7/23/2019 BS 874-3.2.pdf

1/18

BRITISH STANDARD BS 874-3.2:1990

Methods for

Determining thermal

insulating properties

Part 3: Tests for thermal transmittanceand conductance

Section 3.2 Calibrated hot-box method

py

()

7/23/2019 BS 874-3.2.pdf

2/18

BS 874-3.2:1990

This British Standard, havingbeen prepared under thedirection of the Refrigeration,Heating and Air ConditioningStandards Policy Committee,was published under the

authority of the Board of BSIand comes into effect on28 September 1990

BSI 07-1999

First published September 1939

Second edition August 1956

Third edition November 1965

Fourth edition November 1973

Fifth edition (as BS 874-s3.2)September 1990

The following BSI referencesrelate to the work on this

standard:Committee reference RHE/9

Draft for comment 89/76042 DC

ISBN 0 580 18510 9

Committees responsible for thisBritish Standard

The preparation of this British Standard was entrusted by the Refrigeration

Heating and Air Conditioning Standards Policy Committee (RHE/-) toTechnical Committee RHE/9, upon which the following bodies wererepresented:

Autoclaved Aerated Concrete Products Association

British Ceramic Research Ltd.

Chartered Institution of Building Services Engineers

Combustion Engineering Association

Cork Industry Federation

Cranfield Institute of Technology

Department of Health

Department of the Environment (Building Research Establishment)

Department of Trade and Industry (National Engineering Laboratory)

Department of Trade and Industry (National Physical Laboratory)

Electricity Supply Industry in England and Wales

Engineering Equipment and Materials Users Association

Eurisol (UK Mineral Wool Association)

Gypsum Products Development Association

Phenolic Foam Manufacturers Association

Power Generation Contractors Association (BEAMA Ltd.)

Royal Institute of British Architects

Thermal Insulation Manufacturers and Suppliers Association (TIMSA)

Thermal Insulations Contractors Association

The following bodies were also represented in the drafting of the standard,through subcommittees and panels:

Aggregate Concrete Block Association

Association of Lightweight Aggregate Manufacturers

British Board of Agrement

British Cement Association

British Precast Concrete Federation Ltd.

Flat Glass Manufacturers Association

Institute of Refrigeration

Institution of Chemical Engineers

Insulation Jacket Manufacturers Federation

Polyethylene Foam Insulation Association

University of Salford

Yarsley Technical Centre Ltd.

Amendments issued since publication

Amd. No. Date of issue Commentspy

()

7/23/2019 BS 874-3.2.pdf

3/18

BS 874-3.2:1990

BSI 07-1999 i

Contents

Page

Committees responsible Inside front cover

Foreword ii0 Introduction 1

1 Scope 1

2 Definitions 1

3 Principle 1

4 Apparatus 3

5 Test elements 8

6 Test procedure 9

7 Expression of results 9

8 Test report 10

Appendix A Calibration procedures 11

Figure 1 Calibrated hot-box apparatus 2Figure 2 Metering box wall thermopile 5

Figure 3 Example of determination of metering box wall and flankingcoefficients 11

Publications referred to Inside back cover

py

()

7/23/2019 BS 874-3.2.pdf

4/18

BS 874-3.2:1990

ii BSI 07-1999

Foreword

This Section of BS 874 was prepared under the direction of the Refrigeration,Heating and Air Conditioning Standards Policy Committee.

Previous editions of BS 874 summarized the methods available for determiningthermal insulating properties, and indicated which methods were applicable tovarious materials intended for use in a variety of temperature ranges, but themethods themselves were not defined in detail.

The growing importance of energy conservation and the need to have reliableinformation about the insulating properties of materials requires detailedspecifications of the test methods. Accordingly, a complete revision of thestandard is being undertaken, with each test method being fully specified. Thefull revision of BS 874 will take several years to complete, and it has thereforebeen decided to issue each test method as a separate section of BS 874 as andwhen it is completed, thus gradually replacing the 1973 edition.

The method presented in this Section may be used as an alternative to theguarded hot-box method given in Section 3.1 and was not originally in 4.3of

BS 874:1973.

It should be appreciated that experience in the use of the method is limited andrevisions of this Section may become necessary as more experience is gained inthe use of the method.

The accuracy levels and the reproducibility which may be expected from themethod are not well established. Precision experiments, of the type described inBS 5497-1 are necessary before reliable estimates for these can be given.

It is recommended that this Section of BS 874 be read in conjunction withBS 874-1 and BS 874-3.1.

A British Standard does not purport to include all the necessary provisions of acontract. Users of British Standards are responsible for their correct application.

Compliance with a British Standard does not of itself confer immunityfrom legal obligations.

Summary of pages

This document comprises a front cover, an inside front cover, pages i and ii,

pages 1 to 12, an inside back cover and a back cover.This standard has been updated (see copyright date) and may have hadamendments incorporated. This will be indicated in the amendment table on theinside front cover.

py

()

7/23/2019 BS 874-3.2.pdf

5/18

BS 874-3.2:1990

BSI 07-1999 1

0 Introduction

The method described in this Section is intended for

tests on large building elements such as parts orsections of walls, roofs, floors, etc., where thethermal properties may not be uniform due to themethod of fabrication and the materials used.

The method may be used as an alternative to theguarded hot-box method of BS 874-3.1. A slightlymodified procedure, given in Appendix A ofBS 874-3.1:1987, enables tests to be made onsmaller elements such as window systems.

The design and construction of an apparatussuitable for application of the method is a complexsubject and needs to be based on sound scientific

principles. While the calibrated hot-box apparatusis simpler in design and operation than the guardedhot-box of BS 874-3.1, detailed calibrationprocedures are required in order to producemeasurements of comparable accuracy. This Sectiongives guidelines for apparatus design and outlinesappropriate calibration procedures.

