How to Read a Mold Steel Mill Test Certificate | Chemical Composition, Hardness, and Ultrasonic Testing

In mold steel procurement, I have seen far too many purchasing clerks treat the Mill Test Certificate (MTC) shipped with the steel as a routine document. It is often stamped, filed, and forgotten at goods receipt.

Yet when a mold cracks after 5,000 shots, or when a CNC roughing tool wears out unusually fast, the same MTC is usually the first document that should be checked again. A small chemistry deviation, a missing UT class, or a heat number that cannot be traced may not prove the failure alone, but it can become an important risk signal.

A mill certificate is not a formality. It is the technical map that connects chemistry, hardness, UT results, heat number, and real mold performance.

A truly usable mill certificate contains four core modules:

I have personally processed 200+ MTCs for H13, P20, and 1.2344 mold steel. In 2025 alone, I rechecked 47 annealed H13 module certificates and 18 pre-hardened P20 MTCs with HB 280-310 delivery hardness.

The deepest lesson is this: the certificate should be read as engineering evidence, not as an administrative file. It connects the steel mill's spectrometer readings, tensile testing data, ultrasonic inspection results, and heat-treatment decisions.

Most public deep-dives on mold-steel mill certificates stop at the catalogue layer of standards such as ISO 6506, EN 10204, and ASTM A681. They often list standard numbers, but they seldom explain what each number means in the workshop.

For mold steel buyers, the practical value is in reading the numbers together. A certificate value such as 0.32% C, 4.81% Cr, HRC 50, HB 480-490, or FBH 1.5 mm only becomes useful when it is connected to the steel grade, delivery condition, product size, sampling position, and mold application. ^[1]

Chemical Composition

What C, Mn, and Cr Mean

The first page of a mill certificate typically has two columns headed "Heat Analysis" and "Product Analysis". These columns usually list the main alloying elements, such as C / Mn / Cr / Mo / V / Ni.

• Heat Analysis: the chemistry measured from molten steel during steelmaking.
• Product Analysis: the chemistry rechecked from the finished bar, plate, or block.
• Allowable deviation: the permitted difference between heat analysis and product analysis, depending on the standard, element, product size, and purchase contract.

Take H13 (4Cr5MoSiV1) as an example. The common composition limits of AISI / UNS H13 are C 0.32-0.45%, Mn 0.20-0.50%, Si 0.80-1.20%, Cr 4.75-5.50%, Mo 1.10-1.75%, V 0.80-1.20%, Ni 0.30% max, Cu 0.25% max, P 0.03% max, and S 0.03% max.

Do not treat one wider product-analysis window as a universal rule. Always check the exact standard and contract requirement named on the MTC, such as ASTM A681, DIN EN ISO 4957, AISI / UNS T20813, or a customer-specific die-steel specification. ^[2]

Chromium Cr is one of the key corrosion- and temper-resistance elements in H13. When measured Cr is close to the lower limit, the mold is not automatically unqualified, but it should be reviewed more carefully for high-temperature die-casting applications.

When I select H13 mold steel, the most recent 1.2344 module with 305 mm thickness on my desk reported Cr at 4.81%. This value was still inside the common H13 range, but it was close to the lower side, so I treated it as a risk signal rather than a simple pass.

In the workshop, I tend to compare C, Mn, Cr, Mo, V, and Ni with the midpoint of the relevant specification range. For major alloying elements such as Cr, Mo, and Ni, a deviation of about 0.30 percentage point from the midpoint is enough for me to flag the heat for review.

For small elements such as C, P, and S, this 0.30 percentage-point rule should not be used directly. Their limits are much smaller, so they need separate and stricter checks.

1. If measured C in H13 is above the high side of the agreed range, review retained austenite, quench response, and dimensional-change risk.
2. If Cr is only at the lower bound, review whether the mold is used in severe hot-work or die-casting conditions.
3. If the mold is a high-value cavity insert, consider adding a purchase-contract clause such as "Cr shall be no lower than 4.85%".
4. If the part is only a low-risk support block, do not blindly demand tighter chemistry, because it may increase cost and lead time.

Further reading: H13 / P20 / 1.2344 mold steel comparison.

Trace Element Limits

The second page of the MTC usually lists trace and residual elements. These include P, S, Cu, Sn, and As.

For mold steel, small numbers can still matter. P and S may look minor, but they can affect toughness, fatigue resistance, crack initiation, and polishing quality.

