Machining In China

This page is your guide to machining in China. Zhuohang Precision Machining Factory holds milling to ±0.003 mm and turning to ±0.002 mm on machining in China work, backed by 100+ Mazak, Brother, TSUGAMI and Sodick machines — read on for process detail, tolerance logic, and supplier verification.

What is CNC machining in China?

CNC machining in China is the production of machined parts — brackets, housings, fittings, shafts, structural components, prototypes, and production runs — by Chinese manufacturers operating computer-numerically-controlled milling, turning, and grinding equipment, supplied to overseas buyers against a customer drawing. The discipline is the same as machining anywhere: a cutter removes material from a workpiece along a programmed toolpath to a drawing-defined geometry, tolerance, and surface finish.

What is specific to CNC machining in China is the supply chain behind it: domestic stock of aluminium 6061-T6 / 7075-T6 / 6082 in standard sizes, a network of casting foundries for near-net blanks, surface-treatment subcontractors for anodising and powder coat, and a certification infrastructure (ISO 9001 / IATF 16949 / ISO 14001) that supports export to U.S., EU, and Asian automotive and industrial buyers. The machining is universal; the supply chain and the documentation are what differ.

The output of CNC machining in China is a part that meets the drawing — geometry, tolerance, surface finish, and material certificate — and is documented to a quality plan (FAI, PPAP, IMDS) plus the export paperwork (Certificate of Origin, mill certificate, REACH/RoHS declarations). That is the answer to "what is CNC machining China": numerically-controlled subtractive manufacturing by a Chinese supplier, with the supply-chain and export documentation that lets the part land in the destination country.

How should you diligence a CNC machining supplier in China?

Due diligence is the process of turning a supplier's claims into evidence you can check from your own desk, and it fails in one predictable way: buyers run it in the wrong order. The core mistake is to diligence price first and proof never. The right order is the opposite — establish the entity, confirm the certificates, verify the machines, check the material, then demand tolerance evidence, and only then talk price. Each step removes one layer of uncertainty, and each is cheap to do.

Think of the whole exercise as turning belief into evidence. A brochure says "manufacturer since 2005", that is a belief. A business registration number you can look up, certificate numbers a registrar can confirm, an XRF read of the incoming alloy, and a live video of the floor are evidence. The buyer who treats the first as if it were the second is the one who later discovers their "factory" was a reseller with a rented logo.

The order matters because the cost of each check climbs as you move down the list. The entity check costs a web search. The certificate check costs a two-minute registrar lookup. The machine check costs a video call. The material and tolerance checks cost the supplier real work to produce, which is precisely why a trading desk cannot hand them over. Run the checks in order and you spend the least money to learn the most, stopping the moment a supplier fails. The second-order effect is the one buyers miss: a supplier that volunteers evidence at the top of the list is usually honest at the bottom too, because the habit of documentation is the same habit everywhere in the shop.

Apply this framework to any supplier in China — including the one that built this page, which is exactly the point. A supplier that volunteers evidence is the one you should actually be buying from. A supplier that deflects the first question is telling you, in thirty seconds, what the next thirty questions will look like.

What documents prove a machining shop is legitimate?

Legitimacy is established with three documents, and none of them is a brochure. The first is the business registration number — the legal identity that ties a trading name to a registered company. The second is the certificate number for each ISO or IATF claim, verifiable with the issuing registrar. The third is a named, mappable address that matches the registration.

A registration number is not a random string; it is the unique identifier a Chinese company files with the government, and it is the first thing a broker cannot improvise. When you look the number up, it should resolve to the legal name and the address on the quote. A mismatch anywhere in that chain — a number that resolves to a different name, an address that is a mailbox, a legal entity that is a trading company — is the finding, not an anomaly to explain away.

Document	What to ask for	How to verify it
Business registration	The registration number and legal entity name	Look it up against the company's legal name and address
ISO / IATF certificates	The certificate number, standard, and registrar	Check the number with the registrar (URS for this facility)
Factory address	A full street address, not a city name	Find it on a map and confirm it is a manufacturing zone

This page's verifiable example is Dongguan Zhuohang Ltd, trading as Zhuohang Precision Machining Factory and Zhuohang Precision Machining Factory — registration number 91441900MA4UWT6X39, at No. 8, Shuibian Songyuan Road, Hengli Town, Dongguan, Guangdong, China 523470. Those are the kind of specifics a trading desk cannot easily produce, and the kind a real factory lists without being asked twice. The satellite-map check is the quiet tiebreaker: a manufacturing zone with loading docks and roof-mounted equipment reads differently from a shared office floor, and you can see the difference without leaving your chair.