It is important that users of this method are fullyconversant with the principles of heat flow and withprecise thermal measurement techniques. Allinstruments used in this method are required to betraceable to the National Physical Laboratory(NPL) or to a laboratory acceptable to NPL.

1 Scope

This Section of BS 874 describes a method fordetermining the steady-state thermaltransmittance and thermal conductance ofconstruction elements using the calibrated hot-boxapparatus.

The method is suitable for construction elementswith thermal transmittances and conductances inthe range 0.1 W/(m2K) to 15 W/(m2K) for testingwithin a temperature range of 50C to 50 C.

Apparatus designed in accordance with this Section

is capable of measurement to an accuracy of 5 %on test elements of uniform thermal conductance(i.e. elements with no thermal bridges betweensurfaces) in the conductance range 0.5 W/(m2K)to 5 W/(m2K)

NOTE 1 The potential for achieving these levels of accuracy willdepend upon the apparatus, its calibration particulars, and thetype of test specimen being measured. The determination ofthermal transmittances or conductances below 0.5 W/(m2K)requires measurements of high precision, and therefore may beless accurate. Determination of thermal transmittances orconductances in the range above 5 W/(m2 K) may also be lessaccurate as there may be higher surface temperature variationsand difficulties with measurement of representative surfacetemperatures. Additionally, greater inaccuracies are to beexpected for test elements of nonuniform conductance(i.e. elements with thermal bridges between surfaces).

Some building elements may pose difficult problemsof measurement, and careful consideration is

required before a test method is chosen. The methodis considered particularly appropriate for elementswith surface projections or thick metal faces, whichcould cause difficulties if measured in theconventional guarded hot-box configuration ofBS 874-3.1.

Thermal conductance measurements are notappropriate on elements of particulary non-uniformconstruction (see clause 7).

The method does not provide for measurementswhere there is transport of air through the testelement during the test.

NOTE 2 The titles of the publications referred to in thisstandard are listed on the inside back cover.

2 Definitions

For the purpose of this Section of BS 874, thedefinitions given in BS 874-1 apply.

3 Principle

The basis for the method is the measurement, atequilibrium, of the heat flux through the testelement and the corresponding temperaturedifference(s) across it. The apparatus is shownschematically in Figure 1 and consists of two

principal items, a metering box and a cold box,between which the test element, usually assembledin a support frame, is placed.

Heat supplied to the metering box passes throughthe test element to the cold box, which is maintainedat a constant low temperature. The heat fluxthrough the test element is determined from thetotal power supplied to the metering box, correctingfor losses or gains through the metering box wallsand the flanking loss (see 4.4) to the cold boxoccurring around the perimeter of the specimen(see Figure 1). These corrections are determinedfrom prior calibration measurements usingspecimens of known thermal properties.

It is important that both of these corrections be keptsmall. Heat flow through the metering box walls isrestricted by making them of high thermalresistance, and, where appropriate, controllingexternal ambient temperature fluctuations(see 4.2.2). To reduce flanking losses, the testelement is usually well insulated around itsperimeter (see 4.4.2).

NOTE More complicated versions of this type of apparatus havebeen built, in which the metering box is enclosed in atemperature screen or guard (to control metering box wall heatexchanges) and in which the test panel edges are also guarded (to

minimize flanking losses). Although not dealt with specifically inthis Section of BS 874, such an apparatus, when properlydesigned and constructed, may be taken to meet all therequirements of the present method.

py

()

7/23/2019 BS 874-3.2.pdf

6/18

BS 874-3.2:1990

2 BSI 07-1999

Heat may be supplied to the metering box at aconstant rate, in which case the apparatus slowlyreaches the desired equilibrium condition.Alternatively, the temperature in the metering boxmay be kept constant by suitable control of the heatinput. When thermal equilibrium is achieved,i.e. when the temperatures on each side of the testelement and the heat flux through it are essentiallyconstant, then the final measurements are taken.

Heat exchange at the hot and cold surfaces of thetest element involves both convective and radiativecomponents. The former depends upon airtemperature and air speed, while the latter dependsupon the temperatures and emissivities of surfacesseen by the element surface. The effects of heattransfer to or from a surface by convection andradiation are conveniently combined in the conceptof an environmental temperature and a surface heattransfer coefficient. Thermal transmittance isdefined between two environmental temperaturesand therefore suitable temperature measurementsare required to enable these to be determined

(see Appendix B of BS 874-3.1:1987). This isparticularly important with test elements of lowthermal resistance for which the surface resistancesform a significant fraction of the total resistance.

NOTE The energy flux through test element,is given by the following equation.

= P W F

where

P is the power supplied to metering box (in W)

W is the heat flux to laboratory surrounds (in W)

F is the flanking loss(in W)

Figure 1 Calibrated hot-box apparatus

py

()

7/23/2019 BS 874-3.2.pdf

7/18

BS 874-3.2:1990

BSI 07-1999 3

Strictly, the thermal conductance of a buildingelement is a meaningful term only if its faces are

isothermal. However, if the test element issufficiently uniform, it is acceptable to takearea-weighted surface temperatures for thecalculation of thermal conductance even thoughsmall temperature variations are present. If thethermal properties are sufficiently non-uniform toproduce significant surface temperature variations,then the attribution of a thermal conductance is tobe avoided (see clause 7of BS 874-3.1:1987).

Normally the test element will be the same size asthe open face of the metering box (see Figure 1), butin some instances, in particular for the testing ofwindow units, it may be smaller. For this type of

test, the test element is surrounded by a panel ofknown thermal properties, enabling the heat fluxthrough the test element to be deduced from thetotal heat input. The necessary adaptations to themethod for measurements on small test elementsare described in Appendix A of BS 874-3.1:1987.

4 Apparatus

4.1 General

The apparatus shall comprise the componentsshown schematically in Figure 1, together with thenecessary instrumentation to control and measure

the temperatures on each side of the test elementand enable the heat flux through it to bedetermined.