The lower the better for both phosphorus P and sulfur S in premium mold steel. General H13 specifications commonly allow P and S up to 0.030% max, while premium and superior H13 grades may use lower impurity limits to improve toughness and thermal-fatigue resistance.

In published H13 technical data, premium-grade H13 examples may limit S to 0.005% max, while superior-grade H13 examples may limit P to 0.015% max and S to 0.003% max. The exact limit should always follow the named NADCA grade, mill specification, and purchase contract. ^[3]

Once measured P is high, transverse impact toughness may drop. Once S is high, MnS inclusions can become stress-concentration sources.

My own field rule is to treat P + S as a combined risk signal. When P + S together exceed 0.025%, I do not immediately reject the steel, but I mark it for additional review before quenching, especially for large H13 blocks and high-stress mold inserts.

This is not a standard rejection limit. It is only an internal screening threshold for high-risk mold applications.

Residual elements Cu, Sn, and As come mostly from scrap recycling and may remain in the steel before ESR. When Cu, Sn, and As appear unusually high compared with the buyer's normal supplier baseline, especially when their combined level approaches about 0.20%, the heat should be reviewed for residual-element and high-temperature embrittlement risk. ^[4]

When sourcing high-purity ESR H13 blocks, I once handled a second-hand 1.2343 module whose MTC showed Cu 0.12%, Sn 0.045%, and As 0.022%. The total was 0.187%, which was close enough to deserve attention for a high-temperature mold.

The customer reported cracks in the cavity after 12,000 die-casting shots. Failure analysis pointed to segregation-related grain-boundary weakness, so the certificate problem was not one single element, but the combined residual-element pattern.

For residual elements, do not only ask whether one value is over the limit. Ask whether the combined impurity level is suitable for the mold's temperature, stress, and life requirement.

If residual-element data are not on the certificate, ask the mill for an ESR-lot-specific purity report or inclusion rating report. For H13, Cr is a main alloying element, not a residual impurity, so it should not be mixed into a "residual total" calculation.

Further reading: ASTM E45 inclusion rating.

Heat Number Traceability

The heat number is the most overlooked but highest-value field on the certificate. It typically consists of a mill code, a date or batch code, and a heat sequence.

For example, a heat number such as "TATA-H13-20260315-03" may indicate the steel grade, production date, and the third heat of that day. The exact format varies by mill, so consistency is more important than format.

The heat number's real job is to bind the chemistry, mechanical properties, and UT data on the MTC to the actual steel supplied. It also links the material to steelmaking, refining, casting, rolling or forging, and heat-treatment records.

• Check whether the heat number on the MTC matches the steel body marking.
• Check whether the heat number matches the invoice, packing list, and purchase order.
• Check whether one heat number is used across unrealistic sizes or delivery states.
• Check whether the mill can provide the original certificate instead of a retyped trader copy.
• Check whether the certificate number and heat number can be verified through the mill, when verification is available.

Behind a heat number, the real production chain may include EAF melting, refining, vacuum degassing, ingot or continuous casting, rolling reduction, forging ratio, and heat-treatment batch. Hydrogen in quality steel is normally controlled at very low ppm levels, often in the single-digit ppm range depending on mill practice and product requirement.

The moment a batch of molds fails, the heat number is the key for backward tracing. Without the heat number, failure analysis becomes much weaker because the buyer cannot connect the broken mold to the original steelmaking heat.

EN 10204 splits mill certificates into several types, and the most common for mold steel are 3.1 and 3.2. A 3.1 certificate is issued by the manufacturer's authorized inspection representative and includes test results for the supplied material.

A 3.2 certificate adds a stronger witness level by involving the manufacturer's representative and the purchaser's representative or an officially designated inspection body, depending on the order requirement. ^[5]

For mold-steel orders exported to the EU, customs and buyers may treat traceability documents more strictly. Since restrictions on certain Russian-origin iron and steel products tightened from 2023 onward, many buyers now ask the MTC to show country-of-melt or equivalent origin evidence linked to the heat number.

The European Commission has stated that an MTC is an example that can be regarded as sufficient evidence for the country of origin of iron and steel inputs, while the final decision on acceptable evidence belongs to the relevant national competent authorities. ^[6]

I once had an export order of H13 round bars to Turkey held up at customs because the MTC omitted "Country of melt: India". A revised 3.1 certificate had to be issued before the cargo was released.

For any export order, the purchase contract should explicitly state "MTC per EN 10204 3.1 with country of melt" when this information is required. The standard EN 10204 3.1 certificate steel template can be applied directly.

