Scientific Report 1996- 1998


Building Physics


Content


Introduction

Since its start in 1978, the Laboratory and Entity of Building Physics has been involved in basic applied research, consultancy and educational activities in the fields of heat, air and mass transfer in porous materials and building elements, durability, energy consumption in buildings, sustainable construction, building acoustics and room acoustics. To manage that broad field of activity, the entity consists of two units: acoustics and heat&mass. In the period 1996- 1998, the acoustics unit, headed by G. Vermeir, intensified its co-operation with the Entity of Acoustics and Thermal Physics of the Department of Physics within the Faculty of Sciences. This allowed a better usage of the extended infrastructure at the Physics Department. With the aid of the university, the industry and the Flemish government, the heat&mass unit, headed by H. Hens, realized a unique new construction for field testing, called the VLIET Test Building. In attendance of a new laboratory building, which should be constructed close to the principal building of the department, all other laboratory facilities are parceled over three locations.


1. Heat & Mass

Current research and consultancy activities are concentrated on the performance based design of highly insulated envelopes, risk analysis, durability and service life, low energy buildings and sustainable construction. The approach contains three levels: materials, building components and whole buildings.

At the material level, the measurement and modeling of the thermal, moisture and airflow properties and the analysis of damage development have the largest share in the overall activity. At the component level, the research activity focuses on the combined heat, air, moisture and salt transfer in highly insulated wall, roof and floor systems and its relation with durability of the outside cladding and finish. The methodology includes modeling, accelerated laboratory testing, hot box measurements and field testing. At the building level, energy consumption and sustainable construction are the main key interests. Here, field work in buildings plays a vital role.


Research Activities


1.1. Single Mode Nonlinear Resonant Acoustic Spectroscopy (SIMONRAS) for damage detection in quasi-brittle materials
Koen Van Den Abeele and Jan Carmeliet

Due to the complex compliance of micro-inhomogeneities, the elastic behaviour of brittle materials is manifest by strong nonlinearity, hysteresis in stress- strain relation, and discrete memory. In order to quantify the presence of mesoscopic features (micro-cracks, low aspect ratio pores, grain-to-grain contacts, etc) and damage in quasi-brittle materials, we developed a new tool for non-destructive testing, which focuses on the acoustic nonlinear (i.e., amplitude dependent) response of the material's resonance modes when driven at relatively small wave amplitudes. Granular and micro-cracked (or damaged) materials always show softening of the elastic modulus and generation of harmonics with increasing drive levels in resonance experiments at strains as low as 10-8. For example, the included figure shows the nonlinear resonance behaviour (measured acceleration versus frequency) of the first mode in the case of two slate beams. For an intact sample the resonance frequency is constant and there's no evidence of harmonics generated by the material. In a damaged sample the resonance frequency shifts significantly as a function of drive amplitude and the level of harmonics increases dramatically.

Figure 1.1: Examples of resonance curves and harmonics

The observed nonlinear behaviour is used as a diagnostic method to quantify the degree of damage in a material. The technique has been applied to various quasi- brittle (building) materials in order to discern damage due to mechanical, hygrothermal, chemo-mechanical loading. In all cases, we observed a rapid and significant increase of the nonlinear parameters early in the loading process and only a small to modest variation of the linear parameters (wave speed, attenuation). We are currently investigating the influence of heat-rain and frost-thaw cycles on the nonlinear behaviour of slate plates.


1.2. Nonlinear wave modulation spectroscopy
Koen Van Den Abeele

It is known that materials containing structural damage have a far greater nonlinear elastic response than materials with no structural damage. In order to discern damage in materials of arbitrary geometry, a technique called Nonlinear Wave Modulation Spectroscopy (NWMS) has been developed. This method specifically interrogates the nonlinear response of a sample by exciting it with continuous waves of two separate frequencies simultaneously, and inspecting the harmonics of the two waves, and their sum and difference frequencies (sidebands). Undamaged materials are essentially linear in their response to the two waves, while the same material, when damaged, becomes highly nonlinear, manifested by harmonics and sideband generation. The approach has proved to be time efficient and effective in discerning damage to materials in our experience. The method has been successfully applied to corrosion detection in reinforced concrete, investigation of delamination in bonded composites, and single crack detection in plexiglass, sandstone and automobile engine components (figure).

Figure 2.1: Examples of nonlinear wave modulation spectroscopy.

Nonlinear Wave Modulation Spectroscopy on an intact and damaged engine component. The two injected frequencies are 6.7 and 127.4 kHz. The intact sample does not show modulation, whereas the damaged sample has multiple sidebands (127.4kHz ( n*6.7 kHz).


1.3. Multi-scale network model for simulating moisture transfer properties of porous media
J. Carmeliet, F. Descamps, G. Houvenaghel

A multi-scale network model is presented to model unsaturated moisture transfer in hygroscopic capillary-porous materials showing a broad pore-size distribution. Both capillary effects and water sorption phenomena, water vapour and liquid water transfer are considered. The multi-scale approach is based on the concept of examining the porous space at different levels of magnification. The conservation of the water vapour permeability of dry material is used as scaling criterion to link the different pore scales. A macroscopic permeability is deduced from the permeabilities calculated at the different levels of magnification. Each level of magnification is modelled using an isotropic non- planar 2D cross-squared network. The multi-scale network simulates the enhancement of water vapour permeability due to capillary condensation, the hysteresis phenomenon between wetting and drying, and the steep increase of moisture permeability at the critical moisture saturation level. The calculated network permeabilities are compared with experimental data for calcium silicate and ceramic brick and a good agreement is observed.

Figure 3.1: Comparison of the liquid water filling distributions for wetting (a) and drying (b) at a capillary
pressure pc = 4.9 104 Pa. In drying a fingering effect of the dry pores is observed.


1.4. Continuum and discrete modelling of isothermal water and air transfer in porous media
F. Descamps

Water imbibition into porous building materials is a two-phase flow process in which the imbibing water displaces the air that is initially present in the pore space. Accurate non-destructive measurements of transient moisture content profiles during imbibition into calcium silicate brick indicate that the air outflow boundary condition has a predominant influence on the water imbibition process. Significant differences are observed in the water sorption coefficients and the inflow surface moisture contents of water imbibition experiments with free and bounded air outflow boundary conditions.

