How to Choose Qualified Bolts

Nov 07, 2026 Leave a message

 

An in-depth guide from the production floor of a fastener manufacturer

 


202611071025322471

 

      Almost every purchase order for fasteners is decided on price per thousand pieces, and almost every fastener failure is decided somewhere else - in the steel, in the furnace, or in the plating bath. Between those two points sits a question that a catalogue cannot answer: is this bolt actually qualified for the joint it is going into?

      This guide is written from the inside of a fastener plant rather than from the outside of a price list. It explains what the marking on a bolt head really declares, which standards turn that declaration into a testable specification, how quality is created step by step on the production line, how it is proven in the laboratory, and how a buyer can audit a supplier in a single day.

202611071026062481

1 Reading the Mark Before Reading the Price

      The first piece of evidence is already in your hand. Every metric hex bolt carries a head marking, and that marking is the only statement the part makes about itself. It normally contains two pieces of information: the property class, and the manufacturer's identification symbol. A bolt with no marking, an illegible marking, or a marking that changes between cartons from the same order is not a cheap bolt - it is an unidentified one.

1.1 The property class sets everything downstream

      The two numbers are not a quality score. They are a mechanical specification. The first digit is one tenth of the nominal tensile strength in megapascals; the second digit is one tenth of the ratio between nominal yield strength and nominal tensile strength, expressed as a percentage. So 8.8 means a nominal tensile strength of 800 MPa and a nominal yield strength of 640 MPa, which is 80 percent of the tensile value. 10.9 means 1040 and 940 MPa. 12.9 means 1220 and 1100 MPa.

      That single number decides the steel, the heat treatment, the hardness window and the price, which is why it is also the number most often misrepresented.

 

Table 1 Property classes and what they require

202611071027382491 -

Nominal design values per ISO 898-1. Hardness ranges are typical practice; acceptance limits for a delivered lot are set by the standard's own test clauses.

1.2 The material behind the number

      Class 8.8 is generally produced from medium-carbon steel, frequently with boron additions to improve hardenability. Classes 10.9 and 12.9 require alloy steel - 35CrMo, 42CrMo or SCM435/SCM440 grades - with tighter limits on phosphorus, sulfur and residual elements, because those elements decide whether the bolt fails by yielding or by cracking. A supplier quoting 12.9 at 8.8 pricing is either using the wrong steel or skipping the tempering step. Both are invisible on an invoice and both are obvious on a spectrograph and a hardness tester.

1.3 Thread accuracy is a load path, not a detail

      Metric threads are specified by tolerance class: 6g for external threads and 6H for internal threads, a medium fit that covers the overwhelming majority of commercial fasteners. The class defines the allowances and tolerances on pitch diameter, minor diameter and flank angle. A thread that is merely close enough to assemble will still tighten, but load will concentrate on the first few engaged threads instead of distributing across the engagement, and the joint will lose preload earlier than the design intended. This is checked with GO and NO-GO gauges on the line and with a thread micrometer in the laboratory - two measurements that a serious supplier performs routinely and a broker does not perform at all.

1.4 Coatings are the first thing to fail, so specify them by system

      Three coating families dominate: electroplated zinc to ISO 4042, hot-dip galvanizing to ISO 10684, and zinc-flake coatings to ISO 10683. They are not interchangeable. Electroplating is the cheapest and the most common, but the process introduces hydrogen into the steel, which is why classes 10.9 and 12.9 require a documented de-embrittlement bake after plating. Hot-dip galvanizing gives the thickest barrier and is standard for outdoor structural work, at the cost of thread fit that must be accounted for at the drawing stage. Zinc flake systems deliver high corrosion resistance with no hydrogen risk and are the usual choice for automotive and high-strength applications. Quoting a coating by colour - silver, yellow, black - specifies almost nothing. Quoting it by system and thickness class specifies the corrosion life you are actually buying.

2 The Standards That Turn a Claim into a Specification

      A finished bolt is governed by four different families of standards at the same time: the product standard that fixes geometry, the material and mechanical property standard that fixes strength, the thread standard that fixes fit, and the coating standard that fixes corrosion performance. Buying to a product standard alone - DIN 933, for example - specifies a shape and says nothing about strength, thread tolerance or coating. This is the single most common gap in purchase specifications, and it is where most quality disputes begin.

 

202611071028522501 -

A purchase specification that names only the product standard is incomplete. The drawing should name one standard from each family.

 

      The drawing, not the catalogue, is where qualification starts. A line that reads only hex bolt M12 x 60, galvanized can be satisfied by almost anything on the market, and it will be. A line that reads hex bolt M12 x 60, ISO 4014 grade B, class 10.9 to ISO 898-1, thread 6g to ISO 965-2, zinc flake to ISO 10683 with salt spray hours stated, describes a part that can be accepted or rejected on evidence. The difference between those two lines costs nothing to write and prevents most of the arguments that follow delivery.

3 Where Quality Is Actually Created

      Certificates are issued at the end of the process. Quality is created in the middle of it. A fastener plant turns coiled wire rod into a finished bolt through six stages, and every stage can either protect or destroy the property class that will be stamped on the head later.

