A193 B7, B7M and A320 L7: Specifying 4140 Bar That Will Actually Pass Bolting Certification

A193 B7, B7M and A320 L7: Specifying 4140 Bar That Will Actually Pass Bolting Certification

A fastener shop orders 4140 round bar, turns it into B7 studs, sends them for testing, and a batch comes back at 36 HRC. The chemistry was inside A29, the quench was fine, the temper was ten degrees short. Nothing about the bar was "wrong" in the mill-certificate sense, and yet the studs cannot be stamped B7. This is the everyday failure mode in bolting steel: the bar is bought to a bar specification, but it has to satisfy a bolting specification, and the two documents do not ask for the same things.

The three grades below cover most of the chromium-molybdenum bolting sold into refineries, petrochemical plants and LNG terminals, and all three are cut from the same family of bar. This article works through what each one demands of the starting stock, where the size bands bite, why the hardness ceiling matters more than the tensile figure, and what belongs on the purchase order if the studs are going to certify first time.

The Short Version

B7 and L7 are metallurgically the same material. Same chemistry band, same 125 ksi tensile and 105 ksi yield up to 2½ inch, same 321 HBW ceiling. What separates them is a test, not an alloy: A320 L7 makes Charpy V-notch impact testing at −101 °C mandatory, A193 B7 does not require impact testing at all. B7M keeps the B7 chemistry and gives up strength to buy hardness — 235 HBW (22 HRC) maximum instead of 321 HBW — which is the only version of the family that walks into sour service under NACE MR0175 / ISO 15156.

Read the table as a bar buyer and one line stands out: there is no alloy difference to argue about. Every decision that separates a good B7 heat from a rejected one happens in steelmaking, in the temper, or in how the bar was tested — not in the grade name on the drawing.

B7 Is a Bar Specification Wearing a Bolt Costume

A193 does not name 4140. It specifies a chromium-molybdenum chemistry band wide enough to swallow both 4140 and 4142, and leaves the mill free to work inside it. Put the two side by side and the difference is immediately visible:

Two consequences follow. First, a heat at 0.48 % carbon is a legal B7 heat and an illegal 4140 heat. That is not a technicality — carbon at the top of the band raises as-quenched hardness and makes the 321 HBW ceiling harder to hit at any given tempering temperature, so a mill that habitually runs high carbon produces bar that turns into non-conforming studs. Second, a heat at 0.66 % manganese is a legal B7 heat and an illegal 4140 heat. Manganese is there for hardenability, and the wider A193 window lets the mill buy cheaper scrap; the price is a carbon equivalent that moves around from heat to heat.

In practice this means the purchase order should not stop at "4140 to A29". If the studs are going to be certified to A193, the bar is best bought as a heat-controlled 4140 with a carbon target in the lower half of the band — say 0.38–0.42 % — and a manganese ceiling of 1.00 %. That single sentence removes most of the under-tempering scrap before the first bar is cut. The full composition ranges for the base grade are on our 4140 / 42CrMo4 round bar page.

The Size Bands: A Four-Inch Bar Cannot Claim 125 ksi

A193 steps the strength requirement down as diameter grows, and the steps are steeper than most quotations assume. The reason is mechanical rather than bureaucratic: a large section cannot be quenched fast enough at the centre to develop the same tempered structure as a small one, so the specification moves the target instead of pretending the metallurgy does not exist.

The trap sits between the mill and the machine shop. If you buy 5-inch bar and machine it down to 2-inch studs, which row applies? The answer depends on where the test specimen comes from and what the order says, and this is exactly where a shipment stalls at the witness point. The safe route is to state the finished stud size on the order and require the mechanical test to represent the finished section — either from the finished part or from a coupon cut from the same bar and given the same austenitising, quench and temper cycle. A coupon that skipped the quench proves nothing about the stud.

Reduction ratio deserves the same attention on forged or rolled bar. Machining a 200 mm bar down to 120 mm does not create 120 mm metallurgy; the centreline segregation and the porosity of the original ingot or bloom are still in there, only closer to the surface. For bolting that will be impact tested, a minimum reduction ratio from the cast section — 4:1 is a common floor, higher for critical service — is worth writing into the order explicitly, because it is invisible in any certificate that only reports chemistry and tensile results.

