Service Life Prediction

CIB W080: Test Methods for Service life PredictionCIB Publication 331ISBN: 978-90-6363-062-1

Indentification of degradationagents - mechanisms - effects

Agents intensities: critical values(ageing thresholds)

Pre-design of proportion betweenageing phases (ageing subcycles)

Ageing cycle pre-design

Climatic dataanalysis

Standardreference

Correct proportion between ageing cycles: rate ageing cycles / years obtained with comparison between effects achieved with

short-term and long-term exposure

CIB W080

WG3 TEST METHODS FOR SERVICE LIFE PREDICTION

CIB REPORT: PUBLICATION 331

STATE OF THE ART REPORT ON

ACCELERATED LABORATORY TEST PROCEDURES AND CORRELATION BETWEEN LABORATORY TESTS AND

SERVICE LIFE DATA

ISBN: 978-90-6363-062-1

Prepared by: Bruno Daniotti, Fulvio Re Cecconi Building Environment Science and Technology Via Ponzio, 31 I-20133 Milano Italy

Accelerated Laboratory Test Procedures and Correlation Between Laboratory Tests and Service Life Data B. Daniotti – F. Re Cecconi

page 2 of 102 CIB W080 Prediction of Service Life of Building Component and materials

INDEX

Acknowledgements .................................................................................................................. 41. General principle of service life prediction tests ....................................................... 5 2. Best practice............................................................................................................. 7 2.1 Artificial weathering of wood materials ..................................................................... 9 2.1.1 Scope of the test method ......................................................................................... 92.1.2 Description of the tests method................................................................................ 92.1.3 Conclusions............................................................................................................ 102.1.4 References ............................................................................................................. 112.2 Durability of metals ................................................................................................. 12 2.2.1 Introduction............................................................................................................. 12 2.2.2 Test methods.......................................................................................................... 12 2.2.3 Comparison of accelerated tests with service performance................................... 15 2.2.4 Methods of correlating with service conditions....................................................... 162.2.5 Conclusion.............................................................................................................. 17 2.2.6 References ............................................................................................................. 182.3 Evaluation of concrete resistance to carbonation and chloride penetration ........... 19 2.3.1 Scope of the test method ....................................................................................... 192.3.2 Description of the test methods.............................................................................. 202.3.3 Conclusion.............................................................................................................. 202.3.4 References ............................................................................................................. 212.4 Sulphate resistance of cements ............................................................................. 23 2.4.1 Scope of the test method ....................................................................................... 232.4.2 Description of the tests method.............................................................................. 232.4.3 Conclusions............................................................................................................ 232.4.4 References ............................................................................................................. 242.4.5 Standard test methods ........................................................................................... 242.5 Freezing resistance of concrete ............................................................................. 25 2.5.1 Scope of the test method ....................................................................................... 252.5.2 Description of the tests method.............................................................................. 252.5.3 Conclusions............................................................................................................ 252.5.4 References ............................................................................................................. 252.5.5 Standard test methods ........................................................................................... 252.6 Test methods for masonry walls............................................................................. 27 2.6.1 Scope of the test method ....................................................................................... 272.6.2 Description of the tests method.............................................................................. 272.6.3 Conclusion.............................................................................................................. 322.6.4 References ............................................................................................................. 332.6.5 Standard test methods ........................................................................................... 332.7 ETICS Cladding: Degradation and Loss in Hygrothermal

Performance Achieved with Accelerated Laboratory Ageing ................................. 35 2.7.1 Scope of the test method ....................................................................................... 352.7.2 Description of the tests method.............................................................................. 362.7.3 Conclusion.............................................................................................................. 422.7.4 References ............................................................................................................. 452.7.5 Standard test methods ........................................................................................... 482.8 Performance assessment of external renders on facades ..................................... 49 2.8.1 Scope of the test methods...................................................................................... 492.8.2 Description of the tests method.............................................................................. 51


CIBW080 Prediction of Service Life of Building Materials and Components page 3 of 102

2.8.3 Conclusion.............................................................................................................. 532.8.4 References ............................................................................................................. 582.8.5 Standard test methods ........................................................................................... 582.9 Test methods for the durability evaluation of pitched roof ...................................... 59 2.9.1 Objectives of the experiment .................................................................................. 592.9.2 Description of the tests method.............................................................................. 592.9.3 Conclusion.............................................................................................................. 692.9.4 References ............................................................................................................. 702.9.5 Standard test methods ........................................................................................... 712.10 Test methods for the durability evaluation of flat roof............................................. 73 2.10.1 Scope of the test method ....................................................................................... 73 2.10.2 Description of the tests method.............................................................................. 74 2.10.3 Conclusions............................................................................................................ 822.10.4 References ............................................................................................................. 822.10.5 Standard test methods ........................................................................................... 842.11 Durability of the external load bearing walls ........................................................... 85 2.11.1 Scope of the test method ....................................................................................... 852.11.2 Description of the tests method.............................................................................. 862.11.3 Conclusions............................................................................................................ 872.11.4 References ............................................................................................................. 882.11.5 Standard test methods ........................................................................................... 883. Synopsis of the Best Practice................................................................................. 89 4. Conclusions............................................................................................................ 99 CIB Brochure........................................................................................................................ 103 Disclaimer ............................................................................................................................ 105



Acknowledgements The authors gratefully acknowledge all the CIB W080 commission members who have contributed to and worked on this report, without you this work wouldn’t be possible. We also want to give a special thanks to Dr. Riccardo Paolini for his precious work. Finally, we would like to thank Dr. Jean-Luc Chevalier and Dr. Ivan Cole, chairs to the CIB W080, for their effort to sustain this initiative.



1. General principle of service life prediction tests Service life prediction is a (the most important?) part of service life planning of building and building components. The knowledge of building components service life is crucial in many design phases, for example it is essential in operation and maintenance cost estimate, but not only during design stage, for example it has been time since each construction product sold in the European Union must satisfy six essential requirements for its entire service life.

It has been deeply investigated by various researchers how to predict service life of building materials and components both before and after that the ISO 15686-2 “Buildings and Constructed Assets - Service Life Planning - Part 2: Service Life Prediction Procedures” was first published in 2001. Although this standard clearly state a systematic methodology for service life prediction of building components, an investigation of most used service life procedures highlighted quite unexpected results.

Figure 1.1: Systematic methodology for SLP of building components (ISO 15686-2)



Chapter 2 and 3 of this report will guide the reader though eleven different tests methods for service life prediction and allow him to acknowledge that:

• there still is a problem in identifying the difference between building material, component and assembly;

• the structure of the test procedure may be differ from the proposed ISO 15686-2 structure in many points;

• the duration of the tests are very different;

• the correlation between aging agent and climatic data is often weak.



2. Best practice Although there is a well know standard on service life prediction methods, the ISO 15686-2 “Buildings and constructed assets - Service life planning - Part 2: Service life prediction procedures”, the best practice all over the world does not always follow the standardized procedures.

An in-depth research of test methods used all over the world highlighted two different approaches to service life prediction:

• the first approach derives from studies on materials and usually provides for a test method that can be used on a single material, no matter how the material can be used in real project, and characterised by accelerated aging agents that aren’t always related to actual climate agents and are often limited to one or two at the same time;

• the second approach is more related to the ISO method and is more used on building components then on materials. This second class of method usually has a performance approach to service life prediction and often gives information not only on the service life but also on performance decay over time.

Examples of the first type of test methods are the ones used to test concrete, woods and metals, paragraphs from 2.1 to 2.5, while the second approach is shown in paragraphs from 2.6 to 2.11.





2.1 Artificial weathering of wood materials 1 This test method has been designed for artificial weathering of different materials, not only wood materials. It is however applied quite much just for that type of materials, both untreated wood and wood with various types of impregnation or/and surface treatment. The method has been developed throughout a Nordic cooperation, and it is published as a Nordtest method [1]. the test equipment is in use in Norway and Denmark.

2.1.1 Scope of the test method The test method is intended for exposing materials and components used in a building envelope to UV light, heat, water and frost. Thereby, the method seeks to simulate the main strains of a natural climatic exposure. A test specimen is exposed to each of the mentioned strains on a consecutive basis in a number of cycles, and altogether this exposure programme is meant to simulate a natural climate exposure in an accelerated way. The test method is therefore a short-term test, and the test results can be used as a basis for estimation of the technical or aesthetical service life of a building material or component.

2.1.2 Description of the tests method The test method is described in detail in NT BUILD 495, and a figure illustrating the equipment layout is shown in Figure 2.1.1.

The test equipment has a cylinder with a vertical axis and where the test specimens are mounted on the outside of the cylinder. The maximum diameter of the cylinder is 3,0 m and the exposed area for mounting of test specimens is approximate 1,5 m x 2,5 m. Around the cylinder are mounted three exposure chambers where the test specimens are exposed to UV light and heat, water and frost on a consecutive basis. The cylinder carrying the test specimens is rotated 90o every hour, and thereby the specimens are exposed to UV and heat radiation, water spray, freezing and a normal laboratory climate for one hour every fourth hour. The exposure programme can be run as long as the laboratory or client specifies. For each of the exposures, the following specifications are given in the method description:

• UV radiation: UV tubes with a relative spectral distribution in the UV band close to that of global solar irradiance.

• Heat: Black panel temperature is normally 63 ± 5 oC, but if required, the temperature may be chosen to be 35 ± 5 oC, 50 ± 5 oC or 75 ± 5 oC. The temperature is controlled by means of infrared halogen lamps.

• Wetting with a spray of demineralised water. A strain of 15 ± 2 L/(m2 h) is suggested.

• Cooling and freezing to an air temperature of -20 ± 5 oC.

• Thawing at ambient laboratory climate of about 23 ± 5 oC and 50 ± 10 % RH.

During the thawing period, the test specimens may be inspected, rearranged and changed.

1 Professor Per Jostein Hovde - Norwegian University of Science and Technology (NTNU), Dept. of Civil and Transport Engineering



Figure 2.1.1. Principal drawing of apparatus for accelerated weathering. From [NT BUILD 495, 2000].

It is not specified in the method description how much the climate exposure and degradation processes of the materials or components are accelerated. That will also depend on the type of material being exposed. However, based on long experience with application of the method in Norway (SINTEF Building and Infrastructure), it is normally estimated that the acceleration factor is about 10-12. That means that an exposure in the test equipment of one month simulates about one year exposure of outdoor climate in Norway. However, the interpretation of the test results has to be based on sound experience.

The test method can be applied for testing of both single material specimens as well as smaller parts of building components (façade products, panels, window frames, etc.)

The evaluation of the tested specimens can be done by visual inspections or by testing of various characteristics of the specimens (mechanical strength, colour changes, gloss changes, chemical changes, etc.).

Application of the test method for studying a climate exposure of wooden windows is presented in [Gjelsvik, T, 1986].

The test results can be used as input to the service life prediction procedure as described in ISO 15686 Part 2.

2.1.3 Conclusions • The test method comprises exposure of material or component specimens by the main

strains of a natural outdoor climatic exposure. It is therefore a valuable test method for artificial weathering of wood materials as well as other building materials.

• Based on the long experience achieved by application of the test method, it is possible to give an estimate of the acceleration factor of the degradation. This is of great value in the application of test results for service life prediction.



When testing wood materials, the test method does not give information about the resistance to rot fungi, because no fungal attack will appear due to the quite extensive UV radiation exposure on a regular basis.

2.1.4 References Hovde, P. J.: Needs for service life prediction of passive fire protection systems. 8th International Fire Science and Engineering Conference, Edinburgh, Scotland, 29. June - 1. July 1999.

Östman, B., Voss, A., Hughes, A., Hovde, P. J. and Grexa, O.: Durability of fire retardant treated wood products at humid and exterior conditions. Review of literature. Fire and Materials 25 (2001) 95-104.

Hovde, P. J. and Moser, K.: Performance based methods for service life prediction. State of the Art Reports Part A & Part B. CIB Report Publication 294. CIB, Rotterdam, The Netherlands, March 2004. ISBN 90-6363-040-9.

Lauter, P., Time, B., Hovde, P. J. og Nore, K.: Influence of material quality and climate exposure on moisture condition of a wooden facade. Proceedings, 10th International Conference on Durability of Building Materials and Components (10DBMC), Lyon, Frankrike, 17.-20. april 2005, Publication TT2-129.

Hovde, P. J., Jacobsen, B., Jelle, B. P., Larnøy, E. and Vestøl, G.: Enhanced service life of coated wooden facades. Submitted to 11th International Conference on Durability of Building Materials and Components (11 DBMC), Istanbul, Turkey, 11.-14. May 2008.

Jelle, B. P., Myklebost, I., Holme, J., Hovde, P. J. and Nilsen, T.-N.: Attenatuated Total Reflectance (ATR) Fourier Transform Infrared (FTIR) radiation studies of wood rot decay and mould fungus growth on building materials. Submitted to 11th International Conference on Durability of Building Materials and Components (11 DBMC), Istanbul, Turkey, 11.-14. May 2008.

Jelle, B. P., Rüther, P., Hovde, P. J. and Nilsen, T.-N.: Attenatuated Total Reflectance (ATR) Fourier Transform Infrared (FTIR) radiation investigations of natural and accelerated climate aged wood substrates. Submitted to 11th International Conference on Durability of Building Materials and Components (11 DBMC), Istanbul, Turkey, 11.-14. May 2008.

Leicester, R.H.1., Foliente, G.C.1., Mackenzie, C.2., Cole, I.S.1., Wang, C.-H.1., Nguyen, M.N.1., & Cookson, L.3. 2005, The development of durability models for engineered timber construction, Australian Structural Engineering Conference, ASEC 2005 [CD-ROM], Newcastle, N.S.W., September 11-14, 2005

McGeachie, M., Cole, I.S., Zhang, J. and Ganther, W.D. (1999). Characterisation of the Sydney Climate in Relation to Corrosivity, Timber Degradation Risk Factors and the Corrosion of Nails in Timber. Forest Research Bulletin 212

NT BUILD 495. Building materials and components in the vertical position: exposure to accelerated climatic strains. Nordtest, Espoo, Finland, November 2000

Gjelsvik, T.: Accelerated and natural weathering of wooden windows with organic coatings. Paper in Project Report 9, Norwegian Building Research Institute, Oslo/Trondheim, Norway, 1986.

ISO 15686-2. Buildings and constructed assets – Service life planning – Part 2: Service life prediction procedures. First edition. International Organization for Standardization, Geneve, Switzerland, 2001.



2.2 Durability of metals2

2.2.1 Introduction The definition of test methods for the durability of metals is complex for two reasons:

• A detailed hierarchy of corrosion standards has developed both at national and international levels. This hierarchy has been developed for the definition of metal performance across all applications, and thus contains elements that are not relevant to metals in infrastructure.

• In common with many test methods, there are significant technical issues in defining the correlation between field or laboratory test results and the service life of real components on real infrastructure. A range of different tests and approaches have been developed to bridge this gap.

The supporting standards for corrosion testing cannot be fully outlined in this short section, rather it is only possible to define the types of information available. The relevant types of standards are:

• Those that define the atmospheres (exterior and interior) that metals are exposed to, such as ISO 9223:1992 Corrosion of Metals and Alloys – Corrosivity of Atmospheres –. Classification (defines exterior atmospheres), IEC 60654-4:1987 Operating Conditions for Industrial–Process Measurement and Control Equipment: Part 4 – Corrosive and Erosive Influences (defines classes of pollutants), and ISO 11844-1:2006 Corrosion of Metals and Alloys – Classification of Low Corrosivity of Indoor Atmospheres: Part 1 – Classification of Indoor Atmospheres (defines indoor environments).

• Those that define general considerations for testing or for a particular group of tests, rather than a particular test. Of prime relevance here is ISO 11845:1995 Corrosion of Metals and Alloys – General Principles for Corrosion Testing.

2.2.2 Test methods Currently, the principal accelerated test methods are classified as either constant-condition chamber tests or cyclic tests. Constant-condition chamber tests are well established for a broad range of environments, but lack a systematic methodology for correlation with service life. On the other hand, although cyclic tests are available for a narrower band of conditions, they have been correlated with service life.

Chamber tests Testing with condensation

Under some conditions, the main factor driving corrosion is condensation, and IEC 60068-2-30:2005 attempts to mimic these conditions. The test requires a cycle consisting of high relative humidity (RH 95–100%) for 3 hours, lower humidity (90–96%) for 9 hours, and then high humidity again for 12 hours. Chamber temperature should be maintained at 40°C during the lower humidity phase of testing, and at 25°C during the higher humidity phases. The test is only applicable when condensation occurs on a clean surface and in a pure environment, however, such conditions are relatively rare in atmospheric or building corrosion, as surfaces are often contaminated with industrial/marine aerosols or other contaminants.

2 Ivan Cole - CSIRO – Division of Material Science and Engineering



Salt spray tests

The neutral salt spray (NSS) test, which involves a continuous spray of sodium chloride solution at neutral pH, has been around since 1914 [Capp, J.A. 1914], and this test method and it uses have been standardized in ASTM B117. In addition, the NSS test and the related acetic acid salt spray (AASS) and copper and acetic acid salt spray (CASS) tests are defined in ISO 9227:2006 Corrosion Tests in Artificial Atmospheres – Salt Spray Tests.

The NSS test was designed to approximate the deposition of aerosols in a marine environment. Subsequently, the AASS test was developed to simulate the effects of acid electrolytes in industrial environments, using a solution of 5% sodium chloride and glacial acetic acid to a pH of 3.1–3.3. The CASS test, meanwhile, accelerates the corrosion effects of the AASS test by raising the test temperature and introducing divalent copper ions into the solution. Table 1 list the relevant test conditions for the three salt spray test methods.

Table 2.1. Conditions for salt spray test methods

Controlling parameters NSS AASS CASS pH of solution 6.5–7.2 3.1–3.3 3.1–3.3 NaCl (%) 5 5 5 Temperature (°C) 35 35 50 Angle of specimens to vertical (°) 15–30 15–30 15–30

The various salt spray tests have been introduced into a wide range of performance standards and, indeed, correlations have been established between the performance of some materials in the salt spray tests and in particular environments. However, no general method for correlating the severity of the salt spray tests with the severity of external environments has been established.

Testing in sulphur dioxide

Chamber tests using sulphur dioxide gas (typically 5 mg m–3) were introduced to further simulate the effects of industrial environments [Fomin, G.S. 2003]. Higher concentrations of the same gas are used to assess the quality of coatings. In fact, ISO 6988:1985 Metallic and Other Non Organic Coatings – Sulphur Dioxide Tests With General Condensation of Moisture calls for 2000 mg m–3 of sulphur dioxide, as well as the introduction of condensation into the test chamber to further accelerate corrosion effects. However, such tests are very severe and cannot be readily correlated with service conditions.

Cyclic tests It is accepted that the standard chamber tests described in the previous section are useful for comparative assessments of materials, but not as the basis for predicting service life. In order to develop chamber tests that may be applicable to life prediction, cyclic tests have been developed. The requirements for these tests are summarized in ISO 14993:2001 Corrosion of Metals and Alloys – Accelerated Testing Involving Cyclic Exposure to Salt Mist, ‘Dry’ and ‘Wet’ Conditions.

One of the tests included in ISO 14993 – the cyclic corrosion test (CCT) – is derived from a study of the corrosion of automobiles (see Table 2 for details).

Particular tests have also been developed for paint coatings, and these are defined in ISO 11997-1:2005 Paints and Varnishes – Determination of Resistances to Cyclic Corrosion Conditions: Part 1 – Wet (Salt Fog)/Dry/Humidity, and ISO 11997-2:2000 Paints and Varhishes – Determination of Resistance to Cyclic Conditions: Part 2 – Wet (Salt Fog)/Dry/Humidity/UV Light.



Table 2.2. CCT conditions according to ISO 14993

Aspect Action Conditions 1 Salt fog

Temperature (°C) Salt solution %

35 ± 2 5 ± 0.5

2 Heating Temperature (°C) RH (%)

60 ± 2 <30

3 Moisture Temperature (°C) RH (%)

50 ± 2 >95

4 Duration of one cycle (h) Salt fog Heating Moisture

8 2 4 2

5 Correlation 45 cycles corresponds to 13 months natural testing on the island of Okinawa, Japan

ISO 11997-1 is designed to simulate conditions in a marine atmosphere and encompasses three representative cycles. Cycle A (see Table 3A) shows good correlation with conditions in the USA and Japan, and Cycle B (see Table 3B) is used in European countries. Cycle C (see Table 3C), on the other hand, shows good correlation with water-based emulsion paints and similar coatings. A solution of 50 g L–1 of NaCl at pH 5–8 is used in Cycles A and B, while a solution of 0.31 g L–1 of NaCl and 4.1 g L–1 of ammonium sulphate is used in Cycle C.

The accelerated test method in ISO 11997-2 includes the application of ultraviolet radiation during the wetting cycle. Test are carried out in a UV-capable chamber, and a complete test cycle consists of 4 hours of UV radiation at 60°C and then 4 hours of water at 50°C (repeated to a total of 168 hours), followed by 1 hour of salt spray (0.05% NaCl and 0.35% ammonium sulphate) at 24°C and 1 hour of drying at 35°C (repeated to a total of 168 hours).

Table 2.3A. Conditions for Cycle A of ISO 11997-1

Step Time (minutes)

Temperature (°C)

Conditions

1 10 35 ± 2 Salt fog 2 130 60 ± 2 Dry 3 15 50 ± 2 Dry 4 75 60 ± 2 95–100% RH 5 145 60 ± 2 Dry 6 15 50 ± 2 Dry 7 10 60 ± 2 95–100% RH 8 Steps 5–7 repeated 4 times 9 35 ± 2 Dry 10 Return to step 1



Table 2.3B. Conditions for Cycle B of ISO 11997-1

Step Time (days)

Temperature (°C)

Conditions

1 24 35 ± 2 Salt fog 2 8 40 ± 2 100% RH 3 16 23 ± 2 50 ± 20% RH 4 8 40 ± 2 100% RH 5 16 23 ± 2 50 ± 20% RH 6 8 40 ± 2 100% RH 7 16 23 ± 2 50 ± 20% RH 8 8 40 ± 2 100% RH 9 16 23 ± 2 50 ± 20% RH 10 48 23 ± 2 50 ± 20% RH 11 Return to step 1

Table 2.3C. Conditions for Cycle C of ISO 11997-1

Step Time (minutes)

Temperature (°C)

Conditions

1 210 25 ± 2 Salt fog 2 210 40 ± 2 Dry 3 1470 40 ± 2 75 ± 15% RH 4 102 25 ± 2 Dry 5 210 25 ± 2 Salt fog 6 378 30 ± 2 95–100% RH 7 180 35 ± 2 Dry 8 120 25 ± 2 Dry 9 Return to step 1

2.2.3 Comparison of accelerated tests with service performance The most extensive comparisons of chamber and accelerated tests with field tests have been carried out by the automotive industry, with two studies being particularly important. The American Iron and Steel Institute [Lutze, F. & Shaffer, J.R. 1991] compared a wide range of accelerated tests with on-vehicle tests, and found that the one with the highest correlation was the GM9540P test, which consists of, typically, 40 or 80 cycles (960 or 1,920 hours) of the following sequence:

(a) 10 minutes Salt mist (0.9% NaCl + 0.1% CaCl2 + 0.25 NaHCO3, with a pH of 6–8)

(b) 80 minutes Ambient conditions (25°C, 30 –50% RH)

(c) 10 minutes Salt mist

(d) 80 minutes Ambient conditions

(e) 10 minutes Salt mist

(f) 80 minutes Ambient conditions

(g) 10 minutes Salt mist

(h) 170 minutes Ambient conditions

(i) 8 hours Humidity (95–100% RH) at 49°C

(j) 8 hours Dry off (<30% RH) at 60°C



Strom and Strom [Strom, M. & Strom, G. 1993] found that the Volvo STD 1027 test gave good correlations with on-vehicle field tests for painted and zinc-coated steel. In this test, a cycle consists of high RH (90%) for 8 hours, followed by low RH (45%) for 4 hours, both at a constant temperature of 35°C. This cycle is repeated twice a day, with a five-minute immersion in NaCl solution every Monday and Friday.

While both the automotive tests and that defined in ISO 14993 follow the same principles – interspersing salt dosage, dry periods and periods of high humidity – the chemistry of the salt solutions used varies. For example, the GM test introduces both CaCl2 and NaHCO3 to a NaCl solution, but in fact, greater variation is possible, as indicated with the NSS test. The Japanese Automotive Standards Organization (JASO) Method 609 defines a cyclic test for acid rain conditions, which uses a standard salt solution acidified with HNO3 and H2SO4 to a pH of 3.5. In fact, cyclic tests for environments where there is a significant possibility of the acidification of aqueous deposition onto metals (either acidified rain or aerosol) have not yet been fully developed. In addition, a recent review by Cole et al.[ Cole, I.S., Azmat, N.S., Kanta, A. & Venkatraman, M.] indicates that acidification of aerosol can be quite widespread and can occur at some distance from industrial sources, thus further developments in cyclic test methods to reproduce such environments are required.

