Lime Cements, Plasters, Mortars and Concretes


Terminology

There is a plethora of terms used to describe the various products derived from calcined limestone. A definition of terms used in the context of this report are outlined below. The definitions here are based on features identifiable in hand specimen and are therefore intended for use by the field archaeologist. Consequently these definitions may differ somewhat from those applied by scientists employing microscopic and chemical techniques (i.e. Affonso, 1996). Many terms frequently used as nouns to describe cement-based materials, are derived from their usage. Therefore in an effort to create a limited and ëscientificí terminology, words such as render, grout and to a certain extent, plaster and stucco will be used only as verbs describing the application of the material.
 
 

Limestone

Limestone is the natural rock type from which cements and concretes are derived. A limestone is a sedimentary rock composed of carbonates, namely the minerals calcite (calcium carbonate; CaCO3) and dolomite (calcium magnesium carbonate; CaMg[CO3]2), derived either chemically or organically. Being natural materials, limestones can have a wide range of depositional environment and components and can contain varying amounts of non-carbonate material. The type of limestone calcined to produce lime for the manufacture of cements and concretes can profoundly affect the durability and properties of the material produced.
 

Lime

"Calx", lime in Latin provides the etymological root for calcium, calcite and calcination. Strictly speaking, lime is calcium oxide. This is acquired by burning limestone, which in simplest terms, removes the carbon from the calcium carbonate (calcite). Lime forms the base for all cements and concretes. However the composition of the limestone being variable, the true composition of the lime is also variable, and the term may be generally used to described calcined limestone in general. Lime may also referred to as 'quicklime', 'unslaked lime' or 'lump lime'.
 

Slaked Lime

Lime will only harden into a cement when water is added to it. The process is called ëslakingí, hence ëslaked limeí, also referred to as ëlime puttyí.
 

Cement

"Cement" is derived from caementa, which actually referred to the aggregates mixed with the slaked lime rather than the bonding agent itself. Opus caementicium refers to the masonry constructed from concrete coursework. In the context of this work, the word cement is used exclusively in reference to the hardened binding material of any aggregate. In simplest terms, this is the slaked lime, which in the presence of air reverts to calcium carbonate. This material may also contain finely powdered admixtures, such as ash or fired ceramic. Adam (1994) discourages the use of the term cement in this context, defining it as a modern substance, an ëartificial mixture of lime, clay and metallic saltsí. However, an alternative term for the hydrated lime binder is not put forward. In geological parlance, the term cement is used strictly to define material (often calcite, but occasionally silica) that adheres clasts in a rock, and generally the word ëcementí is accepted in non-specialised use as a glue. Consequently it is defined as the adhesive binder in this report.
 

Aggregate

An aggregate is material added to a cement. In this report, the term does not include finely powdered additives to the cement such as ash. It is usually composed of rock fragments, chosen for their strengthening properties and occasionally for decorative reasons. Carefully chosen aggregates can make a concrete or mortar resemble natural rock. Organic material, including grasses, reeds and also bones can be used as aggregates, often in combination with rock material. aggregate can be sub classed into categories of ëfine aggregateí - that with dimensions less than 5 mm, and ëcoarse aggregateí - that with dimensions greater than 5 mm.
 

Concrete and Mortar

Both concrete and mortar are materials composed of a cement plus an aggregate. The two terms are very simply defined. A concrete is a material where the majority of the aggregate has dimensions greater than 5 mm. A mortar is a material having aggregate with dimensions less than 5 mm (Prentice, 1990). "mortarí from the Latin mortarium, originally referring to the trough in which the material was mixed, is often used wholesale to describe the bonding material of concrete masonry. This is a fair use of the term as this material binding brick-sized blocks, often laid in courses rather than haphazardly poured, is composed of lime cement with fine aggregate. ëConcreteí should be used to describe material containing coarse aggregate generally used to fill formwork.
Hydraulic Cements
 Hydraulic cements are waterproof and will even set underwater. Consequently most materials used in Roman and later periods for lining structures intended to carry water and for construction in marine and riparian environments will be hydraulic. Such materials are identified by the presence of finely pulverised material added to the cement binder which will cause discolouration from white to pale browns or pinks. Common additives, are volcanic ash or crushed ceramic sherds. These materials are known as pozzolanas (see below).
 