How do you verify the machines and floor without a visit?

Capability claims live or die on the machine list, and a machine list is only worth something if you can confirm it against the floor. The check has two parts: first, ask for a specific list — models and counts, not "we have many machines" — and second, ask for a live video walk-through that shows the floor and a nameplate or two on camera.

What you are looking for is specificity. A manufacturer knows its own equipment: this facility runs 100+ production machines — 3-axis milling (50+), 4-axis (20+), 5-axis (20+), 25+ turning machines, and 7+ grinders — across a 6,800 m² plant. A reseller knows only the brochure. The gap between those two answers is the gap between a factory and a trading desk, and it is visible in thirty seconds of video.

The live walk-through is the part most buyers skip, and it is the part that replaces the plane ticket. Three details carry the signal on the call. First, ask to see a nameplate — the make and serial on the side of a machine, not the clean logo in the brochure. Second, ask to pan the camera to show multiple machines running, because one running machine proves one machine. Third, ask to see the inspection bench in the same pass — a CMM and a surface-finish instrument on a table say more about the quality system than any paragraph. A shop that agrees to all three is showing you the same reality an in-person audit would confirm, and a shop that resists any one of them has just told you where its floor stops and its brochure begins.

Which machining process does your part need, and how do you confirm the shop runs it?

Process selection is the first technical question you answer, and it is a diligence question as much as an engineering one: the process your geometry needs must be a process the supplier actually runs in-house. A shop that maps your part to a named machine is an owner; a shop that maps your part to a price is a relay, because it will sub the work to someone else and add a margin on the way.

The four core metal-removal processes each cover a distinct geometry, and the mapping below is the rule of thumb you can apply before sending a drawing. A prismatic housing with pockets, holes, and flats points to milling. A shaft, pin, or bushing points to turning. A part that mixes turned and milled features points to mill-turn, where both are cut in a single clamping. A bearing seat or a roundness-critical surface points to grinding, which holds the tightest size and finish. When the supplier names the process and the machine count behind it, the capability claim becomes inventory.

Part geometry	Process it demands	Counted capacity here	Held tolerance
Pockets, holes, flats, housings	3-axis milling	50+ centers	±0.003 mm
Indexed side features	4-axis milling	20+ centers	±0.003 mm
Compound angles, fewer setups	5-axis milling	20+ centers	±0.003 mm (milling) / ±0.002 mm (turning)
Shafts, pins, bushings, round parts	CNC turning	25+ machines	±0.003 mm (milling) / ±0.002 mm (turning)
Turned + milled features together	Mill-turn	Multi-task centers	±0.003 mm (milling) / ±0.002 mm (turning)
Bearing seats, roundness-critical surfaces	Precision grinding	7+ grinders	±0.002 mm

The confirmation question is the same one from the floor check: "which machine runs this part, and can you show me one now?" A manufacturer answers with a make, a model, and a camera. The process question also sets your tolerance expectations — grinding buys you a tighter band than milling, and a supplier that promises a ground-class tolerance from a mill is promising a spec it cannot hold, which is a diligence red flag in its own right.

Tolerance is also a cost ladder, and a supplier should be able to state which rung your print sits on. A looser band costs less because it needs fewer setups and less measurement; a tighter band costs more because it needs more passes, more in-process checks, and a finishing process the shop may or may not run. The ladder below is the general shape you can expect, and the diligence question is whether the supplier's equipment reaches the rung the drawing demands.

Tolerance band	Typical process	What it demands
±0.1 mm	Standard milling / turning	Minimal setups, standard tooling
±0.05 mm	Careful milling / turning	Controlled speeds, more measurement
±0.01 mm	Precision milling / grinding	More passes, in-process inspection
±0.003 mm (milling) / ±0.002 mm (turning)	Tight milling / turning	SPC control and CMM verification
±0.002 mm	Precision grinding	Dedicated finishing and full inspection

How do you verify the material is the alloy your drawing calls out?

Material verification is the check that separates a machine shop from a metal merchant, because a wrong alloy looks identical to a right one until it fails in the field. The check has two parts: ask for the mill certificate for the exact lot, and ask for an XRF read of the incoming material against that certificate. The certificate says what the alloy should be; the XRF read proves what actually reached the dock.