A second limitation of current cyclic tests is that they are ‘idealised’ and do not necessarily correspond to actual cycles that occur in service. However, recent work by Cole et al [Cole, I.S. & Ganther, W.D. 2006] [Cole, I.S. & Paterson, D.A. 2006] [Cole, I.S. & Holgate, R. 1995] has mapped out the range of wet periods and drying times that actually occur for exposed metal plates in different environments. The development of cyclic tests that more closely follow actually cycle times would improve the correlation between cyclic tests and real service conditions.

2.2.4 Methods of correlating with service conditions As indicated above, good correlations have been made between cyclic test results and service life. However, at present these correlations are purely empirical in that they are derived from comparisons of damage that occurs in field tests and damage that occurs in service.

A number of other methods [Cole, I.S. 2000] for both the design of cyclic tests and for estimating life from accelerated tests are in use, but are not yet standardized. In principle, a cyclic test should reproduce actually service conditions as closely as possible. The cyclic test can then be shortened and accelerated by:

• Eliminating any periods where degradation is halted or very slow. For example, the atmospheric corrosion rate of bare metals when the surface is dry is up to two orders of magnitude less than when the surface is wet, so dry periods can be significantly reduced in defining a cyclic tests. This can result in a one-half to two-thirds reduction in cycle time, or produce an acceleration factor of 2–3.

• Altering the balance between wet periods and drying periods. While significant corrosion occurs both when a surface is wet and when it is drying, the rate of corrosion during a drying period is up to an order of magnitude faster than during a period of constant wetness. Further, in a typical diurnal cycle a surface is likely to be wet for a period of 6–18 hours, with only one drying period of from 30 minutes to 3 hours. Rather than incorporating only one drying cycle per day, a cyclic test may have 4 or more drying cycles per day.

• Increasing the dosage of aggressive agents. A cyclic test may be further accelerated by increasing the concentration or deposition of aggressive agents such as gaseous



pollutants or marine aerosols. However, care must be taken against overdosaging with aggressive agents, which would lead to changes in the degradation mechanism of the material being tested, thus distorting the relative performance of that material.

These methods can be used to accelerate cyclic tests, and an acceleration factor can be calculated from first principles. This first principle factor can then be correlated with observations of the comparative performance of real materials in the designed cyclic test and in service.

Another means of relating accelerated tests to service performance is through modelling. Currently, there are two relevant basic methods:

• In-service conditions can be partitioned into a number of states (e.g. dry, wet from rain, wet from hygroscopic salts, drying), and chamber tests can be used to developed empirical relationships defining the damage that occurs in each state [Cole, I.S., Linardakis, A. & Ganther, W. 1995] [Cole, I., Neufeld, A., Furman, S.A. & Sherman, N. 2000]. A model of service conditions [Cole, I.S., King, G.A., Trinidad, G.S., Chan, W.Y. & Paterson, D.A. 1999] can then be used to estimate the period in which an in-service component is in each of these states, and this estimation can then combined with the damage relationship for each state to determine the overall in-service damage.

• A generic model is developed from basic principles and is calibrated against cyclic tests. A model of the in-service conditions is developed and used as an input into the calibrated generic model to give an overall estimate of service life.

2.2.5 Conclusion Both constant-condition chamber tests and cyclic tests exist and are standardized internationally. Constant-condition chamber tests cover a wide range of environments, but are difficult to correlate to service life. Cyclic tests, on the other hand, cover a limited range of service conditions, but they can be correlated with service life.



2.2.6 References

Capp, J.A. 1914, A rational test for metallic protective coatings, Proceedings of ASTM, 14, p. 474.

Fomin, G.S. 2003, Encyclopaedia of International Corrosion Standards, Maney Publishing, London, UK.

Lutze, F. & Shaffer, J.R. 1991, Assessment of nine accelerated corrosion tests on the cosmetic corrosion performance of AISI materials – interim report, SAE Transactions, 100(Sect. 5), pp. 1170–1182.

Strom, M. & Strom, G. 1993, A Statistically Designed Study of Atmospheric Corrosion Simulating Automotive Field Conditions Under Laboratory Conditions – Final Report on the AISI Cosmetic Corrosion Set of Materials, SAE Technical Paper 932338, SAE International, Warrendale, PA.

Cole, I.S., Azmat, N.S., Kanta, A. & Venkatraman, M. in press, What really controls the atmospheric corrosion of zinc? The effect of marine aerosols on the atmospheric corrosion of zinc, International Materials Reviews.

Cole, I.S. & Ganther, W.D. 2006, Experimental determination of time taken for openly exposed metal surfaces to dry, Corrosion Engineering Science and Technology, 41(2), pp. 161–167.

Cole, I.S. & Paterson, D.A. 2006, Mathematical models of the dependence of surface temperatures of exposed metal plates on environmental parameters, Corrosion Engineering Science and Technology, 41(1), pp. 67–76.

Cole, I.S. & Holgate, R. 1995, The rate of drying of moisture from a metal surface and its implication for time-of-wetness, Corrosion Science, 37(3), pp. 455–465,

Cole, I.S. 2000, Development of a concept and techniques for durability performance evaluation in tropical countries, in Proceedings Second Asia/Pacific Conference on Durability of Building Systems Harmonised Standards and Evaluation, Institut Teknologi, Indonesia, 10–12 July 2000, keynote address, vol. 1, paper 2.

Cole, I.S., Linardakis, A. & Ganther, W. 1995, Controlled humidity/salt dose tests for the estimation of the durability of masonry ties, Masonry International, 9(1), pp. 11–15.

Cole, I., Neufeld, A., Furman, S.A. & Sherman, N. 2000, Response of 55% aluminium–zinc coated steel to well defined salt doses under controlled environments, presented to Symposium E1 ‘Corrosion and Corrosion Prevention of Low Density Metals and Alloys’, 198th Meeting of The Electrochemical Society, Phoenix, Arizona, USA, 22–27 October.

Cole, I.S., King, G.A., Trinidad, G.S., Chan, W.Y. & Paterson, D.A. 1999, An Australia-wide map of corrosivity: a GIS approach, in Proceedings Eighth International Conference on Durability of Building Materials and Components, Vancouver, Canada, 30 May to 3 June 1999, eds M.A. Lacasse & D.J. Vanier, vol. 2, pp. 901–911, NRC Research Press, Ottawa



2.3 Evaluation of concrete resistance to carbonation and chloride penetration3 It was initially assumed that reinforced concrete (RC) could be considered as an intrinsically durable construction material. However the environmental actions on reinforced concrete structures can lead to a progressive degradation of concrete or steel reinforcement. It appeared that very often durability of reinforced concrete structures was limited by corrosion of steel reinforcement.

Steel in sound concrete is protected by the alkaline solution contained in the pores of the hydrated cement paste and, under this condition, corrosion rate is negligible. Corrosion can, however, take place when the passive film is removed or is locally damaged. This may take place due to carbonation of concrete or to chloride penetration. Carbonation is the neutralization of alkalinity of concrete due to carbon dioxide in the atmosphere; when, it reaches the steel surface, the steel bars are no more passive and they can corrode, provided oxygen and moisture are available. When chloride ions, which are contained for instance in seawater or in common de-icing salts, penetrate the concrete cover and reach a critical level at the depth of the reinforcement, a localized attack can take place [Bertolini, L., Elsener, B., Pedeferri, P., Polder, R.B. 2004].

As far as corrosion of steel is concerned, the service life of a RC structure can be defined as the sum of the initiation time and the propagation time. For carbonation induced corrosion the initiation period ends when the carbonation front reaches the steel reinforcement, while for chloride induced corrosion it can be defined as the time required for the chloride ion concentration to reach a critical threshold at the depth of the outermost steel bars. Once the carbonation front or the critical chloride content have reached the steel surface, the propagation period begins and this terminates when a given limit state (i.e. cracking, spalling, or delamination of concrete) is reached beyond which consequences of corrosion cannot be further tolerated and a repair work is needed.

Several parameters, which are influenced by many factors related to both the aggressive environment and the concrete, are involved in the evaluation of the service life of reinforced concrete structure.

Tests methods have been developed especially for measuring the resistance of concrete to the penetration of carbonation and chloride ions.

2.3.1 Scope of the test method In order to predict the service life of a reinforced concrete structure the knowledge of parameters characterising the materials properties is needed. As a result a reliable estimation of the resistance of concrete to the penetration both of carbonation and of chloride is required. The evaluation of the resistance of concrete to the penetration of aggressive substances can be obtained from accelerated laboratory tests. However these values cannot be used directly to make extrapolations on the future behaviour, since they differ from those that would be obtained in real exposure conditions. Hence, design equations and corrective parameters are required for service life calculation.

The use of an accelerated test method to evaluate the service life of a structure with respect to reinforcement corrosion was firstly proposed in a CEB bulletin of the 80’ and further developed in the framework of a European project named DuraCrete and in FIB Model Code. These models are based on a probabilistic approach similar to that used in the structural design: limit states that indicate the boundary between the desired and the adverse 3 Federica Lollini, Luca Bertolini – Polytechnic of Milan.



behaviour of the structure are defined. Environmental factors are considered as loads acting on the structure, while materials properties are considered as resistances. Design equations have been set to calculate the failure probability as a function of time. In these procedures an attempt to correlate results of short-term tests on concrete with the long term performance of the structure has been made and statistically based corrective factors taking into account the role of different variables are provided.

As far as the penetration of chloride into concrete is concerned, different rapid test methods were developed in the past years in order to measure the diffusion coefficient of chloride, i.e. the inverse of the resistance to the penetration of chloride. Among these methods the Rapid Chloride Migration test [5] is used in the Model Code for Service Life Design by the International Federation of Concrete (FIB) [FIB, 2006], as in the previous DuraCrete Model [DuraCrete, 2000] since it revealed to be simple and reliable. In these manuals also a test method to evaluate the resistance to the penetration of carbonation is proposed, which is based on the exposure of specimens to a CO2-rich atmosphere.

2.3.2 Description of the test methods The resistance to the penetration of chloride can be measured by means of the Rapid Chloride Migration test, according to NT-BUILT 492 standard, on cylinder specimens (type C), cured 28 days [Nordtest, 1999]. A plastic tube has to be mounted coaxially to a 50 mm thick concrete cylinder, and a chloride free solution has to be poured inside. The specimen, laid on an inclined plastic support, has to be placed in a container with a 10% NaCl solution. A potential difference of 30 V has to be applied, the initial current has to be measured and, according to its value, the applied voltage has to be adjusted and the duration of the test determined (6-96 hours). At the end of the test, the specimen has to be split axially, and a 0.1 M AgNO3 solution has to be sprayed on its fracture surface. The chloride diffusion coefficient has to be calculated as a function of the measured average chloride penetration depth.

The concrete resistance to the penetration of carbonation can be determined by accelerated carbonation tests, following, for instance, the procedure described in the FIB Model Code [1]. Concrete specimens, 100x100x500 mm, after demoulding, have to be stored in tap water with a temperature of Tref = 20°C for seven days. Then the specimens have to be removed from the water and stored for 21 further days in a standardised laboratory climate (Tref = 20°C, RHref = 65%). At the age of 28 days the specimens should be placed in a carbonation chamber with the standardised laboratory climate (Tref = 20°C, RHref = 65%) and a carbon dioxide concentration of 2.0% vol. for 28 days. After the exposure in the carbonation chamber the specimens have to be split and the carbonation depth have to be measured at the plane of rupture with an indicator solution consisting of 1.0 g phenolphthalein per litre. The inverse carbonation resistance can be determined according to the measured carbonation depth.

2.3.3 Conclusion Parameters describing the concrete resistance to carbonation or chloride penetration obtained from accelerated laboratory tests cannot be used directly to predict the service life of a RC structure. In order to make extrapolation the coefficients obtained from accelerated tests should be correlated to the behaviour under natural exposure conditions. An estimation of the long-term performance of a given concrete under specific exposure conditions can only be obtained through a combined approach which systematically compares parameters relative to accelerated tests with those relative to real exposure conditions. Test methods have been proposed in the FIB Model Code for Service Life Design, which were described in this chapter, and design equations and parameters were proposed. Nevertheless, parameters introduced in the model need to be tested on a large scale; feed back data that



will come in the future from structures designed with the proposed model codes will be useful with this regard.

2.3.4 References Bertolini, L., Elsener, B., Pedeferri, P., Polder, R.B. 2004. Corrosion of Steel in concrete: Prevention, Diagnosis, Repair, Wiley-VCH, Weinheim.

CEB 1997, New approach to durability design, Bulletin d’information N° 238.

The European Union-Brite Euram III, DuraCrete final technical report, Delft, 2000.

International Federation of Concrete (FIB), Model code for service life design, Bulletin n° 34, 2006.

Nordtest, 1999, NT BUILD 492, “Concrete, mortar and cement-based repair materials: chloride migration coefficient from non-steady state migration experiment”.





2.4 Sulphate resistance of cements4

2.4.1 Scope of the test method This method gives comparison of the various Portland cements and their types in sodium sulphate solution. Method gives information about expansion behaviour of the various Portland cements under sulphate and sulfo-aluminate corrosion. Test method does not simulate real conditions of aggressive solution penetration into concrete. This method can not be used on an assembly or a building component.

2.4.2 Description of the tests method By ISO 15868-1 p.6.4.4used method classifies as short-term exposure method for pre-tests.

Rapid test method: Cement-sand 1:3 mortars were made with the water-cement ratio 0.5 (EN 196). Flat prisms hardened at 20°C in the saturated Ca(OH)2 solution for 14 days. Then Wittekindt flat prisms of 1x4x16 cm were exposed into the 4.4% Na2SO4 solution. Relative expansion due to sulfate attack of prisms was measured after three months storage and expressed as the rate of expansion of prisms stored in the saturated Ca(OH)2 solution. Repeatability of the laboratory tests satisfied.

Method gave results only by using sodium sulfate solution as an aggressive medium.

We have also long term test experience, duration 1-10 years of mortar specimen 4x4x16 cm. 7.9%NaCl, 1,5 and 10%Na2SO4; 1%MgSO4 and 1% (NH4)2 SO4 aggressive solutions were used. Compression and bending strength, changes in dynamic module and expansion were measured. These test results in conjunction with X-ray analyses of corroded material and measurements of capillary pore content of mortar. Results convinced us in importance of capillary pore content of specimen material, besides sulfate resistance of cement used.

2.4.3 Conclusions Cements are sulphate resisting or not?

Test results give comparison of sulphate resistance between various cements tested or exposed to the same conditions.

Is there anything in the tests method that can be helpful to identify a method to compare laboratory aging tests with external exposure data or with data from survey on existing building (rescaling)?

Degradation of concrete in real exposure conditions is influenced by various alternate impacts. This method is not suitable for direct life time prediction.

Is the test method bound (or can the test method be bound) to a specific service life prediction method (i.e. factor method, engineering method, stochastic method)?

No.

Does the tests method give data on performance over time of the material/building component?

No

The decision about service life time needs concurred tests, for example, water resistance and adsorption of concrete made with tested cement, besides sulphate resistance

4 Professor Lembi-Merike Raado - Tallinn University of Technology



2.4.4 References Raado,L.M, Hain,T, Effect of Portland Cement Composition on Sulphate Resistance, 1st Baltic Conference on Silicate Materials, Riga, 2004 pp.75-78; ISSN 1407-7353;

Raado, L, Hain T., Sulphate resistance of Various Portland Cements, 15th Internationale Baustofftagung, 2003, Weimar, Tagungsbericht, Band 2, pp.2-0975-0981; ISBN 3-00-010932-3

Raado,L.,Hain T., Corrosion of Cement and Concrete - Methods of Testing and Evaluation, Management of Durability in the Building Process, Milan 2003, Proceedings, CDROM, pp.1-7;ISBN 88.387.2935.2

2.4.5 Standard test methods As data in table show the meaning of sulphate resistance of cements have been under discussion for many years. In process of preparing of the CEN/TR 15 697 the existing methods of testing were discussed. Sulphate resistance properties of cement can be assessed by preparing realistic concrete specimen and placing them in field conditions. It takes several years. In laboratory tests we can only compare sulphate resistance of various cements. Prediction of service life time using only short time testing above described method is not possible

Table 2.4.1: CEN/TC 51- Standards under development, April, 2008

00051092 CEN/TR 15697:2008

Cement - Performance testing for sulfate resistance - State of the art report

Approved 2008-04

00051081 EN 197-1:2000/prA2

Cement - Part 1: Composition, specifications and conformity criteria for common cements; Amendment A2 (Sulfate resisting cement)

Under Approval

2009-04



2.5 Freezing resistance of concrete5

2.5.1 Scope of the test method Freezing resistance of the various distilled water or NaCl solution immersed concrete samples under alternating freezing- thawing cycles by scaling material. The method can not be used on a building component (common with 12390-9)

Which is the scope of the test method? Does it apply on one material, on an assembly or on a building component?

Method applies on concrete, cement based materials, natural and artificial stones. Method does not apply with building component. Method might be applied on drilled specimens.

Our laboratory has experience of testing freezing resistance of concrete based on internal destruction by method of GOST10060 (previous Soviet Standard) measuring

2.5.2 Description of the tests method This method might be part of described structure of ISO 15 686-2. This method characterises scaling of the water immersed surface material during the alternate freezing- melting process – it means one of various factors causing destruction of concrete structure.

2.5.3 Conclusions Is there anything in the test method that can be helpful to refine general principle for service life prediction tests?

This method is one of pre-tests needed for prediction

Is there anything in the tests method that can be helpful to identify a method to compare laboratory aging tests with external exposure data or with data from survey on existing building (rescaling)?

Only empirical with long time experimental results

s the test method bound (or can the test method be bound) to a specific service life prediction method (i.e. factor method, engineering method, stochastic method)?

It is, for example Finnish by 50

Does the tests method give data on performance over time of the material/building component?

Material, the conditions of service and maintenance of the building component are very variable

2.5.4 References EVS 814:2003 Frost resistance of normal weight concrete. Definitions, specification and test methods, Estonian Standard

2.5.5 Standard test methods The standard applies to a single material / an assembly / a component?

Material

5 Professor Lembi-Merike Raado - Tallinn University of Technology



The standard requires both short and long term exposure?

No

Laboratory aging tests use one or more aging agents?

Standard method uses distilled water and sodium choride solution … Table 2.5.1: CEN/TC51 Published standards (April, 2008)

CEN/TR 15177:2006 Testing the freeze-thaw resistance of concrete - Internal structural damage

-

CEN/TS 12390-9:2006 Testing hardened concrete - Part 9: Freeze-thaw resistance - Scaling

89/106/EEC

EVS 814:2003 Frost resistance of normal weight concrete. Definitions, specification and test methods

(12390-9 elements)

Estonian standard

One single method (we have used or standard methods) could not be used for life time prediction. Most of the construction materials have two different destruction causes:

a. Open capillary porosity (migration of water or aggressive solution inside of the material)

b. Reacting capability of material or its constituents with aggressive medium



2.6 Test methods for masonry walls6 The Durability of Building Components Group of Building Environment Science and Technology (BEST) Department at Politecnico di Milano has undertaken some researches to set up methods for RSL’s evaluation, applying them to building materials and components for external walls. In particular, an experimental programme with a technical solution belonging to the external not load-bearing walls’ family has been planned in collaboration with the Swiss technical experimental laboratory of SUPSI in Lugano. Both accelerated ageing proofs and real exposition ones in two different climatic contexts (Milan and Lugano) have been conducted. Thanks to the comparison between accelerated and natural ageing results it was then possible to evaluate the “time re-scaling factor” and to define, in particular, the Reference Service Life of the external protecting covering layer.

2.6.1 Scope of the test method The scope of the test method is to apply the procedure described in ISO 15686-2 to evaluate RSL of the technical solution considered: a double tile masonry with the layer of thermal insulation inside, in which both surfaces are sprayed with plaster and in which the external one has also a protecting paint; the choice of this technical component is due to its widespread, especially in residential buildings.

2.6.2 Description of the tests method The designed experimental programme investigates, both through accelerated ageing proofs and through real outdoor exposures, the behaviour of two different external protecting paints (a vinylversatic one and an acrylic one) applied on double tile masonry walls with thermal insulation inside. For each typology of paint two concentrations of resin (PVC40 and PVC60) and two climatic contexts (Milan and Lugano) of exposure were taken into consideration, as showed in figure 2.6.1.

Figure 2.6.1 – The experimental programme for external walls with different protecting paints.

6 Professor B. Daniotti, M.eng. S. Lupica Spagnolo, Politecnico di Milano, Building Environment Science and Technology (BEST) Department.



The different kind of painting protection (acrylic or vinylversatic) was chosen in order to study the differences both in the degree of waterproofing and in the degradation’s mechanisms, while the different degree of protection was adopted in order to verify the influence of the chemical composition on Service Life: Powder Volume Concentration, for its own definition, measures the powders’ percentage inside the paint, so that the resin ratio in the paint is given by the complement to 100 of the PVC-value.

Thanks then to the double exposure, it has been possible to evaluate the effects of pollution, because Milano and Lugano have quite similar climatic conditions, but also different pollution’s levels [Daniotti & Lupica Spagnolo 2008]; in both sites they were exposed toward South. Eventually, the different slope (90°, 45°) has been adopted because the 45° configuration should accelerate the degradation and make the effects on the masonry more evident.

During the experimental programme, the most significant performance over time decays and, the damage levels due to climatic agents have been monitored: at regular intervals, characterization measures were made in order to evaluate the effects of natural climatic agents (for outdoor exposure) and of reproduced climatic agents such as rain, dry heat, wet heat, freeze or UV rays (for lab tests).

Through, eventually, the comparison between accelerated and natural ageing results it has been possible to quantify the “time re-scaling factor” and to define, in particular, the Reference Service Life of the external protecting covering layer, according to the procedure described in ISO 15686-2.

LAB TESTS

Accelerated proofs in laboratory started first of all with not protected masonries and then with protected (by paints) masonries: the aim was to study the behaviour of external layers of masonries subjected to artificially reproduced climatic stress. For this reason, four specimens at one time were fitted in the climate chamber, with external surface facing the centre of it, just in front of the origin of ageing agents stress. The whole ageing cycle lasted 6 hours and 25 minutes, including transition phases, and it was structured as in the following table 2.6.1.

TABLE 2.6.1 Accelerated ageing in laboratory cycle composition

Phase Temperature (°C)

Relative humidity (%)

Duration (minutes)

Rain 20 95 60 Freeze -20 95 90 Humid heat 55 95 60 Dry heat + UV 30 40 80

Masonries were periodically subjected to not destructive characterization proofs: visual inspections of paint’s aspect and, trough weighting, the measure of residual water; moreover, at the end of ageing period some destructive proofs trough some core borings: determination of plaster’s compressive strength, adhesion between external plaster and brick, plaster’s porosity, water absorption, vapour diffusion resistance and microstructure.

The synthesis of undertaken measurements for monitoring is showed in table 2.6.2.



Table 2.6.2: Characterization tests for monitoring

DISRUPTIVE NOT DISRUPTIVE Time of

execution T0 + every 150 cycles + Tf T0 + every 25 cycles + Tf

Tests

- Compression / bending strength - Adhesion / tensile strength - Young modulus – biax. resistance - Porosity - Water absorption (capillarity test) - Water vapour permeability - Mercury porosimetry - Microscope analysis

- Photos: degradation survey - Weight loss - Karsten: low pressure water absorption

OUTDOOR EXPOSURE TESTS

In the experimental phase of natural ageing, 8 partial samples for each site (Milan and Lugano) have been tested. These specimens are totally analogous to the ones used in accelerated proofs (composed of, starting from the inside, perforated brickwork coated by a layer of plaster and then a protective film of paint), in order to allow the comparison between the two ageing, obtaining the so-called time-rescaling factor.

In order to evaluate the decay during time for the different specimens a classification of decays has been defined; it is based on a photographical analysis and it can be quantified through a dimensional scale (from 0 to 3): DL0: intact protecting layer (there can be micro cracks due to the presence of air during the

applying and drying step of the paint, but there are not cracks or swellings);

Figure 2.6.2 - Photographs which show two examples of degradation level 0.

DL1: local cracks or broken swellings;

Figure 2.6.3 - Photographs which show two examples of degradation level 1 (broken swelling on the left,

local cracks on the right one).



DL2: many micro crackles and cracks;


DL3: torn bubbles, detaching and saline crystallisation of surface.


A synthesis of the found degradation levels from 2001 to 2007 on each specimens during the outdoor exposure (both in Milan and in Lugano) is showed in the three tables below.

