Properties of Refractory Materials: What Each Value Means and How It Is Measured

Refractory materials are judged by 12 measured properties, from apparent porosity and cold crushing strength to thermal conductivity and slag resistance, and each value means something only together with its test method, edition and specimen history. Which number on a data sheet is a requirement, and which is a typical result of one product?

Refractory properties follow from two causes set in production: the raw materials, which fix chemistry and grain, and the bond that holds the grains. The same tests apply to bricks, monolithics and insulation, the three forms of refractory materials.

ASTM Committee C08 and ISO/TC 33 publish separate methods for the same 12 properties, and CEN adopted EN ISO 10059-1:2025, EN ISO 5014:2025 and EN ISO 8894-2:2025 on 2025-12-09. A shaped product counts as dense below 45 % true porosity (ISO 5014:2025). Between two laboratories, cold crushing strength can differ by ±45 % and castable abrasion by ±78 % (precision for earlier editions, Headrick and Chalfin, 2021). Cutting specimens falls under the OSHA respirable crystalline silica limit of 50 µg/m³ (29 CFR 1910.1053, effective 2016-06-23).

Every value names its standard and edition, the product data sheet or the dated source it comes from; see our methodology and fact-checking process.

What Are the Properties of Refractory Materials?#

The properties of refractory materials are the measured characteristics that show how a lining behaves at room and service temperature, grouped as physical, mechanical, thermo-mechanical, thermal and chemical properties. Physical properties, such as porosity, density and abrasion loss, are measured at room temperature.

What makes a material a refractory in the first place? ISO 836:2001 (term 107) defines a refractory as a "non-metallic material or product (but not excluding those containing a proportion of metal) whose chemical and physical properties allow it to be used in a high temperature environment", with no temperature threshold. ASTM C71-12(2024) sets the US boundary at 538 °C (1,000 °F).

Dense and Insulating Products Are Tested Differently#

A shaped product is dense below 45 % true porosity by volume and insulating at 45 % and above (ISO 5014:2025), and the split decides the porosity method: ISO 5017:2025 for dense, ISO 5016:1997 for insulating products. True porosity, not apparent porosity, draws this line.

Insulating products also have their own classification. ISO 2245:2006 groups shaped insulating products by the test temperature at which their permanent change in dimensions stays at or below 2 %, and ceramic fiber products fall outside it. Classes, products and their uses are covered under refractory insulation.

Specimen History: Dried, Fired and Coked Values#

One property gives different values after drying at 110 °C (230 °F), after firing at 816 °C (1,500 °F) and after coking, so a data sheet names the state of the specimen.

Carbon-containing brick adds a third state: ASTM C831-18(2023) measures residual carbon, apparent porosity and bulk density after coking. Castables fired to 816 °C (1,500 °F) carry their own porosity error, explained in the cement section below.

Types of Refractory Properties: 12 Properties and Their Test Methods#

There are 12 core properties of refractory materials, each with its own ASTM and ISO or EN test method, and Table 1 lists them in five groups from porosity to hydration resistance.

Table 1 compares the 12 properties by group, reported unit, ASTM method and ISO or EN method.

Group order follows physical, mechanical, thermo-mechanical, thermal and chemical properties in the text. Side by side means the same property, not equivalent results.

In each row of Table 1, the unit says what a data sheet prints and the two method columns say which test produced the number.

Physical Properties: Porosity, Bulk Density and Abrasion#

Porosity and bulk density are measured by boiling water under ASTM C20-00(2022) or by vacuum pressure under ASTM C830-00(2023), and abrasion is measured as volume loss in cm³. C20 is not for refractories that water attacks, and C830 uses mineral spirits for hydratable materials. ASTM C704/C704M-15(2022)e1 blasts silicon carbide grit at room temperature and reports mass and volume loss in SI units only.

Mechanical Properties: Crushing Strength and Modulus of Rupture#

Cold crushing strength and modulus of rupture share one ASTM standard, C133-24 (CCS in sections 4 to 8, MOR in 9 to 13), and split into ISO 10059-1:2025 and ISO 5014:2025. C133-24 lists specimen size, face condition, orientation and loading rate as factors that change the result.

Hot strength has its own methods. ASTM C583-15(2026) measures hot modulus of rupture in an oxidizing atmosphere at a constant force rate, and ISO 22685:2021 measures compressive strength at elevated temperature.