The reason this matters to you is that aluminum and stainless are bought in grades, and the grade difference carries a strength difference you cannot see. A part cut from the wrong temper of 6061, or from a lookalike alloy, will pass a dimensional check and still fail the application. The XRF check at incoming closes the loop between the paper and the metal, and the record of that check is something you can pull up during diligence. A trading desk has no incoming bench to read and no record to show, which is exactly why this one question separates the two so quickly.

Alloy	Family	Typical yield strength	Density	Typical use
6061-T6	Aluminum	~276 MPa	2.70 g/cm³	General structural parts, housings
7075-T6	Aluminum	~503 MPa	2.81 g/cm³	High-strength, aerospace-grade parts
5052-H32	Aluminum	~193 MPa	2.68 g/cm³	Sheet-metal enclosures, corrosion-prone parts
304	Stainless steel	~205 MPa (annealed)	8.00 g/cm³	General corrosion resistance
316	Stainless steel	~205 MPa (annealed)	8.00 g/cm³	Marine and chemical exposure
4140	Alloy steel	~655 MPa (Q&T)	7.85 g/cm³	Shafts and high-strength parts

The yield-strength and density figures above are public reference values for the common grades, not factory-specific claims — they are the numbers you use to check that a supplier's material selection matches the part's load. Confirm the temper, not just the family: 6061-T6 and 6061-O are the same alloy family with very different strength, and a mill certificate that omits the temper is an incomplete certificate.

What tolerance evidence separates a real shop from a broker?

A tolerance number is meaningless without the measurement behind it. Anyone can type "±0.003 mm (milling) / ±0.002 mm (turning)" on a website; the question is whether they can show you the CMM report or SPC chart that proves the process actually holds it. This is the single fastest filter in the whole checklist, because a broker has no measurement bench to show.

The verifiable numbers for this facility are ±0.003 mm on CNC milling, ±0.002 mm on CNC turning and ±0.002 mm for precision grinding, held to process capability rather than to a best-case claim. On a representative length characteristic (76.1–76.5 mm), SPC across 25 subgroups shows Cpk = 1.851 (one characteristic on one part, not a blanket capability claim), against an internal standard of Ppk ≥ 1.67 at the initial study on key characteristics and ≥ 1.33 on general ones. That chart is the kind of evidence you should ask to see — a real control chart, not a paragraph promising accuracy.

Before trusting a Cpk number, read the band it was measured against. Cpk is the tolerance band relative to the process spread — a ratio, not a precision badge. Our example used a wide 0.4 mm band (76.1–76.5 mm) and held Cpk 1.851; the same spread would fail a ±0.005 mm feature outright. Always request the band, the subgroup count and the sampling window for any capability claim.

The Cpk number deserves your attention because it measures stability, not luck. A single part can pass a tolerance by chance; a Cpk of 1.851 (one characteristic on one part, not a blanket capability claim) means the process mean is centered and the spread is small enough that the process keeps producing good parts, subgroup after subgroup. When you read a Cpk chart, ask for the subgroups behind the headline — a real chart has a measured mean and range, while a made-up one has only the number. The same logic applies to any supplier: demand the chart, then demand the data under the chart.

There is one layer under the chart that completes the tolerance check: the instruments that produced it. A CMM or a micrometer is only trustworthy if it is calibrated against a traceable standard, and a real shop can show the calibration record for the instrument that measured your part. This facility runs 20+ inspection instruments — two Hexagon GLOBAL S CMMs, a KEYENCE laser microscope, and a SURFCOM profilometer among them — and the calibration records are part of the quality system, not an afterthought. Ask to see the calibration record for the instrument named on the report; a shop that keeps its instruments certified keeps its numbers honest, and the record is one more document a broker cannot reach for.

How do you read a CNC machining quote without being misled?

A machining quote is a list of cost drivers, and if you read it as a single number, you are the buyer a broker wants. The price decomposes into a handful of components — material, machine time, setup, finishing, and the shipping term — and each component is a place where cost hides or where a supplier's assumptions surface.

Start with the shipping term, because it silently moves the price. The Incoterm defines who bears freight, insurance, and import duty, and two quotes with the same number can differ by a wide margin once the term is named. EXW prices the part at the factory gate and leaves the buyer to arrange everything downstream; FOB gets the goods to the port; DDP puts the seller's responsibility all the way to your door, including import duty. When a supplier quotes without naming the term, ask for it before comparing anything — the term is part of the price, not a footnote to it.