Table 2.6.3: Synthesis of degradation levels for 90° sloped specimens in Milan

VH 90 VL 90 AH 90 AL 90 2001 DL 0 DL 1 DL 0 DL 0 2003 DL 0 DL 2 DL 0 DL 0 2004 DL 1 DL 2 DL 0 DL 1 2005 DL 1 DL 3 DL 0 DL 1 2006 DL 1 DL 3 DL 0 DL 1

2007 DL 2 DL 3 DL 0 DL 2

Table 2.6.4: Synthesis of degradation levels for 90° sloped specimens in Lugano

VH 90 VL 90 AH 90 AL 90 2001 DL 0 DL 0 DL 0 DL 0 2003 DL 0 DL 0 DL 0 DL 0 2004 DL 0 DL 1 DL 0 DL 1 2005 DL 1 DL 1 DL 0 DL 1 2006 DL 1 DL 2 DL 0 DL 1



2007 DL 2 DL 3 DL 1 DL 2

Table 2.6.5: Synthesis of degradation levels for 45° sloped specimens in Milan

VH 45 VL 45 AH 45 AL 45 2001 DL 0 DL 1 DL 0 DL 0 2003 DL 0 DL 1 DL 0 DL 0 2004 DL 1 DL 2 DL 0 DL 1 2005 DL 1 DL 3 DL 1 DL 2 2006 DL 2 DL 3 DL 1 DL 2 2007 DL 3 DL 3 DL 1 DL 2

This kind of classification allows analytical comparisons during time and can be used in implementing diagnostic cards for defining inspection steps on this type of technical solutions; thanks to the comparison among damaging levels found both in Milan and in Lugano, it is possible to sum up the following considerations:

- ageing of sloped at 90° specimens is quite similar for the two localities of exposition; the only important difference is for VL90 specimens (V= vinylversatic resin, L=low concentration of resin) because they are more damaged in Milan than in Lugano;

- ageing in sloped at 45° specimens is faster than in sloped at 90° ones. In order to correctly interpret the results obtained, climatic data of the two different cities have been analysed during the exposition period, noticing that:

- In Lugano the quantity of rain is higher; - In Milan values of pollution (PM10, SO2 and NO2), solar heat and medium temperature

are higher than in Lugano. As a consequence, in particular, vinylversatic paints degrade themselves more quickly in Milan than in Lugano due to thermal stresses.

COMPARISON BETWEEN ACCELERATED AGEING RESULTS AND OUTDOOR NATURAL AGEING ONES FOR REFERENCE SERVICE LIFE’S EVALUATION

Comparison between accelerated ageing results and outdoor ageing ones (analysed on the surface through optical microscope) has given the possibility to define Reference Service Life for the considered paints. Performance decay evaluation of the external water protecting layer was quantified through water absorption measurements.

In particular, specimens covered by high concentration of vinylversatic resin paints show the same levels of damaging after 150 cycles in laboratory and 4 years of outdoor natural exposition: presence of many torn bubbles and detaches. Time-rescaling factor for vinylversatic resin paints has been, therefore, defined: 4 years of natural exposition correspond to 150 cycles in laboratory.

Moreover, after 4 years of outdoor exposition and 150 cycle of artificial one, specimens covered by acrylic paint show a relevant level of water protection, differently from those covered by vinylversatic paints which have nearly lost their performance levels (end of their service life).

To sum up it can be affirmed that:

- Acrylic paints show a better protection than vinylversatic ones;

- High concentration of resin increases the protection;



- Vinylversatic paints with low concentration of resin end their service life after 4 years of outdoor exposition.

2.6.3 Conclusion The described research undertaken for defining RSL has showed the complexity of the method and, in particular, the importance that the evaluation of the technological system’s time behaviour:

- has to be pursued at the level of classes of technical elements;

- has to be analysed separately for each meaningful technological performance of the specific class;

- has to be referred to specific environmental and in use conditions.

The test method gives important information on the performance over time not for the entire component but for a material: the paint. The end of Service Life for the material can be individualized with the lost of its main functional performance, its waterproof.

In the described research undertaken for defining RSL, the methodology of time re-scaling has been used: this means that when the same degradation level is got, a comparison is set between long-term ageing time (field exposure with specimens of the same size of the ones used in short-term ageing) and the ageing cycles reproduced in laboratory tests, in order to gain the rate of ageing cycles per year. It is important to underline that this kind of comparison is reliable only if there is not a big initial error in pre-designing the cycle (the ageing cycle must reproduces the climatic ageing phases in a similar proportion) or the transport error will produce misleading results.

That’s why the preliminary pre-design step has to face with an utmost attention the individualization of stressing agents which can influence Service Life, among these is then necessary to select those which can be reproduced in ageing cycles: to operate a correct re-scaling it is important to compare specimens subjected to the same agents, so even in external natural exposure the preparation phase has to minimize the influence of those agents which are not reproduced in accelerated ageing. When this is not possible, time re-scaling becomes a very delicate operation which can bring to mistakes of evaluation.

In ISO 15686-8 it is recommended that “data based on reference in-use conditions similar to the object-specific in-use conditions should always be sought”, in order to “keep the modifying factors as unified as possible, thus minimizing the probability of error in the ESL due to uncertainty in the way mechanisms of degradation are affected by the modifications; and minimize the probability that a critical property not encompassed by data becomes the terminal critical property”. This involves that not only one value of RSL for a building component should be sought, but a set of Reference Service Lives may be taken into account and the nearest RSL to in-use conditions should be chosen as starting value for calculating ESL: the use of different ageing cycles can reduce the transported error from RSL to ESL.

The choice of two climatic contexts goes towards this direction: the comparison made between climatic data for the two localities is a first fundamental step not only for a correct time re-scaling, but also for a more precise setting of the accelerated ageing cycle and for a better definition of RSL, according to the environmental influence.



2.6.4 References Daniotti, B. & Lupica Spagnolo, S. 2008, Service Life Prediction Tools for buildings’ design and management, in papers of the conference “11DBMC International Conference on Durability of Building Materials and Components”, Istanbul, Turkey.

Daniotti, B. & Lupica Spagnolo, S. 2008, Climatic comparison to analyze different degradation levels in external walls’ outdoor exposure, in papers of the conference “11DBMC International Conference on Durability of Building Materials and Components”, Istanbul, Turkey.

Daniotti, B. & Lupica Spagnolo, S. 2007, Service Life prediction for buildings’ design to plan a sustainable building maintenance, in papers of the conference “Sustainable construction, materials and practices”, Lisbon, Portugal.

Daniotti, B. & Iacono, P. 2005, Evaluating the Service Life of External Walls: a Comparison between Long-Term and Short-Term Exposure, in papers of the conference “10th DBMC”, Lyon, France.

Cole, I.S., Linardakis, A. and Ganther, W. Controlled humidity/salt dose tests for the estimation of the durability of masonry ties, Masonry International. 9, no.1 , p11-15 1995

2.6.5 Standard test methods Thanks to the described research activity at Politecnico di Milano, at the beginning of 2006, the Italian national standard UNI 11156 “Valutazione della durabilità dei componenti edilizi” (Evaluation of the durability of building components) was published. This standard is structured in three parts: “terminologia e definizione dei parametri di valutazione” (terminology and definition of evaluation parameters) – “Metodo per la valutazione della propensione all’affidabilità” (Methods for the evaluation reliability's tendency) – “Metodo per la valutazione della durata (vita utile)” (Method for the evaluation of service life). It is coherent with the ISO international standard (ISO 15686) “Building service life planning”.

This standard can be applied to building components and requires both short and long term exposures; the choice of the number of ageing agents is left to the decision of the researcher, according to information collected during the preliminary definition phase.





2.7 ETICS Cladding: Degradation and Loss in Hygrothermal Performance Achieved with Accelerated Laboratory Ageing7 The aim of this contribution is to provide information about structure and objectives of the ongoing experimental programme for assessing durability of ETICS (External Thermal Insulation Composite Systems with rendering) undertaken by BEST – Politecnico di Milano in 2003 and about the possible contribution of this specific method developed for ETICS in refining and improving the general procedure for Service Life Prediction portrayed in ISO 15686-2. Full description of the test method is provided in [3], [4] and [5], whilst this contribution focuses more on the guidelines of the experimental programme that lead to the test method.

2.7.1 Scope of the test method The scope of the test method is to study the durability of ETICS (i.e. EIFS in North America) applied on a masonry wall. The test method was designed in order to achieve several objectives indeed:

- Specific ones (i.e. for ETICS) o Measurement and survey of loss in performances (mainly hygrothermal ones) o Measurement and survey of degradation evolution o Bases for qualitative degradation models o Suggestions for technology improvement

- General / methodological ones (i.e. useful in refining SLP methodology) o Continuous improving of methods for service life prediction (test method has been

designed in order to provide redundant information and so compare different measurement techniques for the same aspect)

o Bases for design of a Serviceability Limit State Method for building components based on understanding relationships between performance attributes and physical properties of building components (information was collected in order to compare laboratory results and software simulation ones)

Relevant requirements for ETICS were individuated and their response measurement methods have been set in accordance to International, European or Italian standards. New test methods – for specific aspects not standardized – have been designed too. A synoptic description of requirements and test methods is provided in Table 2.7.2. Its structure was chosen referring to ETAG 004 (i.e. European Technical Approval Guideline for ETICS), which portrays the response in performances to the Essential Requirements stated in the European Construction Product Directive (CEE/CPD 106/89).

This test method is intended as a study between a specific and a general one (according to § 6.1.1 - ISO 15686-2), due to the fact that main performances and aspects (loosely specified set of performance requirements) for the building component were individuated and survey and evaluation techniques were set and because the test method was designed in order to obtain complete performance-over-time functions. On the other hand, it must be specified that the complete set of performances was not tested, because not all failure modes occur contemporarily on the same component with the same exposure in the same climate, and some agents cannot be reproduced in laboratory tests or cannot be accelerated or cannot be included in an ageing cycle with other agents.

7 Riccardo Paolini, Politecnico di Milano, Building Environment Science and Technology (BEST) Department



2.7.2 Description of the tests method The test method is based upon the structure suggested in ISO 15686-2. The description of the Definition Phase is provided in Daniotti & Paolini [1] and more widely in Paolini [2], whereas a complete description of the Preparation Phase of the test method and the description of specimens is provided in Daniotti & Paolini [3]. In the following paragraphs a fast overview about the organization of the test method is given.

DEFINITION PHASE

FROM USER NEEDS TO PERFORMANCE REQUIREMENTS

User needs were identified (see Daniotti & Paolini [1]) starting the design process from Essential Requirements of CEE CPD 106/89 and Users’ needs stated in ISO 6242 and then technological requirements (i.e. water tightness, thermal insulation, etc.) were detected according to ETAG 004 and Italian Standards. Thus specifications concerning the required values for thermal resistance were identified according to Italian law at the time when the experimental programme started (i.e. Dlgs 192/2005 - Italian law which acknowledges the European Directive 2002/91/EC on energy performance of buildings). Hence specifications concerning layers and materials were set as a result of the design process and according to specific product standards for the whole system (e.g. EN 13499 for ETICS with polystyrene) and for each material (e.g. EN 13339 for thin aggregates in the mortar of base coat).

BUILDING CONTEXT

Building context is described in Table 2.7.1 with reference to indoor and outdoor environment and in-use conditions as proposed for the Factor Method (ISO 15686-1).

Table 2.7.1: Building context chosen as Reference for the studied building component

Aspect Description Reference Condition Standard In-use conditions

Residential building

No relevant impacts foreseen or chemical agents used in the building

Stress conditions Frame structure

No relevant stress-strain due to stress-strain of structure / foundations / ground

Structural design according to Eurocodes

Temperature Moderate (T [°C] = 23 ± 2) environment. No large variations EN ISO 7730 Indoor

Environment Relative humidity Medium moisture load (class 3) EN ISO 13788 Climatic zone EOTA GD 3 Rain load ISO 15686-7

Macro-climate of the context of Milan Low wind load ENV 1992-2-4

Low pollutants concentration zone ISO 15686-7 Meso-climate Wind protected zone ENV 1992-2-4 Middle part of tall building Driving Rain prEN 13013

Outdoor environment

Micro-climate No particular position (i.e. main section, not under balcony, etc.)

TYPE AND RANGE OF AGENTS

Types of agents were identified according to EOTA Guidance Document 003, ISO 6241 and UNI 8290, whereas the range of agents has been analysed considering climatic data and standard reference (as described in Daniotti et al. [5]). In Table 2.7.2 the basic ageing cycle is reported (wider discussed in [5]). 25 UV cycles, 10 winter cycles and 25 summer cycles compose complete ageing cycles CX, which are therefore assembled in groups of five so to get a macro-cycle TX.



Table 2.7.2: Basic Ageing or Ageing sub-cycles (Note: set point values in bold type).

Climatic chamber Laboratory Basic

Cycle Repeat N° Phase Tair

[°C]

Tsup

[°C]

TH20

[°C] RH [%]

Top,i

[°C] RH [%]

Time

[min]

UV 25 1.1 UV 35 - - 15 ± 2 26 ± 3 60 ± 5 60

2.1 Rain: 1 [lt/m2] 5 ± 1 - 5 ± 1 100 19 ± 2 60 ± 5 60

2.2 Freeze -20 ± 2 - - - 19 ± 2 60 ± 5 180 Winter 10

2.3 Winter heat 30 ± 2 - - 50 ± 2 19 ± 2 50 ± 5 60

3.1 Dry heat 70 ± 5 70 ± 5 - 15 ± 2 26 ± 3 60 ± 5 60

Summer 25 3.2 Rain 20 - 20 100 26 ± 3 60 ± 5 60

MATERIALS CHARACTERIZATION

Materials and sub-components, as described in Daniotti & Paolini [3], were characterized in accordance with product standards, concerning basic features and the ones that had been detected as relevant for degradation thanks to degradation factors and degradation mechanisms analysis. As a general principle, materials and products CE marked were adopted and they were checked in acceptance too. PREPARATION PHASE

IDENTIFICATION OF DEGRADATION AGENTS, MECHANISMS AND EFFECTS

A degradation factors – actions – reactions analysis specifically for ETICS was developed in order to estimate the main degradation mechanisms (an example is provided in Daniotti & Paolini [1]). The description of the failure chain was structured with an approach very close to the one proposed by CIB W086 - Building Pathology (i.e. a detailed description of the failure mode), but the analysis was actually oriented at pointing out which failure modes are relevant in ageing for the studied building component (not the whole spectrum of failure modes). Thus the process was:

a. failure modes b. effects c. hierarchy of effects d. relevant failure modes e. agents involved in each failure mode f. choice of agents that can reproduced in laboratory and can be accelerated g. choice of agents which can be included in a cycle h. critical threshold, for the Working Life of the class of building components, for the

intensity of the agents involved in each critical event, that is the event that repeated in cycles produce fatigue or ageing (i.e. concept similar to the one portrayed in EN 1993-1-9 for fatigue loading of steel).

Due to the fact that the comparison between accelerated exposure and long-term exposure is performed with the comparison of effects (i.e. degradation and loss in performances), this analysis was considered as crucial and its continuous up-dating, even during exposure, was used as a check technique for exposure performing.



CHOICE OF PERFORMANCE CHARACTERISTICS AND EVALUATION TECHNIQUES

Choice of performance characteristics

The choice of performance characteristics was carried out according to a functional analysis (with reference to UNI 11156-2), which consists on studying the spectrum of fundamental functions provided by the building component, which characterize the overall performance and define its durability (according to the definition of durability stated in ISO 15686-1). Thus the performance characteristics which had been chosen concern exterior aspect, hygrothermal and mechanical behaviour (see Table 2.7.4).

Acoustic performances (i.e. Rw - sound reduction index) had not been considered due to the fact that even if the test method is applied on the whole building component, it is focused on ETICS, whose contribution in acoustic transmittance is indeed not negligible (more or less 5 [dB] for an ETICS with 5 cm of mineral wool), but usually neglected in design (usually in acoustic design only the contribution of the mass is taken into account, whilst the acoustic performance of the whole wall system is measured in laboratory). The difference between the dimension necessary for ageing - hygrothermal tests and sound reduction index measurement (at least 10 m2 for walls and between 10 and 20 m2 for ceilings, according to ISO 140-1) and the transport of the specimens from one measurement apparatus to another one are not indeed trivial difficulties. For the same reasons neither reaction to fire had been considered (reaction to fire of single products was evaluated in design phase).

Choice of evaluation techniques

The choice of the techniques useful in evaluation of loss performance and degradation evolution was performed thanks to the degradation factors and mechanisms analysis and a standards analysis. When a measurement method was available specifically for ETICS, this was favoured, whereas when there was not, the measurement method considered as more diffuse (from International to National ones) was chosen. When there was no standard method (such as for un-steady state thermal response or stabilization time of moisture content in base coat and finish coat), specific test procedures (see Daniotti & Paolini [3]) were designed, in particular thanks to the degradation factors and mechanisms analysis and with reference to literature concerning hygrothermal behaviour and standards concerning in-use loads. Hence the hierarchy in choosing evaluation techniques was:

a. Standard test method or technique specific for ETICS (ISO, EOTA / EN, UNI, other standards)

b. General test method or technique (ISO, EOTA / EN, UNI, other standards) c. Designed evaluation technique suited for this test method

Disruptive tests were individuated (to be performed on non-aged samples and on aged samples when the end of Service Life is reached both on large samples (1m2) and on sets of selected smaller scale samples) concerning mechanical behaviour of ETICS and hygrothermal performance.

Only for certain hygrothermal aspects (in table 2.7.3 are on the same line), both a non – disruptive test and a disruptive one were planned in order to register the complete performance-over-time function (not only start and end point or a tiny number of points for the curve) and obtain a comparison between results from disruptive tests on little samples, disruptive tests on cores from the whole building component (1m2) and non-disruptive tests performed on the whole building component. This was helpful too in understanding the influence on degradation of different position inside the chamber, different dimensions of samples and of constraints given by substrate and plastic anchors of the ETICS.

Certain evaluation techniques were moreover chosen due to the fact that they provide redundant information and so comparison in different solicitation conditions is possible. For



instance both SINa and SINb cycles provide decrement factor and time shift, but the range of temperatures is different. On the other hand, in steady state, Infrared thermografy and determination of thermal conductance are complementary. In this case the thermal conductance was measured for cross section only with a heat-flow meter, but even if both an hot-box apparatus with a heat-flow meter too are used (in order to achieve both the total thermal resistance of the wall and the value for the main section), infrared thermografy is necessary due to the fact that degradation is not expected to be homogeneous (particularly where there are cracks in finishing and base coat) and information concerning the overall development of thermal bridges shall be collected.

Table 2.7.3: Evaluation techniques for samples exposed to accelerated laboratory ageing.

Type of test DISRUPTIVE NON - DISRUPTIVE

Time of execution T0 + Tf T0, T1, …, Tf

Suited small specimens [characterization sample + aged sample] 1m2 sample

Specimens Cores from 1m2 specimen [characterization sample + aged sample]

Microscope analysis Photos: degradation survey

Water absorption [EN ISO 15148] Karsten: low pressure water absorption

IRT – Infrared thermography

Hygrothermal Properties Measurement Cycles (performed in this order): SINa (Summer dynamic conditions – external air temperature) SINb (Summer dynamic conditions – sun-air temperature) TI (thermal inertia – on/off forcing) CON (Thermal Conductance) RHst (stabilization time of relative humidity in base coat and finishing).

Water vapour permeability [EN 12086]

Tensile bond strength of adhesive and base coat to insulator [EN 13494]

Tests

Render strip tensile test [ETAG 004 - § 5.5.4.1]

FEEDBACK FROM OTHER STUDIES

Main feedback for the test method was provided by the experience achieved in the previous experimental programme on building components performed by BEST in collaboration with SUPSI (see Chapter 2.6).

PRETESTING PHASE CHECKING MECHANISMS AND LOADS

Mechanisms and loads were tested in order to assess the entity of transient times in performing the ageing cycles with a climatic chamber.



VERIFYING CHOICE OF CHARACTERISTICS AND TECHNIQUES

Preliminary tests were performed in order to verify the possible use of general evaluation techniques on ETICS and define the order of performing measurement tests to avoid mutual influence between measurements.

Table 2.7.4: Synopsis of requirements, test methods and observed features. The designed cycles (SINa, SINb and TI) are described in Daniotti & Paolini [3].

Ess

entia

l R

equi

rem

ent

Tech

nolo

gica

l re

quire

men

t

Inte

rest

ed

char

acte

ristic

Mea

sure

men

t m

etho

d / t

ools

Obs

erve

d fe

atur

e

Spe

cim

ens

Test

Met

hod

Res

ults

an

alys

is

refe

renc

e

T[°C] and RH [%] sensors UR [%] 1000 x

1000 mm Designed

cycle Interstitial condensation control

Water vapour permeability Mass variation

δ [kg/(msPa)]

5000 mm2

5 samples EN 12086

EN 13788

Surface condensation control

surface T[°C] and RH[%], surface rugosity

T[°C] and RH [%] sensors T[°C], RH [%] 1000 x

1000 mm Designed

cycle EN 13788

T[°C] and RH [%] sensors T[°C], RH [%] 1000 x

1000 mm Designed

cycle

Karsten’s method

kg(H2O)/m2 in 1 h

1000 x 1000 mm

NORMAL 44/93

3

Water tightness

Water absorption

Capillary absorption

kg(H2O)/m2 in 24 h

200 x 200 mm 3 samples

EN ISO 15148

Summer transient response

Designed cycle SINa,

SINb Thermal Inertia

Thermal capacity T[°C] sensors

Time shift, decrement factor, time constant(s)

1000 x 1000 mm Designed

cycle TI

EN ISO 13786

Thermal resistance

T[°C] sensors and heat-flowmeter

T(x,t) [°C] ϕ [W/m2]

1000 x 1000 mm ISO 9869

6

Thermal insulation Thermal

insulation continuity

IRT ∇ϕ [W/m2] ∇TSURF [°C]

1000 x 1000 mm EN 13187

Continuity of protective layers

Photo survey

Kind, diffusion and intensity of degradation

1000 x 1000 mm ISO 4628

Base coat tensile strength

Render strip tensile test Cracks width

600 x 100 mm, 3 samples

ETAG 4 § 5.5.4.1

ETAG 4 § 6.5.4.1

Durability

Base coat adhesion to insulator

Tensile bond strength of the adhesive and of the base coat to the thermal insulation material

Tensile load at failure, mode of failure

(cohesive / adhesive)

200 x 200 mm 3 samples

EN 13494



EXPOSURE AND EVALUATION

SHORT-TERM EXPOSURE

Short-term exposure is an accelerated ageing achieved with a climatic chamber which reproduces rain, UV radiation, temperature and relative humidity variation.

Initial results concerning loss in thermal resistance (conductance measurement), variation in capillary water absorption (Karsten’s method), development of thermal bridges on joints between insulation panels (Infrared Thermography) and degradation evolution (photo survey) are portrayed in Daniotti & Paolini [4].

Other results concern dynamic transient response, which are now being evaluated (comparison with results provided by HMT models and lumped parameters simple models). On the other hand, water absorption test are ongoing, whilst tests concerning water vapour permeability, tensile and adhesion tests still have to be performed.

LONG-TERM EXPOSURE

This is a part of the research project which has still to be developed, but the whole test method has been designed and it includes both field exposure and inspection of buildings.

Field exposure

On specimens (of the same typologies of the ones described in [3]) exposed in field both disruptive tests and non disruptive ones will be performed as described in Table 2.7.5.

Table 2.7.5: Evaluation techniques for samples long-term exposed in field

Type of test DISRUPTIVE NON - DISRUPTIVE

Time of execution T0 + Tf T0, T1, …, Tf

Specimens Cores from 1m2 specimen [characterization sample + aged sample] 1m2 sample

Microscope analysis Photos: degradation survey

Water absorption [EN ISO 15148] Karsten: low pressure water absorption

Water vapour permeability [EN 12086]

Tensile bond strength of adhesive and base coat to insulator [EN 13494]

Tests

Render strip tensile test [ETAG 004 - § 5.5.4.1]

Inspection on buildings

Inspections are forecast on buildings (not experimental ones) and degradation level will be measured only with non disruptive tests:

- Photos - Karsten’s method for capillary low pressure water absorption - Infrared thermography

Actually the aim of these tests on real estate is not to record the performance-over-time-function(s) (information concerning craftsmanship, condition during laying, stress-strain of the structure, microclimatic conditions during exposure, etc. are unknown) , but to search for the relationship between different effects and correlation between different failure modes.



ANALYSIS AND INTERPRETATION

Analysis and interpretation of initial results is provided in Daniotti & Paolini [4], but some points are here highlighted.

Comparison between different results achieved on samples subject to laboratory ageing

The test method has been designed in order to provide redundant information and allow comparison between laboratory ageing and outdoor exposure (both of specimens in field and of components in buildings).

The aim is to compare the values of different measurements (achieved with laboratory and in-situ techniques) of the same performance (e.g. water capillary absorption with partial immersion and Karsten’s method) and to compare the mutual influence of different performances. It is sought, in particular, the correlation between an increase in water content of the sample, the loss in thermal resistance and the variation in dynamic thermal response.