Pozzolana

A ëpozzolanaí is defined as a siliceous and/or aluminous substance that will, in the presence of water combine with lime to form cementitious compounds. Such material include clays, that are rendered active by firing (fresh geological clays would absorb too much water during the curing process, resulting in spallation and ultimately cracking of the concrete), and other substances, including waste products from blast furnaces and even rice husk ash (see Hill, et al., 1992 and references therein). However, natural pozzolanas are derived primarily from volcanogenic products.
 

Plaster

The term "plaster" is one of the most universally used, describing a multitude of products, usually those which will provide a smooth coat to a wall or other surface. Plaster of Paris is powdered gypsum (CaSO4.2H2O, derived from rocks distinctly different from limestones) which, with water added, will harden and set. Lime plaster is simply a mixture of lime, water and sand, or just lime and water in the case of whitewash. To avoid confusion, it is recommended that these terms should be used in full to define compositional types.
 

Stucco

Stucco, like plaster, is a term of complex use. However it is entrenched in all literature as describing two distinct materials. Firstly it is used to refer to a fine white mortar composed of lime, crushed marble and glue-like binding additive (egg-white for example) typically used for making good surfaces for painting frescoes or generally for smoothing walls. Alternatively it is used to describe decorative work in plaster of Paris. In this report, ëstuccoí will be used to define the former case alone.



 

The manufacture of lime cements and their derivatives in antiquity


The phenomenon that limestone, when burnt in a "bright fire", changes its composition to lime, which when mixed with water sets with a hardness similar to that of the original stone was amongst the earliest forms of pyrotechnology utilised by man. Like many innovative discoveries, it is likely that calcination of limestone was discovered several times over a wide geographical distribution. But a widespread adoption of a material or technology is related to several factors, as summarised by Stapleton (1975) "technology is confined within men and not in the materials they use or the objects which their skills create", emphasising that technology transfer is a learning process whereby capable and knowledgeable craftsmen are able to teach their skills to an active and, importantly, receptive culture.

Production of cements is highly energy and labour-intensive. Needing not only a plentiful supply of the raw materials (limestone and aggregate) but also requiring copious sources of fuels, i.e. wood. As a consequence it constitutes a profession for a team of people, at least during times of need. Quarrymen, woodmen, kiln stokers, "mixers" and "slakers", as well as skilled plasterers and builders would be required. Accordingly, the manufacture and usage of lime cements with sufficient refinement of technique, fostered an understood technology that was in operation by the 7th millenium BC within the Mediterranean region. The social effects were contemporary with and accompanied those surrounding the establishment of agriculture, a sedentary lifestyle and the emergence of cities.

Early uses of limes
The earliest records of the manufacture and use of lime cements in the Mediterranean region were to make adhesives for tool making as early as 12, 000 B. C., and for such limited use, lime could be made by simply burning small blocks in a fire. The first documented appearance of a lime kiln, and hence evidence for larger scale production is at the Natufian site, Hayonim Cave, dated c. 10, 4000 - 10, 000 B. C. By the 8th millenia before Christ, the technique had become widespread, at least within the Middle East and the Levant. "A sizeable production effort" in the manufacturing of lime cement was in evidence in the Near Eastern Pre-Pottery Neolithic B cultures (PPNB; c. 7200 - 6000 B. C.) which are characterised by rectangular architecture and (lime) plaster floors. Lime cements and their derivatives were not restricted to structural uses alone. The remarkable lime plaster on a reed skeleton sculptures from the PPNB Levantine site of ëAin Ghazal (Jordan) illustrate an early, (c. 6600 B.C.) and sophisticated decorative use of lime. Vessels made from lime plaster (ëVaiselle Blancheí ware) and pottery lined with plaster have been recorded from a variety of Levant sites. A PPNB skull coated in lime plaster has been retrieved from Jericho and shows careful construction, using different grades of plaster to produce different textural effects. These include and aggregate-rich base coat with an outer coat for an aggregate deficient plaster to give a smooth finish, but with the addition of powdered iron oxide to give a pink ëskiní. A beard was added containing a coarser aggregate to "resemble the appearance of whipped cream overbeaten to form peaks" and containing powdered manganese oxide to give a black colouration. Such a construction shows a sophisticated knowledge of materials, where composition and sculpturing coupled with the understanding of aggregates to impart colours and textures were adapted for specific uses.

Greek Archaic through Hellenistic Periods
The Greeks were familiar with the technology of concrete and mortar construction and used the material, routinely for laying floors and covering walls from at least the Early Bronze Age (i.e. The House of Tiles, Lerna) and as a waterproofing agent for cisterns, reservoirs and baths from a similar period (i.e. at Mycenae). As a structural material, concrete was rarely used although a Hellenistic cast concrete barrel vault is known from Upper Peirene at Corinth.