Incoterm	What the seller covers	Diligence note
EXW	Goods at factory gate	Buyer bears all freight, insurance, and duty
FOB	Goods to the export port	Buyer bears freight and duty from the port on
CIF	Freight and insurance to the destination port	Buyer still clears customs and pays duty
DDP	Delivery to buyer's door, duty paid	Seller bears import duty and clearance

Then check the three places cost hides. First, finishing: anodizing, plating, and heat treatment are separate process steps with their own lead time, and a quote that excludes them is a quote for a machined blank, not the part you drew. Second, inspection: a CMM report and material certificate cost time and labor, and a quote should state whether they are included. Third, setup: a low unit price with a large setup charge is a volume signal in disguise. A quote that names all three assumptions is a cost basis and a schedule; a quote that does not is a guess, and the supplier is asking you to fund the guess.

To read a quote critically, decompose it into the cost drivers that actually move a machined part, then ask which ones your geometry leans on. The table below is your map of where the money goes, in the general shape a supplier should be able to explain for any part.

Cost driver	What moves it	Diligence question to ask
Material	Alloy, grade, temper, and blank size	"Is the grade on my drawing, and is the mill cert included?"
Machine time	Geometry, tolerance, and batch size	"Which machine runs this, and how many setups?"
Setup and programming	Complexity and order quantity	"Is setup amortized over the quantity, or charged once?"
Finishing	Surface finish, plating, heat treatment	"Is finishing in the price, and who runs each step?"
Inspection and documents	CMM, SPC, and certification depth	"Is the CMM report and material cert included?"

A supplier that can walk this table line by line is handing you a cost basis, not a number, and the walk itself is a diligence result. The buyer who gets a single price and nothing else is the one who will later discover the anodizing was a line item, the setup was per part, and the freight term moved the whole quote after the fact.

What surface finish and DFM rules should you diligence against?

Surface finish and design-for-manufacturing rules are where your drawing either gets made cheaply or gets quoted to death, and both are checkable with standard engineering references. Surface finish is measured in Ra — the average roughness of the surface — and each band maps to a process that can achieve it.

Ra value	Typical process	What it means
3.2 µm	Standard milling / turning	Visible tool marks; general surfaces
1.6 µm	Finer milling / turning	Smooth; common for sealing faces
0.8 µm	Precision grinding / fine turning	Near-mirror; bearing and seal surfaces
0.4 µm	Lapping / honing	High-precision sealing and sliding

The diligence lesson in the Ra table is the same one as the tolerance table: a supplier that promises a 0.4 µm surface from a mill is promising a process it does not run, because that band needs grinding, lapping, or honing. When you call out a finish, the supplier should be able to name the process that achieves it, not just the number.

DFM rules are the geometry checks that keep a part machinable, and a good supplier raises them during quoting rather than after the chips are cut. The standard rules of thumb are worth knowing because they explain why a quote comes back with questions. A wall thinner than about 0.8 mm in aluminum or 1.0 mm in steel will deflect under the cutter. A drilled hole deeper than roughly four to five times its diameter needs a special cycle, not a standard drill. An internal corner needs a radius at least as large as the cutter radius — typically 1 mm or more — because a sharp internal corner is physically impossible with a rotating tool. A supplier that flags these before cutting is doing DFM review; a supplier that quotes the print as-is and ships whatever comes out is not.

When a print carries dimensions with no individual tolerance, the default tolerance class matters, and ISO 2768 is the common reference. Its three classes — f (fine), m (medium), and c (coarse) — set looser or tighter bands on the dimensions the drawing leaves un-toleranced. If you call out ISO 2768-f, you are asking for tighter control on everything not otherwise specified, and a supplier should be able to state which class it assumes. This is the kind of standard reference a real shop cites without hesitation and a broker has never heard of.

How do you confirm certifications are real, not photocopied?

A certificate is a piece of paper with a number, and the number is the part that matters. A genuine ISO or IATF certificate carries an identifier you can confirm with the registrar; a logo pasted onto a website confirms nothing. When you diligence a shop, ask for the certificate number and the registrar, then check both.

Standard	Certificate number	Registrar
ISO 9001:2015	116024/A/0001/UK/En	URS
IATF 16949:2016	131941/A/0001/SM/En	URS
ISO 14001:2015	116024/B/0001/UK/En	URS

Honest scope note: the IATF 16949 certificate excludes product design under Clause 8.3. This is a build-to-print manufacturer. A shop that states its scope plainly is one you can diligence cleanly; a shop that hides the exclusions is one whose claims you should already be discounting.

