​‌​​‌​

AISI 303 / (1.4305) / UNS S30300 / AMS 5635 / AMS 5638

Corrosion resistance: The corrosion resistance of AISI 303 is slightly lower than that of 304 stainless. Grades 304 or 316 should be specified for harsher conditions. Grade 303 should not be exposed to marine or similar environments. Like other common austenitic stainless steels, grade 303 is subject to chloride stress corrosion cracking in environments above approximately 60 °C.​‌​​‌​

Temperature capability: 303 has good oxidation resistance in environments reaching 760 °C intermittently and 870 °C continuously. Carbide precipitation occurs in the 425-860 °C temperature range, however.

Weldability: Weldability is poor, so welding is generally not recommended. If it has to be carried out, the use of 308L or 309 electrodes is recommended. The material must be annealed after welding for maximum corrosion resistance.​‌​​‌​

Machinability: This is the grade with the best machinability among the austenitics. Its machinability rating is around 78% and it is known as the free-cutting stainless steel. Although the sulphur (S) addition raises machinability, it lowers corrosion resistance (lower than that of 304) and weakens formability; it is particularly unsuitable for sharp bends.

Heat treatment: Unlike the martensitic stainless steels, AISI 303 does not present a risk of hardening and embrittlement when heat treated. The steel generally does not require heat treatment, because the sulphur it contains already gives it good machinability and it does not display martensitic behaviour. In some cases, however, heat treatment can be applied to 303 in order to obtain greater hardness and strength. Solution treatment: 1010-1120 °C (1850-2050 °F). Ageing: this steel is generally not aged, but where necessary ageing can be carried out in the 500-700 °C (930-1300 °F) range. Stress relieving: 250-370 °C (480-700 °F).​‌​​‌​

Applications: 303 (1.4305) is frequently used in the manufacture of bolts, stainless screws and stainless nuts, in mass-produced stainless steel shafts, in the food sector, in mass-produced stainless shafting, in automotive gears and fasteners, in electronic units and in certain special decorative products — in short, wherever stainless steel and volume production are required together.

What AISI 303 Is — Machinability Bought With Sulphur, and What It Costs​‌​​‌​

AISI 303 (UNS S30300 / W.Nr. 1.4305 / EN name X8CrNiS18-9) is the free-machining derivative of the 18-8 austenitic family. The base chemistry follows the same logic as 304 — nominally 17–19 % Cr and 8–10 % Ni — but one deliberate addition changes everything: sulphur is imposed by ASTM as a 0.15 % MINIMUM. This is not an impurity; it is a design decision written into the specification. It creates thousands of manganese sulphide (MnS) inclusions inside the steel; those inclusions break the chip, act as a lubricant at the tool–chip interface and cut the cutting force.

And exactly the same inclusions are thousands of ready-made pit nuclei underneath the passive film. The whole story of 303 fits in one sentence: it is a manufacturing alloy, not a corrosion alloy. Selling 303 as “the easier-machining version of 304” is a dishonest one-order win; the buyer learns the truth either during fabrication (it cannot be welded) or in the first year of service (pitting).​‌​​‌​

303Se (S30323) — the same idea, with selenium​‌​​‌​

303Se tries to do the same job with selenium instead of sulphur: ASTM A582 defines the grade with Se ≥0.15 % and S ≤0.06 %. Selenide inclusions tend to stay more globular than sulphides; the result is better surface finish and cleaner behaviour in burnishing, thread rolling and tight-tolerance work, and somewhat better cold formability than sulphurised 303. But the corrosion problem does not go away — only the chemistry of the inclusion changes. 303Se is still a free-machining stainless and still below 304. It is also effectively absent in Europe: no W.Nr. equivalent for 303Se is widely published, supply is US/Asian and lead times are long.