The time scale differences between the water and the air flow process and the limited place resolution of the non-destructive moisture content measurement preclude the simultaneous and accurate measurement of both transfer potentials. A direct calculation of the water and air transfer coefficients from an imbibition experiment is thus impossible.

Discrete modelling of the pore space is used as an indirect way to calculate these coefficients. Six parameters are identified that characterise the geometrical and topological properties of the pore space. The values of these parameters for calcium silicate brick are calculated so as to conserve the crucial water transfer properties of the porous medium: (i) the critical moisture content, (ii) the capillary moisture content and (iii) the moisture content dependence of the water diffusivity. Percolation concepts are used to incorporate these macroscopic water transfer properties into the discrete network model. In a second step, the one-phase mass transfer coefficients of water and air are calculated from the discrete network model using a stationary approach. The residual air saturation for different boundary conditions is calculated from a dynamic immiscible displacement approach.

Continuum modelling of the imbibition process using a two-phase flow approach and the calculated mass transfer coefficients confirms the experimental observations.


1.5. Development of a new methodology to analyse the durability of facade repair systems

H. Hens, J.Carmeliet

Brite-Euram
Proposal No.: BE96-3290 / Contract No.: BRPR CT96 0229
Starting date: 1 December 1996 - Duration: 36 months
Co-ordinators: H.Hens, J.Carmeliet, Katholieke Universiteit Leuven (BE).
Partners: Technische Universität Dresden (D), Helsinki University of Technology (SF), Eternit N.V. (B), Cape Building products limited (GB), Ahlborn Messtechnik GmbH (D), Cottbuser und Roxeler Ingenieurgesellschaft (D), Polar Construction ldt (GB), Ct Laastit Oy (SF), Rautaruukki (SF)

The main objective of the project is to establish an overall system of envelope repair and retro-fitting performances, associated design criteria, rules, practices and controls. The deterioration of building envelopes is strongly related to HAMST-effects (heat, air, moisture and salt transfer). The scientific objectives of the project are the enhancement of the basic HAMST-knowledge, the development of an extended database of material properties and testing procedures relevant for HAMST and a HAMST-software programme.

Project tasks:


1.6. Damage model for quasi-brittle non-saturated porous media
Jan Carmeliet

Porous materials like concrete are strongly hydrophilic. They have pores with a large specific surface. As a result, such materials exhibit strong fluid-solid interaction. In cementitious materials, the following microscopic fluid-solid interactions are documented in literature: capillary pressure, disjoining pressures due to hindered adsorption in the micropores, and pressures due to the presence of interlayer hydrate water in the nanopores of the CHS hydrate. The fluid-solid interaction induces high micro-stresses in the skeleton. The level depends on the degree of saturation. Therefore, changes in that degree may significantly affect the macroscopic poro-mechanical behaviour. We distinguish the following phenomena are: change of static and dynamic mechanical properties including damage and fracture, volumetric changes, time-dependent behaviour. Quasi-brittle materials are sensitive to micro-crack initiation, growth and coalescence of micro-cracks, which finally leads to localisation in the form of macro-cracks. The growth of micro-cracks not only induces a degradation of mechanical properties, but it also implies a change in the solid-fluid coupling coefficients.

Advanced plasticity and damage models have been developed to model strain softening behaviour of damaged materials. However, moisture dependence is often included in a simple way by considering the model parameters to be dependent on moisture content. When dealing with the complex interaction of different phenomena, such as damage, time-dependence and moisture behaviour, new constitutive equations, which formulate the coupling in an explicit way are required. In this study a poro-visco-elastic damage model is proposed. The constitutive equations introduce coefficients coupling elastic-damageable, viscous and poro-mechanical behaviour. The coupling coefficients are derived for the non-saturated case from a micro-mechanical model. The identified coupling coefficients account for the following phenomena: elastic behaviour of the undamaged skeleton, elastic opening of the existing micro-cracks, damage growth due to micro-crack development, permanent strains due to an imperfect closing of the micro-cracks and viscous behaviour of the undamaged skeleton. An incremental finite element model is formulated. The model is used to simulate the poro- mechanical response for standard experiments. In the uniaxial tensile test, the influence of drained and undrained boundary conditions on the damage behaviour is analysed. Also the influence of damage and viscous effects on the swelling behaviour of porous materials is studied.


1.7. Inverse identification techniques: optimal estimation of gradient damage parameters from localization phenomena in quasi-brittle materials
Jan Carmeliet

Non-local and gradient enhanced damage models are able to properly model localisation phenomena in quasi-brittle materials. By the introduction of an internal length scale they avoid mesh-size sensitivity in finite element calculations and are capable of describing size effects. To experimentally determine the internal length scale and other damage model parameters, indirect identification methods have to be used. The method presented is based on the Markov estimation procedure, which also permits to evaluate the accuracy and the well-posedness of the identification problem, characterised by the uniqueness and identifiability of the estimated values. This information offers the possibility to compare and optimise the designs of laboratory tests in order to minimise the effect of errors on the estimated values. The indirect determination of the gradient damage parameters based on global response measurements far from the crack on one-size specimens is sensitive to the problem of ill-posedness due the high correlation between the gradient damage parameters, which dominate the localisation process. The size effect method can be used as an accurate identification method for the gradient damage parameters, when information on the tensile behaviour is included and a sufficient size range is considered. Local displacement analysis close to the crack provides sufficient information with respect to the identification problem independently of the size of the specimen and gives the most accurate parameter estimations.


1.8. Water and salt transfer in porous building materials
Geert Houvenaghel, Jan Carmeliet

Salt-induced decay processes are one of the major reasons for the decay of stone buildings and monuments. Salt attack is due to the combined action of hygric, thermal, physical and chemical factors. Efflorescence only causes aesthetic damage. Hydration and crystallisation of salts in the pore system however cause an increase in volume and a pressure build-up, which may lead to damage.