On the other hand the interpretation of results about all tested properties and performances will focus on the correlation between degradation levels and loss in performances, in order to provide objective tools useful in in-situ surveying of real estate.

Evaluation techniques in analysing results

In analysing results of hygrothermal tests (SINa, SINb, TI, RHst) comparison between laboratory output and hygrothermal simulations (by means both of simple calculation methods such as lumped parameter approach simple methods and of Heat and Moisture Transport tools) is being carried on.

The main focus in analysing hygrothermal performance is on the comparison between the performance in cross section and in coincidence with joints between insulation panels. Is evaluated particularly the influence of discontinuity due to joints (in terms of thermal bridges, points of penetration of water and mechanical performance) on overall performance of the ETICS.

Time re-scaling

Time re-scaling is to be achieved with the comparison of effects (i.e. degradation levels and loss in performances) obtained with accelerated laboratory exposure and with in field long-term exposure. The comparison of exposure conditions in laboratory and in field is possible thanks to the presence of a micro-climatic station in the exposure field, which records:

- Air temperature - Relative humidity - Wind velocity, direction and frequency - Rain precipitation - Global horizontal irradiance - Diffuse horizontal irradiance - Global vertical illuminance in the four cardinal directions - Global horizontal illuminance

The comparison in this case will not concern only effects, but even agents (comparison between climatic data that lead to the designed ageing cycles and climatic data recorded during long-term exposure). The re-scaling will be performed in order to gain the rate ageing cycle / years (see Figure 2.6.1) and have so a double control tool (on agents and on effects).

2.7.3 Conclusion In the paragraphs below a fast proposal of some points that may be considered in improving SLP methodology is put forward.



POINTS WHICH COULD BE HELFUL IN REFINING GENERAL PRINCIPLE FOR SLP

The features of the test method for assessing ETICS durability presented in this contribution are discussed in phases according to SLP structure of ISO 15686-2.

DEFINITION PHASE

Type and range of agents

The analysis of agents should require both environmental (climate and pollutants) data analysis and mechanical agents analysis and range of agents should be identified according to climatic data analysis and according to the degradation factors and mechanisms analysis.

In ISO 15686-2 the identification of the degradation agents and their intensity is already stated as required for a SLP procedure, but it is meant that “type and intensity distribution of the expected degradation agents based on the knowledge as compiled in accordance with 6.1.2.1 or 6.1.3.1 shall be identified” [ISO 15682-2 § 6.2.2]. Maybe this part of the standard should be improved. Probably not only the identification of the agents and their intensity, but also a wider analysis of the microclimate taken into account may be suitable in order to enhance the repeatability of the test, especially in the comparison between indoor and outdoor exposure. An analysis procedure suitable for durability evaluation would be suitable.

Material characterization

Not only materials shall be characterized, but also design process (in order to certify that building pathologies have been excluded in design phase), construction process (craftsmanship, environmental conditions during laying, initial moisture content, etc.) and management process for the component (i.e. maintenance forecast, etc.). In synthesis this means that the characterization should completely describe the component in order to provide information necessary for identifying reference conditions. This is due to the fact that the final output is the Reference Service Life of the Component and this value must be corrected with Factor Method or other forecasting models (i.e. Engineering Design Methods, etc.) in order to get the Estimated Service Life.

A second issue is storage time of specimens and initial moisture content: preliminary evaluation could be performed with HMT tools in order to determine the storage time necessary for considering negligible the standard deviation of initial moisture content (construction water, adsorbed water in hygroscopic equilibrium during laying, etc.). This point is relevant in reproducing the experiment and achieving results which can be compared.

PREPARATION PHASE

Identification of degradation agents, mechanisms and effects

This point is a crucial in determining the ageing cycles and a loop could be so introduced in the flow diagram, moving this analysis in Definition Phase, because it is thanks to this analysis that type and range of agents are identified, and calling it back again in preparation phase.

Specifications concerning the use of degradation factors and mechanisms analysis should be pointed out:

- composition of the ageing cycles - agents maximum intensity - order of ageing phases in ageing cycles - possible subdivisions of accelerated ageing into phases or steps in order to: - measure properties / performances - survey degradation evolution

The analysis should include advanced analysis techniques such as:



- Simulations with Finite Elements, Heat Air and Moisture Transport models, HMT & FE, advanced degradation models

- Simple preliminary tests (on single agents)

The scope of this analysis is to pre-design the number of single ageing cycles (whilst the intensity of agents is provided by agents-degradation factors analysis) and a Building Pathology approach could be useful. The aim of Building Pathology is to understand the failure mode at qualitative level (in order to avoid the error that produced it in design or construction phase), whereas in case of design of a test method for accelerated ageing the objective of the analysis is to seek for all parameters which are essential for reproducing the ageing phenomena, which kind of acceleration can be performed and which kind of error is encountered in this acceleration (with respect to natural ageing).

Another point that should be taken into account is the composition of basic cycles or phases in ageing (such as thermal shock, freeze-thaw, rain, etc.) into complete cycles. Also there could be the need for introducing drainage periods or reset (for dynamic thermal response). The influence on performance measurements on the same sample due to the order of execution of several performance measurement tests shall be assessed too.

Figure 2.6.1: Process in determining the ageing cycle (described in [5]).

EXPOSURE AND EVALUATION

Accelerated exposure

One important point is how long should ageing cycles be performed. Whilst in structural cases rupture / yielding point is clear, in case of building components with regard to hygrothermal performance, for instance, the end of Service Life is related to performance limits set in design process and cannot not always be fixed. The end of Service Life, and so when ageing shall be stopped, could be identified taking into account:

- The highest value in years foreseen for the Service Life of the material / component (not if the value in years of ageing cycles is still unknown);

- When properties / performances reach a constant value (e.g. ageing causes no more relevant loss in performances / properties alteration) or are quite stable.

Another topic is the mutual influence between subsequent hygrothermal ageing cycles. In particular was noticed that if water is a relevant degradation agent, moisture storage can influence the effect of agents of following ageing phases or cycles. Generally in accelerating



ageing, the effect of one phase should be over before the next one starts. Hence reset and drainage phases should be considered.

Short-term in the field exposure

If moisture is relevant in degradation process and in performance loss of the material / building component outdoor specimens should not be inclined due to the fact that moisture re-distribution in pores could be strongly influenced by gravity force (dependently to pores structure and dimension). In this case outdoor accelerated exposure cannot be performed.

POSSIBLE CONTRIBUTIONS IN IDENTIFYING A METHOD FOR RE-SCALING

The features of the test method portrayed in this contribution that could be helpful in improving rescaling techniques have been introduced in the previous point and are widely discussed in [5].

LINKS TO SPECIFIC SERVICE LIFE PREDICTION METHODS

The test method is bound to the development of an enhanced Factor Method application using factor grids (see Daniotti et al. [6]) and to engineering methods for evaluation of performances-over-time and indirectly Service Life.

PERFORMANCE OVER TIME OF THE BUILDING COMPONENT

The test method was designed indeed to provide data on several performances over time of the building component and allow the assessment of their mutual influence.

2.7.4 References Experimental Programme to Assess ETICS Cladding Durability

[1] Daniotti, B. & Paolini, R. 2005, ‘Durability Design of External Thermal Insulation Composite Systems with Rendering’, 10th DBMC, Lyon 17-20 April 2005

[2] Daniotti, B. & Paolini, R. 2006,‘La valutazione della durabilità di pareti perimetrali con isolamento esterno a cappotto’, in La valutazione della durabilità di pareti perimetrali verticali, ediTecnica editrice, Palermo, pp. 37 – 74

[3] Daniotti, B. & Paolini, R. 2008, ‘ETICS Experimental Programme to Assess ETICS Cladding Durability’, 11th DBMC, Istanbul 11-14 May 2008

[4] Daniotti, B. & Paolini, R. 2008, ‘Evolution of Degradation and Decay in Performance of ETICS’, 11th DBMC, Istanbul 11-14 May 2008

[5] Daniotti, B. & Lupica Spagnolo, S. & Paolini, R. 2008, ‘Climatic data analysis to define accelerated ageing for Reference Service Life evaluation’, 11th DBMC, Istanbul 11-14 May 2008

[6] Daniotti, B. & Lupica Spagnolo, S. & Paolini, R. 2008, ‘Factor Method application using factors’ grids’, 11th DBMC, Istanbul 11-14 May 2008

Studies about durability of ETICS [7] Hugo Hens & Jan Carmeliet, Performance Prediction for Masonry Walls with EIFS

Using Calculation Procedures and Laboratory Testing, Journal of Thermal Envelope and Building Science 2002

[8] Helmut Künzel, Hartwig M. Künzel, Klaus Sedlbauer, Long-term Perfomance of External Thermal Insulation Composite Systems (ETICS), ACTA Architectura 5 (2006) vol.1, pp. 1124.



[9] Barreira, E. & de Freitas, V. P. 2005, ‘Importance of Thermography in the Study of ETICS Finishing Coatings Degradation due to Algae and Mildew Growth’, 10th DBMC, Lyon 17-20 April 2005

[10] Zirkelbach, D., Holm, A., Künzel, H.M. 2005, ‘Influence of temperature and relative humidity on the durability of mineral wool in ETICS’, 10th DBMC, Lyon 17-20 April 2005

[11] Bronski, M.B. 2005, ‘Wall Cladding System Durability Lessons Learned from the Premature Deterioration of Wood-Framed Construction Clad with Exterior Insulation and Finish Systems (EIFS) in the U.S.’, 10th DBMC, Lyon 17-20 April 2005

[12] Sahal, A. N. & Lacasse M. A., 2004, Experimental assessment of water penetration and entry into siding-clad wall specimen, Internal Report, Institute for Research in Construction, National Research Council Canada, 862, (IRC-IR-862).

[13] Nady, M., Saïd, A., Brown, W., Walker, I.S. 2007, Long-Term Field Monitoring of an EIFS Clad Wall, Journal of Thermal Insulation and Building Envelopes, Vol 20, April 1997, pp 320-338

[14] Bomberg, M., Rousseau, M., Desmarais, G., Nicholls, M., Lacasse, M. 2002, Report from Task 2 of MEWS Project - Description of 17 Large Scale Wall Specimens Built for Water Entry Investigation in IRC Dynamic Wall Testing Facility, Research Report, Institute for Research in Construction, National Research Council Canada, 111, pp. 141, October 01, 2002 (RR-111)

[15] Kumaran, K., Lackey, J., Normandin, N., van Reenen, D., Tariku, F., Summary Report From Task 3 of MEWS Project at the Institute for Research in Construction - Hygrothermal Properties of Several Building Materials, Research Report, Institute for Research in Construction, National Research Council Canada, 110, pp. 73, October 01, 2002 (RR-110)

[16] Lacasse, M.A., O’Connor, T.J., Nunes, S., Beaulieu, P. 2003, Report from Task 6 of MEWS Project : Experimental Assessment of Water Penetration and Entry into Wood-Frame Wall Specimens - Final Report, Research Report, Institute for Research in Construction, National Research Council Canada, 133, February 01, 2003 (RR-133)

[17] Mukhopadhyaya, P., Kumaran, K., Tariku, F., van Reenen, D. 2003, Final Report from Task 7 of MEWS Project at the Institute for Research in Construction : Long-Term Performance: Predict the Moisture Management Performance of Wall Systems as a Function of Climate, Material Properties, etc. Through Mathematical Modelling, Research Report, Institute for Research in Construction, National Research Council Canada, 132, pp. 390, February 01, 2003 (RR-132)

[18] Beaulieu, P., Bomberg, M., Cornick, S., Dalgliesh, A., Desmarais, G., Djebbar, R., Kumaran, K., Lacasse, M., Lackey, J., Maref, W., Mukhopadhyaya, P., Nofal, M., Normandin, N., Nicholls, M., O'Connor, T., Quirt, D., Rousseau, M., Said, N., Swinton, M., Tariku, F., van Reenen, D. 2002, Final Report from Task 8 of MEWS Project (T8-03) - Hygrothermal Response of Exterior Wall Systems to Climate Loading: Methodology and Interpretation of Results for Stucco, EIFS, Masonry and Siding-Clad Wood-Frame Walls, Research Report, Institute for Research in Construction, National Research Council Canada, 118, November 01, 2002 (RR-118)

Standards [19] NORMAL 44/93: 1994 – Assorbimento d’acqua a bassa pressione.

[20] ETAG 004 – Edition March 2000. Guideline for European Technical Approval of external thermal insulation composite systems with rendering



[21] ETAG 014 – Edition January 2002. Guideline for European Technical Approval of plastic anchors for fixing of external thermal insulation composite systems with rendering

[22] EOTA Guidance Document 003 – Assessment of working life of products, 1999

[23] ISO 4628: 2003 – Paints and varnishes – Evaluation of degradation of coatings – Designation of quantity and size of defects, and of intensity of uniform changes in appearance

[24] ISO 6241: 1984 – Performance standards in building – Principles for their preparation and factors to be considered

[25] ISO 6242-1: 1992 – Building construction – Expression of users’ requirements. Part 1: Thermal requirements

[26] UNI 8290 – Edilizia residenziale. Sistema tecnologico: analisi dei requisiti.

[27] ISO 9869: 1994 – Thermal insulation – Building elements – In-situ measurement of thermal resistance and thermal transmittance

[28] UNI 11156: 2006 - Valutazione della durabilità dei componenti edilizi

[29] EN 12086: 1997 – Thermal insulating products for building applications – Determination of water vapour transmission properties

[30] EN 13187: 2000 – Thermal performance of buildings – Qualitative detection of thermal irregularities in building envelopes – Infrared method

[31] EN 13494: 2003 – Thermal insulation products for building applications – Determination of the tensile bond strength of the adhesive and of the base coat to the thermal insulation material

[32] EN 13495: 2002 – Thermal insulation products for building applications – Determination of the pull-off resistance of external thermal insulation composite systems (ETICS) (foam block test)

[34] EN 13497: 2002 - Thermal insulation products for building applications - Determination of the resistance to impact of external thermal insulation composite systems (ETICS)

[35] EN 13498: 2002 - Thermal insulation products for building applications - Determination of resistance to penetration of external thermal insulation composite systems (ETICS)

[36] EN 13499: 2003 – Thermal insulation products for buildings – External thermal insulation composite systems (ETICS) based on expanded polystyrene – Specification

[37] ISO 13786: 2007 - Thermal performance of building components — Dynamic thermal characteristics — Calculation methods

[38] EN ISO 13788: 2001 - Hygrothermal performance of building components and building elements – Internal surface temperature to avoid critical surface humidity and interstitial condensation – Calculation methods

[40] EN ISO 15148: 2003 – Hygrothermal performance of building materials and products – Determination of water absorption coefficient by partial immersion

[41] ISO 15686: 2000. Building and constructed assets. Service life planning. Part 1. General Principles



[42] ISO 15686: 2000. Building and constructed assets. Service life planning. Part 2. Service life prediction principles

[43] Decreto Legislativo 19 Agosto 2005, n° 192 – Attuazione della direttiva 2002/91/CE relativa al rendimento energetico nell’edilizia. (Italian Law)

[44] Decreto Legislativo 29 Dicembre 2006, n° 311 e successive integrazioni (Italian Law)

2.7.5 Standard test methods The only standard test method analysed specific for ETICS is ETAG 004, which cannot be considered a proper durability ageing test suited for Service Life Prediction. First it is not influenced by ISO 15686-2 and it does not require both short and long term exposure. Secondly the hygrothermal cycles (described in [5]) are not composed into a complete ageing cycle (freeze – thaw cycle have to be performed if water absorption is more than 0.5 kg/m2 in 24 h and are not performed on the same sample) and UV radiation is not considered (relevant in degradation of paint or polymer binder of finishing coat). Therefore the number of cycles is fixed and insufficient for achieving noticeable effects and evaluating Service Life.



2.8 Performance assessment of external renders on facades8 This contribution includes Portuguese experimental research output on external cementitious facades rendering from an ongoing PhD study of Flores-Colen (from IST - Technical University of Lisbon) related with in-situ tests, and from two master thesis already finished (Sá and Quintela) related with ageing tests, carried out in the Building Physics Laboratory of the Faculty of Engineering of the University of Porto.

Façade rendering is a common cladding system in Portugal. According to 2001 national statistics, rendering represents 61.6% of 2,561,227 Portuguese buildings built between 1946 and 2001 (81% of all Portuguese buildings), followed by architectural concrete (21.3%), stone (11.3%), ceramics (5.5%), and others (0.8%).

Render mortars used in Portuguese recent buildings are of two types: cement and sand mortars in two or three layers, mixed in-situ (current renders), and factory-made mortars, made of cement, sand, admixtures and additives in a single layer (ready-mixed renders), in which the finishing layer is either the render itself (“one-coat mortars”) or other materials (e.g. ceramic tile). The use of ready-mixed mortars started in the 90’s in Portugal and it has increased. In 2005, about 30 manufacturers produced 1.2 million tons of ready-mixed mortars, with a total amount about 90 million euros. In 2006-2015, the reduction of mortars made on site is expected to increase the production of ready-mixed mortars until around 2.5 million tons.

2.8.1 Scope of the test methods The methodology followed in external renders performance assessment is according to the systematic methodology for Service Life Prediction of ISO 15686, and can be synthesized and illustrated by Figure 2.8.1.

The required hardened properties of rendering mortar (required performance criteria) are established by European standard EN 998-1 (2003) (rendering mortars based on inorganic binders use on walls with different fields of use and exposure conditions), and complemented by other technical issues. These properties are assessed through laboratory tests using standard apparatus and specimens.

In-use conditions evaluation includes visual observation of anomalies in facades, in-situ measurements through expedient techniques and lab measurements with samples collected on-site.

The accelerated ageing cycles characterize the hygrothermic character of mechanisms and agents of degradation, namely: variation in temperature, variation in relative humidity, incidence of solar radiation and precipitation. These actions will have a determining role in the ageing of external facades coverings.

The general main issues that have been subject of research for external facades rendering systems are the following:

• how can required rendering performance established at design stage be related with the supplied performance (condition assessment) during the post-occupancy phase?;

• how can in-use conditions (natural ageing) be related with accelerated exposure, through in-situ testing, lab measurements and accelerated artificial ageing tests?;

8 Inês Flores-Colen*, Jorge de Brito*, Vasco Peixoto de Freitas**, Helena Corvacho**, Ana Vaz Sá** and Marisa Antunes Quintela** - * Department of Civil Engineering and Architecture (DECivil-IST), Technical University of Lisbon, Portugal ** Building Physics Laboratory (LFC), Faculty of Engineering (FEUP), University of Porto, Portugal



• what are the main aspects that influence the reliability of rendering performance over-time evaluation and therefore the application of service life prediction models for these systems?

Figure 2.8.1: Methodology followed to assess performance over-time of external renders



2.8.2 Description of the tests method In this chapter in order to discuss the main ideas only a few tests are described, namely: pull-off test, compressive strength with small samples, open porosity, and ultra-sound test. The ageing cycles performed are also described.

IN-SITU AND LAB MEASUREMENTS

The adhesion strength (fA) is determined by the tensile extraction test (pull-off), Fig. 2.8.2, left, according to the EN 10155-12:2003, taking into account the recommendation MDT.D.3 of RILEM (2004) to evaluate in-situ adherence. The test program included five tensile extraction tests on the small-scale models for each product, using square metallic test piece with dimensions 5 x 5 cm2. The device that was used measured the force necessary to extract the disk. The pulling-off resistance was calculated using equation 1.1. The types of the rupture (cohesive or adhesive) were also observed.

After pull-off test and adequate preparation (remove the metallic piece and remaining glue), three samples were submitted a compressive tests (Fig. 2.8.2, centre). The other samples were used on open porosity determination (Fig. 2.8.2, right).

…..…………………………………..…Eq.1.1

where: Fu (rupture force, in N), Asquare (area of a square test piece, in mm2). Figure 2.8.2: Pull-off test (left), compressive strength test with small samples (centre), and samples for

open porosity determination (right).

The open or apparent porosity (Popen) expresses the percentage of the relationship of the volume of open pores in the test specimen to its exterior volume, and can be measured by microscopy and intrusion techniques. The simplest indirect porosity measurement is total pore volume determination by vacuum saturation, where only pores that are interconnected are measured. This technique does not give any information on size and shape of the pores. In this testing program the open porosity (Popen) was measured by immersion in water (Archimedes’ principle), according to equation 1.2 from ASTM C 830 (2006), but with the following adaptations: four samples with dimensions 2.5 x 2.5 x 2.5 cm3 (cut from the samples that were obtained in the adhesion tests on the models) and process of saturation (samples were measured after having been subjected a higher pressure of 88 cm water column for at least 7 days), using a pressure pump (Fig. 2.8.3, left).

This procedure was set to guarantee that water penetrates in all open pores, even the ones that could be partially occluded, giving probably values close to the absolute porosity.

…..……………………………Eq.1.2



where:

W (saturated weight, when two consecutive weightings do not differ by more than ±0.05 g),

D (dry weight, in the oven at a temperature of 70 ºC),

S (suspended weight, after saturation and while suspended in water, Fig. 2.8.3, centre).

Figure 2.8.3: Pressure pump (left), determination of suspended weight (centre), and transit time

measurement on the surface of the model (right).

The apparent ultrasonic pulse velocity (Vapparent) was calculated by using ultra-sound equipment that generates low frequency ultrasonic pulses and measures the time taken for them to travel from one transducer to the other through the material tested (Fig. 2.8.3, right). The transducers were placed on five points on the surface of the render (indirect or surface transmission) with a distance between them of 100 mm. The apparent ultrasonic pulse velocity was determined from equation 1.3.

…..………………………………Eq.1.3

where:

d (path length between transducers, in mm) and T (transit time, in µs).

ACCELERATED AGEING CYCLES

In the study of the one-coat rendering mortar, brick models covered with this mortar were submitted to three different processes of ageing: the combination of Cycles A; Cycle B and Cycle C. The combination of cycles A corresponds to an adaptation of the process of ageing referenced in the European Standard 1015-21 (2002) and consists of the association of cycles of heating/freezing and of humidification/freezing, separated by a time of rest in the laboratory atmosphere, Table 2.8.1 portrays this combination.

Table 2.8.1: Cycles’ combination A

Cycle A1 heating/freezing

Standard environment (laboratory environment)

Cycle A2 humidification/freezing

T RH D T RH D T RH D 60ºC n.c. 8 h Immersion in water 8 h 20ºC 65 % ½ h 20ºC 65 % ½ h -15ºC n.c. 15 h -15ºC n.c. 15 h 20ºC 65 % ½ h

+ 20ºC 65 % 48 h

+

20ºC 65 % ½ h T: temperature [ºC]; RH: Relative Humidity [%]; D: duration of the established conditions

[h]; n.c.: non-controlled (Relative Humidity)



The second ageing process, which can be observed in Figure 2.8.4, consists of the repetition of a cycle (Cycle B) that involves the variation of the conditions of temperature and relative humidity for 24 hours, determined through the analysis of the records of recent years obtained by the meteorology stations of LFC (Building Physics Laboratory), taking into account, for the definition of the temperature variation, the heating of the surfaces exposed to solar radiation.

Finally, Cycle C consists of the variation of conditions of temperature and relative humidity, associated with the effects of rain (spraying with water) and of solar radiation (Xenon arc lamp), having a duration base of 12 hours. These conditions were achieved using a chamber specialized for this type of testing. Simultaneously, stations were set up for natural ageing consisting of a wall covered with different types of one-coat mortar, both exposed to the external environment. Similarly to that performed in the laboratory, adhesion tests were performed on these walls, in order to determine the time period.

Figure 2.8.4: Cycle B.

2.8.3 Conclusion DISCUSSION OF IN-SITU AND LAB MEASUREMENTS

The application of rendering on-site performance assessment based on quantitative measurements, instead of qualitative assessment based on visual observations, has been difficult. The in-situ techniques that have been used in field assessment (see Figure 2.8.1) do not provide a direct correlation with the performance requirements for rendering in European standard EN 998-1 or other technical documents; most of these criteria result from laboratory tests. Therefore, it is important for in-service performance evaluation that in-situ characterization of each render includes relevant properties that are measurable on-site, complementing (or correlating with) standard performance requirements. The experimental research carried out with the support of a manufacturer, included the study of different types of render systems on brick walls: pre-mixed mortars: PM (one-coat render), PP (render to receive heavy finishing, such as ceramic tile), PL (render to receive light finishing, like paint); and made on-site mortars: PC (cement binder) and PB (lime and cement binders). This testing program included the following hardened rendering properties: apparent dry bulk density (ρA), compressive strength (RC), dynamic elastic modulus (Ed), open porosity (Popen), adhesion strength (fA) and apparent ultrasonic pulse velocity (Vapparent). The first three were determined by standard testing. The remaining properties were obtained by techniques that can be applied on-site (in this testing program those tests occurred on the surface of small-scale models or on small samples collected from those models).