Further reading: Heat-number reverse-lookup.

Mechanical Properties

HRC and HB Hardness Conversion

Hardness is the first mechanical property most buyers check on the MTC. The same mold steel can appear under three hardness scales: HRC, HB, and HV.

ISO 6506-1, ISO 6508-1, and ISO 6507-1 govern Brinell, Rockwell, and Vickers hardness test methods respectively. Engineering conversion between hardness scales is commonly done through ASTM E140, but the result is approximate and should be used with caution. ^{[7a, 7b, 7c, 7d]}

A few common approximate cross-references are:

One thing I see all the time in the workshop is that the MTC gives an HB value, such as H13 annealed condition HB ≤ 229, but the customer's heat-treatment work order requires "post-quench HRC 50-52". These two values cannot be directly converted as if they describe the same microstructure.

The annealed pearlite structure and the as-quenched martensite structure get hardness from different mechanisms. The correct workflow is to use the MTC's HB to confirm the annealed condition, and then verify the final HRC after quenching and tempering.

1. For annealed H13, use HB to check machinability and delivery condition.
2. For pre-hardened P20, use HB or HRC to check whether the supplied block matches the ordered hardness range.
3. For quenched and tempered H13, use HRC to verify final working hardness.
4. For nitrided surfaces, use HV under a specified test load instead of relying on HRC conversion.

A nitrided mold surface is often measured in HV 0.5 with a 500 g load. It cannot be directly converted to HRC, because a 0.3 mm nitrided case may be penetrated by a 150 kgf HRC load and the reading may include the softer substrate.

When shopping for a portable hardness tester, prefer a unit with an HRC + HV dual-mode probe. The test method should match the real surface condition.

Further reading: ASTM E140 conversion table.

Tensile Strength and Yield Strength

The second line of mechanical properties on the MTC is usually tensile strength Rm and yield strength Rp0.2, in MPa. Rm is the maximum tensile stress before fracture, while Rp0.2 is the stress where 0.2% plastic strain begins.

For H13 after 1020°C oil-quench and 580°C double-temper, typical values may be Rm 1500-1800 MPa, Rp0.2 1200-1500 MPa, elongation A 8-12%, and reduction of area Z 30-45%. These values are engineering reference ranges and still depend on section size, sample position, hardness level, and heat-treatment condition.

MTCs often list only Rm and Rp0.2. The elongation and reduction columns may be blank or only marked "meets standard".

In that case, ask the mill for the supplementary test report. I once handled an order of 1.2344 blocks with a 400 × 600 mm cross-section where the mill listed Rm 1680 MPa / Rp0.2 1380 MPa and left the elongation blank.

A recheck produced A = 9.5%, satisfying the minimum elongation requirement for that order.

There is a hidden question in the mechanical property table: where was the representative sample taken? ASTM A370 covers procedures and definitions for mechanical testing of steels, stainless steels, and related alloys, while the product specification or purchase contract should define the required sampling position. ^[8]

For round bars, the sample may be taken at an R/2 location. For blocks, the sample is often taken at T/4 or T/2 of the thickness.

• T/4: one quarter of the thickness from the surface.
• T/2: the center of the thickness.
• R/2: half-radius location for round bars.

For a 400 mm thick H13 large block, the T/2 position is 200 mm from the surface. This is often the last-to-solidify region, where shrinkage, porosity, and segregation can be worse than at T/4.

So when an MTC Rm value differs from the on-site recheck by more than 5%, confirm the sampling location first. Do not draw a conclusion before checking sample direction, thickness position, heat-treatment state, and test standard.

I once had a customer complain that during face milling on the machining center, the block's center zone showed "hardness 3 HRC low + tensile elongation 3% low". Looking back at the MTC, the sampling position was marked as T/4 while the customer's cutting zone was T/2.

The mill's sample was compliant, but the block center was more segregated. The fix is to tighten the next order to a "center sampling" clause when the cavity will be machined deep into the block.

Further reading: Sampling-location rationale.

Impact Toughness

Impact toughness is the property that customers look at least during sourcing but turn to first when something breaks. Mold steel is usually rated by Charpy V-notch (CVN) absorbed energy, in joules.

For H13-type hot-work mold steel, CVN helps evaluate resistance to sudden cracking, thermal fatigue, and crack propagation. For cold-work die steels such as D2, 1.2379, SKD11, and DC53, transverse impact toughness is often even more important.