Thermo-Mechanical Properties: PCE, Load, Creep and Linear Change#

Pyrometric cone equivalent (PCE) is a cone number for fireclay and high-alumina materials only (ASTM C24-09(2022)), while refractoriness under load rises in temperature under constant load up to 1700 °C (3,092 °F) under ISO 1893:2007. Creep under ISO 3187:1989 loads dense shapes at 0.2 MPa (29 psi) and insulating shapes at 0.05 MPa (7.3 psi) for 25 h, extendable to 100 h.

Permanent linear change records the dimensional change that remains after reheating, measured by ASTM C113-14(2025) or ISO 2478:2025; the ISO method excludes carbon-containing products. Thermal shock is tested by ASTM C1171-16(2022)e1 and ISO 21736:2020, whose three methods quench in water, compressed air and natural air.

Thermal Properties of Refractory Materials: Conductivity and Heat Flow#

Thermal conductivity is the main thermal property of refractory materials, reported in W/(m·K) (BTU·in/(h·ft²·°F)) at a stated mean temperature. ASTM C201-24 uses a water calorimeter, while ASTM C1113/C1113M-09(2024), ISO 8894-1:2010 and ISO 8894-2:2007 use a hot wire inside the specimen. Reversible thermal expansion, the second thermal value on many data sheets, is measured under ISO 16835:2014.

Chemical Properties: Slag, Oxidation and Hydration Resistance#

Slag and carbon monoxide tests rank materials against each other under ASTM C874-20(2026) and C288-20(2025), while oxidation and hydration tests expose specimens to steam or moist air. ASTM C863-00(2022) tests silicon carbide refractories in steam, which accelerates oxidation. Hydration of basic brick, magnesite or periclase grain and dolomite grain falls under ASTM C456-18(2023), C544-18(2023) and C492-92(2023).

Standard Requirements, Typical Values and Comparative Results#

A refractory property value is one of three kinds: a standard requirement that a class must meet, a typical value that one product achieved in its maker's tests, or a comparative result that ranks samples tested side by side. A comparative result has meaning only next to its reference.

Which Values Are Standard Requirements?#

Classification standards turn a property into a requirement: ASTM C401-12(2022) calls a castable dense above a dried bulk density of 1.68 g/cm³ (105 lb/ft³) and sets classes A to G by a permanent linear change of at most 1.5 %. The class test runs for 5 h at 1095 to 1760 °C (°C as printed in the standard; 2,000 to 3,200 °F).

ISO 2245:2006 does the same for shaped insulating products: the group number equals the test temperature divided by 10 at which the permanent change in dimensions stays at or below 2 %. ASTM C892-19(2024), which classes high-temperature fiber blanket for use up to 1649 °C (3,000 °F), belongs to Committee C16 on Thermal Insulation, not C08.

Comparative Tests That Rank Without an Absolute Number#

Slag, carbon monoxide and refractoriness-under-load tests rank materials rather than predict a service value: ASTM C874-20(2026) runs a reference material in every test and gives no numeric result, and ISO 1893:2007 results are comparative.

ASTM C288-20(2025) is an accelerated test with a higher carbon monoxide concentration than service, and the standard describes it as suitable for acceptance criteria in a specification. The three ISO 21736:2020 thermal shock methods are each written for a different product type, and their results are not comparable with each other.

Which Properties Matter for Which Service Condition?#

The properties that matter are set by the lining position: load, creep and slag resistance at the hot face, conductivity and shrinkage in backup layers, thermal shock, abrasion and atmosphere tests where the service cycles, erodes or reduces.

Table 3 compares service conditions with the properties to specify and example methods for each.

A specification should name the property, the method and its edition, and the specimen state for every row it uses.

Hot-Face Service: Load, Creep and Slag#

ASTM C832-21(2026) measures thermal expansion under load and 20 to 50 h creep, and the standard names blast furnace stoves and glass furnace checkers as its applications. Carbon-bonded materials are tested under C832 only with atmosphere control.

ASTM C16-03(2025) is the hot-load test used in written US specifications, and it applies a different principle from ISO 1893:2007: C16 records deformation under a set load, temperature and time, while ISO 1893 records the temperatures at set deformations as the temperature rises.

Backup and Insulating Layers: Conductivity and Shrinkage#

Backup layers are specified by conductivity at a mean temperature, measured by ASTM C201-24 for materials up to 28.8 W/(m·K) (200 BTU·in/(h·ft²·°F)) or by hot wire from room temperature to 1500 °C (2,732 °F).