The results have shown that the open porosity of small samples (using Archimedes principle technique and higher pressure in saturation process) can be used as an on-site indicator of render type, and of hardened standard properties (compressive strength Rexp

2 = 0.981 and dynamic elastic modulus, Rlog

2 = 0.917, see Fig. 2.8.5). However, the establishment of a minimum value of this property is difficult to all render systems. Therefore, a reference value



of open porosity for render systems is proposed, distinguishing two main groups. The first one includes the majority of pre-mixed mortars (one-coat renders without finishing or to receive light finishing), like PM and PL mortars. In this group of mortars it is to be expected (in small samples collected on site) values of open porosity higher than 20%, compressive strength lower than 5 MPa (categories CSI, CSII or CSIII according to EN 998-1), and dynamic elastic modulus lower than 10 000 MPa. The second group includes all made on-site renders and some ready-mixed mortars (mostly to receive heavy finishing), and the expected values are open porosity lower than 20%, compressive strength (CSIII or CSIV according to EN 998-1) higher than 5 MPa, and dynamic elastic modulus higher than 10 000 MPa (for example, wall with high exposure to impact loads).

The porosity is related with the compacity of render. The results have also shown the relationship between open porosity and apparent dry bulk density with Rlinear

2 = 0.725, with for example for first group: Popen < 20% ⇒ ρA < 1500 kg/m3.

The results have also shown that the pull-off test can be an useful on-site technique, because its results can be compared with standard values (according to EN 998-1 and LNEC 427/05, it should be fA ≥ 0.3 MPa or a cohesive fracture pattern); it also can be used as an indicator of compressive strength of the render (the increase of compressive strength of in-situ samples when compared with standard samples can be due to application procedure changes, Fig. 2.8.6, left).

The results have also shown a good correlation between apparent ultrasonic pulse velocity and dynamic elastic modulus (R2 = 0.8804, with Vapparent < 3.5 km/s ⇒ Ed < 10 000 MPa for the first group of mortars, see Fig. 2.8.6, right), despite using indirect method (surface transmission). From literature, the ultrasonic waves are directly influenced by its elastic parameters (relation of shear velocities and Young’s modulus for homogeneous material), and elastic modulus depend on porosity, therefore pulse velocity decreases with porosity increase; this tendency was observed with relation R2 = 0.676 between Vapparent and Popen.

The experimental campaign on sound samples allowed the specification of mechanical strength indicators that are measurable through in-situ expedient techniques or in laboratory with samples collected on-site, in terms of apparent ultrasonic pulse velocity, adhesion strength, open porosity and compressive strength, despite the high variability of the material under study. Results have also shown that the extrapolation of standard hardened properties (dynamic elastic modulus, apparent dry bulk density and compressive strength) from previous on-site properties, and their comparison with required performance can be made.

In conclusion, the integration of several on-site measurements can help performance over time and required performance definition at design stage. Finally, it is concluded that there is a potential to follow this methodology (with expedient techniques instead of more advanced techniques that can increase costs and time consuming of inspections) of rendering condition assessment, even though its reliability can only be acquired through wider and longer experimental campaigns. The actual performance in-situ during early age of rendering façade is yet to be fully understood (an important step to maintenance planning), therefore, more studies and further research have been conducted in sound and damaged samples (collected on site).

DISCUSSION OF ACCELERATED AGEING TESTS AND SERVICE LIFE MODEL APPLICATION ON RENDERS SYSTEMS

The main difficulty of accelerated ageing tests consists in the interpretation of the results, in what concerns their correspondence to real time. The correlation between accelerated ageing and natural ageing on mortars lead to the prediction model on Figure 2.8.7.


CIBW80WG3Report_final_rev01.docx Final page 55 of 102

Figure 2.8.5: Relationship between open porosity and compressive strength (left) and dynamic elastic modulus (right) from test results.

Figure 2.8.6: Relationship between compressive strength of in-situ and standardized samples (left), and apparent ultrasonic pulse velocity and dynamic elastic modulus (right) from test results.


page 56 of 102 CIB W80 Prediction of Service Life of Building Materials and Components

In this way, given that the type of degradation resulting from either one or the other test is seen to be similar, it is possible to interpret these results and develop a model of service life prediction of the products and systems being studied through the establishment of a correlation between the number of cycles of artificial ageing and the real time of natural ageing. Knowing the degradation curve of a given characteristic (as an example, the Figure 2.8.7 considers the adhesive strength between the mortar and its substrate) it will be possible to establish a correspondence between the number of accelerated artificial ageing cycles and the number of years in real time. Figure 2.8.7: Correlation between the number of accelerated artificial ageing cycles and real time natural

exposure - Prediction Model.

With the adequate performance characteristics chosen (in this case adhesion strength) and the necessary artificial accelerated ageing tests performed, another problem could still arise: the results obtained for different stages of ageing do not reveal any definable tendency, making it impossible to draw a degradation curve. In this case, the hypothesis that there are other aspects that are more relevant to ageing of the material or system in study becomes plausible: the agents of degradation being considered may not be very relevant for the variation of the characteristic being tested as it is overall dependent on, for example, the execution, on the interaction between different elements of the construction (interfaces), on the type of in-use conditions, among others. In this case, the application of the model as presented becomes unviable.

The experimental results on cementitious adhesive in bonded ceramic coverings have shown the variation in adhesive strength as function of three number of ageing cycles to which the specimens designated as PE0, PE1 and PE3 had been subjected. In these three specimens cementitious adhesive of type C2 (according to EN 12004) was used along with three different types of ceramic tiles. There is a very clear decreasing trend (Figure 2.8.8, left) and it can be predicted that, independent of the type of ceramic tile used, in the set of tests performed on the specimens with class C2 cementitious adhesive, 140 cycles would be sufficient to drop below the critical adhesive strength value of 0.3 N/mm2 and thereby reach the service life term of the product. Regarding the tests performed on the wall of the natural ageing station, given the little time of exposure that passed by until the performance of the first tests, it was only intended to demonstrate the methodology to follow to establish the correlation between the results obtained from accelerated ageing and those obtained with natural ageing.

As regards the durability of the one-coat rendering mortar, Figure 2.8.8, right, displays the variation in the adhesive strength of the single layer covering, applied with 1 or 2 cm of thickness, on the brick models submitted to the accelerated ageing cycles. It can be observe a random variation in the performance of the one-coat rendering mortar in terms of adhesion to the substrate, being impossible to identify any tendency for any one of the ageing


CIB W80 Prediction of Service Life of Building Materials and Components page 57 of 102

processes. At the natural ageing station, two series of in-situ adhesion tests were performed. Between the first series of tests and the second series of tests a tendency of diminished adhesion to the substrate surface was observed. However, the observed tendency does not allow coming to a conclusion because the time that had passed since the execution of the wall was still too short.

In conclusion, for the same accelerated ageing tests, the results for adhesion do not follow the same trend over time for cementitious adhesive and on-coat render. As regards the cementitious adhesive applied in a system of adherent ceramic covering, once it is possible to obtain a degradation curve in the accelerated ageing tests, to be able to reliably apply the model, it becomes necessary to only perform the tests in the natural ageing station, for defined time periods (every six months, for example) in order to be able to establish a relationship between the short term tests and the long term tests. Therefore, it will be possible to collect the performance data for in-service conditions of this system, although total characterization would not always be easy in terms of historical application and utilization. For this case study, it will be possible to completely apply a prediction model in a short time. Figure 2.8.8: Service life term prediction for C2 cementitious adhesive (left), and evolution of the adhesive

strength of the single layer covering (right) in the models submitted to the three ageing processes.

In the case of the one-coat rendering mortar the characteristic that was considered to be significant for the definition of its durability, adhesion to the substrate, presented a behavior that significantly varied with time, unable to result in a degradation curve. The evaluation of the situation led to the conclusion that there are several exogenous factors that significantly affect the performance. In this case, the application conditions are determinant, as well as the roll of the interfaces between layers. This fact made the application of the adopted prediction model unviable in this case.

For the development and the future application of the prediction model to different systems and materials, campaigns of accelerated ageing tests must be taken into consideration accompanied by natural ageing tests and the systematic collection of data from existing buildings. For the adopted strategies to be actually efficient, a rigorous identification of the relevant characteristics to be tested (those that are seen to be critical to durability) is very important and also the performance of previous tests for the gauging of the expected variations and for the identification of the most influential agents. After a service life prediction has been obtained for specified conditions, where we would be able to have the service life as a reference point, it becomes necessary to evaluate the influence of the specific conditions of each real situation. For this, for example, the factorial method proposed in the ISO standard 15686-1 or other equivalent method could be adopted.



2.8.4 References Flores-Colen, I; Brito, J. de; Freitas, V. P - Condition assessment of façade rendering through in-situ testing, 11 DBMC, Istanbul, Turkey, May 11-14, 2008

Freitas, V.P.; Corvacho, H; Sá, A. V.; Quintela, M. A. - Discussing the durability assessment of cement mortars - a contribution for a prediction model, 11 DBMC, Istanbul, Turkey, May 11-14, 2008

Flores-Colen, I.; Brito, J. de; Branco, F. - In-situ adherence evaluation of coating materials, Journal of Experimental Techniques (accepted for publication in 11 April 2008)

Flores-Colen, I; Brito, J. de; Freitas, V. P - Stains In Facades’ Rendering - Diagnosis And Maintenance Techniques’ Classification, Journal of Construction and Building Materials, 22 (2008), 211-221.

Flores-Colen, I.; Brito, J. de; Freitas, V. P. - Expedient in situ test techniques for predictive maintenance of rendered façades, Journal of Building Appraisal, V. 2, No. 2, June 2006, pp. 142-157.

2.8.5 Standard test methods Two aspects should be discussed. First, the minimum requirements of EN 998-1 according to other author’s are not enough to evaluate the global behaviour of external rendering. In fact these test methods are applied on standardized samples and relevant properties are not included (such as dynamic elastic modulus). Auxiliary measurements can significantly help the performance over time of renders systems, such as expedient measurements of open porosity and compressive strength on samples collected on site, and apparent ultrasonic pulse velocity measured on façade rendering (indirect method).

Second, the pull-off technique, despite of recommendation of EN 1015-12 and RILEM MDT.D.3, can still be improved in terms of in-service performance assessment of facades, in the following terms. The case studies analyzed in a experimental campaign on rendering systems (one-coat renders, multi-coat renders, with light paint finishing and with ceramic tile finishing) allowed the identification of the potential of the in situ application of the pull-off technique for degradation cases of loss of adherence / detachment and for quality control. It was found out that only in a few cases the analysis of the average values of the adherence tension by itself was enough to allow a diagnosis of the situation, due to the high scatter of values and different types of rupture modes (adhesive in the interfaces and cohesive within the materials that make up the coating system). As a matter of fact, only the evaluation of the rupture mode (and its incidence in each case) and of the respective tension allowed pursuing complete knowledge of the causes of degradation or of deficient quality, very often associated with execution errors hugely different from one case to the other. On the other hand, since the technique is destructive, this allowed its use as a “probing” technique, leading to the analysis of the specimens and of their background after the pulling-off had been carried through. This observation was revealed as very important to justify minimum tension values essentially due to localized gross execution errors, thus complementing visual observation, the project analysis and the collection of data on site. Finally, it was concluded that the choice of the number of tests to be performed and of their location is crucial. The number of valid tests, between 5 and 10, was enough in most cases for a qualitative diagnosis. However, each case must be analyzed individually, taking into account the evaluation of the balance between human and economic resources available and the achievement of the established goal (with as little damage to the existing coating system as possible).



2.9 Test methods for the durability evaluation of pitched roof9 The research carried out by the Unit of Palermo, “Planning and experimental evaluation methodologies of discontinuous roofing class durability”, becomes part in the more general picture of evaluation and control of the technological quality of complex products for the building and in particular of durability, fundamental to reach the quality in building (ISO 8402, UNI 10838), cause it regards maintenance over time of the performances owned at the moment of the entrance in exercise and the modalities in which them decay during time.

For the research result fundamentals the normative contributions, in international field of CIB W080/RILEM 175 (Service Life Methodologies) and ISO TC59-Buildings Construction, with the ISO 15686 Building and Constructed Assets Service Life Planning, and in national field UNI11156 - 2006, “Evaluation of durability in buildings components” approved recently.

2.9.1 Objectives of the experiment The aim of the experiment is to study the behaviour over time of solutions for sloping roofs, in particular sandwich panels with double sheets and an insulating core, in order to test the methods for predicting the service life of these components.

The roof is a system made of many functional elements, each one defines the performances that it must furnish (watertight, thermal insulation, mechanical strength, etc.).

The configuration of a discontinuous roofing results from different factors interdependent among them: functional factors; climatic; contextual; formal; specific of the seal.

The group of functional elements, every with its own physical and chemical characteristics and the mutual position, determines the behaviour of the roof.

The main requirements that concern the class of the discontinuous roofing and particularly the coverage realized with metallic (UNI 10372) elements are: mechanical strength and stability, phase displacement and damping, thermal insulation, control of surface condensation, sound-proofing, watertight, resistance and reaction to the fire, stain resistance, resistance to frost and thermal shock, resistance to ultraviolet radiation, resistance to biological agents, resistance to weathering.

A repertoire of 24 technical solutions has been studied, which includes also representative solutions of the local uses, as well as solutions "innovative" to study it and to compare its characters in comparison to those "traditional".

Another objective is to study the influence that the different colour of the external surface can have on the deterioration in the performance of sandwich panels and their component parts, as a result of the effects of aging.

2.9.2 Description of the tests method The methodology follows the evaluation with two distinguished and complementary procedures of the two parameters of durability: reliability and natural durability.

Reliability is estimated through an object and functional analysis that permit a qualitative evaluation followed by a comparative procedure lead among elements of a catalogue of technical solutions of the discontinuous roofing class, which is considered out of system and with no intended use.

9 Giuseppe Alaimo, Francesco Accurso, Università degli Studi di Palermo, Dipartimento di Progetto e Costruzione Edilizia.



Natural durability is estimated through experimental tests on samples of technical elements exposed to natural agents and accelerated aging test in laboratory.

We choose sandwich panel, as the building components for our research10.

The reliability has been evaluated by the method proposed from Rejna [1995], inserted in the UNI 11156:2006, based on the functional and technical analysis of the element, proceeds through the evaluation, with precise criterions of judgment, of the four components of the reliability (fig. 2.9.1):

- functional reliability, through the functional analysis, indentifies the distribution of the functions inside the solution, on which the degree of fatigue of the element depends during its operation;

- executive reliability, through the technical analysis, verifies the degree of executive correspondence in comparison to the design intensions;

- inherent reliability concerns the possible dimensional changes of the component in phase of exercise;

- critical reliability considers physical-chemical compatibility between material adjacent of different nature constituent the technical solution.

The propensity to the global reliability (fig. 2.9.2) expresses in qualitative and comparative terms the degree of probability that the technical solutions have to last in time. Such values result from the arithmetic average among the correspondents values of reliability of the four components.

On the basis of the range of the fixed values, it is possible to perform with comparison: the design of a solution of given reliability, or to esteem the reliability of a given solution, different from those to the repertoire.

Fig. 2.9.1 - Methodology for the assessment of reliability

10 For this reason have been developed the contacts with AIPPEG (Associazione Italiana Produttori Pannelli ed Elementi Grecati) that kindly has offered its collaboration.

Class of technological

elements

Technological requirements

Functional analysis

Technical analysis

Dimensional changes

Functional reliability

RELIABILITY

Functional characteristic

s

Technical structure

Chemical-physical compatibility

Executive reliability

Inherent reliability

Critical reliability



Fig. 2.9.2 - Histograms of propensity to reliability of technical solutions

The study has been conducted on samples of measures 100x100, 100x250, 100x1000 mm, cut by standard panels as in fig. 2.9.3.

Fig. 2.9.3 - Cross-section of whole panel and cutting of specimens

The typical panel is constituted by a double metal skin of pre-painted galvanized steel and from an insulating core of polyurethane expanded (PUR) and thickness of 40 mm.

The colours of external surface are: White Grey and Siena Red.

Tab. 2.9.1 - Characteristics of samples

Type 1 (roof) Type 2 (wall) Colour Siena Red White-Grey

Side A Pre-painted galvanized steel (0,45 mm) with polyester (25µ)

Pre-painted galvanized steel (0,45 mm) with polyester (25µ)

Outside metal skin

Side B Primer (5µ) Primer (5µ) Foam 40 mm PUR 40 mm PUR

Colour White-Grey White-Grey

Side A Pre-painted galvanized steel (0,35 mm) with polyester (25µ)

Pre-painted galvanized steel (0,45 mm) with polyester (25µ)

Inside metal skin

Side B Primer (5µ) Primer (5µ) Necessary methodological preamble to the experimental phase for the program of accelerated aging cycle, is the definition of the reference frame of the agents - actions - effects11 and therefore of the correspondents deteriorations that they can rise up in the components on the basis of the specific climatic context.

On the samples (new, aged in external and in laboratory) destructive and non-destructive tests have been conducted to check the following parameters and characteristics: aspect;

11 For term definitions please refer to the glossary of CIB W86 Building Pathology.



weight; colour; thermo-physical characteristics (thermal conductivity and thermal resistance); mechanical (traction, shearing and bending) characteristics.

NATURAL AGING The samples of sandwich panel has been exposed outdoor starting from March 2004, directed toward South and with slope of 30° in comparison (fig. 2.9.4a). Before the exposure, edges of the specimens have been protected with a special polyurethane paint.

On some samples programmable data-loggers have been applied (fig. 2.9.4b), in order to record the superficial temperatures reached (exposed and shade side) during the exposure.

The collected data, between March 2004 and March 2006, allowed us to find the real course of the maximum, minimum and averages temperatures reached on the surface of the samples in Palermo during different seasons of the year.

The definition of the characteristics and the parameters (type, intensity, length) of the climatic context constitutes fundamental preamble for the definition and the debugging of the accelerated aging cycle to use in the climatic chamber.

To such purpose a weather stations have been located nearest to the site of exposure of the specimens and climatic data of last ten years have been processed .

Fig. 2.9.4 - The exposure in outdoor and temperature micro-loggers

a b

ACCELERATED AGING From the analysis of the collected data and from the course of the different climatic parameters, four representative seasons of the climatic context of Palermo city are recognized, and particularly: one "summer season" (dry heat); one "autumn season" (rain); one "winter season" (cold); one "spring season" (warm dampness).

For every season the characteristic values of the four principal parameters have been recognized: temperature, relative humidity, rain and solar radiation.

Fig. 2.9.5 - Climatic chamber and temperature loggers

a b

At the end of this phase, kept account of other factors of influence like the slope of the samples, the superficial colour, the changes of temperature in the last decade and the



existing normative on the thermal shock, we reach the following definitive cycle of accelerated aging (tab. 2.9.2, fig. 2.9.6a).

Tab. 2.9.2 - Accelerated aging cycle structure

Parameter Phase hour T (°C) RH (%)

Rain 3.0 - - Cold 1.5 2 - Warm dampness 5.0 35 87 Dry heat 2.5 83 56 Total 12

Fig. 2.9.6 - Accelerated aging cycle: theoretical and experimental

a b

The climatic chamber simulates the principals climatic agents: rain, solar radiation, temperatures, humidity, through an unity of control that allows to program the sequence desired of the agents, the intensity and the duration.

Once fixed the duration of the cycle of accelerated aging (12 hours), the length of each phases result consequently. It is necessary to remember that in the real cycle, between a phase and the following one there are transitory ones, in which temperature and humidity inside in the room before reaching the fixed values, employ a certain time. Nevertheless, for reasons for opportunity such transitions (of the general duration of around 2 h) have been inclusive in the general duration of the cycle (fig. 2.9.6b).

TESTS AND RESULTS The tests, non-destructive (superficial aspect, weight, colour, thermal characteristics on the whole sample) and destructive (traction resistance, bending and shearing, thermal characteristics on the section of PUR), have been made on: samples before the aging process; external aged samples after 1 and 2 years; artificially aged samples after 30, 60 and 120 cycles corresponding to 360, 720 and 1440 hours of accelerated aging.

At the end of every phase of aging are been performed on the samples test turned to appraise the characteristics and parameters changes, in comparison to the samples virgin.

Every test have been conducted at least, on a number of three specimens.



SUPERFICIAL ASPECT The analysis of the aspect don't show substantial differences among the aged samples in external to 24 months and those aged in laboratory up to 60 cycles. Superficial stain and punctual oxidations are primarily developed.

After 120 cycles the samples, also for their limited dimensions, have shown deformations of the form: increase of thickness of 1-2 mms and separations among metal skin and PUR along the edges that as expectation constitute critical points where this degradation have departed despite the protective paint (fig. 2.9.7).

From the comparison among the detected effects a temporal correspondence can be hypothesized between the 60 cycles of accelerated aging and the 24 months of external exposure .

Fig 2.9.7 - Detachment among PUR and metal skin

WEIGHT The sandwich panels in their complex can be considered impermeable. Nevertheless the vapour and, with it, water can migrate inside the insulating layer in polyurethane causing some dimensional variations that influence the shape, the weight and the same performances of the component.

From the comparison of the results among natural aging and accelerated aging a meaningful analogy emerges both qualitative and quantitative among the variations of weight of the samples, very limited and more for the dark samples.

Also under such aspect the effect of the first two years of outdoor aging it results entirely comparable with that caused by the first 60 cycles of aging in the climatic chamber.

COLOUR The colour was monitored periodically, on day one and after 18, 27, 36 and 45 months outside.

The results show a colour shift in terms of ΔΕ. The light and dark samples can be compared.

In fact, after about 45 months both sets of samples show ΔΕ values of between 6 and 7 units. However, after 18 months the light coloured samples already show a marked tendency to change colour compared with the dark ones (ΔΕ equal to about 3 units compared with 1.13 for the dark samples).

In the same way, after 27, 36 and 45 months this tendency persists, but the difference between the values is less clear-cut.



Fig. 2.9.8 - (a) Natural aging; (b) Accelerated aging

a b

The equipment used to simulate aging is a Solarbox 1500e fitted with a Xenon lamp and a filter of 280nm.

In this case samples from the same production batch were used, but 100x100 mm in size.

The measurements were carried out using a colorimeter and a spectrophotometer, with D65 as illuminant to simulate day light.

Each cycle is made up of the following phases:

4 hours of UV (800 W/m²) at a temperature of 65°C; 4 hours of condensation with the lamp switched off and T= 55°C;

repeated three times in the 24 hours. Measurements were taken every 5 cycles up to a total of 110 cycles, equivalent to 2640 hours of exposure in total.

Shifts comparable to the variation in color recorded on the exterior can be observed from the measurements carried out with the colorimeter. In particular on the Siena Red samples, where there is a variation of about 5 units after 110 cycles (equivalent to 2640 hours of aging in total). Smaller variations have been recorded for light colored samples.

The measurements were repeated using a spectrophotometer, for the purpose of evaluating aspects of the aging phenomenon that are not verifiable with the colorimeter.

THERMO-PHYSICAL CHARACTERISTICS The measure of the thermal conductivity of building materials in stationary state was made, according to the standard UNI/CTI 7745 "Determination of the thermal conductivity with the method of the warm plate with mark ring”.

Measures have been conducted both on the whole samples and on the polyurethane sections (using from the core of every samples two sections of the thickness of 10 mm) (fig. 2.9.9), to the purpose to evaluate the influences of the external metallic elements and the interfaces with polyurethane.

Fig. 2.9.9. Sections of PUR and change of conductivity (natural aging)

a b

Up

Down



During 24 months of external exposure the conductivity of the samples has substantially been constant, both in the whole samples and in the sections of PUR.

In general the thermal conductivity of the whole samples results greater than PUR sections, cause the lower contact resistance due to the cover sheets glued to the PUR.

Fig. 2.8.10- Change of conductivity in the whole sample and in the PUR sections (artificial aging test)

a b There is a substantial constancy among the measures on the samples "Up" with those on the samples "Down”.

The results of the tests in climatic chamber show an analogous behaviour from the qualitative point of view among the whole samples (fig. 2.9.10a) and the sections of PUR (fig. 2.9.10b) both for the samples White Grey and for those Siena Red.

Also in this type of tests after the 60 cycles an inversion of tendency is observed with a substantial constancy of the thermal conductivity both in the first year of natural aging, and in the first 30 cycles of accelerated aging.

MECHANICAL STRENGTH The test (fig. 2.9.11) has the purpose to evaluate the tensile strength of the polyurethane foam and its adhesion to the metallic support (ETAG 016-1, UNI EN 1607, EN 14509) and show an improvement of the traction resistance up to 60 cycles, subsequently such tendency is reversed with the increase of cycles (fig. 2.9.12a).

Fig. 2.9.11 - Tensile test and type of failure (PUR)

It seems us interesting to underline the relationship found among the tensile strength, the percentage of uncovered metal skin and the length of the aging. The aging causes in fact an increase of the percentage uncovered after the failure, effect more accented in the dark samples (fig. 2.9.12b).