Hardness, Not Tensile, Is the Real Gate

A B7 stud at 125 ksi tensile typically lands somewhere around 26–30 HRC. The specification's 321 HBW (35 HRC) is a ceiling, not a target, and it exists for a reason that has nothing to do with strength: hard, under-tempered martensite is what cracks under hydrogen and what fails a stress-corrosion review. Bolting failures in service are almost never "too weak". They are too hard, in an environment that punishes hardness.

That makes tempering temperature the controlling variable, and it makes the mill certificate's tempering record more informative than its tensile number. For 4140-type steel, pulling hardness from the low thirties down to the mid twenties means tempering well above 600 °C, and the response curve is steep in that region — a 20 °C shift in furnace setpoint moves hardness by several points. Three practical consequences for anyone buying bar for bolting:

• Ask for the tempering temperature and the hardness distribution on the certificate, not just a single hardness reading from one end of one bar.
• Require hardness to be measured after the final temper, on a properly prepared surface, at a location agreed in advance — decarburised skin reads soft and hides a hard core.
• If the studs will be hot-dip galvanised or plated, remember that acid pickling and baking cycles can nudge hardness and hydrogen content; the ceiling has to hold after finishing, not only before it.

The same logic explains why "just re-temper the batch that failed" is a bad answer. Re-tempering drops strength as it drops hardness, and a heat that was under-tempered enough to hit 36 HRC will not necessarily still make 125 ksi after a second temper at a higher temperature. The steel has to be designed to pass both limits at once, which starts at carbon in the low half of the band.

B7M and the 22 HRC Wall

Sour service is where the family splits hardest. NACE MR0175 / ISO 15156 caps carbon and low-alloy steels at 22 HRC, and a standard B7 stud at up to 35 HRC sits well outside that. B7M is the answer: same chemistry, hardness restricted to 235 HBW (22 HRC) maximum, and the strength requirements lowered to match — 100 ksi tensile, 80 ksi yield, paired with an impact-tested 2HM nut.

Hitting 22 HRC reliably in a 1-inch stud is straightforward. Hitting it reliably in a 3-inch stud is a different job, because the section size that makes the centre slow to quench also makes it slow to temper uniformly, and hardness scatter across the section becomes the real risk. Two additions to the order pay for themselves here: a hardness traverse across the section rather than a single surface reading, and a carbon target at the bottom of the A193 band. A heat at 0.47 % carbon can be tempered into the low twenties, but it leaves no margin for a furnace with a cold spot.

One more thing that gets lost between the mill and the fastener shop: B7M is not a heat treatment you apply to B7 studs after the fact. Re-tempering an already-finished B7 batch to drop hardness changes dimensions, relaxes thread form and may drop tensile below the B7M floor. The grade has to be planned into the bar purchase.

L7 and the −101 °C Impact Test

A320 L7 does one thing B7 does not: it makes Charpy V-notch testing mandatory, at −101 °C (−150 °F), on specimens representing the finished product. The test temperature, specimen size and minimum absorbed energy come from the edition of A320 invoked on the order — the widely applied figure is a 27 J (20 ft·lbf) average on full-size specimens, but the edition on the purchase order is what governs, and a purchaser who quotes an energy value from a summary rather than from the standard is asking for an argument at the witness point.

Passing that test is a steel cleanliness problem more than a strength problem. At −101 °C the fracture path finds inclusions, and the impact energy is decided long before the tempering furnace — in deoxidation practice, in sulphur level and in inclusion shape control. What actually moves the number:

• Sulphur. A320 permits up to 0.040 %, which is generous for a specification with a −101 °C test attached. Orders that must certify first time commonly call for 0.010 % maximum, and the cost difference is small against a failed Charpy set.
• Inclusion shape. Calcium treatment or rare-earth addition turns stringy type II sulphides into globular inclusions that do not act as crack initiators. This is a ladle practice, not something a certificate shows — it shows up only in the impact result.
• Grain size. Fine austenitic grain size, held by aluminium or niobium, gives the cleavage facet somewhere to stop. Coarse grain from an over-hot forge or a long soak at temperature is the most common reason a heat with perfect chemistry misses the energy requirement.
• Test location and orientation. Specimens from the centreline of a large bar will read lower than specimens from mid-radius. A320 addresses the finished product; when bar is tested as a proxy, the location must be written down.

There is also a heat treatment subtlety specific to L7. To combine 125 ksi with low-temperature toughness, the steel needs a fully martensitic structure before tempering — a sliver of upper bainite or ferrite left by a slow quench will pass tensile and fail Charpy. That is why A320 leans on quench and temper with a controlled tempering range rather than allowing normalised stock. For a mill, this means section size and quench severity have to be matched at the order stage; for a buyer, it means a certificate showing "Q&T" without a tempering temperature is worth very little.