ISO 2245:2006 classes insulating products by the shrinkage limit at their test temperature, and the conductivity values then feed the thickness calculation of each layer in a refractory lining.

Cycling, Abrasion and Furnace Atmosphere#

Cyclic service calls for thermal shock data: ASTM C1171-16(2022)e1 cycles prisms at a recommended 1200 °C (2,190 °F as printed) and reports MOR loss and sonic velocity loss.

European practice uses EN 993-11:2007, which quenches in air by method A or B, while ISO 21736:2020 method 1 quenches in water and applies to dense aluminosilicate products only. Erosion service in refineries makes API 936 the main US user of the C704 abrasion test. Reducing atmospheres call for the C288 carbon monoxide test, and silicon carbide in steam-bearing gas calls for C863.

How Binders and Additives Change Measured Properties#

Binders and additives change both the property and its measurement: cement content shifts porosity readings in castables, and carbon with antioxidants changes how hot strength responds to the test atmosphere.

Cement Content, Rehydration and Porosity Readings#

ASTM C20-00(2022) records a porosity error on castables heated to 816 °C (1,500 °F) and then saturated with water, and the error rises with cement level. The standard names the cause as rehydration of the cement during water saturation.

ASTM C830-00(2023) avoids water by using vacuum pressure with mineral spirits, and it is the preferred method for castables and other hydratable materials. ASTM C401-12(2022) separates regular, low-cement, ultra-low-cement and no-cement castables by CaO content, so the cement level that shifts a C20 reading is also the property that names the castable type.

Carbon, Antioxidants and Hot Strength Results#

The ruggedness study in ASTM C1099-07(2025) shows that argon lowered the hot modulus of rupture of resin-bonded magnesia-carbon brick with about 17 % carbon after coking. For aluminum-containing brick, a higher test temperature and a higher loading rate raised the HMOR, evaluated at 1399 °C (2,550 °F).

Residual carbon, measured by combustion under ASTM C831-18(2023) on coked brick, does not separate the carbon held in silicon carbide from free carbon. The roles of antioxidants and binders are covered under refractory additives and binders.

ASTM, ISO and EN Methods for Refractory Properties Side by Side#

ASTM Committee C08 and ISO/TC 33 publish separate methods for the same 12 properties, and CEN adopts most ISO methods as EN ISO parts; side by side means the same property, never interchangeable results.

Table 2 compares the ASTM, ISO and EN methods for nine properties, with the difference that matters for each.

ASTM C08 holds 71 standards in Annual Book Volume 15.01, organized in subcommittees for strength (C08.01), thermal properties (C08.02), physical properties (C08.03), chemical behaviors (C08.04) and monolithics (C08.09). On the European side, CEN/TC 187, with its secretariat at BSI, maintains the EN 993 series and adopts ISO/TC 33 methods as EN ISO parts.

Methods Replaced in 2025 and Withdrawn Methods#

EN ISO 10059-1:2025, EN ISO 5014:2025 and EN ISO 8894-2:2025 replaced EN 993-5:2018, EN 993-6:2018 and EN 993-15:2005 on 2025-12-09. Earlier replacements moved EN 993-8 to EN ISO 1893:2008, EN 993-14 to EN ISO 8894-1:2010, EN 993-20 to EN ISO 16282:2008, and EN 993-17 and -18 to EN ISO 8840:2024. On the US side, C133-24 is current while the API 936 exam effectivity sheet still lists C133-97(2021).

Two ASTM methods are withdrawn: C768 drip slag in 2004 and C1100 ribbon thermal shock in 2005. ISO 18894 is a coke reactivity and strength method, not a refractory test. Every designation with its edition and status sits in the register of refractory standards.

How to Read a Refractory Data Sheet#

A refractory data sheet is read in five checks: the test method and edition, the specimen state, what each temperature means, the chemistry basis and the precision of the method. A value that fails one of the five checks cannot be compared with a value from another sheet.

ASTM C133-24 states that cold strength indicates suitability for refractory construction and is not a measure of performance at elevated temperatures, so a high CCS on a data sheet says nothing about hot behavior.

Temperatures on a Data Sheet: PCE, Test Temperature and Service Limit#

A PCE value is a cone number for heat-work, not a service temperature or melting point (ASTM C24-09(2022)). EN 993-12:1997, used with the EN 993-13:1995 reference cones, gives a useful range of 1500 to 1800 °C (2,732 to 3,272 °F).