The experimental part of the research concentrates on the changes of the hygric properties of a material caused by the presence of salts in the pore structure. These influences are investigated on a classical red clay brick. At relative humidities, higher then the equilibrium relative humidity at salt saturation the moisture content increases and the hygroscopic curve is lifted up. This results in higher moisture content and a more difficult drying-out of the material. The moisture transport properties of the material change due to salt presence: salt solutions have a higher viscosity and another contact angle, density and surface tension than water and the pore structure is changed due to the presence of salts. The diffusion-dispersion coefficient of salts is calculated from a break-through column experiment.

The main part of the research focuses on the modelling of these salt phenomena: the coupled heat- and moisture transfer, diffusion of salt ions due to a concentration gradient, the transport of soluble salts, crystallisation and dissolution of salts, hydration of salts. The influence of the salt concentration on the moisture properties causes a non-linear coupling of both transfer processes. A finite element code has been developed in order to model the combined heat- and moisture transfer in porous materials. The salt transfer equation is a convection-diffusion equation. The accurate modelling of the interaction between convective and diffusive processes coupled to sharp-front transport problems necessitates special numerical techniques. A stream-line upwind Petrov-Galerkin technique is used to tackle the numerical problems. The combination of element free Galerkin techniques (EFG) with classical numerical techniques for convection-diffusion problems is investigated. The EFG method only requires nodes and no element connectivity, so it allows easily a mesh enrichment at sharp moving fronts.


1.9. Service life prediction of building materials and structures
W. Depraetere (FWO), S. Roels (OT), J. Carmeliet, H. Hens

FWO-project + OT-project

Service life prediction is an important topic within the larger research area of Life Cycle Assessment of buildings. LCA intends to develop rationales for an optimisation of building materials and components with respect to their energetic, economical and environmental qualities.

The service life of a building material or a structure is the space of time during which it fulfils all functional and performance requirements imposed. The scatter in observed service live is generally high due to the complex deterioration mechanisms involved and the random nature of environmental conditions and structural properties. Consequently, there is an increasing tendency to a probabilistic approach, called reliability analysis. In this context the service life is defined as the probability of failure at any stage during life. The analysis is based on a systematic methodology including the following steps:

The project is dedicated to the development of the necessary numerical simulation capabilities and the experimental equipment and procedures, which should allow an in-depth investigation of the novel design philosophy. In the project the service life prediction is applied to quasi-brittle porous media such as cementitious materials, ceramics and natural stones, focusing on the combined deterioration processes related to hygro-thermal load and chemical attack.

This research should result in 2 Ph.D.'s: 1) moisture transfer at interfaces of combined porous materials, carried out by W. Depraetere within the framework of the FWO-project 2) a stochastic approach to heat and moisture transfer in heterogeneous porous media, carried out by S. Roels within the framework of the OT-project.

1) Moisture transfer at interfaces of combined porous materials:
The moisture behaviour of combined porous materials is highly determined by the interaction of the moisture characteristics of both materials and the nature of their hygric contact. Dependent on the porous structure of the different composing materials, water transfer may occur from one material to another material, but not in the opposite direction. Four types of hygric contact can be distinguished: hygroscopic contact (a small air layer), natural capillar contact (a small crack), ideal and real capillar contact. Due to physical and chemical processes (during curing), one material can influence the porous structure of the other one at the interface of a real capillar contact. This often results in a boundary layer at the interface with a high hygric resistance. Also the presence of air cavities in the interface zone may disturb the hygric contact. All these complex phenomena can significantly change the moisture behaviour of combined porous materials. Research at the laboratory has been done on combinations of brick and mortar. A finite element-model is developed to simulate the isothermal moisture transfer in combined porous materials including different contact conditions.

2) A stochastic approach to heat and moisture transfer in heterogeneous porous media:
A 2D Finite Element Model is developed to simulate the transient heat and moisture transport in open porous media. The moisture transfer is characterised by high moisture gradients at the waterfront during water imbibition or drying. For an accurate description of the steep gradients very small elements are needed in a finite element analysis. Due to the moving water front a fine mesh over the whole domain would be necessary. To reduce computational costs a mesh adaptive method is implemented. The method allows to uncouple nodes from both material particles and spatial points and to continuously relocate the nodes around the sharp water front. Hence, the number of elements can significantly decrease. The influence of material properties heterogeneities on heat and moisture transfer is investigated on Savonnières, a French, layered limestone. To study the spatial variability of the transport properties and the moisture retention curves, a plate of limestone has been subdivided in small samples.

Measurements of density, porosity, capillary moisture content, water absorption coefficients and moisture retention curves are performed on these samples. Differences in material properties are linked to the micro-structural composition obtained by scanning electron microscopy (SEM). With all the information a stochastic random field-model of the material properties is developed to come to reliable 1D- and 2D-simulations.


1.10. The influence of water repellant agents on moisture transfer properties of porous media
K. Bertels, C. Van den bossche, J. Carmeliet

IWT-project

Nowadays, a water repellent treatment is part of almost every restoration or renovation project, and even a lot of new constructions are treated with a hydrophobic protection. The use of water-repellents can indeed be very beneficial, both to prevent various types of deterioration and to cure certain damages. Treatment of masonry with hydrophobic agents makes it less penetrable for rain water which is often composed of aggressive constituents. A lowered moisture content in the facade reduces biological growth, efflorescences and frost damage.

Water repellent treatments suppress the capillary forces which suck water into porous media. Rather than blocking the pores, ideal water repellents only influence the surface tension at the pore walls. Water repellent agents contain an active ingredient, dissolved in an organic solvent or emulsified in water. The active ingredient is in most cases based on silicon: silicon resins, siloxanes and silanes. In order to avoid inhomogeneous treatment, often achieved by brush or airless application, samples are treated in the laboratory by means of a surface contact with the water repellent mixture at a fixed contact time. The ideal penetration depth varies between 2 and 10 mm.

The effect of siloxane impregnation on the moisture behaviour of clay brick and sand-lime brick has been examined in comparison with untreated substrates. Important hygric parameters are investigated: the change in water vapour conductivity due to the surface treatment, the water retention curve, the resulting liquid diffusivity's for water absorption, the capillary saturation point and the drying behaviour of the masonry after an effective impregnation.

The test shows that a siloxane treatment, if properly applied, repels rain water quite effectively. The question, however, remains whether the experience gained on rather small test elements in laboratory conditions can be translated to the application on a real building.


