Testing standard
Dentistry — Ceramic materials
Written and technically reviewed by Dak System Inc. engineeringLast reviewed
ISO 6872 is the international specification for dental ceramic materials. It divides ceramics into five classes by intended clinical use and sets each class a minimum flexural strength, from 50 MPa for a single-unit anterior veneer to 800 MPa for prostheses of four or more units, together with chemical solubility limits. Strength may be measured by three-point bending, four-point bending or a biaxial piston-on-three-ball test, and the standard also fixes limits for radioactivity, thermal expansion, glass transition temperature and sintering shrinkage.
From the test method to your testing system
Explore the DAK machines already listed for ISO 6872, then review the grips, measurement and setup requirements below.
Universal Testing MachineSeries 7200Explore the machine →
Universal Testing MachineSeries 9000Explore the machine →01Understand the method
ISO 6872 is a product specification with test methods bound into it, not a pure test method. It sets what a dental ceramic must achieve — flexural strength, chemical solubility, thermal expansion, glass transition temperature and radioactivity — and supplies the procedures for measuring them. Flexural strength decides where a ceramic may be used, and there are three ways to measure it: three-point bending, four-point bending, and a biaxial piston-on-three-ball test on a disc. Materials are designated Type I, supplied as powders, pastes or aerosols, or Type II, every other form.
The class table is what makes this standard unusual. Ceramics are divided into five classes by intended clinical use, each with a minimum flexural strength and a solubility limit. A ceramic is not simply strong or weak; it is qualified for single-unit anterior work, or for three-unit prostheses involving a molar, or for four units and more, and the minimum rises from 50 to 800 MPa.
Two details are easy to miss. The strength requirement is a minimum value for the mean, so it belongs to the set of specimens rather than to every one. And the solubility limit splits by sub-class: a monolithic ceramic is held under 100 µg/cm², while a covered substructure in classes 2, 3 and 4 is allowed under 2000, because it is not exposed in service.
Beyond strength, the specification caps activity concentration at 1.0 Bq/g of uranium-238, requires the thermal expansion coefficient within 0.5 × 10⁻⁶ K⁻¹ of the manufacturer's stated value and the glass transition temperature within 20 °C of it, and fixes the sintering shrinkage factor to ±0.002.
02Prepare the specimen and test settings
Dental bend bars and discs are around a millimetre thick, which puts edge preparation among the leading sources of scatter — a point the committee itself acted on.
The crosshead rate is in clause 7.3 and clause 7.3 is paywalled. No figure is published here, and the reason is worth stating.
03Build the test setup on a DAK machine
ISO 6872 does not list ISO 7500-1 among its normative references, so it states no force accuracy class and none should be assumed.
What the test demands instead is resolution at the bottom of the range. A disc about 1.2 mm thick fails at loads in the tens to low hundreds of newtons, so the useful instrument is a low-capacity, high-resolution load cell rather than a large frame run near its floor.
Crosshead speed likewise sits in clause 7.3. Speeds of 0.5 and 1 mm/min both appear in the literature for the biaxial test and cannot both be the standard's figure, so no rate is published here.
Chemical solubility uses 100 ml of analytical grade 4 percent acetic acid in grade 3 water to ISO 3696 — the one numeric condition Amendment 1:2018 restated.
Three-point flexure on a bend bar is one of the three strength routes ISO 6872 allows. The span sits in clause 7.3 and is not reproduced here, so the fixture must be set from your own copy rather than from a default.
Specifications
Four-point loading puts a length of the bar under constant moment rather than concentrating it under one nose, which samples more of the surface and usually returns a lower, more conservative strength than three-point on the same material.
SpecificationsDak verifies against whichever standard the method names, and where a class applies our frames sit a class tighter than it asks.
| The method asks for | Dak supplies | |
|---|---|---|
| Capacity | Low. A dental ceramic disc about 1.2 mm thick fails at loads in the tens to low hundreds of newtons, so resolution matters far more than capacity. The standard specifies no capacity. | Load cells from 1 kg to 60 ton on the Series 7200, and 0.5 to 100 kN on the Series 9000 |
| Force accuracy | none stated — ISO 6872 does not carry ISO 7500-1 as a normative reference and no accuracy class should be assumed for it | ISO 7500-1 Class 0.5 — the method sets no class of its own |
| Gripping | Three-point and four-point bend fixtures for bars, and a piston-on-three-ball rig for discs. The geometry of each sits in clause 7.3, which is not published openly, and no dimension is reproduced. | Our the four-point bend fixture or the three-point bend fixture, built to the specimen |
| Environment | Chemical solubility uses 100 ml of analytical grade 4 percent acetic acid in water of grade 3 to ISO 3696, in Pyrex glass or plastic bottles. The bath temperature and duration were not confirmed and no figure is given. | 3009 series chambers, −150 °C to +400 °C — temperature only |
04Run the test
The flexural strength requirement is written as a minimum value for the mean. A set whose mean clears the class minimum conforms even if individual specimens fall below it — and conversely, a single strong specimen proves nothing.