Whether the Ancient Greeks used pozzolanas with lime cements specifically as waterproofing agent is a subject of some debate. Certainly suitable materials were available from major volcanic structures such as Santorini (Thira) and Melos, both of which are exploited today for pozzolanas. The Greeks were well aware of the properties of lime cement based concretes, and used them almost exclusively as waterproofing agents for baths, cisterns and flooring. Further research is necessary to evaluate intentional inclusion of pozzolanic material in Hellenistic and earlier cements, mortars and concretes. In the Hellenistic fountain houses of Corinth, waterproofing was attained by "burnishing", rubbing the surface of the ëleather hardí mortar with a stones, to produce a hard skin. Presence of volcanogenic material either as a coarse aggregate or finely powdered in the matrix is not obvious, although chemical analyses of the plaster taken at the time of excavation defined the cement to have a high silica content. Efstathiadis (1978) claims the presence of ëSantorini earthí in concrete lining a cistern in Cameiros on Rhodes, although it is not clear what methods were used for analysis of the cement. The intentional use of pozzolanas by the Classical and Hellenistic Greeks cannot be ruled out, however, it appears that they failed to realise the adaptability of the material.

Roman concrete and the architectural revolution
The Romans almost certainly inherited the technology of cement manufacture from the Greeks, but adopted the material to their own uses and significantly altered and perfected the use of aggregates, including pozzolanas, producing through innovation an extremely versatile construction material. Thus an ëarchitectural revolutioní in the use of lime cements and their derivatives, mortars and concretes, has lead to the (misleading) accreditation that the Romans were the inventors of ëconcreteí. The Romans began to employ lime cement based products towards the end of the third century before Christ. By the middle of the second century B. C., the Romans had an established and sophisticated technology. Cato in his treatise on agriculture (c. 160 B. C.) describes in detail the construction and operation of flare kilns. Some of the Earliest Roman structures using lime cements and mortars have been excavated from Pompeii, where friable and unevenly burned lime cements are used to point masonry and opus africanum walls.

The Roman architect and engineer Vitruvius, writing in the 1st century BC described a powder which produced ëwonderfulí results when added to a simple lime-water mix. This ëpowderí, in fact tuffs derived from the volcanic province of Campi Flegrei (Plegrean Fields) on the Bay of Naples, enabled cements and concretes to set in the presence of sea water and, in addition, to produce stronger structures than those built with lime cement alone. The tuffs were quarried from the vicinity of the modern town of Pozzuoli. The Romans called this material pulvis puteolanis, dust of Puteoli, Puteoli being the Latin name for Pozzuoli, which in turn has given the modern term ëpozzolanaí widely applied to all additives to cements which produce an hydraulic set. The unique properties of this material were probably discovered as the local tuffs, scoria and lavas were the aggregate of choice for the construction of the harbour and the port at Roman Puteoli. However, later they were possibly to become an important export from the area. Vitruvius wrote that ëthe masonry which is to be in the sea must be constructed in this way. Earthy material is to be brought from the district which runs from Cumae to the promontory of Sorrento, and mixed in the mortar, two parts of it to one of lime.í The Romanís knowledge of the volcanic activity associated with the Vesuvius-Campi Flegrei volcanic province was probably limited, despite the construction of heated saunas (sudatoria) in the area around Baiae and Puteoli, utilising the naturally occurring hot springs. Until the catastrophic A. D. 79 eruption of Vesuvius which destroyed the towns of Herculaneum and Pompeii, there is no evidence that the Romans were aware that Vesuvius was an active volcano. Therefore they were perhaps justified in believing that the tuffs around the Bay of Naples were unique in their origin and chemistry and therefore in the hydraulic properties which they imparted to concretes.