Read the scope, not just the number. A certificate is issued against a defined scope of activity, and the scope is where a shop's honesty shows. A scope stretched to cover every process and material is a warning; a scope that says exactly what the shop does — manufacture of CNC metal machining parts, with product design excluded — is the kind of sentence an auditor trusts. The verification itself is a two-minute check, which is precisely why a trading desk avoids it and why you should never skip it.

What questions expose a trading company?

You rarely catch a trading company with a direct question; you catch it with the pattern of its evasions. Ask the five questions below, and watch what happens when the answer is not a document but a deflection. A manufacturer answers each in a sentence; a broker changes the subject to price and turnaround.

Run these five on any supplier in China, including this one. The answers are either evidence or an explanation of why evidence is unavailable, and the second answer is itself the finding.

The deeper signal is how the supplier reacts when you ask for proof, not just what it says. A manufacturer reaches for the document; a broker reaches for a reason. Watch for the specific deflection patterns: the price pivot ("let me get you the best price first"), the schedule pivot ("we can start immediately"), and the trust pivot ("we have been doing this for years"). Each one moves the conversation away from evidence, and the movement itself is the tell.

How do you diligence lead-time and capacity promises?

Lead-time and capacity are the easiest claims to fake and the most expensive to be wrong about. The check is the same as every other step: ask for the number, then ask for the basis. A shop that says "3 days" without qualification is guessing; a shop that ties lead time to complexity is telling you how it actually schedules work.

Here the honest answer is a tier, not a date: roughly 3, 7, or 30 days, set by part complexity and the number of process steps, with MOQ of 1 piece. Simple parts ship in about 3 days; a part running 10–20 process steps or needing surface finishing falls into the 30-day tier. When a supplier gives you a range with a stated reason instead of a flat promise, treat that as a sign it is telling the truth about the rest too.

Capacity is the hidden sibling of lead time, and it is worth a direct question: "how many of these could you run per week, and which machines would you put them on?" A manufacturer maps the job to a machine and a count; a broker maps it to a promise. The capacity answer also tests the machine-list claim — if a shop says it has 50+ mills but cannot say which mill runs your part, the count was decoration. Ask the capacity question the same way you ask the machine question, and hold the answers to the same standard of evidence.

How do cross-industry requirements change your checks?

The diligence framework does not change by industry, but the weight on each check does, because different industries regulate different layers of the same shop. If you know which checks matter most for your parts, you ask the right question first instead of the easy one.

For automotive and e-bike drivetrain work, the process-control layer dominates: SPC, PFMEA, and a control plan are not paperwork but the evidence that a dimension will stay in tolerance across a run, which is what a moving part demands. For medical and fluid-handling parts, the material and traceability layers dominate: the alloy must be the exact grade on the certificate, and the lot record must trace a finished part back to its raw material and machine. For robotics and automation, the tolerance layer dominates, because bearing seats and locating features carry the tightest call-outs on the print and fail as a stack, not one at a time. For semiconductor and instrumentation, the surface-finish and cleanliness layers dominate, because a Ra number on a sealing face is a functional spec, not a cosmetic one.

The takeaway is not to run a different checklist per industry; it is to rank the same checklist by what your part will punish you for. If you know that, you are the one who asks the automotive shop for the control chart, the medical shop for the lot trace, and the robotics shop for the CMM report — before any of them has a chance to lead with a brochure.

The industry lens also changes how you read the shop's own emphasis. A floor that runs SPC and control plans as a matter of course is speaking the language of volume manufacturing, which serves automotive and drivetrain buyers. A floor that pulls up a lot trace on demand is speaking the language of regulated buyers, which serves medical and instrumentation. A floor that leads with its CMM reports is speaking the language of tolerance-driven work, which serves robotics and precision automation. None of these is better than the others; each is a signal of where the shop has built its discipline, and your job is to match that discipline to the part in hand.

How do regional compliance differences change your checks?

Export regions differ in what they ask you to prove about a supplier, and the differences are general and well-known enough to plan for without guessing at any specific tariff or regulation. The diligence framework stays the same; the documentation emphasis shifts.

Buyers shipping to Europe typically face tighter documentation on materials and environmental compliance, so the material certificate and the ISO 14001 record carry more weight. Buyers shipping to North America tend to emphasize traceability and process documentation, so the lot record and the SPC evidence matter most. Buyers shipping to Japan often place the heaviest weight on dimensional evidence and first-article inspection, because the acceptance culture leans on measured proof over narrative. Buyers shipping to Australia and New Zealand commonly focus on the certification chain and the named address, because distance raises the cost of a wrong bet.