From the comparison among the effects of the two types of aging, also in this type of tests a correspondence can be noticed among the effects of 60 cycles and those of 2 years of natural exposure.



Fig. 2.9.12 - Tensile test and uncovered metal skin in tensile failure (artificial aging test)

a b

PREDICTING THE USEFUL LIFE OF A PANEL WITH THE FACTORIAL METHOD USING PERFORMANCE GRIDS The factor method proposed by ISO 15686-2 allows an estimate of the service life of the component (ESL) to be made, starting from a known reference value (RSL).

In its various applications the method has proved to be affected by a high level of subjectivity, which has limited its use, although up to now it is the only one provided for by the international regulations.

One possible application of the factor method involves the preparation of performance grids for the attribution of coefficients to different factors, which can be used by designers who want to use the factorial method to determine the duration of a component.

The performance characteristics to be considered are those stated on day one by the producers and by the regulations that refer to that particular product (resistance to tensile stress, resistance to compression, permeability to water, density).

The method requires the indication of the RSL of the component as a starting point and subsequently the adjustment of each of the factors from A to G.

For the purpose of decreasing the subjectivity of the method, performance classes have been provided for each factor. Each performance class corresponds with a coefficient that can be greater than 1 to indicate an improvement in performance compared with the standard condition, or less than one if worsening conditions of the performance of the panel are referred to.

Although this method is of absolutely general validity, it was applied to the sandwich panel, starting from a reference value for its useful life of 10 years and introducing the necessary adjustments for each factor.

In the case of factor A (Quality of the component), seven sub-factors were provided, but since these factors do not have the same importance with regard to the duration of the component, it is necessary to attribute a different importance to each of them. Using the comparison of pairs matrix method and comparing the various sub-factors two by two, it was possible to attribute a weight of importance to each of the factors considered. The conclusion of this was that the characteristics of resistance to tensile stress and compression are the most important.



Tab. 2.9.3 - Performance grid relative to the factor A

Code Sub-factor Factors levels Score

Rtr < 100 KPa 0,90 100 ≤ Rtr ≤ 120 KPa 1,00 121 ≤ Rtr ≤ 150 KPa 1,10

A1 Traction resistance

Rtr > 150 KPa 1,20 Rtr < 120 KPa 0,90

120 ≤ Rtr ≤ 150 KPa 1,00 151 ≤ Rtr ≤ 175 KPa 1,10

A2 Compression resistance

Rtr > 175 KPa 1,20 δ < 34 Kg/m3 0,90

34 ≤ δ ≤ 37 Kg/m3 1,00 A3 Density δ > 37 Kg/m3 1,10 ≥ 300 Pa 0,90 ≥ 600 Pa 1,00 A4 Water permeability ≥ 1200 Pa 1,10 ΔEab* > 8 0,80

5 ≤ ΔEab* ≤ 8 0,90 A5 Colour changing ΔEab* < 5 1,00

FPC (ISO 9000) 0,80 ETA 0,90 A6 Quality

CE marking of products 1,00

Tab. 2.9.4 - Performance grid relative to the others factors

Code Factor Factors levels Score Any certificate 0,90 B Quality of project

According to best practice indications 1,00 Any certificate 0,90 C1 Quality of construction

According to best practice indications 1,00 Not specialized 0,90 C2 Quality of construction: laying

team Specialized 1,00 Very aggressive 0,70

Aggressive 0,80 Middly aggressive 0,90 Low aggressive 1,00

D Indoor environment

Not aggressive 1,10 Industrial and marine 0,70

Industrial 0,80 Marine 0,90 Urban 1,00

E Outdoor environment

Rural 1,10 Environment with bumps against hard

body 0,80

Environment with bumps against soft body 0,90 Environment with no frequent bumps 1,00

F Conditions of use

Environment with rare bumps 1,10 High (< 10 years) 0,90

Medium (10 < t < 20 years) 1,00 G Level of maintenance Low (> 20 years) 1,10

In this way, on the basis of the characteristics of the component chosen for the new design, the designer attributes an adequate level, so as to correct the score of factor A.



Similarly, for the others factors, the following grid was constructed in tab. 2.9.4. At this point it is possible to estimate the useful life of the component (ESL) under the project conditions, using the formula of the factor method:

Tab. 2.9.5 - Assessment of ESL

Component Sandwich panel on discontinuous roofing RSL 10 years

Factor Level Score A According to technical characteristics 1,18 B According to best practice indications 1,00

C Construction in according to best practice indications and by specialized laying team 1,00

D Not aggressive indoor environment 1,10 E Urban environment 1,00 F Environment with rare bumps 1,10 G High level of maintenance 0,90

ESL RSL x A x B x C x D x E x F x G 12,83 years As can be observed, under the project conditions established, the useful life went from 10 to 12.83 years. With the help of grids similar to the ones presented here, in our opinion, the subjectivity of the method is further reduced.

2.9.3 Conclusion The results, even partial, allow to express some considerations on the behaviour over time of the samples of sandwich panel investigated. After two years of outside exposure characteristics of the external covering (pre-painted metal skin) and the insulating polyurethane core are not so different from those that the component had in origin.

An analogous behaviour is observed, from the qualitative point of view, among tensile strength and thermal conductivity, the decrement of the characteristic up to 60 cycles and a following inversion of tendency with cycles.

Such behaviour somehow could be connected to the variations of weight and to the dimensional gymnastics of the elements of the panel (skin and core) favoured by phenomenon of diffusion and absorption of the vapour in the cellular structure of the polyurethane.

The colour, that deserves more attention from the normative point of view, is the characteristic that has suffered the greatest variations among the observed ones. The two types of finishing paint from the point of view of colour variation behave in a different way. Maintenance the project colour during time, cause the frequently use of panel for external roofs or walls, is an aspect that define the quality of the entire construction and deserve an adequate weight between the parameters to put under control. A more clearly norm could give an important contribution to the final product quality.

As it regards the end of service life of the component, according to the EN 14509 it’s achieved when the value of tensile strength is less than 0,050 N/mm2, as shown from the extrapolation of the diagram of fig. 2.9.13.



Fig. 2.9.13 - Specimen’s mechanical strength

It will be useful continue to check the course of the thermal conductivity of the samples exposed in external to assume new elements on the role of the skin-core interface and on the variability of the insulating property. The study could evidence properties of the aged polymer and the gas contained inside the cells to furnish an explanation to the improvement of the characteristics (tensile strength, thermal conductivity) measured in the initial phase of the panel’s aging.

2.9.4 References Alaimo G., Accurso F., Il degrado delle coperture con pannelli sandwich nel contesto climatico di Palermo, Convegno Internazionale “Involucri quali messaggi di architettura”, Napoli, 2003. Alaimo G., L’affidabilità funzionale delle coperture discontinue, EdiTecnica, Palermo, 2004. Alaimo G., Accurso F., The durability evaluation for sandwich panels: first experimental results, 10th DBMC International Conference on “Durability of Building Materials and Components”, Lyon, 2005. Alaimo G., Accurso F., La stima dell’affidabilità delle coperture discontinue, EdiTecnica, Palermo, 2006. Alaimo G. (a cura di), Valutazione sperimentale della durabilità di coperture discontinue. Un’applicazione al pannello sandwich. EdiTecnica, Palermo, 2006. Maggi P.N., Daniotti B., Alaimo G., Ciribini A., Morra L., Nicolella M., Rodonò U., La durabilità dei componenti edilizi – The durability of building components, EdiTecnica, Palermo, 2008. Maggi P.N. (a cura di), La qualità tecnologica dei componenti edilizi - La durabilità, Epitesto, Bologna, 2000. Maggi P.N. (a cura di), La qualità tecnologica dei componenti edilizi - La valutazione della durabilità, Epitesto, Bologna, 2001. Re Cecconi F., Iacono P., Enhancing the factor method. Suggestions to avoid subjectivity, 10thDBMC, Lyon, 2005. Rejna M., Valutazione della qualità tecnologica utile dei prodotti complessi per l’edilizia, Quaderni DISET vol. 4, Società Editrice Esculapio, Bologna, 1995.



2.9.5 Standard test methods EN 14509 Self supporting double skin metal faced insulated panels This European Standard specifies requirements and tests for double skin metal faced insulating sandwich panels, which are intended for discontinuous laying in roofs, internal or external walls and partitions within the building envelope.

The durability tests are defined in the Annex B by the change in the tensile strength of a test specimen that is subjected to climatic test for one or more weeks under constant conditions of temperature and humidity.

This standard has been used to choose test specimen (type and dimension), the maximum temperature for accelerated aging tests and mechanical testing procedures (tensile and bending tests).

ETAG 016 Self supporting composite lightweight panels This guideline covers self-supporting composite lightweight panels with one or both skins made of various materials assembled, with a core which is bonded to at least one of the skins.

The ETAG is divided into four parts, the first part deals with general aspects, the other parts deal with specific aspects relating to a different field of application (Part 2 is specific for use in roofs).

The durability of the panel shall be assessed as decay of performance characteristics after ageing tests, with reference to ISO 15686. The influence of ageing on panels is tested by measuring changes in the tensile strength on different specimen set subjected to climatic test cycles.

The assessment of the performance of panels under the effect of thermal shock, is carried out with an array of infra-red lamps for artificially irradiating the external skin of the test panel, at temperature according to the surface colour. The same procedure has been considered for the accelerated aging test in the climatic chamber but on little specimen cut from the whole panel.





2.10 Test methods for the durability evaluation of flat roof12 As part of the national research project on durability of building components of the external “enclosures” (a network of research organisations from the Universities of Brescia, Catania, Milan, Naples, Palermo and Turin), the Naples research unit – made up by Alba De Pascale, Patrizia Desiderio and Maria Gabriella Russo - addressed roofing subsystems.

2.10.1 Scope of the test method Roofing is one of the most complex (insofar as normally comprised of numerous materials) components of the external coverings of buildings. Probably the area which has traditionally presented the main maintenance problems, it’s also the one which presents most variety across, and within, its different contexts of application.

Another – but certainly not the last – problem which needs to be addressed is “quality of execution”, a particularly sensitive issue in the roofing sector, which in fact remains one of the few remaining building trades still generally entrusted to specialized firms, without whose expertise the results in terms of performance often leave much to be desired.

In the course of our analysis of the state of the art in the durability of continuous roofing – which included both scientific literature and an examination of Italian and international standards for laboratory testing – it became apparent that - as with so many other building components - while indications on individual products are in plentiful supply (e.g. membrane), there is something of a black hole when it comes to the built-up roof itself – and in fact it’s the latter which is particularly interesting, as it’s in the variety of materials which comprise them, and the diversity of the ways in which they’re combined, that lie the reasons for the problems which can often lead to sudden performance decay and anticipated death.

The fact is, here as in other parts of building, we have to acknowledge that the physical changes undergone by roofing materials during and after their application are decisive for their performance evaluation: returning to the example of prefabricated bituminous membrane (a good example after all, since in a sense it’s a crucial component in the appraisal of the performance of the roofing assembly as a whole), this is a material which as prescribed in the standards is tested in such a way as to ignore the changes it undergoes subsequent to its application at temperature using propane gas; likewise ignored are the differences which manifest themselves over time in accordance with the different protection systems currently in vogue in the country: nowadays acrylic paint instead of slate granules.

Then there are other important factors: the articulation of the different layers of the built-up roofing, the presence or absence of an insulating layer, the nature of this layer: these are all of major significance in the evaluation of roofing life cycles.

We felt, therefore, that it would be useful to go a little beyond the strictures of standard prescriptions. Although standards are an obligatory point of reference for a whole range of conditions, they were of little use when it came to designing the samples to be subjected to accelerated ageing if the ultimate goal of laboratory testing is re-scaling to real conditions.

In other words: in agreement with the other offices, a research approach driven by accelerated ageing testing on the different materials ultimately leads to re-scaling due to externally-sourced sampling. All very well, but if this was to have any bearing on real buildings, i.e. the behaviour of the elements undergoing testing as part of a complex as well as individually, all the more reason to design samples capable of reproducing as faithfully as possible the situations and conditions of the same materials in real buildings.

12 Professor Maurizio Nicolella - University of Naples Federico II, Department of Building Engineering.



2.10.2 Description of the tests method In the light of the above, we decided to work on test pieces made in exactly the same way as for roofing used in real buildings.

This meant we first had to compile a systematic inventory of all the possible roofing technical solutions, on the basis of field research and in the light of the solutions adopted by the designers, without taking into account the possible influence of local building traditions.

We then examined each of these solutions so we could select those which were of most interest for purposes of research.

So we decided to limit our work to a restricted number of samples, the idea being to conduct a number of tests on each sample so we could determine its performance over the artificial ageing cycle.

This selection process resulted in four different roofing solutions, layered as shown below (from top to bottom):

Table 2.10.1: roofing solutions examined

slate-surfaced membrane

cement and coarse calcareous sand screed

expanded clay lightweight cement mortar slope layer a

reinforced concrete and hollow tiles mixed floor

slate-surfaced membrane

bitumen-precoated polyurethane insulation roll


expanded clay lightweight cement mortar slope layer

b


membrane with acrylic paint protection


expanded clay lightweight cement mortar slope layer c


membrane with acrylic paint protection

bitumen-precoated polyurethane insulation panel


expanded clay lightweight cement mortar slope layer

d


We should point out that in making these choices we did take into account the fact that one of the most interesting aspects of research is the deterioration of the uppermost layers (since protective coating of membrane), a phenomenon which invariably triggers deterioration in the layers immediately below.

Therefore, the samples were designed according to criteria partially different from those indicated in national and international standards, so that they could be laid exactly as they are in real buildings, i.e. as shown in the layer diagrams for the roofing solutions selected.



From the morphological point of view, the samples constructed – which we also subjected to flooding tests, given the specific natures of the components under review – have the following characteristics (cf. figure in the next page):

- footprint dimensions 52 x 52 cm;

- total height 28 cm;

- bearing structure of steel sections L 30 x 60 x 5 mm tensing a plane comprising two hollow brick tiles (3 x 25 x 50 cm) on the base and vertical sides.

This arrangement was also designed to take into account the fact that in practice many of the maintenance problems associated with continuous coverings are located in the perimeter flashing, and therefore we devised a configuration which not only allowed us to perform flooding tests but also enabled us to carry out waterproofing tests on vertical surfaces connected to horizontal surfaces.

Obviously, the samples differed in terms of the solutions they represented, though all included the slope layer and cement screed.

The samples exposed outdoors were positioned in exactly the same fashion, since it made no sense to position them at pitches which would accelerate ageing in assemblies which are most commonly used on the horizontal plane.

Flooding, evidently, was not induced as in laboratory testing but occurred as per the natural course of events.

Figure 2.10.1: Plant and sections of the samples

The need to describe the “states” which the samples pass through during induced ageing, and to classify those to which a certain condition of degradation – and therefore a certain approach to repair – is to be associated meant that it was essential to itemize the properties of the various elements of the assembly, and its performance specifications.

There were two objectives here:

• to describe performance decay in terms of a curve (and not simply as a line linking two points corresponding to the “birth” and the “death” of the element) which is as close a representation of reality as possible and which – even if by extrapolation – allowed us to plot future deterioration, repair schedules and a design approach which is truly maintenance oriented;



• to create a dynamic checklist allowing us systematically to associate a given objective condition (conventionally coded performance) to a time threshold, so that both in turn can be assigned a maintenance typology.

Our analysis was conducted on the basis of national and international standards, recommendations in the manufacturers’ manuals, and experience from real building contexts.

We therefore ensured that everything essential for the study was in place as required, defining where possible performance thresholds which on the one hand could provide a quick view of the component’s life cycle and on the other would make it possible to determine a certain typology of maintenance.

We carried out a preliminary evaluation designed to individually identify, for each layer in the roofing assembly, the types of tests required under applicable standards relative to the “well-being” need, taking into account for each layer the requirement class/requirement subclass/expected requirements/test methods employed in determining performance, as shown in the table partially reported below:

CLA

SS O

F R

EQU

IREM

ENT

FUN

CTI

ON

AL

LAYE

R

REQUIREMENT SUBCLASS

EXPECTED REQUIREMENT

TEST METHODS FOR DETERMINING

PERFORMANCE

Resistance to mechanical action

behaviour under traction impact resistance static load resistance ripping resistance resistance to penetration of roots interply bond uplift resistance

UNI EN 12311-1:2002 UNI EN 12691:2002 UNI EN 12730:2002 UNI EN 12310-1:2001 UNI EN 13948: 2001 ASTM E 907 – 96

Wat

er re

sist

ance

ab

ility

to p

reve

nt s

urfa

ce w

ater

from

pen

etra

ting

roof

ing

resi

stan

ce e

lem

ent

Resistance to thermal action

cold flexibility dimensional stability dimensional stability in cyclical temperature conditions slippage when hot stability of form when hot differential dilation linear thermal dilatation coefficient pliability at low temperatures

UNI EN 1109:2002 UNI EN 1107-1:2002 UNI EN 1108:2002 UNI EN 1110: UNI 8202-18 UNI 8202-19 UNI 8202-20 UNI EN 495-5:2002

Another preliminary definition regarded the classification of problems, to which a out of order status could be defined and the need for maintenance determined.

To allow us to devise a program of tests to be carried out on the samples before, during and after application, we conducted an exhaustive survey of detected anomalies and, using standard ASTM E 632 as our reference, classified these faults as defects and/or faults which can be detected visually and defects and/or faults which can only be detected using instruments, as shown in the table below (which also gives the standards applicable to each test).



VISIBLE ANOMALY

PROPERTY TESTED FOR DETECTING PERFORMANCE DECAY

MEASUREMENT METHOD STANDARD

UV resistance of protective layer sight microphotography Change in

colour UV resistance of membrane instrumental UNI EN 1297

scrape resistance of membrane sight microphotography

adhesion of granules (for mineral-coated membranes) instrumental UNI EN 12039

UNI 8202-35 Surface alteration

chemical resistance in contact with common substances/ resistance to chemical agents

instrumental

dimensional stability of membrane instrumental UNI EN 1107-1 dimensional stability of membrane in cyclical temperature conditions instrumental UNI EN 1108

flexibility of membrane when cold instrumental UNI EN 1109 pliability at low temperatures instrumental UNI EN 495-5 dimensional instability of insulation panel instrumental UNI EN 1604

Cracking / wrinkling

incorrect application (excess of bitumen, incorrect adhesion of membrane, irregularities in application surface etc.)

instrumental / sight ASTM D 3617-2

dimensional stability of membrane instrumental UNI EN 1107-1 dimensional stability of membrane in cyclical temperature conditions instrumental UNI EN 1108

flexibility of membrane when cold instrumental UNI EN 1109 pliability at low temperatures instrumental UNI EN 495-5 dimensional instability of insulation panel instrumental

incorrect application (incorrect gluing of membrane, continuity of membrane between horizontal and vertical element etc.)


Detachment of membrane

interply bond instrumental ASTM E 907 -96 resistance of joints to traction instrumental UNI EN 12316 detachment resistance of joints instrumental UNI EN 12317

adhesion between layers instrumental

Detachment of side and end laps

uplift resistance instrumental ASTM E 907 -96 resistance to traction instrumental UNI EN 12311 resistance to ripping instrumental UNI EN 12310 impact resistance instrumental UNI EN 12691 Splitting

resistance to penetration of roots instrumental UNI EN 13948 resistance to ripping instrumental UNI EN 12310 resistance to static loads instrumental UNI EN 12730 Incisions resistance to penetration of roots instrumental UNI EN 13948 flexibility of vapour barrier when cold instrumental UNI EN 1109 resistance to traction instrumental UNI EN 12311 transmission of water vapour in water barrier instrumental UNI EN 1931

thermal ageing of vapour barrier in water instrumental UNI8202-27

Interply condensation

incorrect application (absence of vapour layer barrier, presence of trapped water etc.)


water resistance of membrane instrumental UNI EN 1928 water resistance after low-temperature traction of membrane

instrumental UNI EN 13897

transmission of membrane water vapour

instrumental UNI EN 1931

Presence of water where no fissures / detachment exist

hail resistance of membrane instrumental UNI EN 13583

Given the complexity of the elements tested and the considerable number of tests which would be necessary for an exhaustive description of the performance decay of the individual



layers in the roofing assembly, we decided to simplify the process by classifying anomalies into three “macrogroups”:

1. surface alterations (including changes in colour);

2. detachment (including: detachment of membrane from underlayer; interply detachment; detachments along joints, cracking and/or wrinkling, uplifting and/or rippling);

3. cracking (including: alligatoring, incisions, tears etc.).

Note that we have not included surface vegetation as we decided to exclude from laboratory tests the agents which normally trigger the appearance of this pathology.

Surface alterations, therefore, are detected using microphotography; detachments using thermography followed by pull-off; and cracking by sight.

To plot the following time/performance curves:

1. efficiency of protective layer over time;

2. adhesion between layers and to underlayer over time;

3. water resistance over time,

the instruments and test methods having been analyzed, we established a test table as shown below:

TEST REQUIREMENT ELEMENT TESTED MEASUREMENT METHOD

pull-off adherence membrane instrumental

microphotography protection protective layer sight

UNI EN 12086 (heat insulation) UNI 8202 and 8223 (membrane) tests

water vapour permeability µ entire assembly instrumental

ISO 8301 test heat insulation entire assembly instrumental

As an example, below we provide some images of samples before testing, accompanied by their thermographic images, showing the areas of detachment during and after the test cycle.

INFRARED VIDEOTHERMOGRAPHY INSPECTION CHECKLIST TYPE 1 SAMPLE

Time and date of inspection

13:16:08 27.06.2006

Working conditions

Ambient T (sensor): 37.9 °C

Notes

A sunny, very hot day.

Description: membrane with protective coating of acrylic paint – screed with cement and coarse calcareous sand dosed at 300 kg – cement mortar slope layer dosed at



200 kg with expanded clay – underlayer

Image description Bottom surface

Extent of detachment area

Outcome: detachment of 56.14% from bottom surface (yellow-orange-red areas); the protective layer has deteriorated by around 50%.

Examination of cap sheet skirts

Skirt 1A

Skirt 1B



Skirt 1C

Skirt 1D

Outcome: areas of non-adhesion of cap sheet to underlayer along the skirts; the protective paint is for the most part visibly absent

Ageing cycle determination

A methodological approach which determines the service life of a building component by comparing an artificial ageing test conducted in the laboratory with a natural weathering test on the same sample, accelerating ageing, necessarily raises the problem of how accurately to define the aging cycle to which the sample is subjected so that a reasonably reliable re-scaling is possible.

As mentioned above, both national and international standards merely outline a series of criteria for determining this cycle, leaving a certain leeway for its definition.

We feel, on the contrary, that the obligation to proceed sooner or later – otherwise only comparative tests and studies will be possible, via which we can determine analogous patterns of behaviour between the different ways in which the same element is used – to a re-scaling should influence the conditioning cycle, which must necessarily be differentiated by the diverse contexts for which re-scaling is to occur. In other words: it does not seem possible to consider as reliable a re-scaling performed in significantly different climatic contexts, where a sample is subject to same artificial action. Therefore, we tried to design a cycle which provides an accelerated reproduction of what actually occurs, to different extents, in different environmental contexts.

As part of the Prin 2003 research project the various research units examined problems and phenomena in the design of the various components used in the external “enclosures” of buildings: the logical approach in this project was to attempt to simulate, in the laboratory, the ambient actions which are active in the various climatic contexts, differentiating them for purposes of customized re-scaling.

The Naples research unit therefore proposed to the other units a model for determining the aging cycle. This model uses an algorithm in which the input data comprises a series of climatic data which can be obtained from various sources (principally the air force).



The output provides the duration of the subcycles (each one of which represents, and reproduces, a climatic season), temperatures, humidity, sunlight radiation, and rainfall to which the sample is subjected.

In this way we obtain a kind of “virtual year” comprising the four seasons in succession, with different values for temperature, humidity, sunlight radiation, and rainfall, with the values inserted those for the region and the time of year which represents the season in question, as shown in the table below:

SUBCYCLE ACTION VALUE INSERTED

TEMPERATURE absolute minimum

HUMIDITY mean

SUNLIGHT RAD. - COLD

RAIN -

TEMPERATURE highest among means

HUMIDITY mean

SUNLIGHT RAD. xenon lamp lights MILD DRY HEAT

RAIN -

TEMPERATURE absolute maximum

HUMIDITY 95%

SUNLIGHT RAD. xenon lamp lights INTENSE DRY

HEAT

RAIN -

TEMPERATURE lowest among means

HUMIDITY mean

SUNLIGHT RAD. - RAIN

RAIN splash or flooding

Climatic data was collected in advance from various official sources, covering a period no longer than the last 5 years, on account of the extreme variability in recent times.

The final output, generated by special simulation software, takes the form of a summary table as shown in the example below, which provides values for Naples:

SUBCYCLE DUR.

TEMP.

HUM.

SUN RAD.

