11.03 Durability of stone

Categories: Stone Cladding

Introduction
All natural stones weather and change with time.  The rate of weathering and the form it takes are controlled by the characteristics of the stone and the severity of the environment to which it is exposed.  In order to provide reassurance to the construction professionals testing of such materials is a fundamental requirement, particularly with ‘exotic’ foreign materials.  However, it is also a natural progression to assume that UK materials should not be treated any differently.  Although such materials may have a known service record, stone is accepted as being a naturally variable material and sufficient examples of stone failures in the UK are known for the testing requirement to be extended to all stone proposed for use in buildings no matter what its origin.

With regard to durability it is important to realise that once worked and used as a building material any stone will be liable to change in appearance and undergo at least some decay as a consequence of weathering.  The important question to be answered is whether the degree of weathering is acceptable.  For example where a weathered granite is used in a more traditional form of construction the loss of some 20 mm of material from the face of the stone may be acceptable, but where such a stone is used as a 32 mm thick cladding such decay would be disastrous.  In addition the design life of the cladding may vary and a supermarket may require a shorter design life than a cathedral.  The concept of weathering has to be put into the context of the end use of the stone.

Unfortunately there are many different tests, each attempting to measure durability by measuring different stone properties.  Some of the tests, such as the salt crystallisation test, will not reproduce the environment in which the stone is expected to perform but rather accelerate the effects of isolated aspects of the natural stone environment.  Such effects include, for example:

• Dissolution of calcium carbonate in limestones and calcareous rocks by acidic rain water,
• Damage caused by repeated freezing and thawing of water in certain weak stones,
• Damage caused by repeated expansion and contraction of clay minerals in certain rocks caused by cyclic wetting and drying,
• Cracking and bending induced by thermal changes of particular minerals in certain stones, notably marble,
• Damage caused by the crystallisation of salts within some porous stones.

Current guidance
British Standards
In contrast with other construction materials, such as steel, concrete or asphalt, there are few British Standard tests for natural stone, nor standard specifications defining minimum quality requirements.  Those which do exist are concerned with the more traditional uses of stone such as slate roofing (BS 680), slate sills and copings (BS 5628) and stone kerbs and setts (BS 435), rather than with cladding.
 

Other authoritative guidance
The Building Research Establishment (BRE) has long been a source of information on the use and properties of stone, one of its most notable early works being the study of weathering by Shaffer published in 1932 and reprinted in 1972. More recent works include BRE Digest 269 (1983), by Ross and Butlin, and BRE report 141 (1989) which considered certain durability tests for stone and went some way towards providing at least outline guidance on durability testing.  The test methods recommended are mainly for use on limestone, sandstone or slate and do not help with the evaluation of the many granite, marble and other stone varieties which are available.

In the USA the American Society for Testing and Materials (ASTM) publishes national standards for the testing of a variety of materials including stone for use in construction.  A range of test methods is given by ASTM along with a complementary series of standard specifications covering the main groups of building stones (slate, granite, limestone, marble and sandstone).  In the absence of other guidance these standards are used extensively on an international basis. ASTM, however, does not include any procedures to assist in the direct assessment of durability apart from methods for roofing slate.

Various standards have been produced by certain European countries, for example the German standard for freeze-thaw testing given in DIN 52 104.  In France there are also various tests for the assessment of limestone durability.

For the future, harmonised standards for the European Union are being produced by CEN (Comité Européen de Normalisation); these European Standards will be known as EN’s and include test methods and specifications. These standards should also describe when and how products are to be tested.
 

Testing for durability
Considering the available knowledge from published works and foreign standards a suggested regime of durability tests is given below.
 

Petrographic examination
A petrographic examination may range in scope from a simple description of a hand held specimen to a comprehensive microscopic description based on the examination of thin sections.  There is no specific British Standard for the examination of building stone although specifications for the description of rocks in general do exist, for example BS 5930, which deals with rock descriptions in site investigation and BS 812:Part 104, which deals with the examination of aggregates.  Other standards include the American ASTM C 295 and the Italian UNI 9724.

The basic requirement described in the Code of Practice BS 8298, that the stone should exhibit ‘freedom from vents, cracks, large fissures, sand and clay holes and other defects likely to affect durability’ could be largely assessed macroscopically. However, the examination of a stone using thin-sections is an additional and very valuable technique for the evaluation and classification of a stone in general and a starting point for the assessment of durability in particular. The following attributes of a stone can be determined by petrographic techniques:

• The presence of potentially unstable constituents such as clay-like minerals which may undergo volume change on wetting and drying;
• The presence of iron bearing minerals which may be prone to oxidation and cause discolouration;
• The determination of potentially frost susceptible stones, for example oolitic limestones with particularly fine grained minerals;
• Microbrecciated (fractured) rocks or the presence of particularly weak veins or seams within rocks.