None of these changes the core checks — the entity, the certificates, the machines, the material, and the tolerance evidence remain the spine. What changes is which layer you lead with and which document you ask to see first. The disciplined buyer asks the supplier to state, in writing, which standard it works to for the destination market, and treats a vague answer the same way they would treat a vague tolerance: as a claim that has not yet been verified.

The practical version of this is a single question you can put to any supplier before you even open the drawing: "for parts shipping to my market, what documentation do you provide as standard?" The answer should be a list — material certificate, CMM report, SPC chart, lot trace, certificate numbers — not a sentence. A supplier that names the documents unprompted is a supplier that has shipped to your market before; a supplier that has to think about the answer is one you are teaching on your own invoice.

What does a failed diligence look like — the red-team view?

Every due-diligence framework has a failure mode, and the buyer who knows it in advance stops it before it costs money. The failure is not a supplier that answers every question wrong; that supplier is easy to reject. The failure is a supplier that answers every question almost right, and the "almost" is where the cost hides.

The red-team signs to watch are the ones that pass a casual glance. A machine list that names brands but no models or serials. A tolerance claim that cites a number but no chart behind it. A material list that names alloy families but no temper or mill certificate. A certificate that shows a logo but no number you can call in. Each of these is a claim wearing an evidence costume, and the tell is always the same: the supplier can tell you what it has, but cannot show you how you would know.

The deeper failure is a supplier that hides cost in the fine print of the quote rather than in the brochure. An EXW price that quietly makes you the freight and duty payer. A finishing step listed as "available" rather than included. A lead time stated as a date rather than a tier. These are the ways a shop that is real on the floor still fails a buyer on the invoice, and they are why the quote check and the floor check belong to the same framework. The buyer who treats price as the last step, not the first, is the one who catches the fine print before it catches them.

What does a shop that passes diligence look like?

By the end of the checklist, the pattern should be obvious: a shop that passes diligence is one that volunteers evidence before you ask for it. It states a registration number, lists certificate numbers with registrars, shows a machine list you can confirm against a live video, confirms the incoming alloy against a mill certificate, and hands over a CMM or SPC report from an actual run.

Zhuohang Precision Machining Factory is one example of that tier — the legal entity Dongguan Zhuohang Ltd, founded 2005, running a 6,800 m² facility in Hengli Town, Dongguan, with 100+ production machines and 20+ inspection instruments, certified to ISO 9001:2015, IATF 16949:2016, and ISO 14001:2015. It is not the only shop of its kind in China, and that is the point: the tier exists, and the checklist above is how you find it and confirm it — rather than taking a brochure's word for it.

The passed shop has one more habit that is easy to miss and worth the most: it names what it does not do. A build-to-print manufacturer states that product design is outside its certificate scope. A shop that outsources anodizing or high-temperature heat treatment names the partner. A shop that holds a tolerance only on some processes says which ones. That habit of stating limits is the difference between a supplier you can diligence and one you have to guess about, and it is the clearest single signal that the rest of the claims are built the same way.

How do you protect your drawings and IP before you send them?

Intellectual property is the one asset you hand over before you have verified anything, so it deserves its own check, run before the first drawing leaves your desk. The protection is not a single document; it is a sequence of habits that make a drawing harder to misuse and easier to trace if it is.

The first habit is to send only what the quote needs. A supplier needs the geometry, the material call-out, and the tolerances to price and plan a job; it does not need your assembly context, your end customer, or the adjacent parts the print might hint at. If you send the full product file to get a single bracket quoted, you are handing over more than the quote requires, and the extra context is what turns a bracket drawing into a competitive map.

The second habit is a written non-disclosure agreement before the print. A build-to-print shop is used to NDAs and returns a signed copy without ceremony; a shop that will not sign one is telling you how it treats other people's drawings. The third habit is a paper trail on the drawing itself — a revision number, a date, and a distribution note — so that when a copy surfaces somewhere it should not, you can tell which version leaked and when. None of these habits is a guarantee, but together they make a drawing traceable, and traceability is what changes a lost drawing from a mystery into a finding.

For this facility, the build-to-print scope is itself part of the answer: the IATF 16949 certificate excludes product design under Clause 8.3, which means the shop machines your drawing and takes no design credit. That scope line is not just a quality statement; it is the IP statement. A shop that claims to both build your part and own your design is the one to read carefully, because the two roles sit in tension.