Vitruvius (II.6) suggested that a powder which produced ëwonderful resultsí was to be found in ëthe neighbourhood of Baiae and the municipalities in the vicinity of Mount Vesuviusí. He also believed, that the formation of this substance was related to the ëfervent heatsí of the area. More recently, the volcanic activity of the area has been fully described. The Campi Flegrei Caldera is a Quaternary structure which is surrounded by three other Quaternary centres; Ischia, Vesuvius and Procida, providing abundant potential locales for Vitruviusí pulvis Puteoli. Thus the province of Imperial Rome was surrounded by Quaternary volcanoes of the Roman Comagmatic Province, with Rome itself surrounded by the edifices Vulsini, Vico, Sabatini and Laziale. Consequently, building projects along the Tyrrhenian coast used volcanogenic pozzolanas as the aggregate of choice, ashes, tuffs and scoria being in abundance, local and easily quarried. It is also possible that the Roman engineers observed (naturally) calcite-cemented beach rock composed of scoria blocks and used this as inspiration for their own materials. That the Roman lime-pozzolana cement would set under water was probably a chance discovery. Without access to modern strength-testing technology it is likely that they failed to realise that a better set occurred under water, than when exposed to air. We have no evidence that Romans allowed water to run over terrestrial architecture to aid in the curing process. In the dome of the Pantheon (A.D. 118-125) in Rome, a further property of volcanogenic material was utilised; pumice was used as aggregate in the upper parts of the dome to create a strong but lightweight structure.
 As a result of the legacy of the building projects of Imperial Rome, it has often become reported in the literature, and become an assumed fact amongst classical archaeologists that it is the presence of pozzolana that makes concrete ëRomaní. Away from the abundant sources of volcanic material this is apparently not the case. Roman construction in the provinces did not routinely use natural pozzolanas in terrestrial architecture, preferring to use local aggregate and, when required, using a synthetic pozzolana in the form of crushed and powdered potsherds for waterproofing baths, cisterns and aqueducts. Roman architecture in Germany, utilised the local volcanigenic pozzolana, ëtrassí notably in constructions in Cologne situated in the Rhine Graben, with access to abundant volcaniclastic products, but not in Trier near the Belgian Border. For Roman marine architecture, it would appear that an hydraulic and therefore lime-pozzolana concrete is an essential feature. Concrete at the Imperial Roman Ports of Cosa, Ostia and at Puteoli certainly utilise local volcanic pozzolanas. In the Roman provinces, where volcanigenic material may not occur locally, it was necessary to import aggregates and pozzolanas to provide the hydraulic set, and this was apparently the case at Herodís Harbour at Caesarea Maritima.
 Herod organised a highly technical and magnificent building project to construct his harbour at Caesarea Maritima (Sebastos) importing Imperial Roman architectural styles and apparently, Italian architects and engineers. The technological advances made at the harbour are described by Oleson and Branton, (1992). The harbour moles are characterised by huge blocks of concrete, far larger than would be manageable from quarried stone. These were created by pouring concrete into wooden formwork, the forms floated out as rafts and then sunk to produce the moulds. The concrete was a pozzolana-lime hydraulic set, and Oleson and Branton (1992) believe that the pozzolana was imported from quarries in the vicinity of Pozzuoli for the purpose. Their conclusions were reached after major and trace element analyses were conducted on volcanogenic fragments from the Herodean mortars and these were compared with material from local Palestinian volcanic deposits and those surrounding the Bay of Naples, and also material from Santorini. That Herod would go to the expense of importing shiploads of pozzolana from over 2000 km away is amazing enough in itself. It is believed that perhaps the imported Roman engineers preferred to work with materials that were familiar to them. At present, no other proven examples of trade in Italian pozzolanas are known during the Roman period, but the consequences of the discoveries at Caesarea Maritima are far reaching.
 



 

Technology of lime production and concrete manufacture

The starting material for the production of lime cement is the rock limestone which is composed dominantly of calcium carbonate, CaCO3, which manifests itself as the mineral calcite. Limestone can have a wide variety of forms. Most are formed in marine conditions and are often highly fossiliferous, although freshwater limestones (tufas and flowstones) do exist. The body of knowledge about carbonate rocks is well beyond the scope of these notes, but the interested reader may be referred to Tucker and Wright (1990) for further reading from a geological point of view, and to Boynton (1979) and Wingate (1985) as to their suitability as base products for the manufacture of cements.
 

Non-hydraulic, ëlimeí cements.

The chemical reactions involved in the formation of quicklime, calcium oxide, are deceptively simple, belying the skilled labour involved in producing the product. Quicklime is produced byëburningí blocks of limestone in a lime kiln or pit. The chemical reaction for this process, called calcination, is as follows (equation 1):

CaCO3 + heat -> CaO + CO2     (1)

The temperature required for this dissociation to take place is 898°C at 1 atmosphere pressure. If the temperature of dissociation is not reached, then the lime will be underburned and unreacted carbonate (ëschwachbrandí) will remain. If the lime is overburned and the dissociation temperature is exceeded, the lime will have reduced reactivity, rendering it less workable. At high temperatures, the lime becomes ëdeadburnedí and becomes inert.