 

202611071029052511

Figure 1 Multi-station cold heading on the production line, where wire rod becomes a bolt blank before any heat treatment takes place.

 

3.1 Incoming material

      Every lot of wire rod arrives with a heat number. A controlled plant verifies chemical composition against the grade, checks surface condition for seams and cracks, and confirms that the wire has been annealed appropriately for cold heading. Decarburization - the loss of carbon from the surface layer during wire drawing or annealing - is the defect that quietly removes fatigue life, and it cannot be seen on the shop floor. It is found metallurgically, by measuring the depth of the carbon-depleted layer under a microscope. A supplier who has never measured decarburization on incoming wire is not controlling it.

3.2 Forming: cold heading and hot forging

      Cold heading progressively upsets the wire through a series of dies, producing the head and shank at high speed with good grain flow and useful work hardening. Hot forging is used for larger diameters and for shapes that cold heading cannot produce economically. Neither process creates strength; both create shape and grain orientation. Strength is added later, in the furnace, which is why a beautifully formed bolt can still be an 8.8 part sold as a 10.9.

3.3 Heat treatment is the step that decides the class

      Quench and temper is the whole game. The blanks are austenitized, quenched to martensite, and then tempered at a temperature chosen to land the final hardness in the window the property class requires. Tempering too cool gives a bolt that passes a tensile test and cracks in service; tempering too hot gives one that is soft, yields early and loses preload. Furnace atmosphere matters as much as temperature: uncontrolled atmosphere reintroduces decarburization and can charge the surface with hydrogen. In a well-run plant, hardness is checked at the head, the shank and the thread end of parts from each load, because that spread is the fastest way to detect an uneven furnace.

3.4 Thread rolling rather than thread cutting

      Rolled threads are formed by displacing material, so the grain flow follows the thread profile and the thread root is in compression rather than cut through. Cut threads interrupt the grain and leave tool marks at the root, which is exactly where fatigue cracks start. Rolling also raises fatigue performance and is faster. Cutting remains appropriate where the thread must run hard against a shoulder or on very large diameters, but a 10.9 or 12.9 fastener bought for cyclic loading should be rolled, and the process should be stated on the drawing.

3.5 Surface finishing and the hydrogen question

     Acid pickling and electroplating both generate hydrogen at the steel surface. Some of it diffuses into the metal and stays there. In a high-strength bolt, that hydrogen concentrates at stress raisers and causes delayed brittle failure - the part breaks hours or weeks after installation, under a load it had already carried successfully. The control is a de-embrittlement bake performed within a defined window after plating, typically within one hour and held for several hours at an elevated temperature, with the cycle recorded against the lot. Zinc-flake coatings and mechanical plating avoid the risk altogether. If a supplier cannot produce a bake record for a 10.9 or 12.9 electroplated bolt, the process was probably not performed, and the risk is now in your assembly.

3.6 Sorting, gauging and traceability

      The last stage is where a plant proves it knows what it has made: crack detection on critical parts, optical sorting to remove malformed heads and incomplete threads, sampling to the standard's test plan, and lot marking that links the carton back to the heat number and the furnace load. Traceability is not paperwork. It is the difference between a supplier who can tell you what went wrong and one who can only offer a replacement.

4 The Tests That Prove It

      A test report is only useful if you know what each test is capable of proving. Four groups matter, and a credible report covers all of them for the lot, not for a generic sample of the product family.

202611071029272521

Figure 2 Thread and dimensional inspection. A thread micrometer measures pitch diameter, while GO and NO-GO gauges confirm fit to the tolerance class on the drawing.

 

4.1 Mechanical tests

      The tensile test establishes ultimate strength. The proof load test is the more useful of the two for service performance: the bolt is loaded to the proof stress and must show no permanent elongation, which is precisely the condition a preloaded joint depends on. The wedge tensile test loads the bolt through a inclined washer so the head is forced into bending; it catches head-to-shank defects and material toughness problems that a straight tensile test passes. Hardness testing is the fastest proxy for heat treatment and should be reported as a range across several positions, not a single value.

4.2 Metallurgical checks

      Microstructure confirms that the part was actually quenched and tempered rather than normalized or left as-rolled. Decarburization depth confirms the surface has not lost carbon. Both are cheap to perform and rarely reported unless the buyer asks, which is a reasonable summary of why buyers should ask.

4.3 Dimensional and coating checks

      Thread gauges and micrometers cover fit. Coating thickness must be measured on the thread crest as well as the shank, because electroplating deposits preferentially and a thin crest is where corrosion starts. Salt spray hours to red rust are the number that matters commercially, but only when quoted with the coating system, the thickness class and the test standard, since hours without those three mean nothing.

 

202611071029542531

A test report should state the lot number, heat number, standard, class, measured values against limits, date and the signature of the person who ran the test.

202611071030042541

 

5 How to Audit a Fastener Manufacturer

      Auditing a supplier is not a search for the most certificates. It is a search for evidence that the process is measured, and that the measurement is recorded in a form you could trace six months later.