When designing the apparatus, the requiredheating and cooling capacities shall be determinedhaving regard to:

a) the range of thermal transmittances for whichthe apparatus is designed;

b) the construction of the two boxes;

c) the range of laboratory temperatures likely tobe encountered during tests.

All instruments used in this method shall be

traceable to NPL or to a laboratory acceptable toNPL.

NOTE The apparatus may be designed either for manualoperation or for automatic control. However, automatic control ofboth metering and cold space temperature is recommended, sothat the power input will be automatically adjusted forfluctuations in the temperature of the laboratory environment.Such fluctuations should not be large and in any instance need tobe corrected for. It is also beneficial if the laboratory temperaturecan be controlled.

4.2 Metering box

4.2.1 Requirements of the metering box

NOTE 1 The metering box provides a controlled warmenvironment at constant temperature. It is normally of square orrectangular cross section with one open face.

The open face of the metering box shall be providedwith an air-tight perimeter seal so that when it

contacts the test element system (see Figure 1), airflow between the metering box and the laboratorysurrounds is prevented.

NOTE 2 The contact width of the seal may extend across thefull thickness of the metering box walls. Compressible foamrubber in conjunction with a clamping device has been foundsatisfactory for use as an air-tight seal.

The walls of the metering box shall be impervious toair transfer and sufficiently well insulated to ensurethat any heat flux through the walls is smallcompared with that through the test element.

NOTE 3 A minimum thermal resistance of the metering boxwalls of 2.5 m2K/W is recommended.

NOTE 4 A sandwich construction of two metal or plywood skinswith insulating material between them is considered satisfactoryfor the metering box walls. It is advisable that the wall insulationbe substantially free of cracks, voids, or any other imperfectionslikely to adversely affect wall thermal resistance. Highlyconducting structural members passing through the walls of thebox should be avoided. Likewise, any protective facings crossingthe metering box open perimeter insulation should preferably beof low thermal conduction. Plywood or similar facing used for thispurpose should be kept as thin as possible.

The test area shall be defined as the plane area ofspecimen directly exposed to the internal controlledenvironment of the metering box (or, if the testelement is not plane, then the plane projection ofspecimen area).

NOTE 5 The test area will usually be that of the whole of thetest element.

The linear dimensions of the test area shall be atleast 1 m and not less than five times the maximumthickness of any test element.

NOTE 6 The depth of the metering box is governed by thepractical constraints of its construction and the equipment it is tohouse, but it should be kept as small as possible withoutsacrificing uniform air distribution in order to reduce heatexchange with the laboratory environment.

If transmittance measurements are being made, abaffle shall be fixed in the metering box, parallel tothe surface of the test element, in order to provide aradiating surface of near uniform temperature.

NOTE 7 It is recommended that a baffle be fixed in themetering box for all measurements to assist in producing uniformair velocities and temperature distributions.

When testing vertical construction elements, thebaffle shall extend the full width of the meteringbox, and shall have gaps at the top and bottom toallow air circulation unless the specific testconditions require otherwise.

py

()

7/23/2019 BS 874-3.2.pdf

8/18

BS 874-3.2:1990

4 BSI 07-1999

NOTE 8 When testing vertical construction elements, the bafflecan assist in providing a laminar air stream over the surface, andits distance from the test element can give a measure of control

of the vertical temperature gradient in the air stream. Themagnitude of this gradient will depend on the air speed, on thedistance between the baffle and the surface of the test element,and on the conductance of the element. It is recommended thatthe temperature gradient in the air stream should notexceed 1 K/m for test elements of thermal transmittance upto 1 W/(m2K) and a maximum of 2 K/m for elements of highertransmittance. The temperature gradient can be altered bysuitable combination of air speed and baffle position. It should benoted that the air speed will influence the convective coefficientand thus only limited adjustment of air velocity may be possible.If desired, circulating fans may be installed in the metering boxto assist natural convection if it is required to achieve particularvalues of convection coefficient that cannot be obtained by usingnatural convection alone. In this case means should be providedto measure the air speed across the test element surface.

When testing horizontal construction elements, thebaffle shall extend the full width of the meteringbox; the presence of gaps at either end depends uponthe exact test conditions required.

NOTE 9 When testing horizontal construction elements, theuse of a horizontal baffle in the metering box can assist inachieving the desired boundary conditions. It may, for example,be used in conjunction with fans to produce an air stream acrossthe face of the test element. Alternatively, it may be movedfurther away from the test element, so that natural convectioncurrents are established. In this latter case the baffle may, ifrequired, be extended at each end to close any gaps.

Where conditions of natural convection in themetering box are required, the distance between the

baffle and the test element shall be not lessthan 150 mm.

Where higher air speeds are imposed across the faceof the test element, for example by fans, to assistcontrol of air velocity and air temperature gradient,it is permitted to gradually reduce the distancebetween the baffle and the test element to aminimum of 40 mm at an air velocity of 3 m/s orgreater. The direction of any forced air flow should,where possible, be such as to supplement naturalconvection (i.e. downwards for vertical testelements). When forced air flow is applied, meansshall be provided for measuring the velocity of the

air stream over the face of the test element.NOTE 10 It is important to ensure that air temperature sensorsare placed outside the boundary layer associated with eachsurface. See 4.5.3.

If the fans and their motors are installed inside themetering box, then their power consumption shallbe measured (taking account of any phase angle)and added to that supplied by the heater. If only thefan blades are inside the metering box, then theheat contribution of the blades shall be determined,possibly from measurements of the apparent changein conductance of a suitable test element with andwithout the fans in use.

All surfaces in the metering box that can radiate tothe surface of the test element shall be of matt black

finish, and shall have a total hemisphericalemissivity of at least 0.9.

A low temperature, extended source heater shall beinstalled in the metering box behind the baffle andso shielded that it neither radiates directly to thesurface of the test element nor produces local hotspots on the baffle.