05Calculate, report and interpret
The uniaxial relationships are the ordinary ones. The biaxial expression is deliberately NOT reproduced — see the caution on the page.
σ = 3PL / (2bd²)
Thickness enters squared, so thickness measurement dominates the uncertainty on a 1 mm bar.
Take the expression from the standard, not from a paper
Two different constants for the piston-on-three-ball expression are in print in peer-reviewed dental journals, about 19 % apart, along with differing groupings of the terms. Clause 7.3 is paywalled and neither constant is reproduced here, because publishing the wrong one would be worse than publishing none.
fitted to the strength distribution
Weibull statistics sit in an INFORMATIVE annex of the 2015 edition, so they were guidance rather than a requirement there. Whether the 2024 edition makes them normative was not confirmed.
The biaxial formula is the trap worth naming. Two different constants for the piston-on-three-ball expression are in print in peer-reviewed dental journals, differing by about nineteen percent, along with differing groupings of the geometric terms. A laboratory that lifts the formula from a paper rather than the standard can be a fifth out on every strength it reports, and nothing in its own data will reveal it.
Specimen edges are the second recurring problem, and the committee's own chamfer and grinding revisions are the evidence for it.
The third is fracture toughness, which in the 2015 edition sits in an informative annex rather than the class table — and that annex records that the notched-beam method cannot be applied to 3Y-TZP, because the notch cannot be made sharp enough with a razor blade.
06Compare methods and find answers
Table 1, read from ISO 6872:2015 and independently corroborated by US FDA guidance on dental ceramics. The 2024 Table 1 could not be seen; its foreword lists only the addition of Annex C, which suggests no change, but that is not confirmation.
| Class | Intended clinical use | Flexural strength, min for the mean (MPa) | Chemical solubility (µg/cm²) |
|---|---|---|---|
| 1 a | Monolithic ceramic for single-unit anterior prostheses, veneers, inlays or onlays, adhesively cemented | 50 | <100 |
| 1 b | Ceramic for coverage of a metal framework or a ceramic substructure | 50 | <100 |
| 2 a | Monolithic ceramic for single-unit anterior or posterior prostheses, adhesively cemented | 100 | <100 |
| 2 b | Partially or fully covered substructure for single-unit anterior or posterior prostheses, adhesively cemented | 100 | <2000 |
| 3 a | Monolithic ceramic for single units and three-unit prostheses not involving a molar | 300 | <100 |
| 3 b | Partially or fully covered substructure for the same | 300 | <2000 |
| 4 a | Monolithic ceramic for three-unit prostheses involving molar restoration | 500 | <100 |
| 4 b | Partially or fully covered substructure for the same | 500 | <2000 |
| 5 | Monolithic ceramic or fully covered substructure for prostheses of four or more units | 800 | <100 |
The solubility limit is twenty times looser for a covered substructure than for a monolithic ceramic in classes 2, 3 and 4, because a substructure is not exposed in service. Reading the 2000 figure across to a monolithic material would pass a ceramic that should fail. Fracture toughness criteria circulate attributed to ISO 6872, but in the 2015 edition toughness sits in an informative annex and not in this table — check the current text before treating any toughness figure as a pass criterion.
| Property | Requirement |
|---|---|
| Radioactivity | Not more than 1.0 Bq/g of uranium-238 |
| Coefficient of thermal expansion | Within 0.5 × 10⁻⁶ K⁻¹ of the manufacturer's stated value |
| Glass transition temperature | Within 20 °C of the manufacturer's stated value |
| Sintering shrinkage factor | Absolute accuracy ±0.002 |
| Chemical solubility reagent | 100 ml of 4 % (V/V) analytical grade acetic acid in grade 3 water to ISO 3696 |
The thermal expansion and glass transition requirements are tolerances against what the manufacturer declares, not absolute values. A product with an unusual expansion coefficient conforms provided it matches its own declaration.