In many tool steels, transverse CVN can be much lower than longitudinal CVN because carbide distribution, forging direction, inclusion shape, and segregation bands are direction-dependent. ISO 148-1 specifies the Charpy pendulum impact test method for determining absorbed energy in metallic materials. ^[9]

In three separate cases in our workshop, I have seen D2 and DC53 cold-work tooling fail at the second draw because the transverse CVN on the MTC was never checked. I have a D2 MTC on my desk showing longitudinal CVN 28 J and transverse 9 J.

That gap matched the customer's later failure: punch brittle fracture on the second draw operation. The failure analysis confirmed transverse cracking.

CVN on the certificate must also be read together with test temperature. A cold-work die running in a Northern Chinese winter shop at about -10°C may behave very differently from the same die tested at room temperature.

For some cold-work tool steels, low-temperature impact energy can drop sharply depending on heat treatment, carbide distribution, hardness, and sample direction.

1. For room-temperature service, check whether CVN meets the specified lower limit.
2. For cold-work dies, check transverse CVN instead of only longitudinal CVN.
3. For winter or low-temperature service, ask for a -20°C CVN recheck.
4. For thick blocks, check whether the CVN sample position represents the future cavity zone.

This rule was validated in a DC53 punch fracture case in a 2024 winter shop in Northeast China. The certificate read 25 J at 20°C, but the -20°C recheck was only 7 J, and the DBTT curve had shifted to the right.

When selecting high-toughness D2 cold-work steel, I prefer batches with clearly reported transverse impact data. In our workshop, a transverse CVN around or above 15 J is treated as a stronger candidate for high-toughness applications, depending on hardness, test temperature, and tool duty.

Further reading: DBTT of cold-work dies.

Non-Destructive Testing

UT per SEP 1921

The last page of an MTC is usually the non-destructive testing report. For mold steel, the dominant technique is ultrasonic testing (UT).

SEP 1921 is a German steel-and-iron ultrasonic testing specification that is still referenced in some mold-steel and forged-block purchase contracts, although some accreditation and standards documents list it as a withdrawn document. ASTM A388 is used for ultrasonic examination of steel forgings, and ASTM A578 is used for straight-beam ultrasonic examination of rolled steel plates for special applications.

These standards can be compared as UT-related references, but they should not be called fully equivalent because their product scope, terminology, and acceptance wording are different. The purchase contract should state the exact inspection standard used. ^{[10a, 10b, 10c]}

The reference block is made from a similar grade and heat-treatment state as the tested material. It may contain flat-bottom holes of known diameter, such as 0.5 / 1.0 / 2.0 / 3.0 mm.

The probe records the signal response as it scans over these known reflectors. The signal from the actual steel block is then compared with the reference response.

For mold-steel blocks of 200-800 mm thickness, UT is usually performed with a longitudinal-wave normal probe. Frequencies in the 1-5 MHz range are common, but the correct choice depends on thickness, grain size, attenuation, required sensitivity, and inspection procedure.

Many workshops are not familiar with the operational details of SEP 1921. First, the scan coverage must be defined clearly in the contract, especially for large cavity blocks.

For critical mold blocks, the buyer should require full-volume scanning through all accessible faces, with any restricted or inaccessible zones clearly recorded in the UT report. Second, strong back-wall attenuation should trigger review rather than be ignored.

Third, surface roughness affects coupling. A clean machined surface is usually more reliable for UT than a rough scaled surface. If the contract requires a numerical roughness value, use the specified value; in our workshop, Ra ≤ 6.3 μm is a practical target for many block rechecks, not a universal standard requirement.

I once ran a 4 MHz probe over a 305 × 610 × 2000 mm 1.2344 block as a recheck. I found that the mill's "UT passed" report had actually covered only about 70% of the volume.

A re-scan in the center zone uncovered an FBH 2.5 mm equivalent flaw. It was located exactly where the customer later experienced a cutter crash during CNC machining.

"UT passed" is not enough. A useful UT report should show standard, class, sensitivity, probe, scanning coverage, and acceptance result.

When ordering SEP 1921 UT-passed blocks, require full-volume scanning through all accessible faces when the mold application needs it.

Further reading: SEP 1921 vs ASTM A388.

Class D/d or E/e?

SEP 1921 classification combines an uppercase letter and a lowercase letter. The uppercase class mainly relates to individual indications, while the lowercase suffix is commonly used in purchase language to discuss clustered indications.