NOTE 11 Resistance wire has been found satisfactory for theconstruction of the heater.

The power consumption of the heater shall bemeasured to an accuracy of 0.25 %. The totaluncertainty in the determination of the net powerinput into the metering box due to the heater and

fans, after correcting for heat exchange through thewalls, flanking loss (see 4.4), and any powerdissipation in the leads, shall not exceed 3 %.

4.2.2 Limitation of heat exchange through themetering box walls. As part of the procedure toensure that the heat flux through the test elementis accurately determined, the heat exchangethrough the sides and back of the metering box shallbe measured and corrected for. This correction shallnot exceed 20 % of the total power supplied to themetering box.

NOTE 1 For test elements of high thermal resistance, even withwell insulated metering box walls, this may be difficult toachieve, particularly if the temperature of the externallaboratory can differ significantly from that within the meteringbox. In such instances, additional measures may be necessary,such as the control of laboratory temperature, or the use of atemperature controlled screen on the metering box(e.g. see note to clause 3).

One method of measuring the heat flux through thesides and back of the metering box is to install athermopile to measure the average temperaturedifference between the inner and outer surfaces ofthe metering box (see Figure 2). Thermocouplejunctions installed on the side walls of the meteringbox for this purpose shall be distant from the planeof the open face of the metering box by at least thethickness of the metering box walls. There shall bea sufficient number of thermocouples in thethermopile to ensure that the heat flow can beestimated with sufficient accuracy to meet thenecessary criterion for the net power input specifiedin 4.2.1.

NOTE 2 As a guide, there should be at least one differentialpair of junctions for each 1 m2of metering box wall surface area.

py

()

7/23/2019 BS 874-3.2.pdf

9/18

BS 874-3.2:1990

BSI 07-1999 5

NOTE 3 The product of the average temperature differenceacross the metering box walls, obtained from the thermopile, andthe heat exchange coefficient for the metering box provides an

estimate of the correction to be applied. The heat exchangecoefficient for the metering box may be calculated from itsdimensions and the known thermal conductivities of itsconstituent materials, provided that these properties aresufficiently well defined, and that any contribution from theadjoining specimen frame is included. However, it isrecommended that the heat exchange coefficient for the meteringbox be measured directly in the apparatus itself (possibly incombination with flanking loss calibrations, see 4.4, 4.8and

Appendix A).

4.3 Cold box

NOTE 1 The cold box provides a controlled environment atconstant low temperature, and is usually insulated to reduce theload on the cooling system.

Temperature control shall be achieved using a heat

exchanger and a cooled circulating medium such aswater or air, controlled by a suitable combination ofcooling and intermittent heating. The height andwidth of the cold box shall be not less than thecorresponding dimensions of the metering box.

If thermal transmittance measurements are beingmade, then a baffle shall be installed in the cold box,the design criteria being the same as those already,described for the metering box (see 4.2.1). In caseswhere it is required to assess the performance of aparticular structural form, it is permissible for thedesign of baffle to reflect that form: in that case thedetails of the baffle design shall be included in thetest report.

Where higher surface coefficients are required inthe cold box, as in the case of the simulation of

outside conditions on test elements, this shall beachieved by higher air velocities over the face of thetest element.

NOTE 2 The direction of any forced air flow should, wherepossible, be such as to supplement natural convection(i.e. upwards for vertical test elements).

When forced air flow is applied, means shall beprovided for measuring the velocity of the air streamover the face of the test element.

NOTE 3 It is recommended that the temperature gradient inthe air stream over the face of the test element should notexceed 1 K/m for test elements of thermal transmittance upto 1 W/(m2K) and a maximum of 2 K/m for elements of highertransmittance.

Direct radiation exchange between any coolingelement in the cold box and the surface of the testelement shall be prevented by suitable shielding. Ifthe cold box is to be operated at temperaturesapproaching or below 0 C then consideration shallbe given to potential problems of condensation andicing.

All surfaces in the cold box which can radiate to thesurface of the test element shall be of matt blackfinish, and have a total hemispherical emissivity ofat least 0.9.

There shall be an air-tight seal between the cold boxand the test element assembly (see Figure 1).

NOTE 4 The cold box air-tight seal may be similar to that usedin the metering box.

Figure 2 Metering box wall thermopile

py

()

7/23/2019 BS 874-3.2.pdf

10/18

BS 874-3.2:1990

6 BSI 07-1999

4.4 Flanking losses

4.4.1 General

During a test measurement, heat will flow from themetering box to the cold box around the edges of thetest element outside of the test area. This heat fluxis defined as the flanking loss, and a correction shallbe made for this, along with the metering box wallheat flow correction, when determining the heatflow through the specimen.

NOTE 1 This view of flanking loss is somewhat simplified. Inreality, the correction for flanking loss has also to account foredge effects within the test element itself, and any relevant edgeinteractions with metering and cold box wall heat flows.

NOTE 2 The magnitude of the flanking loss depends upon thethickness and apparent thermal conductivity of the test element

at its boundary, and the type of surrounding construction. Therelative influence of flanking loss diminishes as the ratio of testarea to the test area perimeter increases. Thus, in largerhot-boxes, although flanking losses are usually higher, theirresultant effects are relatively less significant.

4.4.2 Limitation of flanking losses

NOTE 1 Insulating the edges of the test element around theperimeter of the test area will assist in reducing flanking losses.

Consideration shall be given to providing a wellinsulated support frame to house the test element.

NOTE 2 Any such support frame should not be narrower thanthe thickest test element that it will house.

NOTE 3 It is recommended that the thermal resistance of thesupport frame insulation around the edges of the test area be at

least 2.5 m

2

K/W. It is usually convenient for the metering box,cold box, and test element support frame to have nominally thesame cross-sectional dimensions and insulation properties.

