Machining Stainless and Duplex Steels for Tensile Specimens

Stainless is one word covering two materials that behave nothing alike at the tool. An austenitic grade such as 304 or 316 is soft enough to cut easily and stubborn enough to harden while you do it. A duplex grade is roughly twice as strong before you start, and it carries a second problem that austenitic grades do not: its properties depend on a phase balance that heat can change. Treat them as one job and the specimens will tell you the difference later, in the tensile data.

This is not the same discussion as machining a nickel superalloy or a hardened tool steel, which we covered in machining Inconel, tool steel and hardened alloys. The stainless families fail in their own specific ways, and the duplex failure is metallurgical rather than dimensional.

A freshly machined flat steel tensile specimen held up beside the milling machine it was cut on — Every dimension on a specimen like this can check out while the reduced section tells a different story. On stainless, what decides the tensile result is how the cut behaved, not only where it ended.

Austenitic Grades Harden Under the Tool That Is Cutting Them

The defining property of an austenitic stainless is a high work hardening rate. The surface a tool leaves behind is harder than the material the tool met, which means the next pass, the next insert, or the next operation is cutting something the drawing never described. Combine that with low thermal conductivity, and the heat generated in the cut stays in the cut instead of leaving through the workpiece, so the tool runs hotter and the surface layer takes more of the thermal history than it should.

The visible symptoms are familiar to anyone who has cut 316 on an under-powered setup. Chips come off long and stringy rather than breaking, material welds to the cutting edge and builds up there, and the finish deteriorates as that built-up edge forms and sheds. On a production part, a poor finish is cosmetic. On the gauge section of a tensile specimen it is a measurement problem, because the reduced section is where the specimen is meant to fail and the surface condition is part of what decides when.

Light, hesitant cuts make it worse rather than better. A tool that rubs instead of cutting work hardens the surface without removing it, and the next pass then has to cut through that hardened layer. The practical rule that follows is counterintuitive for anyone trained on cosmetic finishing: take a definite cut with a sharp edge and a rigid setup, and do not let the tool dwell in the material.

Austenitic grades

The Light Pass Is the Expensive One

A work hardening grade turns hesitation into a cycle. A dull edge on a light pass rubs the surface instead of separating a chip, the surface left behind is harder than the bar was, and the next pass has to cut through that. Break the cycle by cutting rather than by easing off: a sharp edge, a rigid setup and a real depth of cut send the hardened layer away with the chip, and what stays in the reduced section is the material rather than a record of how it was machined.

Scroll the diagram sideways to see all of it.

Duplex Adds a Problem That Does Not Show on the Surface

Duplex grades are two-phase, austenite and ferrite together, and that structure is where their strength and their corrosion resistance come from. It is also what makes them sensitive in a way austenitic grades are not. The machining difficulty itself is well documented: a high work hardening rate, high strength, low thermal conductivity, fracture toughness that resists chip separation, and a tendency to form a built-up edge, which together put duplex below the austenitic grades on any machinability comparison.

The part that matters for specimen preparation is what heat does after it goes in. Thermal cycles above roughly 550 degrees Celsius promote a hard, chromium rich intermetallic constituent, and once it is present in more than a small fraction it takes both toughness and corrosion resistance down with it. Lower down, in the region around 425 to 475 degrees, a different embrittling phase forms. Neither one announces itself on the surface of a bar, and neither is visible in a dimensional inspection.

Duplex grades

Two Temperature Bands That Do Not Show on the Surface

Duplex owes its strength and its corrosion resistance to a balance between two phases, and heat moves that balance. Above roughly 550 °C a hard chromium rich intermetallic constituent forms, and past a small fraction it takes toughness and corrosion resistance with it. Around 425 to 475 °C a different embrittling phase does its own damage. The upper band is drawn open because the source sets a floor for it and no ceiling. A gauge section taken through either band may no longer represent the bar it was cut from, and every dimension on the specimen will still check out.

Scroll the diagram sideways to see all of it.

The industry treats this seriously enough to have written a standard around detecting it. ASTM A923 exists specifically to find detrimental intermetallic phase in duplex austenitic and ferritic stainless steels, to the extent that toughness and corrosion resistance are significantly affected, and it offers three routes to the answer: a sodium hydroxide etch, a Charpy impact test, and a ferric chloride corrosion test. A specification that calls for A923 acceptance on the material is telling you, in advance, that the thermal history of that material is not a detail.

Put those two facts next to each other and the conclusion for a specimen shop is uncomfortable. If a cutting operation is allowed to heat the gauge section of a duplex specimen, the specimen may no longer be metallurgically representative of the bar it came from, and the tensile result then describes a material that exists only in the reduced section. This is a harder failure to catch than a taper or a burr, because every dimension will check out.

What a dimensional report cannot tell you about a duplex specimen

The gauge section saw a thermal cycle the bar never did. A cut allowed to heat the reduced section can move the phase balance in that section alone. The diameter, the taper and the radius all conform, and the material under test is no longer the material that was ordered.

The acceptance criteria may already say so. ASTM A923 exists to detect detrimental intermetallic phase in duplex austenitic and ferritic stainless steels, through a sodium hydroxide etch, a Charpy impact test or a ferric chloride corrosion test. A specification that calls for it is telling you in advance that thermal history is not a detail.