The common mold-steel agreements are Class D/d and Class E/e. In simplified purchasing language, Class D/d is usually stricter than Class E/e, but the exact allowable indication size, cluster rule, and back-wall attenuation must be checked against the agreed SEP 1921 table.

The exact numbers must always be checked against the SEP 1921 table, the product form, the section size, and the purchase contract.

A few practical notes on class selection:

1. The mill's default contract is often Class E/e, so the purchase contract must state "D/d or E/e" explicitly.
2. Class D/d is usually better for high-precision cavities, mirror-polish molds, and deep-machined blocks.
3. Class E/e may be acceptable for lower-risk die bases or non-critical support parts.
4. A stricter class can increase material cost, rejection rate, and lead time, so it should match the mold risk.

The UT sensitivity must also match the block thickness and material attenuation. It is not correct to say that a thicker block always needs a higher-frequency probe.

In practice, probe frequency, DAC correction, reference sensitivity, and scanning method should be selected by a qualified NDT procedure. For very thick blocks, TOFD or other supplemental methods may be considered when the contract or risk level requires it. ^[11]

The case I cite most often is a 1.2343 die-casting die base with a 500 × 800 mm cross-section. It was shipped under the mill's default Class E/e agreement.

A recheck under D/d revealed three clusters at FBH 1.5 mm. These positions later corresponded to early cracks appearing in the cavity after 8,000 shots.

Insist on Class D/d agreement mold steel when the cavity accuracy, machining depth, and failure consequence justify the stricter level.

Further reading: TOFD for thick blocks.

How to Identify a Forged Report

Mill-certificate forgery is not rare in the mold-steel trade, especially after the material passes through several layers of traders. The problem is not always a fully fake certificate.

More often, the risk is a mismatched certificate: a real MTC from another heat, a copied third-party stamp, a retyped trader document, or a UT class changed from E/e to D/d.

Because an MTC may be used as documentary evidence in export, origin, customer-audit, and quality-claim situations, its authenticity and heat-number consistency should be checked before the steel is accepted into a high-value mold project. ^[12]

In our internal supplier-screening work, MTC inconsistencies appeared more often when the material passed through several trading layers. The common problems were heat-number mismatch, retyped certificate pages, unverifiable third-party stamps, missing original mill verification, and UT class inconsistencies.

The common tricks include:

• Copying the MTC of a different batch and overstamping the heat number.
• Faking third-party inspection stamps.
• Substituting annealed-state data for actual mechanical-property test values.
• Reporting surface hardness while hiding center hardness.
• Rewriting a Class E/e agreement as Class D/d.
• Using a genuine mill name on a non-original PDF.

There is a three-step method for spotting a forgery.

1. Cross-check identity: compare the mill's original MTC serial number, heat number, steel body marking, bundle label, invoice, and packing list.
2. Check chemistry logic: look for contradictions between elements, such as Cr matching H13 while Mo or V is far outside the expected H13 pattern.
3. Check thickness-versus-class logic: a very thick block claiming an extremely strict UT class needs a detailed and credible UT report.

Major mills such as TATA, Voestalpine, Bohler, and Daido may provide online or official verification channels, depending on product route and sales channel. When available, enter the MTC number and heat number to compare against the genuine certificate.

A more advanced forgery-detection move is to back-derive from element ratios and heat treatment. For H13, the Cr/Mo ratio is commonly about 2.7-5.0 based on standard composition ranges, and V is usually around 0.80-1.20%.

If the MTC shows a very unusual Cr/Mo ratio or a missing V value, the chemistry may have been edited or the steel may not be the claimed grade. ^[13]

In three separate cases, I have dealt with forged or mismatched MTCs in our factory. The patterns were similar each time.

In early 2025, I helped a customer spot a forged "Daido DC53 MTC". The forger had taken a genuine certificate of the same grade but a different heat, then relabeled heat A as heat B.

The element data looked reasonable, but the CVN transverse column was identical to the longitudinal column. A real certificate usually shows a difference between longitudinal and transverse values, especially for cold-work die steel.

The UT grade was also listed as AA even though the block thickness read 600 mm. This was not impossible in theory, but without a detailed UT method, probe, sensitivity, calibration block, scanning coverage, and acceptance record, it was a strong technical red flag.

A forged MTC is often exposed not by one typo, but by contradictions between chemistry, heat number, hardness, impact direction, UT class, and block thickness.

When sourcing high-purity traceable-certificate steel, require the mill to include its internal traceability number on the MTC for online or official verification.

Further reading: Five common MTC forgery tells.