The product CaO or quicklime, also known as 'unslaked lime' or 'lump lime', is converted to cement by the addition of water, in a process called slaking. Quicklime is highly reactive when put in contact with water, consequently when slaking lime (in a tank or pit) the lime is added to the water and not the other way round. The reaction for this process in equation (2) is:

CaO + H2O -> Ca(OH)2     (2)

It is at this point in the process that the aggregate is added to the cement to make mortar or concrete. Aggregates are mixed with the slaked lime ëputtyí and the constituents are stored together, protected from the air as wet 'coarse stuff' for as long as possible to mature.

 For simple, non-hydraulic lime cement, the coarse stuff should be allowed to sit under water, thus protected from the air, for a minimum of two weeks, but may be stored there for several months before it is used. This prolonged slaking process improves the workability and plasticity of the cement. Before use, it is thoroughly remixed; this process is called ëknocking-upí.
 Ca(OH)2, calcium hydroxide, converts during curing (setting) when exposed to carbon dioxide in the air, to the starting compound CaCO3, producing a recrystallised ësyntheticí limestone, lime cement. This is shown in equation (3) below:

Ca(OH)2 + CO2 -> CaCO3 + H2O     (3)

 This reaction, called carbonation, should be a slow process as rapid drying out will form a weak material. The resulting calcium carbonate, whilst retaining an identical composition to the naturally occurring mineral calcite, develops a new structure not encountered in nature. This is only apparent when the two substances are compared with a scanning electron microscope (SEM). Calcite grows naturally as crystalline ësparí with a variety of crystal habits. In lime cements, CaCO3 forms tiny spherules less than 1 µm in diameter. This polymorph of calcite has similar optical properties to the naturally occurring finely crystalline variety of calcite known as micrite, and is often indistinguishable from micrite when observed under a polarising microscope, the spherules being visible only when using an SEM.

Hydraulic cements
 The processes outlined above are those used for the manufacture of non-hydraulic lime cement. Hydraulic cements, those that cure as a result of a reaction with water and pozzolanic additives is the result of a different manufacturing process. The unique property of pozzolanic additives were recognised by the Romans in their use of pozzolana (volcanic ash) and substances such as ceramic sherds and crushed fired brick. It is the reactive alumina and silica which form new crystals on setting in aqueous environments.

 The process of calcination is the same as outlined above for simple lime cement. The pozzolanic additives are added with the aggregate at the time of slaking, after which the resulting concrete should be used within a few hours as setting immediately begins to take place. The simplified reaction is shown in equation (4) below:

CaO + {xSiO2 + yAl2O3} + H2O -> Ca3Si2O7.3H2O + Ca3Al2O6.6H2O + Ca(OH)2 (4)

 It is the presence of silica (SiO2) and alumina (Al2O3) compounds within volcanic tuffs which react with lime (CaO) which produce an hydraulic set in concretes. Lime, when mixed with water, forms a paste which will on curing take in carbon dioxide from the atmosphere and revert to a synthetic calcium carbonate. Hydraulic sets may be produced in two ways; either an hydraulic lime is produced by calcination of a argillaceous limestone producing a material not dissimilar to modern Portland Cement (see below), or alternatively a pozzolana is added to the dry lime prior to the slaking process. It is the latter process that was adopted in Antiquity. The set is produced with the addition of water (H), whereby available silica (S) and alumina (A) from the pozzolana will undergo complex reactions with the lime (C) to produce amorphous, low-solubility phases. Hydration of lime will produce the mineral portlandite Ca(OH)2, CH in the shorthand of cement chemistry. The presence of pozzolana-derived S and A will promote the formation of calcium silicates and aluminates which hydrate to form an amorphous gel, known as C-S-H and hydrogarnet, C3AH6 during the highly exothermic slaking reaction (see Gani, 1997 and references therein). The resultant material is a strong, water-resistant and durable mortar as is illustrated by the many examples of Roman architecture still standing.

 Portland Cement produces a hydraulic set by the calcination of argillaceous limestones (those containing high proportions of clays, rich in alumina and silica) or a mixture of limestone and clay, which is found to produce the same compounds on slaking (first patented by James Frost in 1811, and called ëPortlandí because of the similarity in appearance of the set concrete to Portland Stone). Additional pozzolanic additives may well be added with the aggregate on slaking. The calcination temperatures required for this process are somewhat higher than those used in the techniques outlined above, requiring ~ 1300°C at 1 atmosphere pressure.
 


