CATRA numbers only mean something when the sample is pinned down. A higher score on a CATRA test knife does not prove the kitchen knife will cut better in your customer’s hand. Edge angle matters. So do 15° per side, 58 HRC steel, the final belt grit, burr removal, and even whether the lab clamped the blade the same way as last time.
If you buy from China, 9 out of 10 suppliers can show a strong number when the sample is prepared for the report. In our Yangjiang, Zhejiang factory in China, where 240 people build OEM and private-label programs, we run edge retention QC as a specification job: blade profile, grind, HRC band, sample size, and lab method get fixed before anyone compares reports. A normal custom run might start at 1,000 pcs MOQ with 35-45 day lead time, but the CATRA sheet still has to read like a process record. Not a sales poster. We have seen this go sideways when the buyer flagged a “same steel” claim, then QC pulled the sample with the angle gauge and found the edge was 3° thinner than the approved golden sample.
What CATRA actually measures
CATRA knife edge retention testing is a controlled cut test. The blade cuts standard test media, and the machine records how quickly cutting performance drops. Repeatability is the point. Same steel grade and heat lot, same bevel angle in degrees, same edge finish from the grinding belt, same clamp setup on the tester. Then we can compare one knife construction against another without hand-test noise. On our grinding line, we check the apex under a 200x scope before samples leave the shop; if the burr is still there, the CATRA result is not worth much.
The key number you will usually see is TCC, or Total Card Cut. In buyer language, a higher TCC means the knife kept cutting after more card travel, for example 450 TCC vs 320 TCC on the same blade pattern. Simple enough. Some reports also show initial sharpness or early-stage cutting behavior. That tells you how the edge starts, not how it lasts. A blade can feel razor sharp on the first pass and still drop fast if the heat treat misses by 1 HRC or the bevel is too thin for that steel. We saw that on a 58 HRC trial lot, and the buyer flagged a 0.3 mm bevel change right away.
CATRA is a wear comparison, not a quality stamp. It does not tell you how the knife handles bone, resists corrosion, or feels after 200 kitchen cycles. That matters for chef knives and pocket knives. This is the wrong question to ask if you want one score to cover every use. On a 3,000-piece run, the first carton can pass and the last carton can drift if the tempering oven walks 5 °C or the polishing stone changes shape. QC pulled samples from carton 47 once for this exact reason.
Read the report like a buyer
The fastest way to get misled is to compare CATRA numbers without the setup sheet. A TCC score on its own means little if the report leaves out edge angle, grit sequence, or heat-treat batch. We run this test work on export orders all the time: QC pulled two knives off the grinding line, one set at 15 degrees per side and one closer to 17 degrees, and the report looked like a steel change. It was not. A 2-degree shift, or a different heat-treatment lot, makes the comparison dirty. The math does not work.
| Report item | What it tells you | What you should check |
|---|---|---|
| TCC | Total cutting before the edge lost performance | Compare the same blade geometry, same edge angle, and same finish |
| Sample count | The number of knives tested | One sample is a sales demo, not buying evidence |
| Hardness | Usually HRC or a related hardness note | Confirm the full acceptable band, not one nice Rockwell reading |
| Prep method | How the edge was ground and finished | Ask for the grit sequence, buffing step, and final edge angle in degrees |
| Lab details | Who ran the test and which method they followed | Keep the same lab and method across repeat orders |
For product and QC managers, the report should read like a sourcing record. Short. Traceable. If the supplier cannot tell you whether the tested knife came from production or from a hand-finished sample room piece, the number should not support a purchase order. We've seen this go sideways when a buyer approved 3,000 pcs from one polished prototype, then flagged the bulk lot because the production edge came off a different wheel. One PO even had a typo on the edge spec, 15° written as 150°, and the buyer still wanted the report to explain it. It could not.
What drives edge retention most
Edge retention is a stack-up, not a magic steel grade. Steel chemistry matters, but heat treatment usually changes the result faster than buyers expect. On our line, the same steel can cut different at HRC 56 and HRC 60 after quench and temper; QC pulled 6 pcs from the temper rack and the CATRA curve made the point. For mainstream kitchen knives in China, the working band is often HRC 56-62, depending on blade type, intended use, and toughness target. Hardness alone is the wrong question. A harder edge holds more cuts only when the steel and grind can take the load without micro-chipping.
Geometry does equal damage to the test result. A thin, acute edge looks beautiful in a sample room and cuts clean on day one, then loses working life fast if the bevel is too light. A thicker bevel may not look as sexy on a sales sheet, but it often scores better in edge retention QC for heavy prep knives. We see it on the grinding line: a 15° edge beat a finer 12° edge after the buyer flagged rolling on test onions across 8 cartons. Surface finish, burr removal, and the final sharpening pass change the number too. Bad burr, bad week. It can feel sharp out of the box and go dead by day three.
