Mill certificates for sheet metal: what they prove

Three coils arrive on the same truck, each with a certificate to the same grade, and one splits on the first bend. Nothing on the paperwork is false. The certificate describes a heat of steel, the split happened in a coil that sat at the edge of the heat's mechanical range, and the bend radius on the drawing was set from the grade's typical properties rather than its guaranteed minimum.

Mill certificates are the most requested and least read document in a sheet metal purchase. They are worth having, and they are routinely asked to prove something they cannot: that the formed, welded and coated part in the crate meets the drawing.

The standards and compliance hub covers the neighbouring documents, including first article inspection, which is what actually evidences the part.

What mill certificates record, and what they describe

A certificate records the steel as the mill delivered it. The fields that matter are the heat number and the cast number, the chemical composition, the mechanical properties from test pieces taken to a defined sampling plan, the dimensions and tolerances, the delivery condition, the product standard and grade, and the signature of the party that validated the document.

IS 513 covers cold-reduced carbon steel sheet and strip, split into parts by forming duty rather than by strength, and IS 2062 covers hot-rolled structural steel with grades from E250 to E550. ASTM A1008/A1008M does the same job for cold-rolled sheet in the American system. Each standard defines the grade, the tolerances and the tests, and a certificate is only meaningful against the standard it names.

Two things follow. First, the certificate belongs to a heat, and a heat can be 200 t of steel spread across dozens of coils. Second, the mechanical results come from test pieces cut to a sampling plan, so they describe the lot the mill tested, within the scatter the standard allows.

The four document types and when each one is enough

ISO 10474 defines the inspection documents supplied with steel products, and it replaced the older EN 10204 numbering with the same type codes. The distinction is not bureaucracy: it decides whether the numbers on the page came from your coil.

The cost of getting this wrong runs in both directions. A 3.2 requirement on a 500-part bracket order adds an inspection visit that can exceed the value of the material. A 2.2 document on a lifting frame leaves the buyer unable to show that the steel in the frame was the steel tested.

Ladle analysis, product analysis and the residuals that matter

The chemistry on a certificate comes in two flavours. Ladle analysis is taken from the molten steel during casting and describes the heat. Product analysis is taken from the finished product and shows what survived rolling, including the pickup and segregation that ladle sampling cannot see.

Product analysis tolerances are wider than the ladle limits in the product standard, and that is why a certificate can show carbon at 0.08% in the ladle and 0.09% in the product while both remain conforming. For forming, the elements to read first are carbon, manganese, sulphur and phosphorus. For welding, add the carbon equivalent, because a CEV above roughly 0.40 raises the preheat requirement and the risk of cracking in a thick fillet.

Phosphorus is the one buyers skip. It raises strength and drops ductility, and a coil at the top of the phosphorus range bends differently from one at the bottom, with the same grade on both certificates.

The four numbers that decide whether the coil forms

Grade names hide the range that matters. Read yield strength, tensile strength, elongation and the bend test result, and read the direction the test pieces were taken in.

The bend itself is a strain calculation. For a 90° bend, the outer fibre strain is approximately t divided by (2R + t), where t is thickness and R is the inside radius. At R equal to t, that is 3 divided by 9, or 33%. At R equal to twice t it falls to 20%.

Compare that with elongation. A cold-reduced drawing grade with an A80 of 26% has very little margin against a 33% bend strain once you add the anisotropy of a real part and the edge condition left by laser cutting. The same material at R = 2t has 6 points of margin. This is why minimum bend radius is a material property rather than a shop preference, and why the certificate's elongation is the number to check before arguing about the radius.

Yield strength drives a second number. Blanking force is roughly cut length × thickness × shear strength, and shear strength for mild steel runs at about 0.75 of tensile strength. A 1.5 mm CRCA blank 300 mm square has a 1,200 mm cut length, so at 350 MPa shear it needs about 63 t, or 76 t with a 20% margin. A coil at the top of the grade's strength range moves that figure, and a press chosen from the nominal grade can end up at its limit.

Direction matters almost as much. Test pieces are normally taken longitudinal to the rolling direction, and transverse ductility is lower. A part that bends across the coil, which is the common layout on a 4,450 mm brake, is asking the weaker direction to do the work. Ask which orientation the certificate reports.

A weighing check that takes two minutes

Thickness tolerance is where the certificate and the crate diverge most often, and it is easy to test. Cut or take a 300 mm × 300 mm blank and weigh it. At 1.5 mm and 7,850 kg/m³ the mass is 1.06 kg, and the rule of thumb is 7.85 kg per square metre for every millimetre of thickness.