Raw materials

Selection of carbonate rocks

The starting material for the manufacture of lime for cements and concretes is calcium carbonate (CaCO3 - the mineral calcite) which usually manifests itself in the form of limestone, marble (metamorphosed limestone), coral or shell or even a (rare) volcanically derived rock, carbonatite. Limestone is most frequently used as a raw material for the production of lime, not least because it is the commonest source of CaCO3. The natural variation in limestones is immense and the interested reader is referred to the large number of available texts on carbonate sedimentology for further information on the subject (e.g. Reading, 1993; Tucker and Wright, 1990). Selection of a limestone for calcination is important as ëany old limestoneí will not necessarily produce a workable lime. Many limestones contain impurities (sand, clay, phosphates, iron, organic carbon) and a second component in major mineral amounts, namely dolomite (magnesium carbonate, MgCO3). Increase in the dolomite content of the limestone causes a corresponding decrease in the ability of that limestone to produce a ëgoodí lime. Thereby lime burners discuss limestones in terms of their ëavailable limeí, i.e. the percentage CaCO3, MgCO3 or other impurities present as classified below (Wingate, 1985).

 ultra-high calcium limestone >97% CaCO3
 high calcium limestone 95-97% CaCO3
 high purity carbonate rock > 95% CaCO3 + <5% MgCO3
 calcite limestone  <5% MgCO3
 magnesian limestone 5-40% MgCO3
 high magnesium dolomite >43% MgCO3

 To summarise, the best limes will be produced from the purest limestones with minimal or no dolomite present; that is, ideally, ultra-high and high calcium limestone and high purity calcium rock. Dolomite may be tolerated, but the MgO (periclase) formed after calcination takes a very long time to slake, much longer than lime (Pliny the Elder recommended three years!) and unslaked periclase will cause the cement to expand and pop while setting, thus destroying the cement. In favour of dolomitic limestones, the calcination temperature is lower and they will, if thoroughly slaked, make perfectly adequate construction cements.

 Available lime and dolomite content are not the only factors that influences a limestones suitability for calcination to produce high quality lime. Ideally, the limestone should have a regular porosity, and should be breakable into evenly sized lumps before firing. Despite this it should be fairly hard, so that it does not crumble or break up when firing is in process. The lime should leave the kiln in lumps of similar dimensions to the limestone that went in.

 The presence of clay impurities in limestones may be beneficial, for these will produce hydraulic limes when calcined. Clay-bearing (argillaceous) limestones will produce hydraulic limes if they contain up to 25% silica-alumina clays. Greater amounts of clay will decrease the available lime to unacceptable levels. Ideally the limestone should contain clays in a disseminated form through the body of the rock, rather than in discrete nodules or beds. The ability of a cement to set underwater and produce a waterproof material increases with the increase in the acceptable clay content. The following classification of argillaceous limestones and corresponding cements was derived by the French military engineer, Vicat (1837);

 Feebly hydraulic <12% clay
 Moderately hydraulic 12-18% clay
 Eminently hydraulic 18-25% clay

Such limestones produce a grey rather than white coloured clay.

 Marbles are metamorphosed carbonate rocks, those that have been recrystallised by heat and pressure. Many marbles have high CaCO3 content, but also commonly contain dolomite and new mineral phases occurring from the metamorphism of other impurities. Minerals such as olivines and pyroxenes will form as a result of reactions involving Mg from pre-existing dolomite and such minerals will disrupt calcination processes. Also, it is generally found that the compactness and hardness of these recrystallised limestones means that they calcine less effectively, with large lumps of uncalcined limestone being left at the end of the firing.

 An important factor in selection of limestones is the proximity of the source quarries to the location of firing, slaking and construction. Large amounts of rock are difficult and expensive to transport across great distances.
 

Selection of aggregates.
 The purpose of aggregate is essentially to improve the durability and strength of the cement, allowing it to be used for a variety of purposes. For use in the construction of buildings, aggregate is selected to meet rigorous criteria; namely strength, water absorption and shrinkage and resistance to weathering. Possibly the most important factors are the strength and water absorption of minerals.

 The strength of a material is its resistance to compression, shear stress and tension, and in rocks and minerals the strength is related to the hardness and strength of the individual minerals and the strength of the cement binding them together to form a rock. The strength of minerals is also somewhat related to their structure. Well-cleaved minerals (like calcite, which has three cleavage planes) may fail along these cleavages. The strongest materials for use as aggregates are hard igneous rocks such as granite, or metamorphic rocks such as gneiss. As a mineral aggregate quartz, being relatively hard and possessing no cleavage, is an ideal fine aggregate.