Insulating products carry a group number that encodes a test temperature. Under ASTM C155-97(2022), the group number of insulating firebrick times 100 approximates the reheat test temperature in °F, not a maximum service temperature. Under ISO 2245:2006, the group equals the test temperature divided by 10, from group 80 to group 180 (800 to 1800 °C, 1,472 to 3,272 °F).

Precision: How Far Two Laboratories Can Differ#

Between-laboratory precision for earlier editions, as printed by ASTM C08 members Headrick and Chalfin (Am. Ceram. Soc. Bull., 2021), is ±45 % for the CCS of dense brick, ±40 % for MOR and ±36 % for the HMOR of alumina brick.

The spread is wider for abrasion and thermal shock: ±47 % for C704 on high-alumina brick, ±78 % on conventional castable, and ±65 % for the C1171 MOR loss. Specimen preparation adds its own shift, since replacing cardboard packing with the specified fiber board can raise CCS by about 50 %. The same authors call C704 not yet reliable as a specification test.

Units and Chemistry Basis#

Data sheets convert with fixed factors: 1 MPa = 145.04 psi, 1 g/cm³ = 62.43 lb/ft³ and 1 W/(m·K) = 6.933 BTU·in/(h·ft²·°F). The per-foot unit BTU/(h·ft·°F) uses 0.578 and never mixes into a column of BTU·in values. ASTM C704 reports in SI units only, and chemistry is printed in mass % on a calcined or as-received basis, which the sheet states.

Specifying and Sourcing Refractories by Property#

A refractory specification names each required property with its test method, edition, specimen state and limit, and the quote then depends on how many specimens, which atmosphere and how long each test runs.

Testing is priced by quote. The quote factors are the specimen count (test laboratory practice: five for C133, two for C16, three bricks for C113), specimen preparation, the atmosphere (argon for C1099, carbon monoxide for C288), the test duration (50 h for C832, 25 to 100 h for ISO 3187) and the report format.

What a Property Specification in an RFQ Contains#

A property specification in an RFQ contains six items, and a buyer should state each one before suppliers quote.

• Product class and the standard that classes it, with edition.
• Required properties with test method and edition.
• Specimen state for each value: dried, fired or coked.
• Service temperature and lining zone.
• Quantity in metric tonnes, with US short tons accepted.
• Documents requested: data sheet and test report.

Installation Quality Control and Site-Made Test Pieces#

Monolithic linings are tested on site-made pieces under API Std 936, 5th edition (March 2024); brick systems fall under API Std 975, 2nd edition (April 2025), and fiber linings under API Std 976, 2nd edition (October 2024). Site sampling, test pieces and acceptance belong to refractory inspection and quality control.

Site test pieces follow ASTM practices for their preparation: castable specimens are cast to C862-16(2020), fired to C865-22, and plastic refractories are prepared to C1054-24.

Silica Dust When Cutting Test Specimens#

Cutting and grinding test specimens releases respirable crystalline silica, which OSHA limits to 50 µg/m³ as an 8-hour TWA with an action level of 25 µg/m³ (29 CFR 1910.1053, construction 29 CFR 1926.1153, effective 2016-06-23). The EU binding limit is 0.1 mg/m³, respirable fraction (Directive (EU) 2017/2398, Annex III, transposition by 2020-01-17). For quartz (crystalline silica), the classification reads "No Annex VI entry; product-specific, see supplier SDS (ECHA C&L: no majority; not classified 27.5 %, H372 25.2 %, H373 18.8 %, H350 7.7 %)".

The crystalline silica phases and their exposure limits are covered on one page:

• crystalline silica in refractories

What Can Laboratory Property Values Not Predict About Service?#

Laboratory property values do not predict service life: cold strength says nothing about hot performance (ASTM C133-24), RUL and slag results are comparative, and a lining fails through the combination of loads that no single test applies. A lining sees heat, load, slag and cycling at once, and each test isolates one of them, which is why spalling, cracking and corrosion are read as refractory failure modes.

A fired CCS and a dried CCS of the same castable, or a C16 and an ISO 1893 result, describe different tests and are not compared.

Are Property Charts and PDF Tables Comparable?#

Property charts and PDF tables are comparable only where each value names the same method, edition and specimen state, and a chart that mixes methods is not. Hot-wire conductivity can read 10 to 30 % higher than a calorimeter value for the same material, and ASTM C38 panel spalling values are historical, since C38 is no longer on the active list.

Firebrick conductivity has no single value: it depends on the mean temperature and on the method that measured it. The values by class and mean temperature are collected under thermal conductivity of firebrick.