1.11. The influence of the coupling between heat and moisture transfer on the heat transfer between buildings and the ground
H. Janssen, J. Carmeliet, H. Hens

IWT-project

Forty percent of the total annual energy consumption in Europe goes to the heating, cooling and use of buildings. In the efforts to decrease our global energy consumption, buildings cannot be ignored. The design and application of appropriate measures however demands a thorough knowledge of the heat transfer mechanism between a buildings and the environment. The heat loss from buildings to the ground is one of these mechanisms.

The largest share of the currently existing research on heat transfer between buildings and the ground considers the thermal properties - conductivity and capacity - of soil to be constant. Furthermore, a simplified heat balance at the surface generally is applied. In most cases, only convection is considered, ignoring phenomena as radiation, evaporation and others. While this approach may have several advantages, it ignores the important coupling between heat and moisture transfer in soil, and the possible influence of phenomena other than convection on the surface temperature. For instance, the moisture content of the soil is the predominant factor in the thermal conductivity; ground water flow causes enthalpy transfer; evaporation may lead to a substantial damping of the surface temperature.

The aim of this doctoral project is the evaluation of the influence of the coupling between heat and moisture transfer and of the complete heat balance at the ground surface, on the heat loss from buildings to the ground. Therefore, a three- dimensional finite element model for coupled heat and moisture transfer in soils is under developement, based on the description of Milly, whom further elaborated the equations of Philip and De Vries. Complementarily, the complete heat and moisture balance at the surface will be implemented. Comparative calculations with and without coupling between the temperature and the moisture field and with and without the complete heat balance at the surface, should allow conclusions on the importance of the phenomena under consideration. After that, the thermal performance of some standard insulation strategies for building parts in contact with the ground will be reassessed with the coupled model. Perhaps. the effect of the hygric behaviour could bring new insights.


1.12. The VLIET test building: field evaluation of highly insulated building components
A. Janssens, A. Morel, W. Depraetere, H. Janssen, S. Roels, J. Carmeliet, H. Hens

VLIET-project

This research project dealt with the construction and management of a test facility for the investigation of highly insulated building envelope systems exposed to the outside climate. The facility allows to test twenty wall and twenty roof components simultaneously. Eight building product companies, IWT and the university participated in the project. The companies had their materials and systems tested during the first measuring campaign (1996-1998).

The main objective of the first campaign was to investigate whether a better than normal thermal quality is achievable with current construction practices in Belgium. Therefore the thermal and moisture performances of following building components have been monitored in detail: cavity walls, walls with fibre cement cladding, duo-pitched tiled roofs, low slope roofs with zinc sheeting and floors on grade, all with design U-values of 0.2 W/(m2K).

The measuring results revealed physical and technological difficulties to achieve the designed thermal quality in reality. The effective performance of the highly insulated systems was mainly affected by air gaps and thermal bridges characteristic of existing construction methods. For instance, the measured thermal permeance of some tiled roofs increased up to 200% of the predicted value due to wind-driven air infiltration through vents and joints in the underlay. Measurements on the cavity walls showed the sensitivity of the performance of these systems to the workmanship quality of the insulation: the effective thermal permeance varied from 0.21 to 0.86 W/(m2K)), depending on the size of air gaps around the insulation boards. The results further indicated that highly insulated envelope systems need improved and tailored methods for moisture control to ensure a good durability.

The results of the first measuring campaign generated the following rules for the design and construction of envelope systems with high thermal quality and safe moisture performance in the Belgian climate: (1) continuity of the thermal insulation layer to prevent air rotation, (2) air barrier at the inside of the thermal insulation to prevent air exfiltration, (3) vapour permeable air retarder at the outside of the thermal insulation to prevent wind-washing and allow drying and (4) adequately designed structural elements to limit thermal bridging.


1.13. International Energy Agency, Annex 32: Integral Building Envelope Performance Assessment
H. Hens & F. Ali Mohamed

Annex 32's overall objectives are to optimise building envelope performance and to minimise risk for envelope failure. The desired level of envelope performance follows from an integration of the building users requirements with the societal requirement of lessening the building impact on overall energy consumption and environmental load. To achieve both objectives, a methodology has been conceived which provides the design and construction professionals with the elements needed to fill in the gaps in present practice. The methodology includes a framework for identifying in a systematic and hierarchical way the performances embedded in the design criteria. She also helps in matching the relation between the performance requirements specified and the design solutions. The verification of the performances requirements is a core process. Annex 32, therefore, provides a package of design verification tools, design guidelines as well as the tests needed for verifying the constructed envelope. The Annex contains two subtasks:

Subtask A: Building Envelope Performance Assessment Methodology
The objective of this subtask is to develop the Assessment Methodology. The methodology focus on ways to apply new concepts and making use of existing (inter-)national experiences and guidelines. It is of use in the different stages in the design process, and takes into account the relation between the envelope, the climate, the building and all services.

Subtask B: Evaluation of the Assessment Methodology and Design Tools
The objectives of this subtask are to evaluate and refine the Methodology developed; and, to evaluate design tools and guidelines which contribute to a lower energy consumption and a high quality thermal and visual indoor comfort. The activities are divided between three Thematic Groups: (1) Envelop Laboratory & Modelling Characterisation Group, (2) Advanced Envelopes Group, (3) Retrofitting and Traditional Envelopes Group.

National teams involved: Belgium (Operating Agent), Canada, Japan, Greece, Italy, Denmark, Germany, Finland, France, The Netherlands, UK and USA.

Internet address: http://www.bwk.kuleuven.ac.be/bwf/projects/annex32/frameanx.htm


1.14. Reliable control of interstitial condensation in lightweight roof systems
A. Janssens, H. Hens

The general objective of this Ph.D.-research was to develop calculation and assessment methods to improve the reliability of condensation control systems in lightweight roofs, considering the uncertainty to achieve continuity of air-tightness in building practice. An assessment method has been developed based on a stochastic approach to moisture performance analysis and on concepts of industrial risk analysis. To predict the thermal and moisture performance of lightweight systems, a two-dimensional transient model has been developed for the combined heat, air and water vapour transfer in building components. The model allows for a two-domain description in terms of porous media and air channels.