Are 1.4305 and S30300 the same thing?​‌​​‌​

Commercially yes, at certificate level no. Both are “303” and both do the same job, but the composition bands are not identical, and that difference decides whether a mill certificate is acceptable against a given specification. The two sharpest: carbon — ASTM A582 allows ≤0.15 %, EN 10088-3 allows ≤0.10 %; and sulphur — ASTM sets only a LOWER bound (≥0.15 %, no upper limit), while EN sets a band (0.15–0.35 %). So a heat with 0.45 % sulphur complies with ASTM A582 and does NOT comply with EN 1.4305. The phosphorus gap is even wider: ASTM ≤0.20 % against EN ≤0.045 % — more than four times. If your customer buys to a European drawing and you ship an ASTM-certified bar, phosphorus can fail you.

Standards by Product Form​‌​​‌​

The most important part of this table is its empty rows. 303 is a bar and screw-machine grade; it does not appear in sheet, plate, boiler tube, process pipe or pressure equipment specifications. Saying so up front prevents an order that will later be cancelled.​‌​​‌​

ASME and Pressure Code Status — Short and Clear​‌​​‌​

This section is normally a table of numbers; for 303 it is a table of refusals, and it should be published as such.

Product Forms With NO Standard — the Section Your Sales Engineers Should Memorise

Most 303 enquiries are actually for a product form that 303 cannot serve. The rows below prevent a cancellation, a return cost and a loss of trust.​‌​​‌​

Chemical Composition​‌​​‌​

There are two separate systems here, and mixing the rows leads to certificate rejection.

Mechanical Properties

​‌​​‌​

The trap in the mechanical section of 303 is this: the primary US bar specification, ASTM A582, sets NO tensile / yield / elongation minimum for S30300 — only a hardness ceiling. The great majority of published “303 mechanical properties” are typical mill values, not specification minima. The European route is the exact opposite: EN 10088-3 sets real minima.​‌​​‌​

It cannot be hardened. 303 is fully austenitic; it cannot be hardened by heat treatment. Strength rises only through cold deformation — and 303’s capacity for cold deformation is limited. If you need a hardenable stainless that also machines well, the candidates are 17-4 PH and the martensitic free-machining grades; 303 is not on that list.

Physical Properties​‌​​‌​

Warning: 303 physical data circulates in two different “schools” among publishers. Both are given separately below; do NOT average them.

Heat Treatment and Thermal Stability

​‌​​‌​

​‌​​‌​

Welding

The honest heading for this section is: 303 was not designed to be welded. What follows is not “how to weld it well” but “how to limit the damage if it is unavoidable”.​‌​​‌​

Machining​‌​​‌​

303 exists for this section. The figures below are starting values; machine rigidity, tool quality and coolant strategy will move them more than anything else.

Is there really a sulphur-free alternative? Yes, but a limited one. The major mills produce improved-machinability 304 variants: controlled sulphur (well below free-machining levels), controlled inclusion shape, low residual elements and tight grain size. They do not reach 303 speeds, but they are clearly above standard 304 and they weld like 304 and corrode like 304. These variants carry no separate AISI number — the ASTM chemistry is still S30400/S30403. The correct commercial sentence is: “If there is welding or corrosion, do not go to a free-machining grade; ask for improved-machinability 304 and get the supplier to confirm it by mill brand name.”​‌​​‌​

Corrosion — Where It Works, Where It FAILS

​‌​​‌​

One mechanism explains everything: the MnS inclusion. The passive chromium oxide film is essentially as good as that of 304 — but the metal beneath it is not the same. Every sulphide inclusion is a discontinuity between film and matrix, and in chloride the pit starts exactly there: the inclusion dissolves, an acidic chloride-rich micro-environment forms inside, and the pit deepens autocatalytically.​‌​​‌​