A proper petrographic examination of a proposed building stone is therefore a starting point for the evaluation of the stone in general, not just for its durability, and it may dictate which other tests will be relevant.  Obviously the petrographic description will only be specific to a certain bed and a separate description would be required for each specific location during the preliminary selection stage and production testing.
 

Water absorption, porosity and saturation coefficient
There are various methods currently in use for the measurement of water absorption, but only one method is a British Standard, BS 680, which is specifically for roofing slates.  Most water absorption testing of building stones currently undertaken in the UK uses ASTM C 97, which involves soaking the samples in water at room temperature (a similar method is also given by Ross and Butlin, 1989).  Details of the measurement of porosity and saturation coefficient are also given by Ross and Butlin (1989).

Stones which exhibit low water absorption or porosity values are generally found to be more durable.  Water, which may be one of the main agents of weathering (either by freezing or thawing in the pores of a stone or by the transportation and deposition of soluble salts) will be less able to penetrate non-porous stone types, and therefore less able to cause damage.

When considering durability the vulnerability of porous building stones such as limestone and sandstones is more dependent on pore size and distribution than upon the absolute pore volume. The BRE have used the concept of saturation coefficients (also known as Hirschwald’s Coefficient), which is the ratio of the water absorption to total pore volume.  Porous stones with a saturation coefficient of more than 0.8 (1 is the maximum possible) are believed to be more susceptible to frost damage.  The saturation coefficient can be used in conjunction with porosity (although no critical value is defined for porosity) as a comparative guide to durability.  According to the BRE, the saturation coefficient is only of limited usefulness for sandstones.

For slates the water absorption acceptance criteria are given in BS 680.  The maximum permissible value is 0.3%; slates with values above this are thought to be potentially frost susceptible.

Measurement of water absorption by capillarity, which involves partial immersion of a stone in water to allow moisture to move up into the stone by capillary action, is another method that may be of value.  No British Standard method is available but standards such as NFB 10.502 are available.
 

Salt crystallisation
The cyclic crystallisation of salts within a porous building stone has often been used to assess the susceptibility of the material to frost action, although such a relationship is somewhat contentious. The BRE salt crystallisation test is mainly intended for use on limestone or sandstone, although theoretically there is no reason why the method could not be applied to other stone types.  In the test, small cubes (40 mm) of stone are repetitively soaked in sodium sulphate solution and then oven dried for a total of 15 cycles.  The durability of the stone is then assessed by the magnitude of the weight loss which occurs.

By using past observations of stone weathering and the test results from these stones the BRE was able to provide a classification of limestone based on their weight loss, as follows:

Class A weight loss  <1%
Class B weight loss  <5%
Class C weight loss  <15%
Class D weight loss  <35%
Class E weight loss  >35%
Class F shatters before 15 cycles

These durability classes can then be related back to the various basic zones of a building, and the resulting exposure conditions of varying severity, for which the stone might be considered suitable for use.  The classification of stones by BRE has only been applied to limestone and no similar classification is provided for sandstone, as it has traditionally been assumed to be frost resistant.

This has been perhaps the most contested test method in the UK stone industry at present with the Stone Federation of Great Britain advocating the abandonment of the test as a misuse of overall durability.  However, the theory of the tests using either sodium or magnesium sulphates is not restricted to building stones; a similar method exists for aggregates within BS 812.  It would appear that it is the interpretation of the test results which is the most difficult area.  Examples of potential problems with interpretation include:

• Many problems arise from the poor techniques used by some laboratories
• The definition of weight loss; if a cube parts along a natural plane during the test should that be regarded as weight loss, or should weight loss be more closely defined perhaps by grading criteria
• The use of control stones; these will also be subject to natural variation which will not be quantifiable and may penalise the stones being tested in comparison with them
• Are the Classes well defined, for example is the real difference between stones with a weight loss of 0.8% (Class A) and 5% (Class C) sufficient for the stone to be treated so differently in the BRE guidance (for example in an exposed coastal area subject to frost but with no pollution, the Class A stone could be used on any part of a building but the Class C stone could only be used as plain walling - in an area of high pollution the class C stone would be unusable, even as plain walling according to the BRE criteria)
• To what extent do the BRE criteria take into account the required design life of the stone.