How do you verify the first article before committing volume?

The documents and the floor checks prove the shop can make parts; the first article proves it can make your part. A first-article inspection is a full dimensional layout of one part against every call-out on your drawing, and it is the bridge between "this supplier looks right" and "this supplier is right." At MOQ 1 piece, the bridge costs one sample, which is the cheapest insurance a volume order can buy.

Run the first article the way you ran the audit: check it against the print, not against the supplier's summary. Lay the measured values next to your drawing and confirm every tight call-out is inside the band — the same ±0.003 mm CNC milling, ±0.002 mm CNC turning, ±0.002 mm grinding figures you verified on the CMM report. Then check the material against the mill certificate, and check the surface finish against the Ra you called out. A first article that passes all three is a process that has reproduced your part; one that passes only the dimensions is a process that has not yet been asked about the material or the finish.

The second-order question is the one buyers skip: does the first article come with the same documents the audit promised? A real first-article submission carries the dimensional layout, the material certificate, and the SPC or CMM evidence behind it. If the part arrives alone, the sample proved the geometry but not the system that produced it, and it is the system, not the sample, that you are about to buy in volume. The first article is the moment the paper and the metal meet; demand both, and the volume order that follows is no longer a bet.

How do you protect the wire transfer itself?

The title of this page is a promise: verify the factory before you wire. The wire is the one step that turns a bad diligence decision into a loss, so it gets its own check, and the check has three parts. First, wire only to the entity you have verified — the bank account name should match the legal entity on the registration and the invoice, not a personal account or a third party. Second, structure the payment so the money trails the evidence: a deposit first, then the balance against the documents you have already learned to demand. Third, keep every payment on the same legal entity, so the paper trail and the money trail are one chain.

None of this requires trusting the supplier; it requires making the payment structure do the diligence for you. A deposit paid against a pro-forma invoice, with the balance against the shipping documents, means the supplier has to produce the same records you checked in the audit before the last payment moves. If you wire the full amount up front, you have surrendered the one lever that keeps a supplier honest through the last process step.

The red flags in payment are the same red flags as everywhere else, just priced higher. A request to wire to a personal account. A bank name that does not match the registered entity. Pressure to pay before you have seen the first article. Each one is the money version of a brochure photo, and each one is your cue to stop and re-run the checks. The discipline that carries you through the audit — demand the document, not the adjective — is the same discipline that protects the wire.

your diligence checklist in one pass

The full framework, in the order to run it, is a list you can hold against any supplier in China. Each item is a question, and each answer is either evidence or an explanation of why evidence is unavailable.

• Entity. Ask for the registration number and legal name, then look both up against the address. A real number resolves to a real factory.
• Certificates. Ask for each certificate number and its registrar, then confirm the number and read the scope. A logo is not a number.
• Machines. Ask for makes, models, and counts, then a live walk-through with a nameplate on camera. A count you cannot see is a brochure.
• Process fit. Ask which machine runs your part and why. The process must be in-house, not subbed to an unnamed floor.
• Material. Ask for the mill certificate and the XRF read of the incoming lot. The alloy on the drawing must be the alloy at the spindle.
• Tolerance. Ask for a CMM report or SPC chart from a real job, and for the subgroups under the headline. A number without a chart is a claim.
• Quote. Ask for the Incoterm, the finishing and inspection inclusions, and the lead-time tier. A quote that names its assumptions is a cost basis.
• Limits. Ask what the shop does not do — what is out of scope, what is outsourced, where a tolerance does not apply. A supplier that states limits is one you can diligence.
• IP. Sign a non-disclosure agreement, send only what the quote needs, and keep a revision note on the drawing. A build-to-print shop that excludes product design from its scope is stating the IP boundary in writing.
• First article. Order a sample at MOQ 1 and check the measured values, material, and finish against your print. The sample should arrive with the same documents the audit promised.
• Payment. Wire only to the verified entity, with a deposit against the invoice and the balance against the shipping documents. A request to pay a third party or a personal account is the money version of a brochure.

Run the list top to bottom, stop the moment an item fails, and treat the speed of each answer as part of the answer. A genuine manufacturer finishes the list in about a business day and hands you a short stack of documents you can file. A trading desk stalls somewhere in the first three questions, and the stall is the finding.