Where it works well

Dry indoor environments. Switchgear and panel components, instrument bodies, furniture and architectural interior hardware, internals of office and laboratory equipment.
Mildly corrosive atmospheres. Urban indoor conditions, lubricated machine elements, gears, bushings, shafts, keys, set screws, lock cylinders.
Oil and grease contact. An oil film keeps both chloride and oxygen away; 303 shafts and bushings run for decades in lubricated bearings.
Neutral, chloride-free internal volumes — provided there is no crevice geometry.​‌​​‌​

Where it FAILS — this list is not negotiable

1. Seawater and marine atmosphere. FORBIDDEN. The producer statement is explicit: “not recommended for marine or similar environments.” Even an outdoor application one kilometre from the coast is a risk for 303. Using 303 on a boat, a jetty or a coastal plant is an invitation to pitting and stress corrosion cracking.
2. Pitting and crevice corrosion in chloride-bearing aqueous media. 303 pits before 304 in the presence of chloride. Gaskets, interference fits, thread roots, blind holes, under-deposit and under-paint areas are all crevice geometries, and they are the weakest point of 303.
3. Chloride stress corrosion cracking (CLSCC). The published threshold: susceptible above ~50 °C in chlorides — lower than the ~60 °C threshold normally quoted for the austenitic 18-8 family. An independent safety-authority report says it directly: the free-machining grades (303, 303Se) show increased susceptibility to CLSCC because the sulphide inclusions act as nuclei for localised corrosion. Welded or cold-worked 303 is worse still — the residual stress is already there.
4. Intergranular corrosion. A producer table marks the intergranular corrosion resistance of 1.4305 as NO in the as-delivered condition and NO in the as-welded condition. Carbon up to 0.15 % and any pass through the 425–860 °C band produce it.
5. Transverse sections and end faces. Because the inclusions are aligned with the rolling direction, the end face of a bar is more vulnerable than its cylindrical surface. The most attacked area of a 303 part is often the faced end or the cut surface — and this appears in field reports again and again as “unexpected”. It is not unexpected.
6. Hygienic / food / pharmaceutical lines. Cleanability and chloride-bearing CIP chemicals make 303 unsuitable.
7. High temperature followed by aqueous service. Every hour spent in the 425–860 °C band lowers the corrosion resistance available later.​‌​​‌​

Honest Comparison — 303 or Something Else​‌​​‌​

Frequently Asked Questions​‌​​‌​

Is the difference between 303 and 304 only machinability? Our parts run dry — should we buy 303?

The difference is not only machinability, but in the conditions you describe 303 may well be the right choice. Let us make the distinction clean.
On chemistry the two look like siblings: both are 17–19 % chromium, 8–10 % nickel. The divergence is sulphur: ASTM imposes a 0.15 % minimum on 303, while in 304 sulphur is an impurity capped at 0.030 % maximum — at least a five-fold difference, in practice often ten-fold. That sulphur sits in the steel as manganese sulphide inclusions which break the chip and lubricate the tool. The machinability rating rises from 40–45 % to 69–78 %, roughly 1.7–1.8 times. On a high-volume turned part that is a genuine cost difference.
The price comes in three items. First, corrosion: those same inclusions are pit initiation sites in chloride; 303 is not recommended for marine and similar environments and is susceptible to stress corrosion cracking above ~50 °C in chlorides. Second, welding: 303 is effectively unweldable. Third, forming: it is unsuitable for cold bending, heading and deep drawing.
The decision rule for your case: if the part runs in a dry or lubricated indoor environment, will not be welded, will not be cold formed and will not see chloride, then 303 is the correct and economical choice — gears, bushings, shafts, set screws and panel hardware fit that description exactly. If even one of those four conditions fails, the machining time you gain is smaller than the cost of one field failure. And there is a middle path: improved-machinability 304 variants machine clearly faster than standard 304 without reaching 303 speeds, and they weld like 304 and corrode like 304.​‌​​‌​

Our supplier says they can TIG weld 303 parts. Can it actually be done?