A blank that weighs 1.12 kg is 5.7% heavy, which corresponds to a thickness near 1.585 mm. That is inside the normal tolerance of several cold-reduced standards, so the certificate is correct and the coil is still not what the process assumed. The consequences are concrete: 5.7% more steel bought, 5.7% more blanking force, and a stack of 100 blanks that stands 158.5 mm high instead of 150 mm, which is enough to upset a destacker or a magazine set for nominal.

Run the check on the first blank of every coil rather than on the certificate. It costs a scale reading and catches the substitution that a paperwork review never will.

What the certificate cannot tell you about the part

Four things sit outside the document, and each one has caused a rejected lot.

Work hardening. Every forming operation raises yield strength and lowers ductility in the worked zone. A 90° bend at R = t leaves the outer fibre at 33% strain, and a second bend in the same area compounds it. The certificate describes the annealed condition.

Welding. A fillet weld creates a heat affected zone that is harder and less ductile than the parent metal, and a spot weld creates a nugget with its own properties. Neither appears on a mill certificate, and neither is covered by the material standard.

Coating. Pretreatment and powder add a surface that changes corrosion behaviour, appearance and, at the tolerances of a tight assembly, fit. Dry film thickness and adhesion are the evidence, not the steel certificate.

The finished dimensions. Thinning at a draw, springback after a bend and distortion after welding all move the part away from the nominal the certificate's thickness implies.

What replaces a certificate once the part exists

The honest answer is a chain rather than a document. The certificate starts it, and three other records finish it.

• Traceability from heat to part. The heat number recorded against the coil, the coil against the sheet, the sheet against the nest, and the nest against the lot. Without that link, a certificate proves that some steel of the right grade exists somewhere in the building.
• First article inspection. Every drawing characteristic measured on parts from production tooling, with material and finish documents attached. Ganesh offers first article inspection and PPAP where a programme requires them.
• Material verification testing where the risk justifies it. For stainless and alloy work, positive material identification confirms chromium, nickel and molybdenum in seconds and catches a mixed coil. Handheld XRF cannot read carbon, so it will separate 304 from 316 and will not separate 304 from 304L.
• Lot inspection records. Hardness or a tensile coupon from the same heat, coating thickness, weld inspection results, and the dimensional record for the lot.

What Ganesh runs

Ganesh forms 0.8 mm to 40 mm material on presses from 20 T to 300 T, a 250 T punching line, a 12 kW fibre laser with a 40 ft × 16 ft bed and a 200 T, 4,450 mm 5-axis brake rated to 10 mm. Tools, fixtures and checking gauges are built in the in-house tool room, which is where a material substitution usually gets caught: a die that was cut for a 1.5 mm coil will not close correctly on 1.585 mm sheet, and the first-off check finds it before the lot does. The stamping capability covers the press range and the tool room in more detail.

The works run under ISO 9001:2015 and ISO 14001:2015, and supply pressings and fabrications to Mahindra, Mahindra Powerol, Indian Railways and others where material records are part of the acceptance.

The limit is worth stating plainly. We pass on the mill's document and keep the heat-to-lot link. We do not issue a mill certificate, and where a drawing demands spectrochemical analysis of every heat, or a 3.2 endorsement, that is a third-party inspection scope that has to be agreed before the order is placed, not after the parts are made.

What to put on the drawing and in the purchase order

• The grade and the standard, written as a pair: E250B to IS 2062, or D to IS 513.
• The certificate type, 2.2 or 3.1, and whether 3.2 is required with a named inspector.
• The impact test temperature, if the service temperature is below ambient.
• The carbon equivalent limit if the part is welded and thick.
• The test piece orientation, if the part bends across the coil.
• The traceability level: heat, coil or lot, and how the marking survives cutting and coating.
• Any positive material identification requirement, and the elements to be checked.
• The retention period for the records.

The bending tolerance and inspection article covers the dimensional side of the same acceptance conversation, and the welding traceability guide covers what happens to the record once the certificate's steel becomes an assembly.

When the certificate requirement is the wrong one

A 3.2 requirement on a part that is not safety-related buys an inspection visit and delays a programme. A 2.1 declaration on a structural weldment buys nothing at all, because it carries no results.

There is also a category where no mill certificate helps: bought-in components. Fasteners, gas springs, hinges and casters arrive with their own documents and their own standards, and a purchase order that asks for a mill certificate on a bought-in hinge will get a confused reply. List those items separately and name the document each one needs.

Send the drawing with the grade, the standard and the certificate type on a quote request, and the engineering team will confirm what can be issued with the lot before the material is ordered.