RAIN

COLD 6 h 4 C° 70% No No

MILD DRY HEAT 9 h 25 C°

60% Yes No

INTENSE DRY HEAT 9 h 40 C° 95% Yes No

RAIN 6 h 10 C° 75% No Yes

These figure is represented in graph form below.



2.10.3 Conclusions The model proposed is to be considered helpful:

• to refine general principles of service life prediction tests, because in this manner, it is possible to have simulations specifically made for re-scaling, and homogeneity for all the researchers, in all contexts;

• to compare data obtained from accelerated ageing tests, external exposure, or existing buildings.

The test method is not bound to a specific service life prediction method. It is being possible to use it with the same goal and the same effectiveness in the Factor Methods, in Engineering Methods and in Stochastic Methods, also.

Data on performance over time of building components can be provided only after a reliable period of trialling, and so, experimentations on different components are needed for testing the method.

2.10.4 References Architectural Institute of Japan (1993) Principal Guide for Service Life Planning of Buildings, AIJ, Tokyo.

Beech, J. and Beasley, J. (1995) Effects of Natural and Artificial Weathering on Buildings Sealants, ASTM.

Brandt, E. (ed.) (1990) Feedback from Practice of Durability Data — Appendix — Examples of Field Investigations of Buildings and Structures, CIB report 128, International Council for Building Research Studies and Documentation (CIB), Rotterdam.

Cole, I.S., Norberg, P. and Ganther, G. (1996) Environmental factors promoting corrosion in building microclimates, in Proceedings of the 13th International Corrosion Congress, Melbourne, Australia, 1996.

Cole, I.S., Ganther, G. and Norberg, P. (1996) Sensors for measuring factors promoting corrosion within the building envelope, in Proceedings of the 13th International Corrosion Congress, Melbourne, AU, 1996.

Caluwaerts, P., Sjöström, C., Haagenrud, S.E. (1996) Service life standards – background and relation to the European Construction Products Directive, in Proceedings of the 7th International Conference on Durability of Building Materials and Components, Stockholm, 1996, (ed. C. Sjöström), E & FN Spon, U.K.



CIB W080 / RILEM TC 140-TSL Committee on Service Life of Building Materials and Components, Guide and Bibliography to Service Life and Durability Research for Building Materials and Components, 2004.

Croce, S. (1996) Deterioration propensity of modern facades: analyisis criteria, in Proceedings of the Conference Defects in buildings, Varenna.

Haagenrud, S.E. and Henriksen, J.F. (1996) Degradation of built environment – Review of cost assessment model and dose-response functions, in Proceedings of the 7th DBMC, Stockholm, 1996, (ed. C. Sjöström), E & FN Spon, U.K.

Haagenrud, S.E., Henriksen, J.F and Skancke, T. (1996) Modelling and mapping of degradation of built environment from available data and GIS based information tools, in Proceedings of the 7th International Conference on Durability of Building Materials and Components, Stockholm, 1996, (ed. C. Sjöström), E & FN Spon, U.K. Hovde, P.J. (2004) Task Group Performance Based Methods for Service Life Prediction - State of the Art Reports - Part A Factor Methods for Service Life Prediction, CIB Report: Publication 294. Hovde, P.J. (2002) The factor method for service life prediction from theoretical evaluation to practical implementation, in Proceedings of 9DBMC.

Jernberg, J., Sjöström, C., Lacasse, M.A., Brandt, E. and Siemes T. (2004) Guide and Bibliography to Service Life and Durability Research for Buildings and Components, Part I – Service Life and Durability Research, CIB: Publication 295.

Maggi, P.N. (2000) La qualità tecnologica dei componenti edilizi: la durabilità, Epitesto, Milano.

Maggi P.N. (2001) La qualità tecnologica dei componenti edilizi: la valutazione della durabilità, Epitesto, Milano.

Maggi, P.N., Daniotti, B., Gottfried, A. and Morra, L. (1996) Control of project pathologies of buildings components: methodology for the evaluation of the reliability, in Proceedings of the Conference Defects in buildings, Varenna.

Moser K., Task Group Performance Based Methods for Service Life Prediction - State of the Art Reports – Part B: Engineering Design Methods for Service Life Prediction, CIB Report: Publication 294, 2004.

Nicolella, M. (2000) Methodology for calculation of constructive elements life cycle, in Proceedings of the International Conference Mantenimiento y gestion de los edificios, Barcellona, 2000

Nicolella, M.; De Pascale, A. (2005) Service life of building components. Analysis and proposals of definition of the modifying factors, in Proceedings of 10DBMC.

Nicolella, M.; De Pascale, A. (2008) A Model for Determining Accelerated Ageing Cycles in Durability Research: a Case study on Continuous Roofing, in Proceedings of 11DBMC.

Saunders, S.C., Jensen, M.A. and Martin, J.W. (1990) A Study of Meteorological Processes Important in the Degradation of Materials through Surface Temperature, National Institute of Standards and Technology, Technical Note 1275.

Sjöström, C. and Brandt, E. (1990) Collection of in-service performance data: state of the art and approach by CIB W080/ RILEM 100-TSL, in Proceedings of the 5th International Conference on Durability of Building Materials and Components, Brighton, 1990, (eds. J.M: Baker, P.J. Nixon, A.J. Majumdar and H. Davies), E & FN Spon, U.K., 7-9 November.

Sjöström, Ch. and Brandt, E. (eds.) (1990), Feedback from Practice of Durability Data — Inspection of Buildings. CIB report 127, International Council for Building Research Studies and Documentation (CIB), Rotterdam



2.10.5 Standard test methods ASTM G 7-89, Standard Practice for Atmospheric Environmental Exposure Testing of Nonmetallic Materials

ASTM G 90-91, Practice for performing Accelerated Outdoor Weathering of Nonmetallic Materials Using Concentrated Natural Sunlight

ASTM E 632–82 (Reapproved 1996), Standard Practice for Developing Accelerated Tests to Aid Prediction of the Service Life of Building Components and Materials

ASTM G 166–00 (Reapproved 2005), Standard Guide for Statistical Analysis of Service Life Data

ASTM G 169, Guide for the Application of Basic Statistical Methods to Weathering Tests

ASTM G 172–03, Standard Guide for Statistical Analysis of Accelerated Service Life Data

British Standards Institution (1992): BS 7543:1992 Guide to Durability of Buildings and Building Elements, Products and Components. British Standards Institution, London, UK.

BSI – British Standard Institution (1992) Guide to Durability of Buildings and Building Elements, Products and Components, BSI, BS 7543

European Organization for Technical Approval (1999a): Assumption of working life of constructional products in guidelines for European Technical Approval, European Technical Approvals and harmonized standards. EOTA Guidance Document 002, European Organization for Technical Approval, Brussels, Belgium, December.

European Organisation for Technical Approvals (1996) Assessment of Working Life of Products, EOTA, Draft TB 96/23/9.3.1.

European Union (1994): Interpretative Documents of Council Directive 89/106/EEC. Official Journal of the European Communities, C62. Vol. 37, 28. February 1994.

International Organization for Standardization (1984) Performance Standards in Buildings – Principles for Their Preparation and Factors to be Considered, ISO, Geneva, ISO 6241-1984 (E).

ISO 15686 Buildings and constructed assets - Service life planning: Part 1 (General principles), Part 2 (Service life prediction procedures), Part 3 (Performance audits and reviews), Part 4 (Data dictionary - Technical Report), Part 5 (Whole-life costing), Part 6 (Procedures for considering environmental impacts), Part 7 (Performance evaluation for feedback of service life data from practise), Part 8 (Reference service life), Part 9 (Guide on service life declarations for building products), Part 10 (Serviceability), Part 11 (Terminology).

UNI 11156-1 Valutazione della durabilità dei componenti edilizi - Parte 1 (Terminologia e definizione dei parametri di valutazione), Parte 2 (Metodo per la valutazione della propensione allʼaffidabilità), Parte 3 (Metodo per la valutazione della durata (vita utile)).



2.11 Durability of the external load bearing walls13 This paragraph deals with an experimental research programme performed into the ICATA Department (Dipartimento di Ingegneria Civile, Architettura, Territorio e Ambiente) of the University of Brescia within the integrated research programme called “Methodology for design and durability evaluation of the building components in sustainable production processes: experimental evaluation of the standardised durability and their correction to use the component in special service condition in accordance to a building maintenance programme”.

In close co-operation with the other Research Units, the Unit from Brescia developed the experimental tests concerning a building component coded PP5 - external load bearing wall built by means of lightened bricks - (see the DISET index of the Politecnico di Milano: Daniotti/Maggi – Valutazione della qualità tecnologica caratteristica dei prodotti complessi per l’edilizia – 1993).

The Research Unit investigated the behaviour of this building component because of its large use in Italy, assessing samples without protective layers (painted or coated) to simulate supporting walls with any painted surfaces. It has been considered that the function of the paint layer could be useless due to the failure because of a poor quality product – which can be occurred during the construction phase. The Unit took into account chiefly the variation of the two main characteristics of the building component: the load bearing capacity and the thermal insulation.

Finally, the Research Unit reported the experimental results of destructive and non destructive tests after performing accelerated ageing tests in laboratory and some preliminary tests after natural ageing in outside places for the temporal re-scaling to estimate the Reference Service Life. It has been evaluated that the minimum period of natural ageing cannot be less than four years.

2.11.1 Scope of the test method The building component PP5 is a wall made up by lightened bricks (25x30x24 cm) with plastered surfaces (1,5 cm). On the external surface any protective painting system has not applied.

Making up the samples, the rules suitable to obtain reproducibility in testing and results have been assumed.

Generally, the experimental phase of the research on a building component is needed to obtain data for:

• the initial characterization of the performances of the technical solution; • the assessment of the initial performances and the accelerated or natural degrade,

surveying the relative position regarding to the minimum performance level. The experimental programme included a set of tests after accelerated ageing into the weathering chamber and after natural ageing on external sites.

The results obtained from the accelerated tests must be validated through a comparison with those ones obtained in natural condition.

This test method is applicable on separated or assembled materials, or on a building component.

13 A. Ciribini, E. Donini, F. Turla - University of Brescia, Faculty of Engineering, ICATA Department



2.11.2 Description of the tests method The experimental programme has been implemented complying with the standard ISO 15686 - “Buildings and constructed assets – Service life planning “ Part 1 - “General principle” and Part 2 - “Service life prediction procedures”.

To define the experimental programme the natural agents (chemical, climatic, artificial external, artificial due to the use, biological) applicable to the specific technical solution have been previously stated, choosing those agents that could be simulated by means of a weathering chamber and that will constitute the accelerated test cycle.

To define the phases of the cycle to be set by the chamber we took into account the maximum intensity of the single agents obtained from the climatic data provided by the City of Brescia (gathered over the last 16 years).

To set the length of the single phase it was necessary to perform preliminary tests of calibration to verify the time span to reach the stated goals.

These preliminary tests showed that to reach the minimum winter temperature (-5°C) and the maximum summer temperature (35°C) in the deepest layers of the sample the researchers have to assume during an accelerated cycle a temperature of 15°C lower or higher than in natural condition. So the Unit set during the accelerated ageing a temperature of -20°C for winter condition (frost climate) and 50°C for summer condition (hot humid climate).

After the calibration tests, the reserachers decided to use an accelerated ageing cycle with the following four stages:

• Phase of rain In this phase lasting 30 minutes the wall is sprayed with demineralised water.The air

inside the chamber has been kept at 20° C.

• Phase of frost A rapid decreasing of the temperature to -20°C is set and maintained over 90 minutes.

• Phase of hot humid climate The temperature (55°C) and the humidity (80%) are maintained over 60 minutes. A

transition period of 80 minutes has been managed before starting such a phase. • Phase of hot dray climate In the chamber a temperature of 30°C and a humidity of 40% are maintained over 80

minutes.

In the recorded diagram concerning a whole cycle the temperature of the plastering layer, after that the chamber temperature has been kept at -20°C for 60 minutes, reached a value of about 0°C. Moreover, the frost front went beyond the interface plaster/bricks.

At the end of the hot humid phase the temperature of the plaster and of the interface area has been assessed about at 40-45°C.

The phase of raining is needed to obtain an acceptable level of wetting on the surface and in the layers under the plaster. Viceversa, the phase of dry climate is necessary to obtain a sufficient drying of the same layers.

From the climatic data provided by the City of Brescia (monthly average value of the minimum temperature and of maximum temperature) it has been observed that only in December, January and February, during the night, the temperature decreases under the zero degree, but, from a statistical point of view, only 50 times over three months (20 times in December, 23 times in January and 7 times in February). In summer the temperature exceed the 30°C in the month of June, July and August but, from a statistical point of view, it



happens only 50 times during the three months (10 times in June, 20 times in July and 20 times in August).

After these considerations the Unit reckoned that 50 cycles of accelerated ageing were equivalent to one year of natural ageing, assuming that the climate of autumn and spring has no influence on the ageing of the samples.

As far as such a preliminary phase of test, it has been decided to perform tests after every 100 cycles and for 300 cycles total of artificial ageing, equivalent to two years and to six years of natural ageing.

After the featuring of the test cycle for the artificial ageing the researchers defined the type and the amount of the characterization tests to be done on the samples at time zero and at stated intervals on the samples submitted to artificial ageing procedures.

The ageing tests were performed also on samples of lightened bricks and of plaster prisms used for the making up of the samples. These samples were put in the chamber and submitted to the same cycles of the wall samples.

On July 2005 the Unit made up the samples of walls to submit to artificial and natural ageing. In the same period also the time zero sample was made up to be submitted to characterizing tests.

On September 2005 the sample of the wall submitted to the artificial ageing was put on the metallic support and fixed in the same place of the chamber door preventively removed.

The sample of the wall for the natural ageing was put on a metallic support 45° lopsided in an outdoor site. According to the standards, such procedures allow a more fast ageing (factor 6-8). Such a way of exposure does not influence the effects of the decay and in the same time allow a faster comparison with the results of the accelerated tests.

The sample of wall for characterization tests (time zero) was made up in the same time and with the same procedures of the other samples. It was put in an indoor place.

The experimental phase lasted a time span from October 2005 to November 2006. To complete it tests have been on the sample put in the outdoor site for the re-scaling of the results (October 2008). A further test programme will be performed on October 2011.

2.11.3 Conclusions As far as non-destructive tests are concerned, it has been purported that with the increase of the cycles the Unit did find a superficial decay (narrow and large splits) that, in terms of amount, reaches the stabilization around the 150 cycles.

Moreover, it has been found an accumulation of the matrix constituted by the plaster surface, that affects the decrease of the value of the water absorption.

The values of the plaster/brick adhesion show a decreasing ageing grade and could compromise the difficulty for the execution of the tests and the reliability of the results to be obtained.

The mechanical strengths to the bending and to the compression show important increments concerning the number of cycles. This phenomenon can be due to the increment of the curing grade but it must be studied through a further testing programme because of its difficult understanding.

The obtained values of the thermal transmittance showed an increment of transmittance versus the number of cycles due presumably to the saturation of the brick material base because of the rain water and the consequent humid front that more and more affects the sample in depth.



More recently, the reaerchers carried out some tests on the sample put in the outside place in order to carry out, if possible, the necessary temporal re-scaling. After 15 months of natural ageing they obtained 20 ml/h for water absorption and 1,80 W/m2°C for thermal transmittance.

The data obtained from the tests performed on the sample exposed to the natural ageing for 15 months show a good compliance with the results of the tests on the artificial aged sample after 50 cycles (interpolation of the values referred to 10 and 100 cycles). This fact must be confirmed by the results of the tests on the natural aged sample scheduled on October 2011.

The researchers believe that these preliminary results can be helpful to tune the general principles for service life prediction tests in the process that starts from the gathering of the climatic data. Such results allowed them to define the cycles of the artificial ageing. They hope that their method could be helpful to compare laboratory ageing tests with external exposure data (rescaling).

Eventually, when emphasizing the temporary features of the tests campaign that has been performed, the resrarchers believe that their test method gives data on performance over time of the load bearing external walls.

Therefore, the building component which has been taken into account needs the protecting painting layer if the decay of the characteristics concerning the thermal insulation that featured such a component is to be avoided.

2.11.4 References B. Daniotti, P. N. Maggi Valutazione della qualità tecnologica caratteristica dei prodotti complessi per l’edilizia, Progetto Leonardo, 1993

P.N. Maggi, La qualità tecnologica dei componenti edilizi – La durabilità Epitesto, 2000

A.CIRIBINI, E.DONINI, F.TURLA, The durability evaluation of external load bearing walls Editecnica, 2007

2.11.5 Standard test methods NFT T 30-049 “Peintures et vernis – revêtements à usage extérieur: essai de vieillissement artificial”.

For the tests: NORMAL 44 - UNI 7357 - EN 196 part 1 - EN 1015 parts 12 and 19 - EN 772 part 1.



3. Synopsis of the Best Practice In this part a synopsis of all the introduced test methods is addressed. In particular, aspects which may be useful in refining the general Service Life Prediction procedure (ISO 15686-2) are taken into account. Issues specifically dealt are:

• the correlation between in situ measurements and laboratory measurements;

• the correlation between in situ tests and design performance requirements;

• the application of ISO procedure;

• the correlation between laboratory time and real life time (acceleration factor or time re-scaling);

• bind to Service Life Estimation methods and adopted standards.

The structure of this synopsis is presented in Tab. 3.1. Table 3.1: Structure of the synopsis of all testing methods presented for building materials

Test name and chapter Reference to the part in chapter 2 where the test method is reported

1. SLP based It is discussed whether the test method is meant for providing information regarding Service Life or it is simply a durability test (pass or fail)

2. Material/component Is the test method to be applied to a specific material / component or is it general?

Aims

3. Agents Which agents are reproduced

4. Refining SLPP Specificities of the test method which can be useful in refining the general procedure for Service Life Prediction (ISO 15686-2)

5. Lab exposure Is ageing to be achieved by means of laboratory procedures?

6. Ageing cycle

Brief description of the ageing cycle. In particular description of the type of cycling:

- Extended exposure to the same agent - Simple cycling (reproduction of the same cycle) - Cycles grouped in phases

7. Outdoor exposure - Ageing achieved through accelerated outdoor exposure - Is outdoor exposure prescribed in the test method (in

addition to lab exposure)?

Procedure

8. Rescaling Is rescaling (i.e. proportion between accelerated and natural ageing) part of the procedure?

9. Link to ESL method Is the test method linked (i.e. built in order to provide data useful for) to any method for Service Life Estimation (i.e. Factor Methods, Engineering Methods, Stochastic Methods)?

10. Performance(s) over time

Does the procedure provide data regarding performance(s) over time? Are performance test(s) prescribed? At which intervals?

Outcomes

11. RSL Does the procedure provide the Reference Service Life for the studied material / component?

12. Standards Standards referenced or adopted in the test method with regard to the ageing procedure (ageing cycle especially)

13. Fully adopted Are these standards fully adopted? Otherwise, does the test method partially adopt the standards and modify some aspects? Why?

14. Agents considered Which agents are considered in the standard test method? Are they the same taken into account in the proposed test method?

Standards for ageing procedure

15. 1 or more agents Does the standard test method take into account one or more agents?

Performances evaluation 16. Standards

Are in the test method referenced the standards to be used for performance(s) assessment (if applicable)?



Tables from 3.2 to 3.17 refer to each single test method as described in each paragraph of chapter 2, in particular:

• Table 3.2 deals with wood materials weathering;

• Tables from 3.3 to 3.8 deal with metal materials;

• Tables from 3.9 to 3.12 refer to concrete aging;

• Table 3.12 is for masonry wall test procedure;

• Table 3.13 deals with ETICS test method;

• Table 3.14 deals with rendering test procedures;

• Table 3.15 refers to pitched roofs building components test method;

• Table 3.16 refers to flat roofs building components test method;

• Table 3.17deals with building components for external load bearing walls.

Table 3.2: Wood materials weathering

Test name and chapter NordTest NT Build 495 - § 2.1 SLP based Yes Material / component Both (in § 2.1 example applied on wood) Aims Agents 4: UV light, heat, water and frost.

Refining SLPP

- Required the registration of preparation of samples - Sequence in ageing phases - Moveable apparatus optimising the exposition and allowing

ageing achieved with different chambers.

Lab exposure Yes Specimens dimension: up to 1,5 m x 2,5 m

Ageing cycle Simple cycling Method of determination of the cycle: not declared

Outdoor exposure Required, procedure not referred nor described in NT Build 495

Procedure

Rescaling Yes Link to ESL method Yes Performance(s) over time Not specified in NT Build 495 Outcomes

RSL Yes, by re-scaling

Standards

- NT BUILD 495 – NordTest Method - DS/EN ISO 4892-3, 1999 “Plastics – Methods of exposure

to laboratory light sources – Fluorescent UV-lamps”. - ASTM G 53 – 96,“Standard Practice for Operating Light-

and Water-Exposure Apparatus (Fluorescent UV-Condensation Type) for Exposure of Non-metallic Materials”.

Fully adopted NA Agents considered NA


1 or more agents More agents Performances evaluation Standards Not specified in NT Build 495



Table 3.3: Metals. Testing with condensation

Test name and chapter Metals – chamber tests: testing with condensation – § 2.2 – p1 SLP based No

Material / component Metal surface (applicable only for clean surface in pure environment) Aims

Agents Variations in temperature and relative humidity Refining SLPP NO Lab exposure Yes

Ageing cycle

Simple cycling: - 3 hours (RH 95–100%), T=25°C - 9 hours RH (90–96%), T=40°C - 12 hours (RH 95–100%), T=25°C Method of determination of the cycle: not declared

Outdoor exposure Not declared

Procedure

Rescaling Not declared Link to ESL method No Performance(s) over time No Outcomes

RSL No Standards IEC 60068-2-30:2005 Fully adopted Yes Agents considered Same


1 or more agents Same Performances evaluation Standards Not declared

Table 3.4: Metals. Spray salt tests

Test name and chapter Metals – chamber tests: spray salt tests (NSS, AASS, CASS) - § 2.2 – p2

SLP based No Material / component Metal surface

Aims Agents

- NSS: 5% NaCl (in neutral pH solution), 35 °C - AASS: 5% NaCl and glacial acetic acid (pH 3.1–3.3), 35 °C - CASS: 5% NaCl and glacial acetic acid (pH 3.1–3.3), 50 °C

Refining SLPP No Lab exposure Yes Ageing cycle Extended exposure to the same agent Outdoor exposure Not specified

Procedure

Rescaling No method has been proposed for re-scaling Link to ESL method No Performance(s) over time No Outcomes

RSL No

Standards - ASTM B117 - ISO 9227:2006 Corrosion Tests in Artificial Atmospheres –

Salt Spray Tests. Fully adopted Yes Agents considered Same


1 or more agents Same Performances evaluation Standards Not specified



Table 3.5: Metals. Chamber tests – Testing in sulphure dioxide

Test name and chapter Metals – chamber tests: testing in sulphur dioxide – § 2.2 – part 3

SLP based No Material / component Metal surface Aims Agents Sulphur dioxide gas Refining SLPP No Lab exposure Yes Ageing cycle Extended exposure to the same agent Outdoor exposure Not specified

Procedure

Rescaling Not possible (torture test) Link to ESL method No Performance(s) over time No Outcomes

RSL No

Standards ISO 6988:1985 Metallic and Other Non Organic Coatings – Sulphur Dioxide Tests With General Condensation of Moisture

Fully adopted Yes Agents considered Same



Table 3.6: Metals. Cyclic tests – ISO 14993

Test name and chapter Metals – cyclic tests - § 2.2 – p 4.1 SLP based Yes Material / component Metal surface Aims Agents Salt fog, variation in temperature and relative humidity Refining SLPP No Lab exposure Yes Ageing cycle Simple cycling Outdoor exposure Not specified

Procedure

Rescaling Yes, defined for Okinawa, Japan (no other locations indicated) Link to ESL method No Performance(s) over time No Outcomes

RSL Yes

Standards ISO 14993:2001 - Corrosion of Metals and Alloys – Accelerated Testing Involving Cyclic Exposure to Salt Mist, ‘Dry’ and ‘Wet’ Conditions.