With very heavy specimens, such as masonry walls,it will usually be impractical to provide sufficientsupport for these within a specimen frame solely byusing insulation. However, any highly conducting(i.e. metal) support members or facings used for thispurpose shall be confined to the exterior (laboratory)edge of the frame insulation, unless they aredesigned specifically for edge-guarding purposes(see note to clause 3). Otherwise, on the interiorside, around the perimeter of the test area, the useof a combination of foam insulation material facedwith plywood or similar intermediate conductivitymaterial is permissible for load spreading purposes.Such additional facings shall be kept as thin aspossible to reduce flanking losses.

4.5 Temperature measurement equipment

4.5.1 General. For the measurement of thermalconductance, the mean surface temperature overthe metering area on each face of the test element isrequired. Measurement of thermal transmittancerequires that air and mean radiant temperaturesare measured in the hot and cold boxes, so that thetwo environmental temperatures may be calculated.

Thermocouples generally will be the most commonlyused thermometers for this test method

(see clause 6of BS 874-1:1986 for more detaileddiscussion) but the use of other suitablethermometers is not excluded.

Thermocouples shall be made from a stock ofcalibrated wire or wire which has been certified bythe supplier to comply with BS 4937 to a toleranceof 0.4 % otherwise individual thermocouplecalibration shall be required. A potentiometer fordigital voltmeter with a resolution of 14V or bettershall be used to measure the output from thethermocouples. The uncertainty in themeasurement of temperature difference betweenthe hot and cold faces of the test element at a point

shall not exceed 1 % of the temperaturedifference. The uncertainty in the measurement ofthe voltage output from a thermopile shall notexceed 44V.

4.5.2 Surface temperatures. Thermocouples madefrom wire not exceeding 0.25 mm in diameter shallbe used for measuring surface temperatures tominimize heat conduction along the wires.Thermocouples shall be in intimate contact with thesurface for at least 25 mm from their junctions.They shall be cemented or taped securely to thesurface, following isothermals where possible.

A sufficient number of thermocouples shall be usedto enable the mean surface temperature to bedetermined, as follows. For homogeneous testelements there shall be a minimum of nine on eachface, uniformly distributed over the test area. Forlarge test areas, this number shall be increased asnecessary such that there is at least onethermocouple per 0.5 m2of test area. If the testelement is non-homogeneous, then sufficientadditional thermocouples shall be placed to providerepresentative temperatures for each discreteregion of the element, the average surfacetemperature being determined on the basis of the

proportionate areas of each region. Within eachdiscrete region the variability of temperature shallbe explored to determine the position of thethermocouples.

For the measurement of mean radiant temperaturesof surfaces seen by the test element in meteringand cold boxes, there shall be a minimum of ninethermocouples, appropriately distributed on eachsurface to take account of its relative radiantinfluence on the test element.

py

()

7/23/2019 BS 874-3.2.pdf

11/18

BS 874-3.2:1990

BSI 07-1999 7

4.5.3Air temperatures. For the measurement of airtemperatures in the metering and cold boxes there

shall be a minimum of nine thermocouples in the airspaces on each of the test elements, uniformlydistributed in relation to the test area. For large testareas, this number shall be increased as necesssarysuch that there is at least one thermocoupleper 0.5 m2of adjacent test area.

Thermocouples shall be positioned outside theboundary layer associated with each surface. Whereair movement is by natural convection,thermocouples shall be placed not closerthan 75 mm to a surface. Where an air velocity isimposed it is permitted gradually to reduce thisdistance, but it shall not be less than 20 mm at an

air velocity of 3 m/s or greater. Thermocouples shallbe shielded from radiation if their junction diameteris greater than 1 mm.

4.6 Temperature of the apparatus

When equilibrium has been reached, anyfluctuations in the average air temperature on thehot and cold sides of the metering area shall notexceed 1 % of the air-to-air temperature differencebetween hot and cold over a period of at least 8 h.

4.7Performance checking

After construction of the apparatus, the followingperformance checks shall be carried out in order to

establish its satisfactory operation. Check:

a) that temperatures and temperature gradientsare sufficiently uniform;

b) the effects of air velocities in the boxes;

c) that the control systems are satisfactory;

d) that, during calibration procedures (4.8), theheat flux through a test element and thetemperature difference across it for a given meantest element temperature are proportional toeach other.

4.8 Calibration procedures

4.8.1 GeneralIn order to determine the heat flux through the testarea from the power supplied to the metering box,allowance shall be made for any heat exchange withthe laboratory environment (see 4.2.2), and also theflanking loss around the perimeter of the testelement (4.4).

NOTE The magnitude of these corrections to the power inputdepend not only on the constituent insulation properties of themetering box, and the test element and its support frame, butalso on the presence of additional conducting members such ascables, facings, etc., which may cross the various insulatingregions. It is not usually practical to provide estimates ofcorrections for these effects based wholly on calculation.

4.8.2 Calibration tests and calibration test elements

Estimation of the corrections required to account for

these effects shall be provided by carrying out priorcalibration measurements on the apparatus usingtest elements of known thermal conductance. Thesecalibration elements shall be effectively opaque tothermal radiation and impermeable to air transfer.

NOTE 1 It is recommended that the calibration elementsshould not contain internal cavities or voids where significantinternal convection may occur, and that their properties do notchange significantly with time.

The conductances, thicknesses, mounting andpositioning of the calibration elements shall as faras possible cover the likely ranges of those forproposed test elements.

Conductances of the calibration elements shall bedetermined either from the known thermalconductivities of their components, as determinedusing the guarded hot-plate apparatus(see BS 874-2.1), or by prior measurement using theguarded hot-box method (see BS 874-3.1). Suchequipment used for this purpose shall be traceableto NPL, or to a laboratory acceptable to NPL.

NOTE 2 It is particularly convenient for conductancedeterminations to be carried out using the guarded hot-boxmethod if the calibrated hot-box apparatus undergoingcalibration can be alternatively set up as a guarded hot-box forthis purpose.