Physically the arc strikes and a bead forms. As engineering, that is not a weld — it is a planned crack.
The mechanism: sulphur and phosphorus form very low melting point liquid films in the molten pool. As solidification advances these films are pushed to the grain boundaries, and as the bead cools the shrinkage stress opens exactly those liquid films. The result is hot cracking — often invisible to the eye, found by dye penetrant or radiography, and sometimes only in the field. In 303 these two elements are not impurities but a specification requirement; we are not talking about accidental contamination but about the definition of the material.
The institutional side is equally clear: a national welding standard does not pre-qualify welding of 303. You cannot claim “welded to a standard procedure”; each joint needs separate qualification, and passing is not guaranteed. A European mill marks intergranular corrosion resistance for 1.4305 as NO as delivered, NO as welded.
If it really is unavoidable, the way to limit damage is to create δ-ferrite in the weld metal: ferrite dissolves sulphur and phosphorus far better than austenite and interrupts the crack path. That is why the recommendations point to high-ferrite fillers — E312 gives the most ferrite, 309 sits between, 308L gives the least. Autogenous (no filler) TIG must not be used, because then the pool is pure base-metal chemistry. Keep heat input low, beads small and interpass temperature cold. Post-weld solution annealing is needed for maximum corrosion resistance; but in the producer’s own words the result still remains poor both mechanically and for corrosion.
The correct commercial answer: make the welded part from 304 or 304L. The minutes lost on the machine come back from NDT, scrap and field failure.​‌​​‌​

Our 303 attracts a magnet — did we get the wrong material?

Almost certainly not, and this is physically expected behaviour.
303 is fully austenitic in the solution-annealed condition and effectively non-magnetic; published relative permeability values range from ~1.02 to ≤1.3. But austenite in 18-8 chemistry is thermodynamically metastable. Cold deformation — bar drawing, rolling, the turning operation itself, thread rolling, centre pressure — converts part of the austenite to strain-induced martensite. Martensite is ferromagnetic. So a cold-drawn 303 bar attracting a magnet lightly is normal, and the machined surface itself may be more magnetic than the bulk.
There are two practical consequences. First: “stainless does not attract a magnet” is not a material rule but a folk belief; for the 18-8 austenitics it is close to true only in the annealed condition. Second: if the buyer specification really contains a magnetic permeability ceiling (instrumentation, sensor proximity, medical imaging environments), that must be discussed before the order. The solution is then either solution annealing the finished part — which brings dimensional and surface risk — or moving to a more stable austenitic grade from the outset. Remember: this magnetism is not a chemistry error but a trace of processing history; a mill certificate verifies chemistry, not magnetic response.​‌​​‌​

Shafts we machined from 303 bar pitted within two years. The certificate looked compliant. What happened?

The certificate probably was compliant. The cause of failure was not chemistry but microstructure and geometry.
A mill certificate verifies dissolved chemistry: chromium, nickel, carbon, sulphur. The weakness of 303 is invisible in those rows, because the weakness is the form the sulphur takes — thousands of manganese sulphide inclusions. Each inclusion is a discontinuity beneath the passive film. When chloride loads that film, the pit starts exactly at those inclusions: the inclusion dissolves, the inside of the pit turns acidic and chloride-rich, and the pit deepens autocatalytically.
Look at two things. First, direction: sulphide inclusions are strung out along the rolling direction, which is why corrosion resistance is markedly lower in cross-section. The faced end, groove root and shoulder radius of a machined shaft are usually attacked first. Second, crevices: bearing seats, o-ring grooves, interference fits, keyways and thread roots are all crevice geometries; inside a crevice oxygen is depleted and chloride concentrates, and 303 is at its weakest there.
Also ask: where did the chloride come from? Very often the answer is not the process fluid but the environment — coastal atmosphere, road salt, condensation, wash water, moisture under insulation, or a cleaning agent used during assembly. If the temperature exceeds ~50 °C and the part is stressed (a machined and torqued shaft is), then the risk on the table is not only pitting but stress corrosion cracking.
What to do: if the same geometry will see chloride, change the material — 316 or 316L move up a class in pitting and crevice resistance thanks to molybdenum. If the geometry cannot change, seal the crevices and radius the sharp corners. What you must not do is ask the supplier for “a better 303”: the corrosion limit of 303 is not a quality problem, it is the definition of the grade.​‌​​‌​