Table 3.7: Metals. Cyclic tests – ISO 11997-1

Test name and chapter Metals – cyclic tests – Marine Atmosphere - § 2.2 – p 4.2 SLP based Yes Material / component Metal surface

Aims Agents

Salt fog (NaCl in cycles A and B, NaCl and ammonium sulphate in cycle C), variation in temperature and relative humidity

Refining SLPP - Cycles specific for product types - Cycles specific for Climate/exposure zone

Lab exposure Yes

Ageing cycle Simple cycling Three possible cycles: A) for USA and Japan, B) Europe, C) specific for water-based emulsion paints

Outdoor exposure Not specified

Procedure

Rescaling Yes, for defined sites in USA, Japan, Europe Link to ESL method No Performance(s) over time No Outcomes

RSL Yes

Standards ISO 11997-1:2005 Paints and Varnishes – Determination of Resistances to Cyclic Corrosion Conditions: Part 1 – Wet (Salt Fog)/Dry/Humidity




Table 3.8: Metals. Cyclic tests – ISO 11997-2

Test name and chapter Metals – Cyclic tests - § 2.2 – p 4.3 SLP based Yes Material / component Metal surface Aims Agents UV Radiation, water, salt spray (NaCl and ammonium sulphate),

variations in temperature and relative humidity

Refining SLPP - Cycles specific for product types - Cycles specific for Climate/exposure zone

Lab exposure Yes

Ageing cycle Simple cycling Three possible cycles: A) for USA and Japan, B) Europe, C) specific for water-based emulsion paints

Outdoor exposure Not specified

Procedure

Rescaling Yes, for defined sites in USA, Japan, Europe Link to ESL method No Performance(s) over time No Outcomes

RSL Yes

Standards ISO 11997-1:2005 Paints and Varnishes – Determination of Resistances to Cyclic Corrosion Conditions: Part 1 – Wet (Salt Fog)/Dry/Humidity






Table 3.9: Concrete. Resistance to carbonation

Test name and chapter Concrete resistance to carbonation - § 2.3 – part 1 SLP based No Material / component Reinforced concrete Aims Agents CO2 Refining SLPP Storage procedure before ageing

Lab exposure Yes Specimens 100x100x500 mm,

Ageing cycle 28 days at Tref = 20°C, RHref = 65%, carbon dioxide concentration of 2.0% vol.

Outdoor exposure Prescribed natural exposure

Procedure

Rescaling Rescaling methods are being developed Link to ESL method No Performance(s) over time No Outcomes

RSL No Standards FIB Model Code Fully adopted Same Agents considered Same


1 or more agents Same Performances evaluation Standards Included in FIB Model Code

Table 3.10: Concrete. Resistance to chloride penetration

Test name and chapter Concrete resistance to chloride penetration - § 2.3 – part 2 SLP based Yes Material / component Reinforced concrete Aims Agents Cl- Refining SLPP - Lab exposure Yes Ageing cycle Extended exposure to the same agent Outdoor exposure Required

Procedure

Rescaling Not set up Link to ESL method - Performance(s) over time Yes Outcomes

RSL Yes Standards NT-BUILT 492 – Rapid Chloride Migration test Fully adopted Yes Agents considered Same


1 or more agents Same Performances evaluation Standards Not declared



Table 3.11: Concrete. Resistance to sulphates

Test name and chapter Sulphate resistance of cements - § 2.4 SLP based No Material / component Material: Portland cements Aims Agents Sodium sulphate (4.4% Na2SO4 solution) Refining SLPP No

Lab exposure

Yes Specimens: 1x4x16 cm Also long term exposure (1-10 years): 7.9%NaCl, 1,5 and 10%Na2SO4; 1%MgSO4 and 1% (NH4)2 SO4

Ageing cycle Extended exposure to the same agent Outdoor exposure Yes

Procedure

Rescaling No Link to ESL method No Performance(s) over time No Outcomes

RSL No

Standards

- CEN/TR 15697:2008 - Cement - Performance testing for sulfate

- resistance - State of the art report - EN 197-1:2000/prA2 - Cement - Part 1: Composition,

specifications and conformity criteria for common cements; Amendment A2 (Sulfate resisting cement)

Fully adopted Same Agents considered Same


1 or more agents Same Performances evaluation Standards -

Table 3.12: Concrete. Resistance to freezing

Test name and chapter Freezing resistance of concrete - § 2.5 SLP based No

Material / component Material: concrete, cement based materials, natural and artificial stones Aims

Agents Variations in temperature (freeze – thaw) Refining SLPP No Lab exposure Yes Ageing cycle Freeze – thaw cycling (simple cycling) Outdoor exposure -

Procedure

Rescaling Not considered in the test method Link to ESL method Finnish by 50 Performance(s) over time No Outcomes

RSL No

Standards

- EVS 814:2003 Frost resistance of normal weight concrete. Definitions, specification and test methods, Estonian Standard

- CEN/TR 15177:2006 - Testing the freeze-thaw resistance of concrete - Internal structural damage

- CEN/TS 12390-9:2006 - Testing hardened concrete - Part 9: Freeze-thaw resistance - Scaling



1 or more agents Same Performances evaluation Standards -



Table 3.12: Masonry walls

Test name and chapter Masonry walls - § 2.6 SLP based Yes Material / component Component. Paints applied on masonry walls Aims Agents Rain, Freeze-thaw, variations in temperature and relative

humidity, UV radiation

Refining SLPP

- Time re-scaling technique (rate accelerated -. natural ageing cycles)

- Climate analysis - Exposure in different sites

Lab exposure Yes Ageing cycle Simple cycling Outdoor exposure Yes

Procedure

Rescaling Yes Link to ESL method Yes, Performance Limit Method Performance(s) over time Yes Outcomes

RSL Yes Standards No Fully adopted NA Agents considered NA


1 or more agents NA Performances evaluation Standards Yes

Table 3.13: ETICS

Test name and chapter ETICS - § 2.7 SLP based Yes Material / component Building component Aims Agents UV, variations in temperature and relative humidity, rain

Refining SLPP

- Specimen construction standard procedures (recording) - Curing procedures before ageing - Climate data analysis and critical thresholds analysis so as

to pre-design the ageing cycle - Assessment of the influence of performance tests on ageing

Lab exposure Yes

Ageing cycle Composition of basic cycles (determined after climate data analysis). Attempt to reproduce the ageing part of actual reference year.

Outdoor exposure Prescribed

Procedure

Rescaling Yes Link to ESL method Yes, Performance Limit Method and Performance(s) over time Yes Outcomes

RSL Yes Standards ETAG 004 Fully adopted No Agents considered Variations in temperature and relative humidity, freeze


1 or more agents More agents, but cycles performed on different specimens (ETAG004)

Performances evaluation Standards

Water absorption [EN ISO 15148], Water vapour permeability [EN 12086], Tensile bond strength of adhesive and base coat to insulator [EN 13494], Render strip tensile test [ETAG 004 - § 5.5.4.1]. Hygrothemal performances tests not standardised.



Table 3.14: Renders

Test name and chapter External renders on façades - § 2.8 SLP based Yes Material / component Renders Aims Agents Variation in temperature and relative humidity, water

(immersion), rain, solar radiation Refining SLPP Multiple performances indicators for assessing SL Lab exposure Yes

Ageing cycle

Composition of cycles - Cycle A: heating/freezing, lab conditioning,

humidification/freezing - Cycle B: variations in temperature and relative humidity - Cycle C: variation in temperature and relative humidity, rain

and solar radiation Outdoor exposure Inspection on buildings

Procedure

Rescaling Yes, prediction model

Link to ESL method Performance based engineering methods and Factor Method (enhanced)

Performance(s) over time Yes Outcomes

RSL Yes

Standards EN 1015-21 RILEM MDT.D.3

Fully adopted No, adapted Agents considered Same


1 or more agents More Performances evaluation Standards EN 10155-12:2003, ASTM C 830

Table 3.15: Pitched roofs

Test name and chapter Pitched roofs - § 2.9 SLP based Yes Material / component Building component: sandwich panels for roofing Aims Agents Rain, UV, variations in temperature and relative humidity

Refining SLPP Strong link with Service Life Estimation methods (performance simple tests which can be performed on well - cores)

Lab exposure Yes

Ageing cycle Simple cycling Climate data analysis

Outdoor exposure Yes

Procedure

Rescaling Yes Link to ESL method Factor Method Performance(s) over time Yes Outcomes

RSL Yes

Standards EN 14509 Self supporting double skin metal faced insulated panels ETAG 016 Self supporting composite lightweight panels

Fully adopted No, adapted Agents considered Same


1 or more agents More Performances evaluation Standards UNI/CTI 7745



Table 3.16: Flat roofs

Test name and chapter Flat roofs - § 2.10 SLP based Yes Material / component Building component Aims Agents UV, variations in temperature and relative humidity, rain Refining SLPP Climate data analysis for ageing cycle determining Lab exposure Yes

Ageing cycle Cold, mild dry heat, intense dry heat, rain Climate data analysis

Outdoor exposure Yes

Procedure

Rescaling Yes Link to ESL method Not specifically Performance(s) over time To be developed Outcomes

RSL Yes Standards - Fully adopted - Agents considered -


1 or more agents - Performances evaluation Standards UNI EN 12086, UNI 8202, UNI 8223, ISO 8301

Table 3.17: External load bearing walls

Test name and chapter External load bearing walls - § 2.11 SLP based Yes Material / component Building component: masonry walls Aims Agents Rain, variations in temperature and relative humidity Refining SLPP Climate data analysis for ageing cycle design Lab exposure Yes Ageing cycle Rain, frost, hot humid climate, hot dry climate Outdoor exposure Yes

Procedure

Rescaling Yes Link to ESL method - Performance(s) over time Yes Outcomes

RSL Yes

Standards NFT T 30-049 “Peintures et vernis – revêtements à usage extérieur: essai de vieillissement artificial”

Fully adopted - Agents considered -


1 or more agents - Performances evaluation Standards NORMAL 44 - UNI 7357 - EN 196 part 1 - EN 1015 parts 12 and

19 - EN 772 part 1



4. Conclusions Starting from test methods reported in chapter 2 and the synopsis of chapter 3 it is possible to compare different approaches used for each components/materials in a table that will be useful in order to point out some conclusions. This is done in Table 4.1 for building materials and in Table 4.2 for components, in these tables it is possible to compare every test methods in term of aims, procedures and outcomes.

Table 4.1: Synopsis of all testing methods presented for building materials M. C = Component, CS = Composition of more phases within the cycle, CDA = Climate Data Analysis, EE = Extended Exposure to

the same agent, Y = Yes, M = Material, N = No, NS = not specified, S = Simple Cycling r = Required

Test

nam

e

Woo

d

Met

als: c

onde

nsat

ion

Met

als: s

pray

salt

tests

Met

als: s

ulphu

re

dioxid

e

Met

als:

ISO

1499

3

Met

als:

ISO

1199

7-1

Met

als:

ISO

1199

7-2

Conc

rete

: ca

rbon

ation

Conc

rete

: chlo

ride

pene

tratio

n

Conc

rete

: sulp

hate

s

Conc

rete

: fre

eze

-thaw

1. SLP based Y N N N Y Y Y N Y N N

2. Material/ Component

M/C M M M M M M C C M M/C Aims

3. Agents 4 2 1 1 3 3-4 5 1 1 1 1-2

4. Refining SLPP Y N N N N N Y Y N N N

5. Lab exposure Y Y Y Y Y Y Y Y Y Y Y

6. Ageing cycle (Simple / Composed)

SC SC EE EE SC SC SC EE EE EE SC

8. Outdoor exposure

R NS NS NS NS NS NS NS NS Y NS

Procedure

9. Rescaling Y NS NS NS Y Y Y N N N N

10. Link to ESL method

Y N N N N N N N N N Y


Y N N N N N N N Y N N Outcomes

12. RSL Y N N N Y Y Y N Y N N

13. Standards Y Y Y Y Y Y Y Y Y Y Y Standards for ageing procedure 14. Fully adopted Y Y Y Y Y Y Y Y Y Y Y

Performances evaluation

15. Standards NS NS NS NS NS NS NS Y NS NS NS



Table 4.2: Synopsis of all testing methods presented for building components. C = Component, CS = Composition of more phases within the cycle, CDA = Climate Data Analysis, EE = Extended Exposure to

the same agent, Y = Yes, M = Material, N = No, NS = not specified, S = Simple Cycling

Test

nam

e

Mas

onry

wall

s

ETIC

S

Rend

ers

Pitch

ed ro

ofs

Flat

roof

s

Load

bea

ring

walls

1. SLP based Y Y Y Y Y Y

2. Material/ Component

C C M/C C C C Aims

3. Agents 4 4 5 4 4 3

4. Refining SLPP Y Y Y Y Y N

5. Lab exposure Y Y Y Y Y Y

6. Ageing cycle (Simple / Composed)

SC CS SC SC SC SC

7. Outdoor exposure

Y Y Y Y Y Y

Procedure

8. Rescaling Y Y Y Y Y Y

9. Link to ESL method

Y Y Y Y Y Y


Y Y Y Y Y Y Outcomes

11. RSL Y Y Y Y Y Y

12. Standards N Y Y Y NS Y Standards for ageing procedure 13. Fully adopted - N N N NS NS

Performances evaluation

14. Standards Y Y Y Y Y Y

The first point under discussion is the general scope of the durability test, in order to evaluate the actual availability of test methods useful for service life prediction.

It appears clearly form item n.1 “SLP based” of both table 4.1 and 4.2 that most of accelerated tests for materials are simply a pass or fail durability test, while the test method for building components are mainly meant for providing information regarding Service Life, representing then the innovative performance based test methods.



This is also evident when considering outcomes (item n.9,10 and 11 of table 4.1 and 4.2): test procedures for building components are usually linked to some method for Service Life Estimation (i.e. Factor Methods, Engineering Methods, Stochastic Methods), allowing to evaluate performances over time and Reference Service Life; on the other side a few test methods for building materials have such outputs, as in the case of wood and only some reinforced concrete test methods, related with some specific action (i.e. Chloride penetration).

Related with the pass or fail durability test approach, it may be highlighted that in general such tests take into account the effects produced by only one or two agents, while of course the service life evaluation purpose implies to consider different agents synergic effects as it is in the actual service life (see item n.3: agents).

We may conclude that new research is needed for those test methods which don’t allow service life prediction, while the available test methods for service life prediction may be considered as actual best practice to refine the general procedure for service life prediction (ISO 15686-2) (see item n.4, Refining SLPP)

The accelerated test procedure All considered tests are based on laboratory ageing procedures (see item n. 5), most of them using the iteration of the same ageing cycle composed by different stages (SC), especially the already mentioned innovative test methods, developed for service life prediction (see item n. 1,6), thus simulating different agents; on the other hand some traditional test methods used for single material (for example: metals or concrete) uses extended exposure (EE) to the same agent (see item n. 3).

Test methods developed for service life prediction of building components (see tab.4.2) aims at estimating service life through the comparison of lab results with outdoor exposure through time rescaling, i.e. finding the proportion factor between accelerated and natural ageing (see item. 7, 8), while the traditional materials testing procedures seem not to make clear reference to outdoor exposure (see item. 7), but refer to some proportion factor, based on “experience” or data from different industrial sectors (like automotive industries for metals).

Standards for ageing procedure Most of the analyzed test methods refer to specific products durability standards (see item 12) but while for materials the standards are fully adopted, in the case of building components the test method has been partially adopted (usually for the ageing cycle), while the whole procedure had to be renewed as for the whole ageing timing and data elaboration, focusing on Service life and performance evaluation (see item16).

This report on test methods for service life prediction reveals a big international effort on the item, dealing with a relevant number of the most used building materials and components but established that some materials test methods, traditionally used to check durability with a “pass or fail” approach need to be revised in order to obtain service life and performance data.

In any case the new approach used for the recently developed test methods for service life prediction, could be used as a guide to develop new test methods for other building components with the same intended use.

Another upgrade should be considered about materials test methods, that can be used on a single material, no matter how the material can be used in actual project, and characterised by accelerated aging agents that aren’t always related to real climate agents and are often



limited to one or two at the same time; the development should focus to obtain specific test methods related with the specific intended use of the material in the building component.

In conclusion, despite a big effort by scientists and researchers, future developments, both on vertical objectives (for example: ETICS-EIFS, external renderings, roofing, concrete metals etc) and on horizontal objectives (for example: the methodology to define accelerated ageing cycle using statistical climatic data , the method to compare lab ageing results with outdoor exposure or building condition assessment data, in order to evaluate time re-scaling factor, the method to elaborate experimental data in order to define RSL and performance decay data) seem to be needed.

International Council for Research and Innovation in Building and Construction

CIB’s mission is to serve its members through encouraging and facilitating international cooperation and information exchange in building and construction research and innovation. CIB is en-gaged in the scientific, technical, economic and social domains related to building and construction, supporting improvements in the building process and the performance of the built envi-ronment.

CIB Membership offers:• international networking between academia, R&D organisations and industry• participation in local and international CIB conferences, symposia and seminars• CIB special publications and conference proceedings• R&D collaboration

Membership: CIB currently numbers over 400 members origi-nating in some 70 countries, with very different backgrounds: major public or semi-public organisations, research institutes, universities and technical schools, documentation centres, firms, contractors, etc. CIB members include most of the major national laboratories and leading universities around the world in building and construction.

Working Commissions and Task Groups: CIB Members participate in over 50 Working Commissions and Task Groups, undertaking collaborative R&D activities organised around:• construction materials and technologies• indoor environment• design of buildings and of the built environment• organisation, management and economics• legal and procurement practices

Networking: The CIB provides a platform for academia, R&D organisations and industry to network together, as well as a network to decision makers, government institution and other building and construction institutions and organisations. The CIB network is respected for its thought-leadership, information and knowledge.

The CIB has formal and informal relationships with, amongst others: the United Nations Environmental Programme (UNEP); the European Commission; the European Network of Building Research Institutes (ENBRI); the International Initiative for Sustainable Built Environment (iiSBE), the International Or-ganization for Standardization (ISO); the International Labour Organization (ILO), International Energy Agency (IEA); Inter-national Associations of Civil Engineering, including ECCS, fib, IABSE, IASS and RILEM.

Conferences, Symposia and Seminars: CIB conferences and co-sponsored conferences cover a wide range of areas of interest to its Members, and attract more than 5000 partici-pants worldwide per year.

Leading conference series include:• International Symposium on Water Supply and Drainage for Buildings (W062)• Organisation and Management of Construction (W065)• Durability of Building Materials and Components (W080, RILEM & ISO)• Quality and Safety on Construction Sites (W099)• Construction in Developing Countries (W107)• Sustainable Buildings regional and global triennial conference series (CIB, iiSBE & UNEP)• Revaluing Construction• International Construction Client’s Forum

CIB Commissions (April 2010)TG53 Postgraduate Research Training in Building and ConstructionTG57 Industrialisation in Construction TG58 Clients and Construction Innovation TG59 People in Construction TG62 Built Environment Complexity TG63 Disasters and the Built EnvironmentTG64 Leadership in ConstructionTG65 Small Firms in ConstructionTG66 Energy and the Built EnvironmentTG67 Statutory Adjudication in ConstructionTG68 Construction MediationTG69 Green Buildings and the LawTG71 Research and Innovation TransferTG72 Public Private PartnershipTG73 R&D Programs in ConstructionTG74 New Production and Business Models in ConstructionTG75 Engineering Studies on Traditional ConstructionsTG76 Recognising Innovation in ConstructionTG77 Health and the Built EnvironmentTG78 Informality and Emergence in ConstructionTG79 Building Regulations and Control in the Face of Climate Change W014 Fire W018 Timber Structures W023 Wall Structures W040 Heat and Moisture Transfer in Buildings W051 Acoustics W055 Building Economics W056 Sandwich Panels W062 Water Supply and Drainage W065 Organisation and Management of Construction W069 Housing Sociology W070 Facilities Management and Maintenance W077 Indoor Climate W078 Information Technology for Construction W080 Prediction of Service Life of Building Materials and ComponentsW083 Roofing Materials and SystemsW084 Building Comfortable Environments for All W086 Building Pathology W089 Building Research and Education W092 Procurement Systems W096 Architectural Management W098 Intelligent & Responsive Buildings W099 Safety and Health on Construction Sites W101 Spatial Planning and infrastructure Development W102 Information and Knowledge Management in BuildingW104 Open Building Implementation W107 Construction in Developing Countries W108 Climate Change and the Built Environment W110 Informal Settlements and Affordable Housing W111 Usability of WorkplacesW112 Culture in ConstructionW113 Law and Dispute ResolutionW114 Earthquake Engineering and BuildingsW115 Construction Materials StewardshipW116 Smart and Sustainable Built EnvironmentsW117 Performance Measurement in Construction

PAGE 1

International Council for Research and Innovation in Building and Construction

Publications: The CIB produces a wide range of special publications, conference proceedings, etc., most of which are available to CIB Members via the CIB home pages. The CIB network also provides access to the publications of its more than 400 Members.

Recent CIB publications include:• Guide and Bibliography to Service Life and Durability Research for Buildings and Components (CIB 295)• Performance Based Methods for Service Life Prediction (CIB 294)• Performance Criteria of Buildings for Health and Comfort (CIB 292)• Performance Based Building 1st International State-of-the- Art Report (CIB 291)• Proceedings of the CIB-CTBUH Conference on Tall Buildings: Strategies for Performance in the Aftermath of the World Trade Centre (CIB 290)• Condition Assessment of Roofs (CIB 289)• Proceedings from the 3rd International Postgraduate Research Conference in the Built and Human Environment• Proceedings of the 5th International Conference on Performance-Based Codes and Fire Safety Design Methods• Proceedings of the 29th International Symposium on Water Supply and Drainage for Buildings• Agenda 21 for Sustainable Development in Developing Countries

R&D Collaboration: The CIB provides an active platform for international collaborative R&D between academia, R&D organisations and industry.

Publications arising from recent collaborative R&D ac-tivities include:• Agenda 21 for Sustainable Construction• Agenda 21 for Sustainable Construction in Developing Countries• The Construction Sector System Approach: An International Framework (CIB 293)• Red Man, Green Man: A Review of the Use of Performance Indicators for Urban Sustainability (CIB 286a)• Benchmarking of Labour-Intensive Construction Activities: Lean Construction and Fundamental Principles of Working Management (CIB 276)• Guide and Bibliography to Service Life and Durability Research for Buildings and Components (CIB 295)• Performance-Based Building Regulatory Systems (CIB 299)• Design for Deconstruction and Materials Reuse (CIB 272)• Value Through Design (CIB 280)

An example of a recent major CIB col-laborative activity is the Thematic Net-work PeBBu Performance Based Building: a four-year programme that included 50 member organisations, that was co-ordinated by CIB and that was funded through the European Commission Fifth Framework Programme.

Themes: The main thrust of CIB activities takes place through a network of around 50 Working Commissions and Task Groups, organised around four CIB Priority Themes:• Sustainable Construction• Clients and Users• Revaluing Construction• Integrated Design and Delivery Solutions

CIB Annual Membership Fee 2007 – 2010

Fee Category 2007 2008 2009 2010

FM1 Fee level 10526 11052 11605 11837FM2 Fee level 7018 7369 7738 7892FM3 Fee level 2413 2534 2661 2715AM1 Fee level 1213 1274 1338 1364AM2 Fee level 851 936 1030 1133IM Fee level 241 253 266 271All amounts in EURO

The lowest Fee Category an organisation can be in depends on the organisation’s profile:

FM1 Full Member Fee Category 1 | Multi disciplinary building research institutes of national standing having a broad field of research FM2 Full Member Fee Category 2 | Medium size research Institutes; Public agencies with major research interest; Companies with major research interestFM3 Full Member Fee Category 3 | Information centres of national standing; Organisations normally in Category 4 or 5 which prefer to be a Full MemberAM1 Associate Member Fee Category 4 | Sectoral research & documentation institutes; Institutes for standardisation; Companies, consultants, contractors etc.; Professional associations AM2 Associate Member Fee Category 5 | Departments, fac- ulties, schools or colleges of universities or technical Institutes of higher education (Universities only)IM Individual Member Fee Category 6 | Individuals having an interest in the activities of CIB (not representing an organisation)

Fee Reduction: A reduction is offered to all fee levels in the magnitude of 50% for Members in countries with a GNIpc less than USD 1000 and a reduction to all fee levels in the magnitude of 25% for Mem-bers in countries with a GNIpc between USD 1000 – 7000, as defined by the Worldbank. (see http://siteresources.worldbank.org/DATASTATISTICS/Resources/GNIPC.pdf)

Reward for Prompt Payment:All above indicated fee amounts will be increased by 10%. Mem-bers will subsequently be rewarded a 10% reduction in case of actual payment received within 3 months after the invoice date.

For more information contact CIB General Secretariat:e-mail: [email protected]

PO Box 1837, 3000 BV Rotterdam, The NetherlandsPhone +31-10-4110240;Fax +31-10-4334372Http://www.cibworld.nl

PAGE 2

DISCLAIMER

All rights reserved. No part of this book may be reprinted or

reproduced or utilized in any form or by any electronic,

mechanical, or other means, now known or hereafter

invented, including photocopying and recording, or in any

information storage or retrieval system without

permission in writing from the publishers.

The publisher makes no representation, express or implied,

with regard to the accuracy of the information contained in this book

and cannot accept any legal responsibility or liability in whole or in part

for any errors or omissions that may be made.

The reader should verify the applicability of the information to

particular situations and check the references prior to any reliance

thereupon. Since the information contained in the book is multidisciplinary,

international and professional in nature, the reader is urged to consult with

an appropriate licensed professional prior to taking any action or making

any interpretation that is within the realm of a licensed professional practice.

CIB General Secretariatpost box 18373000 BV RotterdamThe NetherlandsE-mail: [email protected]

CIB Publication 331