The framework's real product is not a stack of documents; it is the ability to compare suppliers on evidence instead of adjectives. Two shops can both say "ISO certified" — only one can produce the certificate number and a registrar that confirms it. Two shops can both say "±0.003 mm (milling) / ±0.002 mm (turning)" — only one can produce the CMM report and the subgroups under it. When you run the checklist, the comparison stops being a feeling and becomes a table of checkable facts, and that table is what a rational sourcing decision is built from. Buy the evidence, not the brochure, and the wire transfer that follows is no longer a leap of faith.

What this shop actually runs

A buyer asking for capability evidence should get a specific answer. The facility behind this guide operates 100+ CNC machines across a 6,800 m2 plant in Dongguan, China, with 100+ staff. The machine list by category:

• 5-axis machining centres (10+) - Mazak VARIAXIS I-600 (510/910/510 mm travel, 12,000 rpm, 22 kW, 71.6 N·m, 30-tool BT40) and Mazak INTEGREX I-250H S (B-axis 0.0001°, 24 kW, 124 N·m, 38-tool HSK-A63). Geared to compound-angle parts, single-setup multi-face prismatic work, and aerospace fittings where datum transfer loss is unacceptable.
• 3-axis / 4-axis machining centres (50+) - Brother SPEEDIO S500Z2N / S700Z2N (10,000 rpm, 14/21-tool ATC, 0.7 s tool-to-tool, Z-axis 2.2 G) and Mazak VCE-430A L / VCE-570C L (10,000 rpm, 118 N·m, 24-tool ATC). Geared to prismatic parts, plates, brackets, housings.
• Swiss-type turning centres (10+) - Citizen Cincom A20-3F7 (Φ20 mm, 10,000 rpm), TSUGAMI B0206-III / B0205-III (Φ20 mm, 0.1 µm), STAR SR-32JII / SR-32J Type N (Φ32 mm, 8,000 rpm). Geared to long, slender rotational parts at ±0.002 mm.
• CNC turning centres (20+) - HEADMAN T70M/1250 (720 mm swing, 117 mm bore, 3,000 rpm, BMT75 12-station turret). Geared to shafts, bushings, flanges up to ±0.002 mm.
• Wire EDM (5+) - Sodick AQ535L (550/350/300 mm travel, ±20° taper, 800 kg workpiece). Geared to hardened tool steel, internal splines, sharp internal corners that milling cannot reach.
• Inspection: Hexagon CMM + 20+ gauges - Hexagon CMM for coordinate measurement, plus vision system, profilometer, hardness tester, coating thickness gauge, XRF for material verification. The 10:1 rule (gauge 10x more accurate than the band) is applied to every feature.
• Machine highlights - VARIAXIS I-600: one-setup multi-face 5-axis + MAZA-CHECK auto-calibration; INTEGREX I-250H S: DONE-IN-ONE mill-turn, B-axis 0.0001° + Ai thermal shield; SPEEDIO S500Z2N / S700Z2N: Z-axis 2.2 G + 0.7 s ATC; Swiss-type: guide-bush switchable + 0.1 µm resolution.

Tolerance bands held: ±0.003 mm on CNC milling, ±0.002 mm on CNC turning, ±0.002 mm on precision grinding and wire EDM. Capability indices: Ppk >= 1.67 initial on key characteristics; Cpk >= 1.33 ongoing in production.

Certifications: ISO 9001:2015 (116024/A/0001/UK/En), IATF 16949:2016 (131941/A/0001/SM/En, design clause 8.3 excluded - disclosed openly), ISO 14001:2015 (116024/B/0001/UK/En). Certificates are issued to Dongguan Zhuohang Ltd (USCC 91441900MA4UWT6X39, established 2005).

Service area: Austria, Belgium, the Czech Republic, Denmark, France, Germany, Greece, Hungary, Ireland, Italy, the Netherlands, Spain, Sweden and the United Kingdom in Europe; Brazil, Canada, Mexico and the United States in the Americas; China, Israel, Japan, Singapore and Sri Lanka in Asia; Australia and New Zealand in Oceania; and South Africa in Africa. Based in China, serving customers worldwide.

Contact: Felipe Iglesias, [email protected], +86 13126536955.

Frequently Asked Questions

Standards referenced on this page

• ASTM B211/B211M-19 — Standard Specification for Aluminum-Alloy Rolled or Cold Finished Bar, Rod, and Wire
• ASTM A276/A276M-17 — Standard Specification for Stainless Steel Bars and Shapes
• ISO 9001 — Quality management systems — Requirements