Common datasheet errors — check these before you order

1. Selling 303 with 304 corrosion data — the most common and most expensive error. Many distributor pages print the chemistry of 303 and then copy the 304 text into the corrosion section: “excellent corrosion resistance”, “food industry”, “broad chemical resistance”. That is not true. The producer’s own wording: the resistance of 303 is “significantly less than Grade 304 due to the sulphur addition”. Do not trust any 303 page whose corrosion paragraph was copied from 304.
2. Offers for “303 sheet / plate / pipe”. There is no sheet, plate, strip, pipe or tube product standard for S30300; ASTM A240, A312, A213, A249 and EN 10088-2 do not include 1.4305. Such an offer is either the wrong grade or a product with no standard behind it.
3. Believing ASTM A582 places an UPPER limit on sulphur. A582 sets only the S ≥0.15 % minimum; there is no ceiling. EN 10088-3 gives a 0.15–0.35 % band. If you do not write a sulphur band into the order, machine behaviour will vary from lot to lot.
4. Forgetting phosphorus. ASTM ≤0.20 % against EN ≤0.045 % — more than four times. An ASTM-compliant heat may fail 1.4305. If you want dual certification, write carbon and phosphorus into the order.
5. Printing mechanical values as if they were specification minima. ASTM A582 sets no tensile / yield / elongation minimum for S30300 — only ≤262 HBW. The “tensile 650 MPa, yield 300 MPa” figures circulating on datasheets are typical values. The real minima come from the EN 10088-3 route: Rp0.2 ≥190, Rm 500–750, A ≥35 %.
6. Publishing a single “303 strength”. In cold-drawn bar diameter governs: at ⌀ ≤40 mm Rm is ~600–950 MPa, at ⌀ >63–100 mm it is ~500–750 MPa. A strength figure given without diameter and delivery condition is meaningless.
7. Density and melting point conflicts. Density is printed as both 7.9 and 8.03 g/cm³; melting as both a 1400–1450 °C range and a single ~1455 °C point. Do not average them; state which source you used.
8. A unit error: “density 8.03 kg/m³” — the real value is 8030 kg/m³.
9. Column-shifted tables. One widely circulated 303 PDF shows modulus of elasticity as 62 GPa, sulphur as ≤1 % and tensile as 398 MPa. All three are wrong: modulus is ~193 GPa, sulphur 0.15–0.35 %, annealed tensile 500–750 MPa.
10. Phosphorus printed as “0–0.4 %” by one technical portal. No standard allows this: ASTM ≤0.20 %, EN ≤0.045 %.
11. The claim “303 can be hardened by heat treatment”. False. It is fully austenitic; hardness comes only from cold deformation. There is no quench-and-temper recipe.
12. Confusing ISO 3506 class A1 with A2. 303 fasteners fall in A1, and the standard flags A1 as the low-corrosion-resistance free-machining class; A2 is 304-based. Buying A1 bolts for outdoor or wet service is a direct specification error.
13. Presenting 303Se as “303 with the corrosion problem solved”. False. Selenide inclusions improve surface finish and some formability; the corrosion disadvantage remains. It is also effectively unavailable in Europe.
14. Publishing an ASME P number for 303. Unverified, and it should not be published. 303 is not a WPS basis; there is no allowable stress in any code requiring a pressure boundary.
15. Defending 303 with PREN. The PREN values of 303 and 304 are numerically close (~17–21). The PREN formula cannot see inclusions; the real weakness of 303 is in the microstructure. A 303 offer resting on PREN cannot be defended technically.​‌​​‌​

​‌​​‌​