· aggregates according to several concrete prism ... aggregate and geological types found across Europe and calibrated the results of ... Aggregate Origin Comb.

Copyright © PARTNER-project-GRD1-CT-2001-40103 All rights reserved

3

Preface This report is one of a series produced as output from PARTNER, a project funded by the European Community under the “Competitive and Sustainable Growth“ programme. The overall objective of this project is to provide the basis for a unified test procedure for evaluating the alkali reactivity of aggregates across the different European economic and geological regions. It will enable CEN TC-154, Aggregates, to fulfil the requirements of the Aggregates Mandate, M125, which identifies durability against alkalis as a necessary performance characteristic in the specification of aggregates for concrete (EN 12620) to meet the Essential Requirements of the CPD for Strength and Safety. The project will achieve this by:

• Evaluating the tests developed by RILEM, and some regional tests, for their suitability for use with the wide variety of aggregate and geological types found across Europe.

• Calibrating the results of these accelerated tests against behaviour in concrete in real structures and in field sites.

• Producing an “atlas” of the geology and petrography of European aggregates. • Educating European petrographers and testing organisations in the effective use of these

methods. • Making recommendations, based on the above work, to CEN for suitable CEN methods of test

and specifications to ensure durability against alkalis. The project has 24 Partners from 14 countries, covering most of Europe, from Iceland to Greece. Partners BRE ……………………………...……………………………………………………………….UK PC Laboratoriet A/S ....................................................................................................................... DK SINTEF ......................................................................................................................................... NO SP - Swedish National Testing and Research Institute .................................................................. SE RAMBOLL .................................................................................................................................... DK ISSeP – Institut Scientifique de Service Public.............................................................................. B LCPC Laboratory .......................................................................................................................... F VDZ – German Cement Works Association ................................................................................. D TITAN Cement Company, S.A...................................................................................................... EL LABEIN ......................................................................................................................................... ES CRIC – Centre National de Recherches Scientifiques et techniques pour L’Industrie Cimentière .................................................................................................................. B IMBiGS .......................................................................................................................................... PL NORCEM A.S................................................................................................................................ NO CESI ............................................................................................................................................... I VOZ – Austrian Cement Association............................................................................................. A IBRI – Icelandic Building Research Institute STATS............................................................................................................................................ UK NCC AB ......................................................................................................................................... SE DANISH TECHNOLOGICAL INSTITUTE ................................................................................ DK AIDICO.......................................................................................................................................... ES ANEFA........................................................................................................................................... ES Cemex............................................................................................................................................. ES HOLCIM ........................................................................................................................................ B Hönnun........................................................................................................................................... IS


4

List of available reports in the PARTNER project WP 2: Aggregates and structures (leader: RAMBÖLL) Report

no. Report title Author

2.1 State-of-the art report: Key parameters influencing the alkali aggregate reaction SBF52 A06018 / ISBN 82-14-04078-7 / 978-82-14-04078-7

Hönnun, RAMBÖLL,

SINTEF

WP 3: Test methods (leader: SINTEF) Report


3.1

Experience from using petrographic analysis according to the RILEM AAR-1 method to assess alkali reactions in European aggregates SBF52 A06019 / ISBN 82-14-04079-5 / 978-82-14-04079-5

SINTEF

3.2 Experience from testing of the alkali reactivity of European aggregates according to the RILEM AAR-2 method SBF52 A06020 / ISBN 82-14-04080-9 / 978-82-14-04080-9

PC-lab

3.3 Experience from testing of the alkali reactivity of European aggregates according to several concrete prism test methods SBF52 A06021 / ISBN 82-14-04081-7 / 978- 82-14-04081-7

BRE

3.4 Experience from testing of the alkali reactivity of European aggregates according to two Danish laboratory test methods SBF52 A06022 / ISBN 82-14-04082-5 / 978- 82-14-04082-5

RAMBÖLL

3.5

Field site tests established in the PARTNER project for evaluating the correlation between laboratory tests and field performance SBF52 A06023 / ISBN 82-14-04083-3 / 978- 82-14-04083-3

VDZ

WP 4: Precision trials (leader: SP) Report


4.1 PRECISION TRIAL – Determination of repeatability and reproducibility of the amended RILEM methods SBF52 A06024 / ISBN 82-14-04084-1 / 978-82-14-04084-1

SP

WP 5: Dissemination (leader: PC-lab) Report


Final results and recommendations of the PARTNER project. Paper to be published at ICAAR 2008, Trondheim, Norway

BRE + several co-authors

Database / atlas of the alkali reactivity of European aggregates Published by Geological Survey of Belgium see www.aarig.org/webatlas/atlas.htm