In a Yangjiang, Zhejiang production environment, we separate design intent from marketing wording. For a premium chef knife, we tune the steel and final edge. For an outdoor knife or pocket knife, we put toughness and corrosion resistance ahead of the peak CATRA score. One PO asked for a "laser edge" in the remarks column, then the buyer sent back photos of chipped tips after carton drop testing from 80 cm. We have seen this go sideways. The right answer comes from the use case, not the headline number.
How to spec CATRA in your RFQ
If purchasing needs a CATRA number they can defend, put the test method in the RFQ before price talk. Lock it early. Do not let the supplier choose the knife build after sampling. We have seen this go sideways: a sales sample touched up by hand on a 1000 grit belt gets compared with a blade from the grinding line, then the buyer calls it a quality problem.
- State the exact product type: chef knife, kitchen knife, pocket knife, or outdoor knife, because blade thickness and cutting use change the result.
- Lock the steel grade and heat-treatment target, such as 14C28N at HRC 57-59 or a stainless kitchen blade at HRC 56-58.
- Define the edge angle and bevel style, then name the final sharpening grit, such as 15 degree per side with a 1000 grit belt finish.
- Ask for the lab name, sample count, and whether the test knife came from the production line.
- Require the full report, not a cropped chart image sent by sales on WeChat.
Set the acceptance rule too. For a 1,000 pcs lot, ask for the minimum average TCC and the allowed spread across samples, not the top result from one clean-looking knife. The best sample is the wrong question to ask. We run this tighter when the buyer confirms the CATRA setup before tooling starts: steel, HRC, 15 degree per side edge, belt grit, sample qty, and whether QC pulled the sample from bulk packing. One buyer flagged a PO typo that said HRC 56-68 instead of HRC 56-58; catching that before production saved 12 days of arguments after shipment.
Build edge retention QC on the line
AQL 2.5 is fine for finish marks, handle gaps, logo position, and carton scuffs. It does not prove the edge will keep cutting. We have seen a lot pass visual inspection while the Rockwell reading sat 2 HRC below spec because the tempering oven drifted during the night shift. Looks clean. Cuts poorly. For edge retention QC, we run a separate engineering check beside the normal visual inspection, with the hardness tester and cut-test sample logged against the approved blade number.
A practical factory flow is simple. First, match the incoming steel coil or sheet to the heat-treatment record, including grade, furnace batch, and quench date. Then check hardness across the batch: for a 3,000 pc run, QC should not test only one blade from the top tray. We usually pull 3 positions, first rack, middle rack, and last rack, with each position marked in the report. After that, pull a small production sample for a cut test knife check and compare the result against the approved reference. If the knife family is a kitchen line, watch edge wear and early chipping under a 15 degree per side edge. If it is an outdoor or pocket knife, the wrong question is “how sharp is it on day one?” The better check is whether the edge rolls or chips after harder use.
In our own Yangjiang, Zhejiang factory in China, we treat this as a gate, not a sales photo. QC checks the first article before packing starts, pulls a mid-run sample from the grinding line, and reviews the finished lot before the cartons are sealed. One buyer once flagged a PO typo that changed 58-60 HRC to 56-58 HRC; that small number would have changed the whole edge retention result. The math doesn't work if the first sample is good but the last 800 pcs drift softer. For a buyer, the point is simple: one good CATRA score does not guarantee a stable run. The process has to repeat.
CATRA versus real kitchen use
CATRA is a solid screening tool, but it is still a bench test: clean media, fixed pressure, and a fresh edge clamped in the machine. Real kitchens are messier. Tomato skin, carton tape, wet herbs, sweet potato, chicken joint contact, sink washing, towel drying, then loose storage in a drawer all hit the edge in different ways. A blade can post a good TCC number and still bring complaints if the edge is too thin behind the bevel, or if the grinding line runs 17° per side while the PO says 15°. We have seen QC pull a chef knife sample at 58 HRC with good TCC, then the buyer flagged micro-chipping after 2 days in a prep kitchen. CATRA alone is the wrong acceptance rule.
The better way is to pair the lab number with a short field trial. For a chef knife, we run a controlled prep test: 5 kg tomato, 5 kg carrot, 3 kg onion, then a paper-slice check and a re-sharpen check after the fixed workload. Simple. For a pocket knife, use cardboard and rope cutting, with the same operator, same cutting board condition, and the same 18 mm carton board if that is what the buyer sells against. For an outdoor knife, add harder contact and check the edge under a 20x loupe, because TCC will not show whether the bevel rolled, chipped, or dulled cleanly.