| ISO 6872 | ASTM C1161 | ASTM C1499 | |
|---|---|---|---|
| What it is | Product specification with methods | Test method | Test method |
| Property minima | Yes, by clinical class | None | None |
| Geometry | Bend bars and discs, about 1 mm thick | Bars typically 3 × 4 × 45–50 mm | Discs or plates |
| Biaxial loading | Piston on three balls | — | Concentric ring on ring |
| Covers solubility, expansion, Tg | Yes | No | No |
The biaxial pair is the one to be careful with. Piston-on-three-ball and ring-on-ring produce different stress fields and have different closed-form solutions, so an ISO 6872 result and a C1499 result are not interchangeable even on the same material. C1499 also explicitly accommodates as-fired specimens with limited warpage, which three-ball supports handle differently.
EN ISO 6872:2024 is the identical European adoption. ASTM C1161 measures flexural strength of advanced ceramics on much larger bend bars, with no material classification and no property minima — a method where ISO 6872 is a specification. ASTM C1499 is the American equibiaxial method but loads ring-on-ring rather than piston-on-three-ball, a different stress field with a different closed-form solution, so the two are not interchangeable. ISO 23146 is the notched-beam fracture toughness method ISO 6872 cites; ISO 20501 covers Weibull statistics.
It is the international standard for dental ceramic materials, currently ISO 6872:2024 and adopted identically in Europe as EN ISO 6872:2024. It is a specification rather than a pure test method: it classifies ceramics by intended clinical use, sets each class a minimum flexural strength and a chemical solubility limit, and supplies the procedures for measuring those and several other properties.
Five, by intended clinical use, with the flexural strength minimum rising with the demand: 50 MPa for class 1, covering single-unit anterior prostheses, veneers, inlays, onlays and coverage of a framework; 100 MPa for class 2, single-unit anterior or posterior; 300 MPa for class 3, single units and three-unit prostheses not involving a molar; 500 MPa for class 4, three-unit prostheses involving a molar; and 800 MPa for class 5, prostheses of four or more units.
No. The requirement is written as a minimum value for the mean, so it applies to the set of specimens rather than to each one. A set whose mean clears the minimum conforms even where individual specimens fall below it. It also means a single strong result proves nothing on its own.
ISO 6872 allows three — three-point bending, four-point bending and a biaxial piston-on-three-ball test on a disc. The biaxial method is widely used for dental ceramics because it loads a central area rather than an edge, so it is less sensitive to the edge finish of a very thin specimen. Four-point puts a length of bar under constant moment and usually returns a lower, more conservative figure than three-point on the same material.
It is in clause 7.3, which is not published openly, and it is not reproduced here for a specific reason: two different constants for the piston-on-three-ball expression are in print in peer-reviewed dental journals, differing by about nineteen percent, along with differing groupings of the geometric terms. A laboratory that takes the formula from a paper rather than from the standard can be a fifth out on every result it reports, with nothing in its own data to reveal it. Take the expression from your purchased copy.
They are in clause 7.3 and are not published openly. We do not reproduce them, because the secondary literature disagrees: disc diameters of 12, 14 and 15 mm and ball diameters between 3 and 4.5 mm all appear in peer-reviewed papers describing ISO 6872 testing. Several sources agreeing with one another is not the same as one source that is the standard.
In the 2015 edition it sat in Annex B, marked informative, which makes it guidance rather than a requirement. Whether the 2024 edition changes that was not confirmed. What laboratories routinely report under it is the Weibull modulus and the characteristic strength, the latter being the strength at which cumulative failure probability reaches 63.2 percent. ISO 20501 covers Weibull statistics for strength data and appears in the ISO 6872 bibliography.
Far less than people expect, and resolution matters much more than capacity. A dental ceramic disc about 1.2 mm thick fails at loads in the tens to low hundreds of newtons, so the right instrument is a low-capacity, high-resolution load cell rather than a large frame working near the bottom of its range. The standard specifies no capacity, and it does not carry ISO 7500-1 as a normative reference, so it states no force accuracy class either.
It is the current published edition, but not a settled one. Its ISO catalogue stage already reads "International Standard to be revised" and a committee draft, ISO/CD 6872, is in development. The 2024 edition itself added Annex C on the hydrothermal stability of Y-TZP. Anyone building a long-lived procedure around it should expect to revisit it.
Discuss your specimen, test requirements and reporting needs with DAK engineering.
This page describes the method as practised. The governing text is the current edition from the issuing body. Tell us what you are testing and we will answer with the machine, the fixture and a quotation.