The range of applied conditions for the series ofcalibration tests shall cover that likely to beexperienced in test element measurements.

NOTE 3 This applies to the external laboratory environment aswell as to the applied conditions within metering and cold boxes.

4.8.3 Correction factors for test elementmeasurements

(see Appendix A)

Since sample properties are known, estimates of thecombined influence of laboratory temperature andflanking loss can be calculated from the results ofeach calibration measurement. These shall be usedto build up a directory of correction factors forapplication in the subsequent measurements on thetest elements.

NOTE 1 It is usually convenient to define correction factors toaccount for laboratory temperature effects and flanking lossesseparately. Appendix A considers some aspects of this procedurein more detail.

In any subsequent test element measurement, theparticular flanking loss correction applied shall bethat most appropriate for the thickness,conductivity and edge support details of the elementat its boundary, as inferred from the calibrationresults.

NOTE 2 It is permissible to use numerical computational

procedures to assist in the interpretation of flanking effectsprovided that such procedures have been demonstrated as beingsatisfactory on the basis of previous calibration measurements.

py

()

7/23/2019 BS 874-3.2.pdf

12/18

BS 874-3.2:1990

8 BSI 07-1999

The accuracy assigned to measurements made ontest elements using this apparatus shall take into

account the results obtained from the calibrationtests.

4.8.4 Calibration verification checks

Regular calibration verification checks shall becarried out to ensure that the performance of anapparatus does not change significantly with time(e.g. due to insulation ageing, cracking). Anyobserved change in calibration aspects of theapparatus shall be carefully investigated, andcorrective measures taken as appropriate.

NOTE It is recommended that such checks are carried out atleast annually.

5 Test elements

5.1 General

NOTE 1 It is clearly desirable that the construction element tobe tested is representative in terms of materials and constructionof those used in practice. However, considerable caution shouldbe exercised before assuming that the results of measurementson test elements can be directly applied in practice.

There shall be no pathways for air leakage from thehot to the cold side of the element during a test. Testelements with continuous vertical cavities, such ascavity walls, shall be sealed at their outer edges toprevent air exchange with the externalenvironment, and the edges shall also be insulatedto provide a thermal resistance of atleast 2.5 m2K/W.

NOTE 2 It is recommended that the test elements be mountedin an insulated support frame (see 4.4.2) and sealed around theperimeter of the test area adjoining the support frame. Mounting,positioning, and sealing procedures should be consistent withthose of appropriate calibration measurements (see 4.8).

NOTE 3 Test elements such as windows are normally tested ina special surround panel. Details of the procedure in this case aregiven in Appendix A of BS 874-3.1:1987.

Careful consideration shall be given to the relevanceof the test conditions selected (temperatures and airvelocities) to those applicable to the test element inpractice. For example, high air velocities may affectthe thermal transmission properties of some lowdensity, open textured, insulation materials bycausing air movement within them; for suchmaterials the boundary air velocities in normal useshall be considered. In addition, the thermalproperties of some materials are temperaturedependent and consideration shall be given to therelevance of the hot and cold face temperatures topractical conditions.

In the case of non-homogeneous test elements suchas walls built of concrete blocks, the test area shallas near as possible be representative of an infinite

panel of the same construction. The relative sizes ofthe metering area, the individual components of thewall and the bonding arrangement shall beconsidered carefully.

NOTE 4 The importance of this factor increases as the ratio ofthe component area to metering area rises, and failure toconsider this may give a result that is not representative of

full-scale constructions.

5.2 Test elements smaller than the meteringarea

The thermal transmittance of elements smallerthan the metering area shall be measured with thecalibrated hot-box by following the procedures givenin Appendix A of BS 874-3.1:1987.

5.3 Conditioning of test elements

If, after assembly, test elements are likely toundergo any change that will affect the thermalconductance or transmittance, e.g. variation ofmoisture content of porous material or recovery

from compression of fibrous materials, then theelement shall be conditioned prior to the test unlessthe purpose of the test is specifically to look for sucha change.

Where moisture considerations are important, thefollowing shall apply for any test measurement.

a) The moisture content shall be known.

NOTE 1 A measurement on, for example, a masonry wall ismeaningless without this information.

b) There shall not be any significant change inmoisture content during the course of a test.

NOTE 2 The change in moisture content is dependent on the

material and environmental conditions. A change in moisturecontent of 0.2 % V/Vcan give rise to a 2 % change in thermalconductivity of some low density masonry materials. Also, thetransport of moisture and associated latent heat effects mayaffect the measured result.

c) Dependent on the purpose of the test, themoisture content shall be at a realistic level[see item i) of clause 8].

Conditioning shall be undertaken to satisfy b)and c).

Test elements containing moisture shall beconditioned in well-ventilated ambient indoorsurroundings. The conditioning period shall be

sufficiently long to achieve effective equilibrium forthe purposes of the test.

NOTE 3 A progressive change in mass of 0.1 % or less per week,with random variations, is normally taken as a criterion forestablishment of effective equilibrium. (This will usually implymoisture levels of 5 % or less by volume in most conditionedmasonry specimens.)

Accelerated conditioning techniques are permittedprovided that the thermal properties of thespecimen are not adversely affected (e.g. therequired curing periods for concrete or mortarshould be allowed prior to any accelerated drying). Ifaccelerated conditioning is used, then it shall befollowed by natural conditioning in the laboratory toensure that effective equilibrium is established.

py

()

7/23/2019 BS 874-3.2.pdf

13/18

BS 874-3.2:1990

BSI 07-1999 9

NOTE 4 Moisture contents of test elements are normallydetermined gravimetrically by drying to constant mass at 105 C(or such other temperature as is appropriate). But alternatively,

indirect methods of measurements are acceptable provided theyare of proven accuracy.