Not withstanding the criticisms of the test and its interpretation it is accepted that the crystallisation test itself does have some merit as a measure of salt resistance and further work into the interpretation of the results and the classification of the stone is currently being undertaken.  Despite the problems of interpretation the salt crystallisation test is still a valuable aid to the determination of the durability of the stone in salty environments.
 

Acid immersion testing
The effect of pollution can lead to the precipitation of slightly acid water (acid rain) and in some areas this may be a major cause of damage to stone on the buildings.  All limestones (which are predominantly composed of calcium carbonate) will be particularly vulnerable to such damage, leading to the erosion of the surfaces and profiles and in some cases to the delamination of planes of stone due to the crystallisation of reaction products within the stone itself.  Other stones which have calcite as a cement between sand grains will also be vulnerable to acid attack.

The effects of acid attack on sandstones is assessed using the Acid Immersion Test outlined by Ross and Butlin (1989); this test can use various acid strengths depending upon the durability period required.  Such a test is not used on limestone as the results would be meaningless given their chemical composition.

Slates are also tested to determine their resistance to acids using the methods described in BS 680 (for roofing slates) and BS 5642 (for slates used as sills and copings).
 

Freeze-thaw testing
The frost resistance of stone is not covered by any British Standard test method; some believe that the salt crystallisation test is an indirect measure of frost resistance.  However, various overseas methods can be used to determine the frost resistance;  for example DIN 52 104 is sometimes used in the UK to determine the frost resistance of some limestones and sandstones.  There are no pass or fail criteria specified but the condition of the stone is visually assessed throughout the test and any defects noted.  The test, which involves cycling the specimens of stone between -20 and +20^oC, is certainly able to cause breakdown in some weak limestones, for example, which may show cracking or spalling after as few as five cycles of freezing and thawing whereas more dense, less porous, limestone can survive the test visually unaffected.  As with the salt crystallisation test it is useful to use control stones for comparison purposes.  Different criteria can be used to interpret the effects of freeze-thaw; in addition to visual examination French and American methods use the measurement of sonic velocity carried out on the specimens at certain stages of the test to determine the effects of the freeze-thaw cycles.
 

Thermal stability
There is no specific test method to cover thermal stability of stones as most building stones are thought to be dimensionally stable.  Although this may be true in many cases it is not true of certain marbles when used as external cladding or even paving at certain thickness.  Bowing or dishing of marbles, when used as external cladding panels, is thought to be due to a process known as thermal hysteresis whereby the thermal expansion of the calcite crystals in the marble causes irreversible deformation;  others consider the effect may be due to the heat releasing built-in geological strain.  Whatever the cause the effects were measured in experiments by Erlin in the USA and more recently by Messrs Sandberg in the UK (unpublished data) where the effects were found to be markedly different for varying thickness’ of stone.  The most important factors appear to be a temperature of between 20 and 60^oC and occasional wetting of the stone panels under test.  After only 50 cycles in the tests carried out by Messrs Sandberg, one commonly used marble had cracked and shown an overall expansion of 0.5%; such performance on a building may lead to a failure of the stone in service.
 

Dye testing
One of the tests that has significantly improved the visual inspection of cut stone is dye penetration testing.  This technique works very well with polished stones such as granites and marbles.  These are particularly prone to induced cracking due, for example, to blast damage caused by explosives during quarrying, or to mechanical damage during handling and transportation of blocks and cut panels.

Measurement of water absorption by capillarity, which involves partial immersion of a stone in water to allow moisture to move up into the stone by capillary action, is another method that may be of value.  No British Standard method is available but standards such as NFB 10.502 are available.

Dye testing allows induced cracks to be seen more easily and also enables them to be distinguished from the normal pattern of micro-cracks on the basis of size, width, orientation and overall morphology.
 

Interpretation of results
The concept of a single test which could address all the various mechanisms of weathering will almost certainly be impossible to realise, so the assessment of durability will have to be taken on the basis of the results of a series of tests such as those described in the previous sections.

The tests required for each stone group are given in the table above and the results may be interpreted to some extent using the criteria given below.
 

Stone acceptance criteria
 

* These ASTM Standards were reapproved in 1996.

This table is for guidance only and should not be used as a substitute for a specification.  The current test used to measure the flexural strength of a stone is the four point loading test and it is the preferred test for assessing strength of the cladding panels.