The evidence you will be asked for is the route, not the result. When a number is questioned, a shop that can state the power, the depth of cut, the tool change policy and what was done to keep heat out of the reduced section is in a different position from one that can only produce a dimensional report.

What This Changes in the Preparation Setup

Everything above points the same way: the cut has to be positive, the heat has to leave, and the process has to repeat without an operator nursing it. That is a setup question more than a technique question.

Rigidity comes first, because deflection turns a planned cut into rubbing. Power comes with it, since an under-powered spindle forced into a light pass is the exact condition that work hardens the surface ahead of the tool. Our flat specimen preparation machines are built around that requirement, with the Classic Upgrade running a 3.2 kW spindle, positioning accuracy of 0.13 mm and repeatability of 0.02 mm, and a rated capability on material up to 60 HRC. Duplex sits comfortably inside that envelope, which means the limiting factor is the cutting strategy rather than the machine.

Batch behaviour matters as much as single-part capability. The triple clamping fixture takes up to three stacked sets of one inch material, so a set of specimens is cut under the same conditions rather than one at a time with the tool in a different state each time. Consistency across a set is what allows scatter in the tensile results to be read as material scatter instead of preparation scatter.

Flat specimen blanks clamped together in the fixture of a CNC milling system before the reduced sections are cut — A set cut in one setup shares one tool state. Cut the same specimens one at a time and the insert is in a different condition for each of them, which puts preparation scatter into results that will be read as material scatter.

Round work has its own constraint. On a turned specimen the gauge diameter, the shoulder transition and the concentricity of the bar all have to hold while the material resists the tool, which is what our TensileTurn CNC is built for. And where the surface finish of the gauge length is specified rather than assumed, a directional finish applied with the automatic longitudinal polisher removes the circumferential marks a turning operation leaves behind, which on a work hardening material is the difference between a specimen that fails in the gauge and one that fails at a scratch.

Close-up of a round tensile specimen with a circumferential mark at the shoulder of the reduced section — A circumferential mark sits broadside to the load. On a work hardening grade it becomes the place the specimen chooses to fail, and the result then describes the mark rather than the material.

Tooling is consumable here in a way it is not on mild steel. An edge that has started to dull is no longer cutting a work hardening material, it is burnishing it, and the specimen surface records that. Planning insert changes by part count rather than by appearance is the discipline that keeps a batch uniform, and it is worth stocking for, which is why we keep consumables and spare parts available alongside the machines.

A set of end mill cutting bits used for preparing tensile specimens — On a work hardening grade the edge is a process variable, not a spare part. Change inserts by part count rather than by appearance, because a dulling edge stops cutting and starts burnishing long before it looks worn.

Preparation equipment

What a Work Hardening Grade Asks of the Setup

Rigidity so the tool cuts instead of rubbing, power so a definite pass is available, and a route that repeats across a set rather than being nursed one specimen at a time.

TensileMill CNC flat tensile specimen preparation machine for stainless and duplex coupons

Flat Specimen Preparation

The Classic Upgrade runs a 3.2 kW spindle on material to 60 HRC, with positioning accuracy of 0.13 mm and repeatability of 0.02 mm, and a triple clamping fixture that takes up to three stacked sets of one inch material.

Flat coupons

A whole set cut under one tool state
Specimens to 14 by 2 in

Flat tensile sample preparation machines

TensileTurn CNC round tensile specimen preparation machine

Round Specimen Preparation

Turned work holds the gauge diameter, the shoulder transition and the concentricity of the bar while the material is resisting the tool, which is where an under-powered setup starts rubbing instead of cutting.

Round bar

Concentric through the gauge
Clean shoulder transition

Round tensile sample preparation machines

Automatic longitudinal polisher system for the reduced section of tensile specimens

Longitudinal Polishing

A directional finish along the specimen axis removes the circumferential marks a turning operation leaves, which on a work hardening grade is the difference between failing in the gauge and failing at a scratch.

Gauge finish

Along the loading direction
Applied the same way on every specimen

Longitudinal polisher systems

What the Standards Expect From the Finished Specimen

None of the above changes the geometry the test asks for. ASTM E8/E8M and ISO 6892-1 define the specimen forms and the dimensional tolerances that make a result valid, and ASTM A370 is the route most steel product testing arrives through, with the product specification deciding what has to be reported.

What changes is the burden of proof around them. For a duplex order, dimensional conformity is necessary and not sufficient, because the acceptance criteria attached to the material may include an intermetallic phase check that the preparation route can influence. A laboratory that can describe how its specimens were cut, at what power, with what tool life policy, and with what steps taken to keep heat out of the reduced section, is in a much stronger position when a result is questioned than one that can only produce a dimensional report.

The Short Version for a Specimen Shop

Austenitic grades punish hesitation: cut positively, keep the edge sharp, and accept that a rubbing tool is making the next pass harder. Duplex punishes heat twice over, once at the surface and once in the microstructure, and the second one is invisible until somebody runs an impact or corrosion check and gets an answer nobody expected.

Next step

Send us the grade and the acceptance criteria, not just the section size

Tell us which stainless families cross your bench, the section sizes you prepare, the standards you report against, and whether your orders carry intermetallic phase acceptance requirements. We will confirm the machine configuration, the fixturing for those sections and a tooling plan that survives a work hardening grade.

Request a quotation

Related machines

Equipment That Does This Work