 Shrinkage of the cement occurs during the curing process due to the loss of water. Shrinkage is greatest in pure lime cement, with no aggregate present. Use of an aggregate of rocks and minerals such as quartz, flint and marble produce the lowest shrinkages, whereas platy minerals like micas and clays and the rocks that these form should be avoided (mudstones, micaschists etc.), due to their ability to absorb too much water on mixing which is rapidly lost during setting. Excessive shrinkage will cause microcracking, permeating the structure of the cement, along the boundaries of the aggregate and sometimes passing through the aggregate utilising fractures and flaws within the aggregate. This process would eventually result in the disintegration and failure of the concrete.
 The susceptibility of mineral species to chemical weathering is of course very important. Rapidly weathering minerals will causes weaknesses in areas exposed to the weather, which will eventually permeate through the rock. Again quartz or quartz-rich rocks like granite are ideal as aggregates.

 The size of the aggregate should also be graded depending on the intended use of the mortar or concrete. Aggregate can be simply subdivided into the categories of coarse (particles over 5 mm in diameter) and fine (particles less than 5 mm in diameter) aggregate. The former variety is used in concretes for construction purposes, whereas the latter is used in mortars for finishing stonework and for aesthetic use. Coarse aggregate can be further graded by screening. "Fines", particles with diameters less than 75 µm should be screened out of both fine and coarse aggregate. These usually constitute clay minerals which have deleterious effects upon the concrete when curing, causing shrinkage.

 Aggregates used are most likely to come from local sources, except for those specially imported as pozzolanic additives (see below), or for decorative use. However, aggregate is generally carefully selected and graded depending on its use. By the age of Augustus, Romans were sorting aggregate by density with heavier materials used in foundations and lighter materials used in vaults and arches.

 Aggregate is batched to the lime putty in amounts depending on the intended use of the product; usually approximately 3 parts of aggregate are used for 1 part of cement. The consequences of this are that large quantities of aggregate are required. There are two major sources of aggregate. Firstly it may be collected in a naturally comminuted form from residual (i.e. in situ weathering of a rock), colluvial (scree), alluvial (river gravels), marine or glaciogenic deposits. Alternatively quarried rock may be crushed. Both types have their attributes and disadvantages. For example a concrete with natural rounded aggregate has a far superior workability than one with crushed angular aggregate. However, angular aggregate has the advantage of being a good key for the adherence of the cement.
 

Selection of pozzolanic additives.
 The term pozzolana is derived fro the Roman use of pulvis Puteolanus (dust of Puteoli), volcanic ashes (tuffs) and volcanogenic sands from around the town of Roman Puteoli (modern Pozzuoli) used to form an hydraulic set with lime cements. The ashes rapidly weather to clay minerals, essentially aluminosilicates, which react with quicklime and water to form a concrete that will set under water. First utilised in the construction of the port of Puteoli, pulvis puteolanus was extensively quarried around Pozzuoli where it has a sandy appearance, grey-brown in colour. This material is known to have been exported throughout the Mediterranean for exclusive use in underwater constructions (i.e. Herod's Harbour at Caesarea; Oleson and Branton, 1992). The volcanic province of the Mediterranean, characterised by Vesuvian- and Plinian-type eruptions is rich in ash deposits associated with volcanic edifices such as Vesuvius, Melos and Thira, many of which were recognised by the Roman architects and engineers as suitable materials to form hydraulic sets in municipal buildings. However, it is apparently unlikely that the Romans realised that these other ashes were essentially of the same composition as their prized pulvis puteolanu. Vitruvius says (2.6.1) ;

 "There is a kind of powder which by nature produces wonderful results. It is found in the neighbourhood of Baiae [near Puteoli] and in the lands of the municipalities around Mount Vesuvius. This, when mixed with lime and rubble, not only furnishes the strength to other buildings, but also, when piers are built in the sea, they set underwater".

And again in 5.12.2;

 "[In the construction of breakwaters] the masonry which is to be in the sea must be constructed in this way. Earthy material is to be brought from the district which runs from Cumae to the promontory of Minerva [Sorrento], and mixed in the mortar, two parts of it to one of lime"

In fact, hydraulic cement actually set harder in aqueous conditions than when exposed to air. No evidence has been found to suggest that the Romans allowed water to run over terrestrial cement-work indicating that they were also unaware of this phenomenon. This is perhaps not surprising, due to the inaccessibility of submerged masonry for the purpose of testing its strength and durability.

