Gauging Lime Mortars
Lime mortars have been used with certain reactive materials known as ‘pozzolans’ for almost as long as they have been in existence. These additives, which include such common materials as fragments of pottery and certain types of brick, may have been introduced into the mix for various reasons, but their effect was to produce a mortar that had some hydraulic properties, was less permeable and generally more durable than an ordinary lime mortar. The addition of small quantities of pozzolan or any other material that achieves this effect is known as gauging.
A non-hydraulic lime mortar sets by a simple chemical reaction between lime (calcium hydroxide) and airborne carbon dioxide alone, whereas a mortar with some hydraulicity also sets by a reaction between calcium and the reactive components, principally silica. (Hydraulicity is actually defined as the ability to set under water, without any carbon dioxide.)
The current debate is focused on the blending of non-hydraulic lime putty and hydraulic hydrate.
It is the author’s view that gauging with hydraulic hydrate is poor practice. Although it is important to make clear that the principal proponents of this practice can provide many examples of the successful use of these ‘complex mixes’, long-term durability of these mortars has yet to be assessed.
The combination of ordinary lime and hydraulic lime in the same mortar causes concern for the following reasons:
- There is little historical precedent. This practice is likely to be a development of the Portland cement / lime hybrid mortars typified by the 1:1:6 and 1:2:9 mixes used in the 20th century. Lime in these mixes was added as a plasticiser to improve workability, although in a 1:2:9 it was presumably expected that the lime would carbonate and assist the setting process. Analysis of these hybrids regularly finds lime still present after 50 years. The setting cement impedes the carbonation of lime.
- It is a dangerous assumption that the addition of, for example, lime putty to an eminently hydraulic hydrate will produce a moderately hydraulic lime. The chemistry of hydraulic lime is complex and the setting processes are delicate. Data available on this1 is preliminary and suggests significantly reduced compressive strength, increased watervapour permeability, and far worse performance in salt crystallisation tests. It is extremely important to define this practice. The blending of non-hydraulic and hydraulic materials can range from a hydraulic hydrate with five per cent putty added to a putty with five per cent hydraulic hydrate added, and every variable between. The end-product will have varying properties.
- This blending is now not necessary. We are fortunate to have a wide range of hydraulic hydrates, lime putty and pozzolanic additives available in the UK, and there is no application in historic building repair where a blend is likely to outperform the correct grade of hydraulic hydrate, or pozzolanic lime. The only possible exception to this is the addition of less than eight per cent by volume of lime putty to a hydraulic hydrate to improve plasticity. This will improve workability, hopefully not at the expense of durability, although thorough mixing of the hydrate mortar often makes the addition of putty unnecessary. In the UK, specialists first became aware of this practice when Jura-kalk, an eminently hydraulic hydrate from Switzerland became available. It was advised that this material had been routinely blended with lime putty in equal proportions in Denmark and elsewhere in Europe. Jura-kalk is a binder containing very little lime2,3 and is a complex blend of compounds of calcium and silica (principally C2S), calcium and alumina (principally C3A), calcium, alumina and silica (C2AS), calcium, alumina and iron (C4AF) and calcium carbonate. The addition of lime to this is likely to be different to the addition of lime to a material that does contain lime. Many hydraulic hydrates notably from England and France have significant lime content.
BRIEF HISTORY OF LIME
Lime was produced by burning locally available limestone in a coal or wood fired kiln at a temperature rarely in excess of 900 degrees C. The properties of the lime produced were largely influenced by the chemical composition of the limestone burnt, and limestones containing clay minerals produced a lime with weak hydraulic properties. The hydraulicity of these limes was likely to be weak because of the low firing temperatures, and weakened further if the lime produced was stored as putty. Certain compounds, notably the calcium aluminates, but also any di-calcium silicates present will hydrate or begin hydration in the lime pit. This explains why the ‘hot mixes’ where quicklime is mixed with water and aggregate on site have different properties.
The belief that most historic limes were non-hydraulic or only very weakly hydraulic is supported by the fact that certain additives have been added historically to alter the performance characteristics. These ‘heated’ materials contained silica, alumina and iron which became reactive towards alkalis including lime.
The earliest mortars analysed from Jericho in the Jordan Valley, and Tel-Ramad, Syria4, dating from 7000 BC contain stable end-products of pozzolanic reactions although it is not possible to conclude whether the pozzolanic material was deliberately added, or naturally present in the aggregate.
