The comprehensive efforts of the past few decades to improve thermal insulation and save energy have led to an increased use of insulating materials in the external components of buildings. These materials reduce the amount of heat that these components transfer away from the building’s interior. In terms of building physics, this increases the probability that condensation might form on the outer surface of the façade due to the cooling effect of long-wave radiation of heat during the night. The main prerequisite for microbial growth is sufficient moisture. However, the moisture content in exterior wall coatings is not the decisive factor for microorganisms, because according to present knowledge they can start growing only if there is a film of moisture on the surface.
The studies conducted to date indicate that hydrophilic coating systems are less susceptible to surface condensation. But this is true only if they are capable of releasing the absorbed moisture within a short period of time, to avoid accumulation. It therefore follows that it might be possible to reduce the risk of microbial growth by optimizing the hygroscopic properties of the materials used for rendering and in other external wall coatings.
To evaluate what happens when condensation remains on the surface of the material, the researchers developed a test facility to help them simulate the nightly formation of condensation in the laboratory. Their experiments were able to reproduce as closely as possible the conditions to which real building façades are exposed: excursions below the dew point are of the same order of magnitude as those achieved in the open-air test facility.
Because it is difficult to simulate nighttime heat loss by radiation in the laboratory, the specimens are cooled to below their dew point by means of a rear-mounted cooling plate, which causes condensation to form on the specimen’s surface. To optimize heat transfer from the test specimens, which have a surface area of 25 square centimeters, the scientists used a heat-conductive paste to attach them to the water-cooled, copper heat sink. They then measured the temperature on the surface of the specimens with ultra-flat pt100 sensors attached using hot-melt adhesive.
To simulate the behavior of a real façade in the laboratory, temperatures were selected that correspond to the natural drop in temperature associated with the formation of condensation. Similarly, the durations selected for the cooling phase, which range between one and six hours, are the same as might be expected in real-life conditions. These durations are determined by comparing the measured surface temperature with the dew point in the test chamber.
Taking the measurements
The refrigeration unit required to cool the specimens was switched on about one hour before the test was started and set to a temperature between 6.5 and 12.5 °C. The temperature for each specimen was chosen depending on its thickness and thermal conductivity, to ensure that all specimens have the same surface temperature at the start of the experiment. To record the environmental conditions in the test chamber, a humidity sensor and a pt100 temperature sensor were placed in the vicinity of the specimen. Environmental conditions were kept constant throughout the measurements, with the air temperature set to 23 °C and the relative humidity to 65 percent.
Five separate measurements were taken for each test setup, with cooling durations of one, two, three, four and six hours respectively, and approximately the same excursion below the dew point. The amount of surface moisture was determined by dabbing the surface with a cellulose pad and comparing its weight before and after the experiment. A similar method is used in open-air testing. A new test was started after each measurement, using a fresh specimen to avoid the accumulation of moisture. The previously used specimens were stored for at least one week in the test chamber to allow them to return to their original condition, while the tests continued using other specimens.
The measurements revealed that system C, which was coated with a layer of highly hydrophobic silicon resin paint, produced a high amount of surface moisture. The three other systems with more marked hydrophilic properties produced results, with barely any moisture picked up by the cellulose pads. These findings correspond closely to the results of experiments conducted by the Hygrothermics department on the same coating systems. Coating system E is a standard weatherproof system consisting of a silicon resin-based render and a silicon resin paint, comparable to system C but with the addition of two coats of renovation paint. After a cooling duration of up to four hours, this system exhibited a very low amount of surface moisture. This proves that a suitable hydrophilic coating, even when applied to a hydrophobic substrate, can significantly reduce the amount of surface moisture after overnight condensation. On the other hand, if the cooling phase lasts longer than four hours, the paint coating is no longer capable of absorbing the condensation, and water droplets become clearly visible on the surface.
The novel hydrophilic silicon resin paint used in system F was not capable of reducing surface moisture to the same extent as the mineral-based variants used in systems A, B and D, but nonetheless reduced the amount of surface moisture by about a factor of four. This demonstrates that the ability of paste-type surface coatings to regulate surface moisture can also be improved by enhancing their hydrophilic properties.
The concordance of the results of these laboratory experiments with those obtained through open-air testing has convinced the researchers at Fraunhofer IBP to establish this new method as a rapid, low-cost means of assessing the ability of coating systems to limit the accumulation of surface moisture. It is particularly adapted to the needs of companies wishing to enhance the properties of their coating products, enabling them to compile a shortlist of suitable candidates for further testing in the open-air facility, which demands more time and effort but ultimately delivers more statistically relevant data.