 Alternatives to pulvis puteolanus and other pozzolanas often used in Roman cements are crushed, fired clay, essentially of similar composition to the clays in volcanic ash and consequently imparting similar properties. Typically these alternatives were recycled broken, fired ceramic tiles and fragments of pottery, or crushed fired clay brick. As noted above, fresh clays when added to the lime would absorb water and cause shrinkage while the plaster is curing. However, fired clays do not show this property and are hence an ideal pozzolanic additive.
 

Selection of fuels
 Fuels cannot be disregarded when discussing the manufacture of concrete. Although coal, oil and gas are generally used in modern calcination processes, wood is unsurpassable in terms of flame quality and controllability. The low temperature of woods ëlong-flamesí also make it virtually impossible to overburn the lime: temperatures greater than 900°C are rarely achieved. However large quantities of wood are required if wood is to be used as a fuel in lime kilns. To calcine around 1 tonne of limestone requires around 300 kg of wood (Wingate, 1985), and this would produce approximately 0.5 tonnes of lime. These values are probably conservative. 19th century lime kilns probably used equal amounts of fuel to limestone charge and open pit firing would require approximately twice as much fuel as charge (Kingery et al., 1988, and references therein). To summarise, plentiful supplies of fuel must be readily available in areas proximal to the kilns. In fact, the much documented deforestation of southern Britain during the middle ages has been blamed on the requisition of forests to fuel the royal lime kilns! Other fuels commonly used for small-scale lime burning include brushwood, kernels and shells of nuts and fruit, pine cones, and orchard prunings which were all readily available in the Mediterranean region, and are the fuels used for lime burning in the region at present.



 

Selected Bibliography


Books

*Adam, J-P., 1994, Roman Building: Materials and Techniques. Batsford. INST ARCH CDC 398 ADA or YATES QUARTOS K 30 ADA

*Ashurst, J and Ashurst, N., 1990, Practical building conservation, volume 3. Mortars, plasters and renders. English Heritage technical handbook, Gower Technical Press., 85 pp. INST ARCH LC ASH

Gani, M. S. J., 1997, Cement and Concrete., Chapman and Hall, London. ENGINEERING CS 52 GAN

*Hill, N., Holmes, S. and Mather, D. (Eds.), 1992, Lime and other alternative cements. Intermediate technology publications, London. ENGINEERING CS 52 HIL

Lamprecht, H., O., 1993, Opus Caementitium: Bautechnik der Romer. Beton-Verlag, Dusseldorf. YATES K 30 LAM

*Prentice, J. E., 1990, Geology of construction materials., Chapman and Hall, London., 202 pp. GEOLOGY F 50 PRE

Wingate, M., 1985, Small-scale lime burning, a practical introduction., Intermediate Technology Publications, London. (N.B. Not in UCL Libraries).
 


Journal Articles

Freestone, I. C., 1995, Ceramic Petrography. American Journal of Archaeology., 99, 111-115.

Grissom, C. A., 2000, Neolithic statues from ëAin Ghazal: construction and form. American Journal of Archaeology., 104, 25-46.

Kingery, W. D., Vandiver, P. B. and Prickett, M., 1988, The beginnings of pyrotechnology, part II: production and use of lime and gypsum plaster in the pre-pottery Neolithic Near East., Journal of Field Archaeology, 15, 219-244.

Lechtman, H. N. and Hobbs, L. W., 1986, Roman concrete and the Roman architectural revolution. in: W. D. Kingery (ed.), Ceramics and Civilisation, vol. III; High Technology ceramics: past, present and future. 81-124.

Oleson, J. P. and Branton, G., 1992, The technology of King Herod's harbour, in: Vann, R. L., (Ed.)., Caesarea Papers: Stratonís Tower, Herodís Harbour and Roman and Byzantine Caesarea., Journal of Roman Archaeology, Supplementary Series No. 5, 49-67.

Rollefson, G. O., 1990, The uses of plaster at Neolithic 'Ain Ghazal, Jordan., Archaeomaterials, 4, 33-54.

Siddall, R., 2000, The use of volcaniclastic material in Roman hydraulic concretes: a brief review. in McGuire, B., Griffiths, D. & Stewart, I. (eds) The Archaeology of Geological Catastrophes. Geological Society , London, Special Publications 171, 339-344.

Whitbread, I. K. 1986, The characterisation of argillaceous inclusions in ceramic thin sections. Archaeometry, 28, 79-88.

Whitbread, I. K., 1989, A proposal for the systematic description of thin sections towards the study of ancient ceramic technology. Archaeometry, 31, 127-138.


Back

R. Siddall, 2000