The comparative in situ hygrothermal performance of Hemp and Stone Wool insulations in vapour open timber frame wall panels

An in situ experiment in a full scale timber frame test building was carried out to compare the hygrothermal performance of Hemp and Stone Wool insulations of identical thermal conductivity. Hemp and Stone Wool insulations were installed in timber frame wall panels without vapour barrier. The comparison was made in terms of heat transfer properties, likelihood of mould growth and condensation. Step changes in internal relative humidity were performed to explore the effect of high and normal internal moisture load on the wall panels. No significant difference between the average equivalent thermal transmittance (U-values) of the panels incorporating Hemp and Stone Wool insulations was observed. The average equivalent U-values of the panels were closer to the calculated U-values of the panels based on the manufacturers’ declared thermal conductivity of Hemp and Stone Wool insulations. It was observed that the placement of heat flux sensor along the depth of the insulation had significant influence on the measured equivalent U-value of the panels during high internal moisture load. The frequency and likelihood of condensation was higher in the interface of Stone Wool and Oriented Strand Board (OSB). In terms of the parametric assessment of mould germination potential, relative humidity, temperature and exposure conditions in the insulation-OSB interfaces were found to be favourable to germination of mould spore. However, when the insulations were dismantled, no mould was visually detected.

Quasi steady state and dynamic hygrothermal performance of fibrous Hemp and Stone Wool insulations: Two innovative laboratory based investigations

Moisture and heat management properties of Hemp and Stone Wool insulations were studied by mounting them between a hot and a cold climate chamber. Both insulations were exposed to identical hygrothermal boundary conditions. Quasi steady state and dynamic tests were carried out at a range of relative humidity exposures. The likelihood of interstitial condensation was assessed and equivalent thermal conductivity values of the insulations were determined. The adsorption-desorption isotherms of the insulations were also determined in a dynamic vapour sorption (DVS) instrument. It was observed that the likelihood of condensation was higher in Stone Wool insulation than in Hemp insulation. Hemp insulation performed better in managing moisture due to its high hygric inertia and water absorption capacity. It was observed that the equivalent thermal conductivity of Stone Wool insulation was dependent on enthalpy flow and phase change of moisture. The equivalent thermal conductivity of Hemp insulation was close to its declared thermal conductivity in dynamic conditions when high relative humidity exposures were transient. In quasi steady state boundary conditions, when the insulation was allowed to reach the equilibrium moisture content at ranges of relative humidity, there was a moisture dependent increase of thermal conductivity in Hemp insulation.

MUSA: A Scalable Multi-Touch and Multi-Perspective Variability Management Tool

Variability management is one of the main activities in the Software Product Line Engineering process. Common and varied features of related products are modelled along with the dependencies and relationships among them. With the increase in size and complexity of product lines and the more holistic systems approach to the design process, managing the ever- growing variability models has become a challenge. In this paper, we present MUSA, a tool for managing variability and features in large-scale models. MUSA adopts the Separation of Concerns design principle by providing multiple perspectives to the model, each conveying different set of information. The demonstration is conducted using a real-life model (comprising of 1000+ features) particularly showing the Structural View, which is displayed using a mind-mapping visualisation technique (hyperbolic trees), and the Dependency View, which is displayed graphically using logic gates.