When the Job Goes Colder Than L7

Below about −101 °C the chromium-molybdenum answer runs out. A320 L43 brings nickel into the chemistry — a 4340-type Ni-Cr-Mo steel at the same 125 ksi / 105 ksi level — and the nickel is doing real work: it lowers the ductile-to-brittle transition temperature and keeps the Charpy shelf energy up where a plain Cr-Mo heat would already be brittle. The trade is cost, and the availability of bar with a documented low-temperature impact history behind it. For a look at how nickel changes the section behaviour of this family, our comparison of 34CrNiMo6 against 4140 covers the same trade from a shaft rather than a fastener angle, and the chemistry and properties of the nickel grade itself are on the 4340 / 1.6582 round bar page.

The substitution rule is worth stating plainly, because it is asymmetric. L7 will satisfy every mechanical requirement of B7, with documented impact data on top; B7 cannot substitute for L7 at any temperature below the code limit, no matter how good the individual heat happens to be. If a particular B7 heat would pass a −101 °C Charpy test, that is a pleasant accident — the specification does not require the test to be performed or reported, so the stud is non-compliant without the paper. Purchasing by "we tested it once and it was fine" is the most expensive way to buy bolting.

What to Put on the Bar Purchase Order

Most bolting-steel disputes are decided before any steel is melted, in the wording of the order. The list below is what we ask for when bar is destined for B7, B7M or L7 studs — it is short, and every line on it maps to a failure mode described above.

Two of those lines are worth a sentence more. Decarburisation is the quiet one: a hot-rolled 4140 bar can carry a decarburised layer deep enough that a stud machined from it has a soft skin over a hard core, and the surface hardness check then reports a comfortable 28 HRC on a part that has no business being in sour service. Peeled or turned bar removes the argument. And the certificate type is not paperwork theatre — a 3.1 certificate is issued by the manufacturer's inspection representative, a 3.2 is countersigned by an inspector independent of the manufacturer, and refinery and LNG specifications routinely name one or the other.

Frequently Asked Questions

Can we use 4142 instead of 4140 for B7?

Yes. A193's carbon band spans 0.37–0.49 %, which covers both 4140 and 4142, and both are standard starting stock for B7. The practical difference is that 4142 sits at the upper end, so it reaches a given hardness at a lower tempering temperature and leaves less room under the 321 HBW ceiling. Where the same shop also makes B7M, 4140 is the friendlier base.

Does a B7 stud need impact testing if the site is cold?

Not by the base specification — A193 imposes no Charpy requirement. If the design temperature falls below the limit the piping code allows for B7, the correct answer is to change the grade to A320 L7 rather than to add an impact test to a B7 order. Adding the test gives you data; it does not give you a compliant designation.

Why did a heat that met 125 ksi fail the 35 HRC ceiling?

Almost always tempering temperature or time, occasionally carbon at the top of the band. Hardness and tensile are not the same measurement: a heat can clear 125 ksi comfortably while sitting at 34–36 HRC, because the strength floor and the hardness ceiling are independent limits. The fix belongs in the heat treatment cycle, starting with a proper tempering soak rather than a short pass through the furnace.

Can bar be supplied annealed and hardened after machining?

Yes, and it is common when the fastener shop has its own quench and temper line. In that case the bar order should specify annealed condition with hardness ceiling and a microstructure free of excessive banding, and the shop takes responsibility for the Q&T cycle and the resulting certificates. What must not happen is an annealed bar arriving with a certificate that reads like a finished B7 stud — the two documents describe different states of the same steel.

Buying the Bar, Not the Grade Name

B7, B7M and L7 are three answers to three different service questions, built from one family of chromium-molybdenum bar. The steel that certifies reliably is the one whose carbon sits in the lower half of the band, whose sulphur was controlled for toughness rather than for cost, whose tempering temperature is on the certificate, and whose test specimens came from a location that represents the finished stud. None of that is exotic; all of it has to be asked for.

We supply 4140 / 42CrMo4 and 4340 / 1.6582 round bar in peeled, turned and Q&T condition from 20 mm to 500 mm, with heat-level traceability, EN 10204 3.1 certificates and third-party inspection arranged at order stage. If you are qualifying bar for a bolting line, send the finished stud size, the grade designation and the design temperature — the bar specification follows from those three numbers. Write to enquiry@qilumetal.com.