Pozzolanic materials in the form of crushed brick and tile were deliberately added in large quantity in mortars of Minoan Crete of c 1000 BC, ancient Greece, and the Roman period. Evidence suggests that the Romans used crushed brick and tile before they discovered the naturally occurring pozzolanic aggregates from around Vesuvius.
The practice, not so much ‘gauging with pozzolans’ but more the deliberate inclusion of pozzolanic materials as aggregate, was lost in Britain after 400 AD but continued in Europe as demonstrated by the analysis of samples from the Byzantine Empire5, Venetian renders6, Sistine Chapel plasters6, and recently analysed samples from the 13th century Moorish castle in Gibraltar which have the ingredients of ancient mortars - carbonated lime, calcium silicate hydrate, brick particles, quartz sand and limestone particles.
Further evidence that limes were generally non- or only very weakly hydraulic is demonstrated by Vicat7 in early 19th century France, and his frustration with the limes available, and his exhaustive trials to find a binder that would prove durable for hydraulic engineering works. His work, and others’, notably John Smeaton in late 18th century England, led to the recognition of natural hydraulic lime, the manufacture of artificial hydraulic limes, the invention of Parker’s ‘Roman Cement’, and in 1824 to the first Portland cement patent.
THE SPECIFICATION OF REPAIR MORTARS
This is a complex issue and each building and its particular condition and problems must be considered individually. There are two considerations that are of paramount importance:
1. The ‘Like-for-Like’ philosophy.In most cases it is technically and aesthetically appropriate to carry out repairs using a mortar to match the existing or original material, replacing like for like. This requires proper analysis to ascertain exactly what material was used, and demands a detailed understanding of materials currently available. This does not imply that poor mortars should be matched, particularly where their use might be harmful to original fabric.
2. Mortars should be durable yet sacrificial to the building fabric.This normally entails preparing the mortar from the constituents required and in the right proportions to ensure that the result is both more porous and more permeable than the stone or brick. This is so that mortars age, decay and ultimately fail before the masonry - hence the term ‘sacrificial’.
THE CHARACTERISTICS OF THE VARIOUS LIME TYPES
1. Non-hydraulic Lime Putty (Fat Limes) Many traditional limes were non-hydraulic, as is most modern lime putty. They set by the reaction with atmospheric carbon dioxide in the presence of moisture alone. A non-hydraulic lime mortar is soft, porous, permeable and plastic. They are used for bedding mortars, for internal and external pointing mortars, and for internal plasters. Internal putty plasters have been and still are on occasion gauged with gypsum to accelerate the set and reduce shrinkage. This was commonly done from c 1760 to ceilings and especially run and cast work, but never to walls where there is a risk of damp as gypsum is slightly soluble in water and sulphate salts migrate and crystallise on the plaster surface.
2. Impure Lime Putty (Lean Limes) Most traditional limes, but sadly few (if any) modern limes fall into this category. They contain impurities such as coal or wood ash, unburnt or partially burnt limestone and a small proportion of reactive silica produced by the de-hydroxylation of clay minerals in the limestone. Some contained a small proportion of di-calcium silicate (C2S). The setting process was principally carbonation augmented by a very weak hydraulic reaction as the C2S hydrates and the silica reacts with lime. The un-converted calcium carbonate and fuel ash also played a positive role. These limes were not as hydraulic as today’s ‘feebly hydraulic’ classification. These limes, supplied as both putty and quicklime, were used to build most things in Britain, including much of London. They were also used for base-coat plasters and external renders.
3. Traditional Hydraulic Limes Certain limestones with high clay mineral content produced hydraulic limes that would be today classified as ‘feebly hydraulic’ (or sometimes possibly moderately hydraulic). The principal examples of these are the Lias limestones from Somerset, Devon, and Aberthaw in South Wales. Arden lime in Scotland is another example. These were invariably supplied as quicklime to be mixed with water and sand on site and used immediately. Putty made from these limes would set quickly, and usage advice for Totternhoe lime, a hydraulic chalk lime from Bedfordshire which was slightly less hydraulic than Lias limes, was to slake only enough on a Friday necessary for the following week’s work.
These traditional hydraulic limes produced durable mortars which were used widely. For example, Lias lime from Devon was used at the Tower of London from the 15th century.