The calculation method has been applied to produce a better understanding of the effects of air movement and discontinuities on the performance of lightweight roofs. The development of the assessment method involved the definition of design climate values for the evaluation of condensation risk due to air leakage, the definition of limit state values to assess the risk and the use of `redundant' protective measures to reduce the risk. The research has produced tools and recommendations for the design of lightweight roof systems with reliable control of interstitial condensation.

Figure 14.1: Predicted velocity and temperature distribution in a 60 degree-sloping roof with air gaps: global velocity field (left), porous field (middle, scale x15), condensation profile (right)


1.15. Transparent insulation: an alternative solution for the summer discomfort
G. Verbeeck, H. Hens

VLIET-project

By applying transparent insulation materials (TIM) as an external insulation, as a glazing system or as a solar collector, the energy losses in a building are reduced and even switch to energy gains by passive and/or active use of the incident solar radiation. Characteristically, because of the transparency and the structure of the material, solar radiation is transmitted but convection in and conduction through the material suppressed, so that the heat losses towards the surroundings are reduced.

From June 1995 until June 1998 a research project on transparent insulation went on in co-operation with the Group for Mathematical Analysis of the University of Gent. In the project an early developed 1-dimensional model for the transient heat transfer through TIM was upgraded and extended to a 2-dimensional model for different applications of TIM in buildings. The model can be plugged in as an extra component in the ESP-r simulation program for transient heat transfer in buildings (from the University of Strathclyde, Glasgow). The model has been validated through measurements in situ on the TIM-walls of the VLIET test building. Three different TIM-walls were designed: a traditional wall with TIM as external insulation, a TIM-wall with natural convection in the air gap between the TIM and the massive layer and a third wall, equal to the second, but with extra mineral wool at the cavity side of the massive layer. From earlier research we knew that the traditional TIM-wall without any solar shading gives substantial energy gains during the heating season, but also an unacceptable indoor climate during summer. As traditional outdoor solar shadings are quite expensive, this research was made to examine alternative and cheaper solutions for the summer discomfort: natural ventilation of the air gap between TIM and massive wall and/or extra insulation of the massive wall with opaque insulation at the cavity side of that wall. From the analysis of the winter situation the ventilated TIM- walls seems to be energetically equivalent to the traditional TIM-wall. The lower conductive gains were compensated by the heat which the ventilation air could pick up in the cavity. The summer analysis showed a larger difference between the walls. Stack induced ventilation of the air cavity in wall 2 lowered the indoor temperatures during summer quite a lot, but not enough. Only wall 3, stack induced ventilation in combination with an extra opaque insulation, achieved an acceptable indoor climate in summer.


1.16. Performance-based assessment of active envelopes
D. Saelens, H. Hens

IWT-project

Today, the development of new technologies to improve building envelope performances is highly encouraged and provides a clear challenge for designers and researchers. In this context several typologies of active envelopes have become very popular amongst designers and architects. Typically, active envelopes consist of two panes with in between a cavity, through which air is drawn by means of forced or natural convection.

This research tries to nuance the common favourable impression which exists about active envelopes regarding energy efficiency. Although active envelopes may introduce certain benefits for certain performances, this cannot be generalised for all performances. This research compares the energy efficiency of several active envelope technologies to common solutions.

Analytical and numerical models will be developed and compared to experimental data. The models will be used in energy simulation programs to compare the energy consumption on building level of active envelopes to common choice solutions. The performances of the airflow window of the DVV-Headquarters building in Brussels and a measurement set-up in the Vliet- test-building, Leuven are the subject of a performance-based assessment. The DVV case study revealed that especially bad workmanship and a wrong choice of typology dramatically decreases the achieved performances.

Only if active envelopes save energy on building level they can compensate for the increased built-in energy. A life cycle energy analyses will point out whether the efforts don't lead to disappointing small results causing the overall energetic, economic and ecological performance to be very discouraging.

If, finally, active envelopes prove to be a favourable choice concerning energy efficiency during the whole life span, a set of rules of thumb will be developed to help designers predict the performance of their solutions during design and to control the achieved performances during and after construction.


1.17. CO2-project: Energy use and emissions in the residential sector
G. Verbeeck, B. Verdonck, H. Hens

Electrabel, SPE

Like all industrialised countries, Belgium accepted to diminish its greenhouse gas emissions. Electrabel and SPE started an extended research on that in October 1996, involving researchers from all universities and scientific institutes. Aims of the project are (1) to catalogue the energy consumption and emissions for electricity production in all type of power stations, (2) to evaluate energy consumption in the residential and industrial sector and (3) to analyse the impact of energy efficiency measures on the magnitude of the emissions. The section Building Physics, together with other labs of KULeuven and UCL, focuses on the residential sector.

The first part of the research included an analysis of the annual energy consumption measured in 200 parental houses of students. Based on the database a model was developed to predict energy use for household and lighting. An earlier developed model (VERBRUIK) to calculate energy consumption for heating in a building was integrated in a bottom-up model VERBCO2, which predicts the overall energy consumption and related emissions in the residential sector on the scale of a city, a district, a whole country. The model uses the statistical information on the Belgian building stock of NIS (National Institute for Statistics). After validation we predicted the impact of different energy conservation measures for the whole housing stock in Belgium. Three policy scenario's were considered: (1) restricted retrofit, unlimited expansion, (2) more retrofit, less expansion, (3) more retrofit, no expansion beyond 2010. In a second part the database was extended with information on the annual energy consumption in 1200 social houses and 40 low energy dwellings. The measured and calculated energy consumption in all 1440 dwellings was used to further validate the bottom-up model.

The project is still going on, with a focus now on increased energy efficiency of the heating system and optimal measures for the building as a whole (including LCA). Conclusions after the first two year are that although the residential sector may contribute to the CO2-reduction, major efforts will be needed to reach the Kyoto-objective. Increased energy efficiency of new buildings and retrofits is not sufficient. Also the housing policy should be reviewed, with more emphasis on retrofitting than is the case today. Even then, reductions in emissions will proceed slowly, as it takes at least 70 years before most of the country's housing stock is renewed.


