Moisture transport and convection in building envelopes

Charlotte Gudum

Research output: Book/ReportPh.D. thesis

415 Downloads (Pure)

Abstract

When occurring, convection is known to be the dominant type of moisture transport. The different parts of the building envelope are protected against convection by ensuring that connections are airtight. In ventilated building components the outdoor air passes through a ventilated cavity on the outside part of the insulation, in order to keep the design dry by exploiting the potential of convection to transport moisture coming from inside the building.

Previously light weight facades have been ventilated and used only in one or two storeys. The Danish building regulation were however in 1995 adjusted to permit wooden facades in up to four storeys, provided that the façade was designed without cavity. Another change in façade designs has been a continuing increase in insulation thickness as a function of demands for decreased heat loss. For these reasons the present PhD project has focussed on designing and validating a model for analysis of the effect of ventilation and insulation thickness upon the moisture load of the wall.

On a ventilated 25 mm cavity (height 1650 mm, width 559 mm) placed on a north facing wall of an outdoor test house the cavity air velocity and the wind pressure at the top and bottom of the cavity were measured together with the wind velocity and wind direction 4.8 m above ground.

The cavity air velocity was determined from the air change, using a gas analyser and the constant dose method. Tracer gas was dosed constantly across the cavity and samples were taken upstream and downstream from the dosing tube. The highest tracer gas concentration was used for determination of the flow direction and air velocity. Using the tracer gas the cavity air velocity was measured from –1 to +1 m/s for wind velocities in the interval 0.5-10 m/s 4.8 mabove ground level. A shortcoming of the method was, that erroneous results were observed when changes in the cavity air direction were faster than the methods time constant of 4-5 minutes. The use of a gas analyser also facilitated the measurement of convective moisture transport.

A comparison between the tracer gas method and thermo anemometers showed a satisfactory correspondence between the average velocities over a 9 minute period. The velocities measured with the thermo anemometers in the centre of the cavity were adjusted by multiplication of the factor 2/3 to match the velocities measured with tracer gas. The factor of 2/3 matches the expected value for laminar flow. A similar velocity profile was also observed in parallel with the wall, but this velocity profile was dependent on the wind direction.

The roughness of the surroundings was measured to a high value of 6.1.m, which was attributed to turbulence in the vicinity of one measuring position. The wind pressure coefficients outside the cavity vents were determined with a high standard deviation from pressure measurements as a function of the wind direction. A set of wind pressure coefficients was estimated for the façade (Knoll, 2000). They showed that the pressure difference between top and bottom were highest for side wind and lowest when the façade was in leeside. A model for simulation of coupled moisture and heat transport in a ventilated façade was
designed using Simulink under Matlab. The model simulated one-dimensional coupled moisture and heat transport by conduction and diffusion in the material layers, and cross-flowing onedimensional air stream in the cavity. The model was validated with satisfactory results: By comparison of a non-ventilated case with MATCH and by comparison with a one-year outdoors measurement on four examples of composite ventilated façade designs and a non-ventilated design.

A simulation model, of the coupled heat and moisture transfer in a ventilated wall, was made using Simulink in Matlab.

9 different yearly simulations using the Danish reference year (DRY) as outdoors boundary condition and an in-door climate varying from 21-23ºC and 40-66% RH was used too study the effect of insulation thickness (100 mm, 200 mm and 300 mm); presence or absence of vapour barrier and the degree of ventilation (none, low or high). A critical condition level was defined as simultaneous RH above 80% and temperature above 5ºC. The simulated RH and temperature behind the wind barrier were compared. Simulations showed that the time of critical moisture load was increased with increasing insulation for a ventilated façade with vapour barrier, but not to a critical length of time. For designs with vapour barrier ventilation increased the period of critical moisture load, but again not to a critical length of time. For a design without vapour retarder the period with critical moisture load was longer in the absence of ventilation than in the presence of ventilation, but in either case the period length was critical.

Based on the simulations it was concluded that a non-ventilated wooden façade may be considered a durable design, provided that the vapour barrier is both vapour and airtight. Furthermore a ventilated façade may compensate for a non-perfect vapour retarder.
Original languageEnglish
Place of PublicationKgs. Lyngby, Denmark
PublisherTechnical University of Denmark
Number of pages124
ISBN (Print)87-7877-107-2
Publication statusPublished - Feb 2003
SeriesByg Rapport
NumberR-047

Fingerprint Dive into the research topics of 'Moisture transport and convection in building envelopes'. Together they form a unique fingerprint.

Cite this