Where it is not practicable to weigh the whole of thetest element, conditioning shall be monitored byweighing a smaller sample element constructed atthe same time as the test element and storedadjacent to it over the conditioning period. In orderto reproduce in the small sample the dryingcharacteristics of the test element, considerationshall be given to the path along which moisturediffuses during drying.

NOTE 5 For example, the drying of a test wall can be simulatedwith a separate small section of the same material, sealed at its

edges and the two external faces exposed to the dryingenvironment.

Where appropriate, the mean moisture level over atest shall be estimated by averaging measuredvalues taken before and after the test. Where it isnot practicable to weigh the whole of the testelement, small sections shall be taken from suitablepositions near the edge of the test element.

It shall be ascertained that the latent heatassociated with any moisture change over the periodof the test is small in comparison with the energytransmitted through the test element during thesame period.

NOTE 6 If measurements are required to be made at moisturecontents other than those obtained with these conditioningprocedures, humidity control may be required in the apparatus.

6 Test procedure

Install the test element between the boxes and fix inposition the number of thermocouples in accordancewith 4.5.2, as appropriate for homogeneous ornon-homogeneous surfaces. Clamp the boxes to thetest element/support frame taking care that allseals are airtight. Check all fans, thermocouplesand instruments for correct operation.

Establish the required temperatures andenvironmental conditions in the apparatus.

NOTE 1 A minimum temperature difference between the hotand cold air temperatures of 20 K is recommended ifmeasurement of thermal transmittance is to yield accurateresults. Since the thermal properties of some materials aretemperature dependent, the choice of hot and cold temperatures,needs to reflect the conditions of use. It is customary in thermalconductivity tests for building control purposes on masonrysamples for the equilibrium temperatures on the hot and coldsides of the apparatus to be 27 3 C and 10 3 C respectively.See BS 874-2.1.

Where appropriate, control the external laboratorytemperatures to minimize metering box wall heatexchanges (4.2.2).

Take readings only when the apparatus has reachedequilibrium.

Once equilibrium has been reached, i.e. when thetemperatures, the power supplied, and the

computed results begin to vary randomly ratherthan continuously increasing or decreasing,continue the test for at least a further 8 h and do notterminate the test until measurements of thermaltransmittance or conductance, averaged over atleast two successive 4 h periods, differ by lessthan 1 %

NOTE 2 The time required for the apparatus to reachequilibrium can vary from a matter of hours to a week or so. Itwill depend upon the apparatus, the test element and the methodof control, it will take longer with test elements of high thermalcapacity or with substantial insulation on them.

7 Expression of results

The results to be reported in clause 8shall be theaverage of the final, successive sets ofmeasurements. If there are small periodicvariations in temperature or power in the apparatusdue to the operation of the control systems (butwithin the limits specified in 4.6), then a greaternumber of sets of data shall be averaged, which willlead to a more reliable result.

The thermal conductance, [in W/(m2K)], shallbe calculated as follows.

A thermal conductance value shall not be quoted ifthere are local variations in temperature on eitherface (associated with thermal bridges)exceeding 20 % of the difference between averageface temperatures, or if the test element is a window

system.

where

is the heat flux through the metering area(in W), i.e. the rate of heat supplied to themetering box, including power supplied toany fans and corrected where necessary for:

(a) heat flux through the metering boxwalls;

(b) flanking loss around the perimeter ofthe test area;

(c) any power loss in the leads to the

heater or to the fans in the metering box.A is the test area defined in 4.2.1(in m2);

Ts1 is the average surface temperature of thetest element in the metering box (in C);

Ts2 is the average surface temperature of thetest element in the cold box (in C).

A Ts1

Ts2

( )----------------------------------=

py

()

7/23/2019 BS 874-3.2.pdf

14/18

BS 874-3.2:1990

10 BSI 07-1999

The thermal transmittance, U[in W/(m2 K)], shallbe calculated as follows.

The surface coefficients h1and h2[in W/(m2K)] are

given by the following equations:

1) surface coefficient (hot face)

2) surface coefficient (cold face)

NOTE The surface resistance at each face is the reciprocal ofthe appropriate suface coefficient.

For transmittance measurements, where it is notpossible to determine surface heat transfercoefficients due to the nature of the test element

(e.g. because of irregular surface geometry,excessive local surface temperature variations), thesurface coefficients for a reference calibrationelement of similar thermal resistance shall bemeasured under the same conditions of temperatureand air velocities as those applied for the testelement. These surface coefficients shall beassumed to also apply for the test elementtransmittance measurements.

8 Test report

The test report shall include the following

information:

a) the name and address of the laboratoryundertaking the test;

b) the method of test, i.e. calibrated hot-boxmethod, BS 874-3.2, including dimensions anddisposition of temperature sensors;

c) the orientation of the test element and thedirection of heat flow;

d) a description of the test element, including allrelevant dimensions and materials used in itsconstruction;

e) whether natural or forced convection is usedand, if forced, what air speeds;

f) where applicable, the methods of weightingused to calculate mean surface temperatures;

g) the measured thermal conductance and/ortransmittance and the uncertainties in thesemeasurements;

h) if thermal transmittance measurements arereported, the measured heat transfer coefficientsand their method of determination;

i) details of any conditioning treatment and,where appropriate, the moisture content by

volume before and after the test and the meanmoisture content during the test;

j) the average environmental, air, and surfacetemperatures (as appropriate) on each face of thetest element;

k) details of corrections (metering box walls,flanking, etc.) applied to the metering box heatinput and brief outline of the calibrationprocedures used to obtain these;

l) the external laboratory temperature (average,maximum and minimum) and mean relativehumidity over the duration of the test;

m) the date of start of the test and its duration;n) details of any special baffle design (see 4.3).

where

Te1 is the environmental temperature in themetering box (in C);

Te2 is the environmental temperature in thecold box (in C).