This is where honest sourcing matters. If you are buying from China, ask the supplier for the CATRA report and the production build sheet: steel batch, heat-treatment curve, target HRC, bevel angle, and final belt grit from the grinding line. A knife from Yangjiang, Zhejiang can be tuned for excellent lab retention, but the real question is whether that result holds across 500, 1,000, or 10,000 pcs. We have seen this go sideways when one golden sample passed and bulk stock drifted by 2 HRC or 0.4 mm behind the edge. The math does not work. The best answer comes from the lab result, the line record, and a field sample that actually gets used.
Common red flags in supplier claims
Red flags show up fast when you ask for the test setup behind the headline score. Be careful when a supplier shows 1 CATRA score with no blade count, no edge angle, and no steel batch record. Be more careful when they cannot confirm 15° per side, the steel grade, or the hardness band in HRC. We have seen a report taken from a hand-ground prototype while the PO called for a stamped blade with a different bolster; QC later measured the production edge at 18° on the left side and 21° on the right. That number is marketing, not sourcing control.
A second warning sign is a strong-looking report that fails on the next batch. The process is drifting. Same issue when the supplier talks about sharpness in soft words but cannot provide the CATRA test knife report, sample count, or hardness data. Ask for one repeat test from the same setup, with QC pulling the sample from the grinding line, not from the showroom drawer. We run this check with the lot card on the bench, because one buyer once flagged a PO typo where “14°” became “14mm.” If the result moves from 420 TCC to 260 TCC, the buyer has a process issue, not a sales issue.
For buyers, buy on reproducibility. One peak number is interesting. 3 consistent results within 8% are worth discussing. The wrong question is “what is your best CATRA score?” Ask what the normal production lot delivers at MOQ, after heat treatment, grinding, polishing, and final sharpening. We’ve seen this go sideways when the golden sample came from a senior sharpener, while mass production ran through 6 operators on two shifts. A stable production line in Yangjiang, Zhejiang is what keeps a program profitable after launch.
Frequently asked questions
There is no universal number because CATRA depends on steel, edge angle, hardness, and geometry. For a kitchen knife, a score that repeats across three samples is more useful than a single high result. In practice, a buyer should compare against the same blade family. If one sample is 15% to 20% higher than another with the same setup, that usually points to process variation. Ask for the full report, not just the headline number, and confirm whether the knife was tested at the final production edge angle. A good score that cannot be repeated is not a sourcing spec.
Higher TCC usually means better edge retention in the CATRA test, but only within the same setup. A blade with a thinner edge may score well and still chip in real use. A tougher blade may show a lower TCC and still be the better choice for heavy prep, outdoor work, or a pocket knife that may see rough use. Treat TCC as one piece of the specification. If you are buying for retail, ask how the knife behaves after re-sharpening, after a wash cycle, and after contact with harder materials. One number alone does not make the product right.
Yes, if the 58 HRC blade has better geometry, cleaner burr removal, or a more suitable steel for that application. Hardness is important, but it does not act alone. A 61 HRC blade can still underperform if the heat treatment is uneven or the edge is too fragile for the test. That is why buyers in China and Europe should ask for the steel grade, HRC band, and the exact sharpening method together. In a real sourcing decision, a stable 58-59 HRC with repeatable TCC may be more valuable than a loose 61 HRC claim that only appears on one sample.
For an engineering check, one sample is not enough and three are better than one. A practical starting point for a 1,000 pcs lot is a small production sample from the first run, then another from the middle or end of the batch. If the program is important, add periodic lab checks across different lots. AQL 2.5 is fine for appearance inspection, but performance needs its own sampling logic. The key is consistency: compare the same steel, the same edge angle, and the same production process. If the supplier cannot repeat the result, the sample count was too small or the process is unstable.
Ask for the full CATRA report PDF, the lab name, the sample count, the steel grade, the HRC target, and the final edge angle. Also ask whether the knife came from production or from a hand-finished prototype. If the supplier refuses to share those basics, the chart is not useful for procurement. For product and QC managers, the goal is to make the test reproducible. Once you have the real data, you can compare factories in Yangjiang, Zhejiang, or anywhere else in China on the same basis instead of buying the strongest sales pitch.
Send your knife spec for a CATRA check
We can review the steel, edge angle, HRC target, and QC plan before you commit to tooling, MOQ, or a production run.
Request a Quote