1.18. Optimisation of energy systems in residential buildings, a life-cycle based analysis
B. Verdonck, H. Hens

IWT-project, SPE-Electrabel

As the concern about global warming grows, countries around the world, including Belgium, make efforts to reduce the burden upon the environment. The Kyoto-agreement aims at a significant reduction in CO2-emissions. As residential buildings are responsible for at least 25% of these emissions, it is important that the residential sector contributes to the reduction. To minimise the cost of this contribution and to prevent any deteriorating of comfort levels that might result from it, it is necessary to look for optimal solution.

A thorough knowledge of the energy consumption in buildings and its dependency of envelope, habitants, domestic appliances and heating equipment is a precondition for the study of the energy related aspects in the life-cycle assessment of residential buildings. As one of the elements within the strategy to gain a better understanding of residential energy consumption, fifth year civil engineering students and fourth year architecture students, attending the lectures on building services, were asked as part of the exam to perform an analysis of the energy and water consumption in their homes. These data were used later to develop tools for estimating the energy consumption for heating, warm water supply and domestic appliances.

These estimates form the boundary conditions of the optimisation problem. The goal function reflects the optimisation criteria. The criteria can be economic, environmental or energy related. The ultimate goal is to construct a flexible tool for architects and designers in order to be able to make optimal choices in developing a project. The best solution depends after all heavily upon the boundary conditions. These tools are not developed yet, but the results have been used to evaluate governmental policies (insulation regulation, subsidies,..) and strategies of energy companies.


1.19. EpiGoon
H. Hens, G. Verbeeck

Vliet-bis project

After the energy crises of the seventies, most industrialised countries adopted mandatory standards for the thermal insulation of buildings. The first generation of codes focused on the U-factor of the different envelope parts. For each part, a maximum value was imposed. A second generation used the overall U-factor for the envelope as indicator. This factor was calculated as an area weighted average of the U-factors of all individual parts. Fixation of a threshold value was based to principles such as: loss by conduction should be a constant per cubic meter of heated volume. This ended in reference curves as given in the figure (Flemish decree, 1993).

Figure 19.1: Overall U-factor as a function of compactness, i.e. the ratio between heated volume and encapsulating surface

At the end of the eighties the goal to lower energy consumption was bypassed by an ever growing environmental concern on global warming and ozone depletion. This lead to decreasing U-factors. The U-factor however governs only part of the energy consumption in buildings, the one due to conduction losses and gains. Ventilation, hot water production, lighting and appliances, solar gains, are not covered at all. The lower U, the lower the conduction losses and the less their share in the overall energy consumption. With other words, restricting the regulations to U-factors has a decreasing impact on total consumption with increasing severity of the value imposed. This tended the developed world to adopt a new approach: the energy performance. This number represents the allowable standardised energy consumption per m2 of heated surface, per m3 of heated volume or per m2 of encapsulating surface. The project aims to develop a energy performance standard for Flanders. Partners are: WTCB, Physibel, Daidalos, Wenk.


Industrial Projects & Consultancy

During the period 1996-1998, the laboratory proceeded through 99 smaller and larger industrial projects and demands for consultancy. These projects and demands concerned (1) extended measurements of hygro-thermal material properties, (2) analysing and upgrading the performance of new designs for roofs, walls and windows, (3) evaluation of industrial processes (f.e.: accelerated drying of wet formed judo mats), (4) damages on roofs and walls (among them the Opel industrial plant in Antwerp, a subtropic swimming pool in Herfort (Germany), the new metro station near the Erasmus bridge in Rotterdam, several cases of exterior insulation in the Netherlands, a huge fitness centre in Limburg, a modernist school building in Vlaams Brabant, dwellings in Belgium and abroad (i.e. Berlin), a museum in Gouda, the Netherlands. Especially metallic roofs entered the damage files. These files were of great help in feeding the doctoral study of Arnold Janssens), (5) low energy dwellings and office buildings, (6) the MIRA environmental reports Flanders, (7) large scale energy consumption prognosis for the tertiary sector in Flanders, (8) the competition 'Zonnewoningen' of Electrabel, where the lab. was involved in a technical evaluation of all participating projects

Each project ended with an extended report. These reports can be used for research purposes.


Building and Room Acoustics

Since its start in 1978, the Laboratory of Building Physics has been involved in research and educational activities in the field of acoustics in the Faculty of Applied Sciences. There are courses on building acoustics, room acoustics and noise control. The lessons are taught in Dutch and are attended by civil engineering students of the building construction and architecture departments, and by postgraduate students in environmental engineering.

The research activity concerns the sound insulation of sandwich panels, the flanking transmission in building structures, room acoustical prediction and measurement methods, including applications in the field of indoor noise control, and the study of the acoustic comfort in buildings.

In the research on sound insulation of sandwich panels, the basic phenomena related to airborne and structure-borne sound transmission through layered plates are studied. In the theoretical modelling, panels of infinite and finite extent are considered. The model of an infinite panel allows the study of the influence of the different wave types in the layers, of the boundary conditions between the layers and of the mechanical properties of the layer materials. The model of a finite panel allows the study of the modal vibrational response and the sound radiation of the panel. The experimental study relies on a combination of vibration and sound intensity measurements. These experiments are complemented by the measurement of the elastic parameters of the materials.

In the research on flanking transmission in buildings structures, the vibrational sound transmission at junctions between plates is studied. In the theoretical modelling, plates of infinite and finite extent are considered, and different joint designs including rigid, pinned and elastic joints are studied. The experimental study is performed on a scale model of a plate structure, using vibration and structural intensity measurements.

In the room acoustics research, methods to predict, measure and control the room acoustical properties in closed spaces are studied. The prediction method is based on cone/ray tracing, and allows the control of most relevant room acoustical quantities for the case of single and multiple sound sources. By comparison with analytical models and experimental results, the validity of the model is investigated. The model has been applied in the design of lecture and concert halls, and in the study of the sound propagation and sound reduction in factory halls. The experimental room acoustics research focuses on the development and the use of signal analysis methods that allow for an accurate and repetitive measurement of the room impulse response and related quantities.

In the study of the acoustic comfort, the relation between environmental sound sources, the sound insulation of building constructions and the (dis)comfort resulting from the residual sound exposure is investigated. The purpose of the research is the establishment of guidelines to realise sufficient acoustic comfort indoors and in the direct surrounding of the building.

