 38%. Testing to ASTM D3529 for resin content (and ASTM D3530 for volatiles, more on that in a minute) is fine — but the test method only tells you how to measure. It doesn't tell you what tolerance to accept. That's your call, and it belongs in your purchase order.

The practical move: put your tolerances in the PO and get written confirmation. If they can't hold them, you find out before signing, not three months later in the scrap report.

Step 3: Require Batch-Level COAs on Every Shipment

A sample data sheet and a batch COA are different documents serving different purposes. The data sheet is a sales tool. The COA is a quality record.

My rule: every batch needs a COA tied to a real lot number, with test results, test date, and an operator sign-off. At minimum, it should cover areal weight, resin content, volatiles, and gel time. A COA without a lot number is a piece of paper, however pretty the numbers on it.

If you can pull the last 10 COAs into one spreadsheet, that table tells you more about a supplier than any polished sample report. A lot of buyers skip this step because it requires a supplier willing to share traceable data — but willing vs. unable is itself a useful filter.

Step 4: Confirm Volatiles Content Control

This is the step most people miss on their first pass, and it's one of the details I weight most heavily. Volatiles content — the fraction of residual solvent and moisture that gets driven off during heat-up or layup — affects porosity, delamination, and post-cure properties. Get it wrong and no amount of good tensile data saves you.

It's measured under ASTM D3530, and the limit depends on your resin system and end use; for many industrial systems, roughly 1–2% is the controlled range. The problem is that many suppliers either omit it from the COA or write "low." Then ask: low based on what test, taken when?

The checkpoint for this step: every COA has a volatiles result with a test method behind it. If a supplier says "we've always been stable" without numbers, that's a claim, not data. (Unfortunately, this is one of the last things sales teams think to prepare, so push for it early.)

Step 5: Check Storage and Transport — Cold Chain Is Not Optional

Carbon fiber prepreg is not ambient freight. Most industrial prepreg stores around −18°C, and shipping typically uses dry ice or refrigerated transport. The working out-time between pulling from the freezer and layup is also a real spec, not a footnote.

I've seen quotes that default to "ambient shipping" and explain it away as "short haul, no problem." A short haul may well be fine — but if you don't have data, don't bet the batch on it. The step here is simple: get written confirmation of storage temperature and shipping method, and ask whether they provide transport temperature records. If they can't, you may be receiving material that looks fine and is already outside its working window.

Step 6: Review Batch-to-Batch Consistency, Not Just the Sample Data

This is the step that takes the most time and is, honestly, the one that pays back the most. Ask the supplier for the last 10 batches of key parameters — areal weight, resin content, volatiles, gel time — and look at the spread, not the average.

A clean average with wide spread is worse than a slightly-behind average with tight spread. I learned this the hard way. I had a sample batch with beautiful numbers — tensile and flexural both over spec. The first two production batches were fine. The third was 8 gsm off on areal weight, and the parts it went into were a write-off. Looking back, I should have asked for batch-to-batch data before the first PO. At the time, the rep said "our data is all pretty standard," and I took it at face value. It took me three years and a few hundred thousand dollars in scrap to internalize something simple: a supplier willing to perform well on one batch and a supplier able to hold consistency across a hundred are two different companies.

When you ask for the 10-batch table, you're not just checking the numbers. You're checking whether they can produce the numbers. That second signal tells you a lot about how your future POs will go with prepreg carbon fiber suppliers.

Step 7: Match the Certification to Your End Use

ISO 9001 is a floor, not a selling point. The question to ask is: does that certification cover the specific plant I'm ordering from? Aerospace work generally needs AS9100, and certain processes trigger NADCAP approval on top of that. Medical and rail have their own regimes.

My working principle: if the certification doesn't match your end use, it doesn't count. Don't get distracted by a wall of framed certificates. I'm not 100% sure any single audit will catch every gap, but asking "which plant, which scope, which audit year" usually surfaces the answer within two emails.

Notes and Common Mistakes

A list of what goes wrong is more useful than a list of what goes right:

• Treating typical values as guaranteed. "Typical" is sales language; your PO needs quality language.
• Skipping the cold chain. If storage and transport aren't in the contract, the risk is yours, not theirs.
• Judging on the sample batch. Sample performance and production consistency are different problems.
• Treating "epoxy" as a generic term. Different resin systems have completely different process windows.
• No out-time logs. Track when material left the freezer and how long it's been out — internally, on your side.

Last thing, and this is the honest-limitation part. This checklist is calibrated for industrial and commercial prepreg sourcing. If you're buying prototype quantities for development work, steps 3 and 6 can be loosened — at that stage you should be more interested in whether the supplier will engage with your process questions than in COA completeness. If you're heading toward aerospace or medical, the list above is necessary but not sufficient. The ASTM and certification references here were accurate as of early 2025; check current revisions before you quote them in a supplier contract, because test methods do get updated.