U

A Te1 Te2( )----------------------------------=

1

A Te1 Ts1( )----------------------------------=

2

A Ts2 Te2( )----------------------------------=

py

()

7/23/2019 BS 874-3.2.pdf

15/18

BS 874-3.2:1990

BSI 07-1999 11

Appendix A Calibration procedures

A.1 GeneralThese are designed to provide metering box andflanking loss corrrection estimates to apply incalibrated hot-box test measurements. They rely onprior measurements using calibration specimens ofknown uniform conductance.

A.2 Method

In a calibrated hot-box test, the measured power,P(in W) supplied to the metering box can be expressedby the following.

It is convenient to relate , and F, in terms of themeasured surface temperature difference %Tacrossthe calibration element, and wto the averagesurface temperature difference %Twacross the

metering box walls. Equation A1 may then bewritten

NOTE Alternatively, !and may be defined in terms of themeasured environmental temperature difference across thecalibration element, in which case !will then be expressed as theproduct of the calibration element thermal transmittance U[in W/(m2K)] with the test areaA(in m2).

By carrying out calibration measurements withdifferent calibration test elements (i.e. known !), atdifferent applied conditions (i.e.%T, %Tw),equation A2 can be used to inferand behaviour.This forms the basis for the calibration procedures.

Example. One possible approach that can be used isas follows. With a given calibration element

(known !), vary both warm and cold sidetemperatures to produce different temperaturedifferences (%T, %Tw) across the element andmetering box walls, but retain the same meantemperature (i.e. same !) for the element.

Using equation A2 in the form

then simple plotting and/or linear regressiontechniques may be used to extract the appropriate and *coefficients. Figure 3 shows an example of

results obtained using this approach.Supposing that the heat exchange coefficient! =Afor the calibration test elementwas 2.54 W/K (as determined from measuredmaterial thermal conductivity, thickness, andexposed test area). Thus from Figure 3, thecoefficient defining heat exchanges with thelaboratory is

= 1.87 W/K

and the flanking loss coefficient for the particularcalibration test element system measured

= 2.72 2.54 = 0.18 W/K.

This procedure is then repeated for differentcalibration elements.

P= +W +F (A1)

where

is the energy flux (in W) through the testelement;

W is the energy flux (in W) through themetering box wall (including any componentthrough the support frame);

F is the flanking loss (in W).

P= !%T+%Tw +%T (A2)

where

! is the heat exchange coefficient (in W/K) ofthe calibration element, derived from ! =A,

being its conductance [in W/(m2K)] and

A the test area (in m2);

is a heat exchange coefficient (in W/K) forheat flows to or from the laboratory surrounds

(primarily through the metering box walls);

is a flanking heat loss coefficient (in W/K).

(A3)

Figure 3 Example of determination ofmetering box wall and flanking coefficients

P

%T-------- +( )

%Tw%T

------------+=

py

()

7/23/2019 BS 874-3.2.pdf

16/18

12 blank

py

()

7/23/2019 BS 874-3.2.pdf

17/18

BS 874-3-3.2:1990

BSI 07-1999

Publications referred to

BS 874, Methods for determining thermal insulating properties.

BS 874-1, Introduction, definitions and principles of measurement.

BS 874-2, Tests for thermal conductivity and related properties.BS 874-2.1, Guarded hot-plate method.

BS 874-3, Tests for thermal transmittance and conductance.

BS 874-3.1, Guarded hot-box method.

BS 4937, International thermocouple reference tables.

BS 5497,Precision of test methods1).

BS 5497-1, Guide for the determination of repeatability and reproducibility for a standard test method byinter-laboratory tests.

1) Referred to in the foreword only.

py

()

7/23/2019 BS 874-3.2.pdf

18/18

BS 874-3.2:1990

BSI

389 Chiswick High Road

London

W4 4AL

BSI British Standards Institution

BSI is the independent national body responsible for preparingBritish Standards. It presents the UK view on standards in Europe and at theinternational level. It is incorporated by Royal Charter.

Revisions

British Standards are updated by amendment or revision. Users ofBritish Standards should make sure that they possess the latest amendments oreditions.

It is the constant aim of BSI to improve the quality of our products and services.We would be grateful if anyone finding an inaccuracy or ambiguity while usingthis British Standard would inform the Secretary of the technical committeeresponsible, the identity of which can be found on the inside front cover.Tel: 020 8996 9000. Fax: 020 8996 7400.

BSI offers members an individual updating service called PLUS which ensuresthat subscribers automatically receive the latest editions of standards.

Buying standards

Orders for all BSI, international and foreign standards publications should beaddressed to Customer Services. Tel: 020 8996 9001. Fax: 020 8996 7001.

In response to orders for international standards, it is BSI policy to supply theBSI implementation of those that have been published as British Standards,unless otherwise requested.

Information on standards

BSI provides a wide range of information on national, European andinternational standards through its Library and its Technical Help to Exporters

Service. Various BSI electronic information services are also available which givedetails on all its products and services. Contact the Information Centre.Tel: 020 8996 7111. Fax: 020 8996 7048.

Subscribing members of BSI are kept up to date with standards developmentsand receive substantial discounts on the purchase price of standards. For detailsof these and other benefits contact Membership Administration.Tel: 020 8996 7002. Fax: 020 8996 7001.

Copyright

Copyright subsists in all BSI publications. BSI also holds the copyright, in theUK, of the publications of the international standardization bodies. Except aspermitted under the Copyright, Designs and Patents Act 1988 no extract may bereproduced, stored in a retrieval system or transmitted in any form or by anymeans electronic, photocopying, recording or otherwise without prior writtenpermission from BSI.

This does not preclude the free use, in the course of implementing the standard,of necessary details such as symbols, and size, type or grade designations. If thesedetails are to be used for any other purpose than implementation then the priorwritten permission of BSI must be obtained.

If permission is granted, the terms may include royalty payments or a licensingagreement. Details and advice can be obtained from the Copyright Manager.Tel: 020 8996 7070.

py

()