5.1 Certifications, and what they do not prove

      ISO 9001 is the baseline. IATF 16949 is the meaningful one for automotive supply and imposes process capability, traceability and change-control requirements that a general quality system does not. A certificate confirms that a system was audited on a particular date; it does not confirm that the system was running on the day you walked in. Ask for the previous audit's non-conformities and their closure evidence. A supplier who can discuss their own findings openly is usually the safer one.

5.2 The laboratory tells you what the plant can see

      Walk the lab before the showroom. A tensile tester sized for the largest product the plant makes, a calibrated hardness tester, a salt spray chamber, a spectrograph for incoming steel, a metallographic microscope for decarburization and microstructure, a full set of thread gauges with calibration labels. If the equipment is absent, the control is absent, and the test report is being issued by someone else on the basis of a sample you have never seen.

5.3 Traceability from heat number to carton

      Ask the supplier to trace one carton on the floor back to its steel heat, its furnace load and its test results. This is a fifteen-minute exercise for a controlled plant and an impossible one for a trading company. The answer tells you more than any certificate.

5.4 Process capability, stated honestly

      Ask for capability data on the critical characteristics - head height, thread pitch diameter, hardness. A supplier who reports Cpk values and can explain the weak ones is measuring. A supplier who reports only pass or fail is inspecting, which is a different and weaker thing. The same conversation should establish capacity and lead time honestly: a factory that accepts any volume offered is either subcontracting outside its control or planning to ship late.

5.5 A one-day sample audit you can run yourself

      You do not need a laboratory to catch most problems. Take three cartons from different lots and: check that the head marking is identical and legible across all three; measure ten thread pitch diameters with your own micrometer against the drawing tolerance; run a file or portable hardness check on the head and shank of several parts; look at the thread crest under a magnifier for laps, tears and thin coating; and weigh a fixed count of parts to detect mixed sizes. Any one of those checks failing is a reason to stop, and none of them costs more than an afternoon.

6 The Procurement Checklist

      The following ten points convert the preceding sections into something that can be attached to an enquiry. A supplier who can answer all ten with documents is a supplier worth developing; one who answers three is a supplier worth sampling before trusting.

 

Table 4 Ten checkpoints and the evidence to request

202611071031202551

Items 1 to 9 are supplied by the manufacturer; item 10 is the buyer's own check and should never be skipped.

 

7 Failure Modes You Will Actually See

7.1 Grade substitution

      A bolt marked 10.9 that behaves like 8.8 is the most common and most profitable form of misrepresentation in the trade, because the price difference between the two classes is real and the marking costs nothing. It is detected in one minute with a hardness tester, and almost never detected by the buyer who inspects only the paperwork.

7.2 Decarburization

      A soft, carbon-depleted surface layer on a correctly hardened bolt looks normal and tests normal in tension, then initiates fatigue cracks at the thread root under cyclic load. It is a furnace atmosphere problem, and it is invisible without metallography.

7.3 Hydrogen embrittlement

      The failure arrives late and without warning: bolts that were torqued correctly and held load for days or weeks break suddenly, often with no measurable elongation and a brittle, intergranular fracture surface. It looks like a bad bolt. It is almost always a missing bake after plating, which makes it a process failure in a part that was dimensionally perfect.

7.4 Thread galling

      Stainless fasteners and some zinc-flake systems gall under power tools, especially when lubrication is omitted to keep the parts clean for inspection. The joint seizes before it reaches the specified preload, so the assembly is under-torqued while the wrench reports success.

7.5 Coating thickness measured in the wrong place

      Electroplated coatings deposit thinner at the thread crest than on the shank because of current density. A report generated from a shank measurement can pass a specification that the part fails at the crest, where corrosion actually begins.

7.6 Mixed lots in one carton

      Two heats, two coating runs, or two machines in the same box. Individually the parts may be fine; together they make traceability impossible and guarantees that the one failed part you eventually find cannot be tied to its process record.

8 The Real Cost of a Bolt

      A fastener is one of the few components whose cost and consequence are almost completely unrelated. The part costs cents. The joint it holds can cost a production day, a warranty claim, or an inspection finding that closes a project. The price difference between a documented bolt and an undocumented one is usually measured in small percentages; the difference in consequence is measured in orders of magnitude.

      Which is why the useful question is not what does this bolt cost, but what does this bolt declare, and who can prove it. A manufacturer that answers with a standard, a property class, a heat record, a test report and a lot number is selling you a part you can design around. A manufacturer that answers with a price is selling you an assumption. Both are legitimate businesses. Only one of them belongs in a joint that matters.

      Qualified, in the end, is not the adjective on the certificate. It is the chain of evidence that runs from the steel mill to the carton in your warehouse, and it is worth checking before the bolts go in, not after they come out.

 

 


This guide is part of an industry knowledge series for purchasing engineers, structural fabricators and OEM supply-chain teams. Standards cited are the current ISO, DIN, ANSI/ASME, ASTM and GB/T documents governing metric and inch hex bolts; always confirm the edition referenced on your own drawing.

 

 

 

Contact now

 

 

 

 

 

Send Inquiry

whatsapp

Phone

E-mail

Inquiry