The Analysis of Mortar:
The Past 20 Years
Peter Ellis

Where historic buildings are concerned, repairs should be carried out using materials and techniques which match those used originally as closely as possible. There are three main reasons for doing this: firstly, repair materials which match the originals will provide some continuity with the past, keeping intervention to the minimum; secondly, by matching the original materials and techniques closely, the repair materials will age in a similar way as the original; and finally, modern materials and techniques introduced in past repair work have often proved to be incompatible with the original, causing accelerated deterioration of building fabric. Changes are usually made only where the existing materials have been shown to be inappropriate.

In recent decades, specialists involved in the conservation and repair of historic buildings have become increasingly concerned by the damage caused by the use of certain types of mortar on historic brick and stonework, and by the use of ordinary Portland cement in particular. This has sparked renewed interest in traditional lime mortars, which are more compatible with old buildings. Some analysis of the existing mortar is now standard practice, if only to match aggregates. However, a better understanding of historic mortars is clearly required, not least because there have been some cases where modern lime renders have failed and, at the opposite extreme, there is now concern as some types of modern hydraulic lime mortars continue to gain strength as they age.

RECENT DEVELOPMENTS
In parallel with the conservation industry, mortar analysis has progressed dramatically over the last 20 years. The procedures pioneered by Ian Constantinides and others, based as so much was (and still is), on John and Nicola Ashurst's excellent English Heritage Technical Handbooks have been replaced by various more sophisticated techniques.

Many individuals, companies, laboratories, and universities are now offering the service, and much research is being undertaken. Analysis is carried out with differing goals, from a simple colour match to ensure a sympathetic 'matching' repair mix, investigative assessments and performance evaluations, to academic studies that attempt to determine the precise reasons for the remarkable durability of ancient mortars.

Many analytical techniques are being employed, each with particular strengths and limitations. The established techniques are summarised in the following table.

The important, but often underestimated first step is to ensure that any sample taken for examination is representative of the mortar to be analysed. Far too often the method of sampling influences the result, and taking small or insufficient samples can lead to poor assessments or diagnosis. The original 'kitchen sink' tests are no longer considered to be of any use for much more than insoluble aggregate type matching. Generally these tests consisted of dissolving a sample in dilute acid to separate the acid-soluble from the insoluble. The soluble proportion is (often incorrectly) assumed to be the binder, (as it so often includes soluble limestone aggregate and calcareous clays), and the binder is assumed to be carbonated lime (which of course it often isn't). In some cases the hydraulic component is being assessed on the insoluble 'fines' proportion. This is meaningless, as in the vast majority of cases where the proportion of 'fines' is high, the mortar includes unwashed clayey aggregates.

The more chemical tests undertaken, the better the overall understanding of the mortar, and the higher the confidence in interpreting the data on the sample.

The minimum requirement is to carry out recognised standard tests for sulphate (to determine whether the binder is, or contains, gypsum), calcium, soluble silica, and insoluble content. The soluble silica test is critical for assessing the hydraulic proportion as it determines the combined percentage of calcium silicates, calcium silicate hydrates and hydrated silica gel present. It is the amount of this material which determines the hydraulicity of binder, whether it be Portland cement, Roman cement, hydraulic lime (of any grade) or an added pozzolan. In fact, as this result is so critical, the soluble silica test should be carried out to a high degree of confidence and calibrated using two distinct methods, the BS technique listed in BS4551, and the colorimetric method. In common with much in the analysis of mortars, great caution must be exercised, as some granite aggregates will release soluble silica and this could be taken as part of the hydraulic component.

Other chemical tests can also help; a magnesium test for example determines whether the lime was dolomitic.

Physical properties, such as porosity, should be determined. All reliable data indicates a complex interrelationship between porosity, permeability, pore size distribution and durability.

More meaningful mortar analysis should consist not only of chemical tests properly conducted, but these tests should be augmented by one or more collaborative instrumental techniques such as DTA, microscopy, XRD, or SEM.

One of the most effective collaborative techniques is DTA (differential thermal analysis). This technique is particularly useful in determining the calcium compounds present. It positively determines calcium sulphate, calcium hydroxide (lime), calcium carbonate, often distinguishing between carbonated lime and calcareous aggregate, and calcium silicate hydrates. In older mortars, DTA cannot distinguish between hydrated alite (C3S) and hydrated belite (C2S) as these are essentially the same, and so on its own cannot be used to determine whether the binder is cement or hydraulic lime. However, positive DTA identification of calcium silicate hydrates and calcium carbonate, and a porosity test, considered with the quantitative soluble silica test results combine in making an informed assessment of binder type. Only by determining whether unhydrated C3S is present can Portland cement be confirmed, although there is an indicative calcium oxide : soluble silica ratio. The experience of the person interpreting the results is almost as important as the results themselves. In one recent project the same mortar was tested by four laboratories, and each interpreted differently. Conclusions ranged from a cement/lime blend, a hydraulic lime (possibly in the form of a natural cement), a hydraulic lime/non-hydraulic lime blend, to a 'hydraulic or cementitious' binder.

On another occasion, a fresh lime putty plaster sample had been tested because gypsum gauging was suspected. The testing laboratory, using XRD found no gypsum but identified C3A, a constituent of cement, and some hydraulic limes. Cement or hydraulic lime gauging was immediately identified ignoring the facts that the wet putty plaster was not setting and that calcium silicate hydrates had not been identified. This resulted in a high level meeting of client, architects, analysts, main contractor, and plastering sub-contractor, most of us flown in at great expense. DTA was able to prove conclusively that hydrated calcium aluminates were not present. The outcome was that the XRD reflection had been incorrectly identified as C3A, and it was in fact a constituent of the complex igneous aggregate.

The area most underestimated is how mortars age and the complex chemical reactions and changes that occur with time. The fact that the binder in an aged sample is now principally calcium carbonate does not indicate that it was lime originally, as much of the hydrated hydraulic compounds in cement and hydraulic lime will themselves react with CO2 (carbon dioxide), and carbonate. To complicate further, lime still present in a 40-year old sample for example, does not necessarily indicate a lime mortar originally, as lime is a reaction product of the hydration of C3S and C2S. Indeed, the presence of lime in such a sample is more likely to indicate a cement mortar as its non-permeable nature would have impeded or prevented the access of CO2, and carbonation of the reaction product has therefore not taken place.

Mortar analysis is now a very sophisticated business. However, any examination that measures only part of the components, characteristics or properties of a mortar and their relationship with durability or performance must be viewed with caution. As with everything, a little knowledge can be a dangerous thing.