The Building Conservation Directory 2024

126 THE BUILDING CONSERVATION DIRECTORY 2024 CATHEDRAL COMMUNICATIONS events as a result of the changing climate, the need for remedial work to deal with rain penetration looks set to increase, so it seems timely to review the findings of the early phases of the Damp Towers and remind ourselves about the benefits of lime render. Although initially focussed on church towers, the findings are relevant to other building types suffering from rain penetration, particularly in areas of high exposure. INITIAL INVESTIGATIONS In 1989, in response to requests for advice, Iain McCaig of English Heritage investigated seven church towers in Devon with bare masonry towers suffering from rain penetration. His report concluded that the main problems were their exposure and their dependence on the mortar joints alone for moisture buffering of the construction due to the impermeable nature of the stone. He attributed most of the moisture penetration to failure at the joints, which appeared to be exacerbated by shallow pointing using impermeable materials that had cracked and de-bonded. Several potential solutions were presented, including deep pointing, grouting, lime rendering, replacing cement mortars internally with lime and improving ventilation. However, the report noted there was no data to indicate which of these would be most effective. There was little scientific literature that specifically targeted solid walls: most research had focused on cavity wall construction. Furthermore, the historical records were not always clear about repairs and interventions. Weather records were often generalised rather than specific to the individual site and differences in architectural detailing around each building could make the situation more complex within each structure. As a result of these field studies, in 1996 English Heritage commissioned a programme of field and laboratory research. The aim was to determine how water enters and migrates within solid walls and identify which remedial treatments could effectively minimise water ingress, and ideally also maximise drying following periods of wind-driven rain (WDR). It was hoped that the work could also provide information on driving rain through stone, mortar joints and renders, and provide general information on the performance of different mixes for mortar, rendering and plastering. FIELD AND LABORATORY RESEARCH Although the research included field studies, there are obvious difficulties in carrying out detailed investigations on sensitive historic fabric; many of the buildings were still active places of worship and some were listed Grade I. Also, while the towers shared some features such as the use of impermeable stones and composite walls, each was unique. There were considerable variations in the stonework type (ashlar, rubble, slate), joint width and mortar type, state of weathering, tower design, level of maintenance and previous interventions and the height, altitude, and aspect of the buildings. The common repair methods applied to existing structures exposed to WDR were reviewed and, as a result, the research focused on pointing, rendering, plastering and grouting for void filling. A number of other repair options (such as slate hanging, tree screening and internal drainage) fell outside the remit of the work and were not considered. Approaches focused on evaluating moisture content within the masonry walls and the quality and condition of the wall facing and rubble core, followed by lime-based repair using rendering, plastering, pointing and grouting techniques. Field work was also undertaken, including rendering the tower of the church of the Holy Trinity, Challacombe, on the edge of Exmoor. It should be noted that, at the time of testing, natural hydraulic lime (NHL) was considered the optimum choice of binder for exposed mortars, but it is now recognised that its use will have distorted some outcomes. Given the difficulties of field research and the lack of existing research on driving rain, it was decided to develop an intensive programme of laboratory investigations. This research was based at Sheffield Hallam University, where a number of different strands of enquiry were pursued using a combination of methods. LABORATORY SIMULATION A climatic simulator was adapted to enable the simulation of WDR on one side of the test walls and various sensors (existing and innovative) were deployed to collect data. It was accepted that it can be difficult to replicate complex historical structures in laboratory settings, so a skilled stonemason (the late Colin Burns) was called on to build the walls. Pilot phase During the pilot phase of research, five wall panels were constructed. Each measured half a metre by one metre and was built with diorite setts of varied sizes (average dimension of approximately 100 mm x 100 mm x 150 mm). Mortar was wellgraded aggregate mixed with NHL 3.5. Rain simulation delivered water at a rate of 0.5 litres per minute per square metre for six hours at a pressure of 25 mm H2O standing (the equivalent of a wind speed of 20 metres per second). In addition to sensors, visual logging was used to provide qualitative data to aid interpretation. Five wall panels were constructed of an internal and external skin with throughstones, giving a wall thickness of 150 mm. • Panel 1: Panel rendered with NHL mortar, given a smooth finish • Panel 2: Panel with eroded NHL mortar joints • Panel 3: Panel rendered with NHL mortar, given an open textured finish • Panel 4: Panel with defective cementpointed joints • Panel 5: Control panel, with good NHL mortar joints One surprising outcome of this initial work was the unexpectedly quick water penetration through to the rear of the panel. All panels leaked within six hours, and resulted in free water running down the back (interior) face despite the relatively low wind speed and realistic rainfall profile. Two runs of testing confirmed good performance from both rendered walls, with the smooth render performing slightly WALL 1 WALL 2 WALL 3 WALL 4 WALL 5 Fine render Washed out Course finished Repointed Control mortar render in cement panel The five test panels in the laboratory at Sheffield Hallam University (Photo: SHU and Historic England).