Finally, the Laboratory is involved in industrial projects including the measurement of material and construction properties, room acoustics and noise control consultancy, and environmental noise studies.

Preparing the future

Since 1996 great effort has been delivered for the preparation of a merging between the acoustics group of the Laboratory of Building Physics and de Laboratory of Acoustics and Thermal Physics. This resulted in an agreement by the Directory Council of the University in 1997. A budget was attributed for the construction of an extension to the laboratory facilities. The construction will start in august 1999.

Since 1998 the already long lasting collaboration was intensified, resulting in several research, development and consultancy projects.


Research Activities


2.1. Analytical modelling of structure-borne sound transmission and modal interaction at complex plate junctions
I. Bosmans

The investigations at the Laboratory of Building Physics were concerned with the improvement of the accuracy and applicability of Statistical Energy Analysis for predicting airborne and impact sound insulation in buildings. In view of the importance of flanking transmission, which is largely determined by vibration transmission between walls and floors, accurate SEA predictions require accurate coupling loss factors describing the energy flow between plates. The research at the Laboratory of Building Physics aimed at enhancing SEA, by extending the state-of-the-art of calculating coupling loss factors to junctions of orthotropic and point connected plates. While most of the work was concerned with calculating coupling loss factors using the wave approach for semi-infinite plates, much attention has been paid to the evaluation of the restrictive assumptions inherent to SEA by means of a semi-modal superposition technique for finite-sized plates.

The scientific results were obtained in the context of a Ph.D. research. The following paragraphs will summarize the various conclusions while relating the results to the publications generated during the research project.

Structure-borne sound transmission between orthotropic plates
The wave approach for semi-infinite plates and the semi-modal superposition technique for finite-sized plates were applied to junctions of thin orthotropic plates. For this purpose, wave propagation in thin orthotropic plates was investigated by deriving analytical solutions for the equations of motion. The analysis emphasised on the influence of orthotropic plate characteristics on the description of bending and in-plane wave response and on the energy flow associated with plane wave propagation. The calculation model based on the wave approach was integrated in an SEA calculation scheme by deriving expressions for the modal density and the coupling loss factor for subsystems associated with orthotropic plates. The influence of the degree of orthotropy on the structure- borne sound transmission was studied by a parametric calculation on a range of corner, tee- and cross-junctions. The calculation models were verified experimentally on three corner junctions between two grooved plywood plates.

Structure-borne sound transmission between point connected plates
The theory of the wave approach and the semi-modal superposition technique was extended towards the case of point connected plate junctions. Local rigid connections were modelled by introducing an elastic interlayer with a space dependent stiffness. The analysis was concerned with the formulation of the boundary conditions and the implications of these boundary conditions on the solutions to the equations of motion. A parametric calculation for a variety of point connected plate junctions demonstrated the influence of the location, the dimensions and the elastic properties of the coupling elements on bending and in-plane wave transmission. The predictive performance of both calculation models was validated on a set of corner junctions between two PVC plates.

Validation of Statistical Energy Analysis using semi-analytical calculation models
Finally, it was shown that the reliability of SEA can be evaluated by comparing results of SEA and the semi-modal superposition approach. It was found that SEA predictions of structure-borne sound transmission between plates with different properties are unreliable under the condition of low modal overlap. In addition, the accuracy of SEA increases with increasing modal overlap factor, but a tendency towards overestimation of transmission remains, even for a modal overlap factor exceeding unity. The experimental verification (illustration) demonstrated the ability of the semi-modal superposition approach to determine whether discrepancies between SEA predictions and measured data are caused by a violation of the basic SEA assumptions, or by an incorrect model of the junction.

Figure 2.1: SEA experimental verification


2.2. Sound insulation of composite walls
F. Nuytten( +1997) (IWT), I. Roelens, G. Vermeir (FWO Support)

The purpose of this project is to study the sound transmission through double wall constructions with local coupling elements. The modelling uses the wave approach and results obtained with a description of the coupling by introducing local stiffnesses for differential movement of both blades for translation and rotation (see figure).

A holographic measuring set-up is developed for the measurement of the sound transmission on scale 1:4. The test-wall is placed between two rectangular reflecting concrete boxes. The sound field is scanned in a plane close to the radiating wall (see figure).


2.3. In-situ determination of the elastic properties of building materials
I. Roelens, F. Nuytten( +1997), I. Bosmans, G. Vermeir

In this research we looked for methods to measure the stiffness properties of beams and plates. This work resulted in three measuring methods in the range from 100 to 4000 Hz. The first technique the bending wave velocity for acoustically thin beams and is based on the finite difference approximation of the bending wave-number. The other two methods involve the measurement of the plate response in two points in order to determine the bending wave velocity for acoustically thin as well as for acoustically thick plated. Both these methods are convenient to estimate the angle dependent wavespeed of orthotropic plates. The three measurement methods have been verified by experiments on beams and plates of various materials and thicknesses.

Figure 2.3: Measurement set-up used in the experiments on the acoustically thin PVC beam
according to the first measurement technique.


2.4. Low-frequency airborne-sound transmission through single partitions in buildings
A. Ossipov, P. Mees, G. Vermeir
(IWONL, WTCB, CEDIA U.Liège, K.U.Leuven)

The low-frequency (20-250 Hz) airborne sound transmission of single partitions is investigated. Three theoretical models are used for the prediction: an infinite plate model, a baffled plate model and a room-plate-room model.(illustration). The calculation models are verified by detailed comparisons with experimental results obtained in the laboratory. A parametric study is carried out to examine the influence of the dimensions of the room and the partition. The results demonstrate the strong modal behaviour of the low-frequency sound transmission. As a result, the low-frequency sound insulation depends not only on the properties of the test wall, but also on the geometry and the dimensions of the room-wall-room system (see illustration for a measurement and calculation of a test on a window pane). In general, the large variation of the low-frequency insulation compromises the accuracy of a prediction of the sound insulation in situ.

Figure 2.4: CSTC/KUL transmission rooms. Measured SRI and transmission loss (TL).