The hydraulicity of the lime produced from these complex raw materials is determined by kiln temperature, and indeed Blue Circle cement is now made from Aberthaw limestone. Traditional kilns rarely got hot enough for complete combination, and the hydraulicity of these limes was largely due to a pozzolanic silica/ lime reaction together with the hydration of limited C2S and C2F (di-calcium ferrite).
4. Modern Hydraulic Hydrates These are produced from clay mineral rich limestones similar to those used to make the traditional hydraulic limes, but once burnt, the material is passed through a hydrating plant, where sufficient water is added to convert the quicklime to calcium hydroxide but not to hydrate the C2S. However, any calcium aluminates are likely to be hydrated by this process. These range in hydraulicity from feebly to eminently hydraulic depending on factors such as kiln temperature and length of time in the kiln, as well as the chemical composition of the limestone. Some of these materials are subsequently blended with pozzolanic additives and in some cases white cement. The Foresight Project2 has identified C3A (tri-calcium aluminate) and C4AF (tetra calcium alumino ferrite) in most types tested indicating kiln temperatures in excess of 1,000 degrees C, hotter than traditional lime kilns. These modern hydrates are therefore more hydraulic than the earlier materials. Available data also suggests an inversely proportional relationship between hydraulicity and permeability.1,8
These hydrates, in particular the less hydraulic grades, have a part to play in historic building repair in applications where reduced porosity and increased strength are advantageous and where reduced vapour permeability is acceptable. These applications include external mortars and renders especially in exposed or aggressive environments. There is less risk of failure when work must proceed in winter as they set more quickly and are thus vulnerable to frost for a shorter period. They are clearly appropriate for repairs to hard mortars such as ‘Roman Cement’, and a better option than cement based mortars. They are rarely appropriate for pointing mortars or internal plasters.
5. Pozzolanic Lime Mortars The durability of pozzolanic lime mortars of correct mix design is proven beyond doubt. A pozzolan is defined as a material that is capable of reacting with lime in the presence of water at ordinary temperatures to produce cementitious compounds. The essential difference between these and modern hydraulic hydrates is that the reaction takes place in solution. The pozzolanic reaction products and the compounds produced on ageing will differ and depend on the calcium to silica ratio in solution. Modern hydraulic hydrates derive their hydraulic properties from the subsequent hydration of compounds of principally calcium and silica produced by solid state reaction in the kiln. The chemistry is of the same chemical nature, but it is not the same.
Pozzolans vary in reactivity, and historically include naturally occurring volcanic Italian pozzolana and Santorini earth as well as artificial forms including brick and tile powder. The varieties most used in the UK are the metakaolin Metastar 501, certain brick dusts of known reactivity and Trass from Germany, but other forms include HTI (ceramic ‘high temperature insulation’) and PFA (‘pulverised fuel ash’). The ground slags are not true pozzolans as they may themselves be cementitious; these are classed as latent hydraulic binders. The addition of ten per cent pozzolan improves durability and strength and slightly reduces porosity and permeability. The pozzolan reacts with the lime and does not set in isolation as occurs when hydraulic lime or cement is added. Pozzolans are added to lime putty mortars where there is doubt about durability and the reduced porosity is not disadvantageous.
To quote Vitruvius from De Architectura in the first century BC: ‘If to river or sea sand, potsherds ground and passed through a sieve, in the proportion of one third part, be added, the mortar will be the better for use.’
- Jeanne Marie Teutonico et al, International RILEM Workshop Proceedings PRO 12, English Heritage and BRE UK
- Paul Livesey et al, Foresight Project, University of Bristol
- Dave Hughes and Simon Swann, 'Hydraulic Limes: A Preliminary Investigation', Lime News, Volume 6, 1998
- Joseph Davidovits, 'Ancient and Modern Concretes: What is the real difference?', Concrete International, December 1987
- L Binda et al, 'Experimental Study on the Mechanical Role of Thick Mortar Joints in Reproduced Byzantine Masonry', International RILEM Workshop Proceedings PRO 12
- E Charola (USA) and F Henriques (Universidade Nova de Lisboa, Portugal), 'Hydraulicity in Lime Mortars Revisited', International RILEM Workshop PRO 12
- L J Vicat, A Treatise on Calcareous Mortars and Cements, Artificial and Natural, (1837), re-printed Donhead, Shaftesbury, 1997
- P Banfill and A Forster (Heriot Watt University, Scotland), 'A Relationship Between Hydraulicity and Permeability of Hydraulic Lime', International RILEM Workshop PRO 12