2.5. Room acoustical simulation: further development and research
L. De Geetere, G. Vermeir, P. Mees

The prediction of room acoustical properties and sound propagation in closed spaces is a fundamental topic in architectural acoustics and indoor noise control. The problem essentially concerns the calculation of the sound field in a closed space with sound-absorbing boundaries, possibly fitted with arbitrarily positioned objects, and excited by a single or multiple sound sources. An exact solution of this problem without excessive computational effort is only feasible for idealised room geometry's and boundary conditions, or for a limited frequency range. However, it is not strictly necessary to calculate the deterministic room response in order to evaluate room acoustical quantities related to perception or acoustic comfort. For the study of the acoustics of general room shapes over the whole audio frequency range, approximate methods can provide satisfactory results. In this study, the application of an approximate model known as the cone/ray tracing method to the solution of the problem is investigated.

The input data of the model include the definition of the room geometry as an assembly of plane surfaces; the position, directivity and strength of the sound sources for different frequency bands; the position of the receivers; the sound absorption and diffusion coefficients of the room surfaces for different frequency bands. The geometry of a large lecture room is shown in figure 1. In the cone tracing method, a number of image sources related to the actual sound sources are calculated by scanning the interior room volume by means of a small 'light-beam'. The contributions of the individual image sources sum up to the global room response at a receiver point. In the ray tracing method, it is assumed that the wave front emitted by a source can be digitised as energy particles and that wave propagation can be simulated by propagating these particles through the room along straight lines or rays. The sound radiation of each source is modelled by emitting a large number of rays. In both methods, specular or diffuse reflection and absorption on the room boundary surfaces is taken into account. The calculations finally result in the impulse (sound pressure) response or the energy (sound pressure squared) response for the different frequency bands at each receiver position and for each sound source. From these detailed calculation results, other room acoustical quantities are derived: the sound pressure level, the reverberation time, the speech intelligibility, ... Digital convolution of the broadband impulse response with anechoically recorded speech or music allows for an auditory assessment of the room acoustics quality.

The development of the corresponding EPIKUL software was continued. The evaluation is done in the context of thesisworks and consultancy work. The program was applied in the consultancy of the project `Grande Aula' of the UCLouvain (project in collaboration with P. Mees, Daidalos, architect Ph. Samyn, see 3D view of the simulation model) and for the consultancy concerning the retrofitting of the famous concert hall Henri Le Boeuf of the Paleis voor Schone Kunsten / Palais des Beaux Arts (architect Baron Victor Horta, 1929, architect of the renovation: G. Baines). Auralisation facilities where developed and applied in order to be able to listen to the predictions.

Figure 2.5: 3D view of the "Grande Aula", UCL, Louvan-la-Neuve.


2.6. Study of the reflection and diffusion of sound using acoustic holography. Application to room acoustics
W. Lauriks, G. Vermeir, G. Janssens, L. De Geetere
(project supported by the research council of the K.U.Leuven, 1998).

This project, started in 1998. It concerns the modelling of porous materials, the application in the room acoustical modelling and the development of holographic measuring techniques. The fundamental work focuses on the study of surface waves which is of particular interest for long range outdoor sound propagation and for material research.

The experience in this field will be used to improve the quality of the input data for the acoustical simulation program. This is necessary in relation to the auralisation of the first reflections and for understanding and characterising the propagation of the `grazing' sound wave over the seating area. A laboratory set-up is developed for the application of the near-field acoustic holography (see illustration). Acoustic holography refers to the reconstruction of a sound field in three dimensions based on the recording of the sound field on a two dimensional plane (hologram). It has been used whenever there is a need to visualise objects that cannot be observed optically. Although the mathematical principles are identical to optical holography, there are less technical restrictions to acoustic holography. The reason for this is that the phase of the (optical) field can only be recorded by exploiting interference with a reference beam (due to the extremely rapid oscillations of the optical wave), whereas amplitude and phase of the acoustic field can be recorded with great precision with microphones (enabling even broad band recording). Since the data can be digitised, there is no need for a diffraction pattern like in optical holography to reconstruct the field or image. Instead, the acoustic field can be reconstructed using numerical techniques. From the Fourier transform in wave number spectrum the inhomogeneous plane wave components can be determined and the propagation of surface waves can be analysed.


2.7. Distributed construction of a virtual science museum with guide, application to education in acoustics
G. Vermeir, V. Meerbergen, I. Bosmans
(project supported by the Flemish Community in collaboration with 4 partners: RUGent INTEC, VUB Werktuigkunde en Akoestiek, KULeuven, Laboratorium Bouwfysica KAHO St.Lieven Gent)

The aim of this project is to establish a co-operation between acoustical experts of universities and high schools in order to develop a multimedia learning environment. This learning environment will be build similar to a science museum that students can visit alone or together with a guide. The museum will be virtual and will contain a database of high-quality multimedia materials about acoustics. The multimedia materials in the museum will be created by the specialists in the specific fields of acoustics and using internet tools these materials can be held up to date by those experts. The guide is the instructor of the students and still has important didactical tasks, like focussing the students' attention to the most important aspects, instructing the students how to use the multimedia for effective learning, etc.

To realise this project, we focus on:

  • a guide-book containing style rules, technological and didactical requirements for the materials created by the project partners;
  • effective working principles for co-operation;
  • obtaining experience with distributed development;
  • co-operation with new groups in Flanders and in other countries;
  • an evaluation of the extra didactical value of multimedia materials for education in acoustics.

    Specific goals for the building acoustics research group
    Offering visual and auditory illustrations for use in lectures. Moreover it is necessary to put the illustrations at the disposal of the students in order that they are able to look at them and to listen to them at home. These illustrations contribute to a better understanding of the theoretical principals, as explained during the lectures.

    For further information, see: http://educinno.rug.ac.be/onder wijsinnov/


    Consultancy

    The research group was involved in numerous projects on materials research and performed a large series of laboratory measurements on building products concerning airborne sound insulation, sound absorption,.The research in room acoustics was stimulated by several important study contracts. We mention the design of the Aula of the UCLouvain, with the architect Ph. Samyn and the study concerning the retrofitting of the famous concert hall Henri Le Boeuf of the Paleis voor Schone Kunsten / Palais des Beaux Arts (architect Baron Victor Horta, 1929, architect of the renovation: G. Baines.


    Last updated by Erik.Toorman@bwk.kuleuven.ac.be on 24 June 1999