PC-lab, ISSEP, Hönnun


5

Summary This report presents the results of an evaluation of the use of several concrete prism test methods to assess the alkali reactivity of European aggregates. It is one of a series of such evaluations, carried out under Work Package 3 in the PARTNER programme. The R&D programme has evaluated the tests developed by RILEM, and some regional tests, for their suitability for use with the wide variety of aggregate and geological types found across Europe and calibrated the results of these accelerated tests against behaviour in concrete in real structures and in field sites. Five concrete prism tests were evaluated: - RILEM AAR-3 concrete prism method (storage at 38 degrees C) - RILEM AAR-4 concrete prism method (storage at 60 degrees C) - RILEM AAR-4 alternative method (storage at 60 degrees C) - German concrete test method (storage at 40 degrees C) - Norwegian concrete prism method (storage at 38 degrees C) In total 20 combinations of aggregates, reflecting the range of geological types found across Europe, were tested. In the overwhelming majority of cases all the concrete methods correctly identified those aggregate combinations that had been shown by field experience to be involved in cases of damaging alkali reactivity. Similarly all of the methods correctly identified those aggregate combinations that were established from field experience as being non-reactive. However, some uncertainties showed up where the damaging reactions were known to be particularly slow. In general, amongst the RILEM methods, the results from the AAR-4 methods and in particular the AAR-4 reactor method, were the most consistent and most clearly identified the reactivity of also the slowly reactive aggregate combinations. Where the Norwegian and German methods were compared with the RILEM methods, the agreement is good. The most common problems experienced were in ensuring adequate humidity around the specimens. Other concerns were the procedures for working with dry aggregates, the proportion of the fine aggregates in the mix, allowing for the different densities of the aggregates, the resulting mix design and the mixing procedure. Recommendations for improving the methods in the light of these tests are made.


6

Content

1. Introduction ........................................................................................................................... 7

2. Method ........................................................................................................................... 7

3. Work programme .................................................................................................................. 8

4. Results ......................................................................................................................... 11

5. Discussion / interpretation .................................................................................................. 31

6. Comments on the methods.................................................................................................. 33

7 Conclusion ......................................................................................................................... 35

8. References ......................................................................................................................... 35 Annexes Annex 1: Test methods and notes on testing Annex 2: Petrographic composition and field performance of the aggregates tested Annex 3: Fresh concrete properties


7

1. Introduction This report describes the results of an evaluation of the use of several concrete prism test methods to assess the alkali reactivity of European aggregates. It is one of a series of such evaluations, carried out under Work Package 3 in the PARTNER programme. The R&D programme has evaluated the tests developed by RILEM, and some regional tests, for their suitability for use with the wide variety of aggregate and geological types found across Europe and calibrated the results of these accelerated tests against behaviour in concrete in real structures and in field sites. Five concrete prism tests were evaluated:

- RILEM AAR-3 concrete prism method (storage at 38 degrees C) - RILEM AAR-4 concrete prism method (storage at 60 degrees C) - RILEM AAR-4 alternative method (storage at 60 degrees C) - German concrete test method (storage at 40 degrees C) - Norwegian concrete prism method (storage at 38 degrees C)

The report is written by Philip Nixon ([email protected]) and Simon Lane ([email protected]), BRE, UK.

2. Method The concrete prism methods which were assed are summarised in Table 2.1. The references are to the original published test methods while further notes and descriptions specifically for this project are given in Annex 1. The notes also include results sheets for the different methods. Table 2.1: Brief outline of test methods used for Work Package 3.3 (concrete prism tests)

Test method Brief outline of method Criteria, i.e. reactive if:

RILEM AAR-3 (Ref. 1)

Accelerated expansion test for 12 months. Prisms stored in individual containers within a constant temperature room at 38°C and measured at 20°C.

Expansion > 0.05 % after 12 months

RILEM AAR-4 (Ref. 2)

Accelerated expansion test for 20 weeks. Prisms stored in individual containers within a reactor at 60°C and measured at 20°C.

Expansion > 0.03 % after 20 weeks

RILEM AAR-4 Alternative (Ref. 2)

Accelerated expansion test for 20 weeks. Prisms stored in individual containers within a constant temperature room at 60°C and measured at 20°C.

Expansion > 0.03 % after 20 weeks

German Method (Ref. 3)

Test duration of 9 months. Cubes and prisms stored in a fog chamber at 40°C with measurements taken immediately with no cooling down period.


Norwegian method (Ref.4)

Accelerated expansion test for 12 months. Prisms stored in individual containers within a constant temperature room at 38°C and 100% relative humidity and measured at 20°C.



8

3. Work programme 3.1 Aggregates tested In total 20 combinations of aggregates were tested using the concrete prism methods. The full list of aggregate combinations is given in Table 3.1. Not all laboratories tested every combination; neither has every method been applied to all aggregate combinations (see schedule in 3.2). The petrographic composition and field performance of the aggregates tested is given in Annex 2. Table 3.1: Aggregate combinations tested. Sample number Aggregate details Combination B1 Silicified limestone C + NRF B1 Silicified limestone C + F D1 Gravel with opaline flint C + F D2 Sea gravel with semi-dense flint F + NRC F1 Gravel with flint C + NRF F2 Non-reactive limestone C + F F3 Siliceous gravel C + F G1 Gravel with siliceous limestone and chert C + NRF IT1 Gravel with limestone, chert and flint C + F IT2 Gravel with quartzite and gneiss C + F N1 Cataclasite C + NRF N2 Sandstone C + NRF N3 Non-reactive granitic sand C + F N4 Gravel with sandstone and cataclastic rocks C + F N5 Gravel with rhyolite and quartzite C + F N6 Gravel with sedimentary rocks C + F S1 Gravel with porphyritic rhyolite C + F UK1 Greywacke C + F UK2 Gravel with quartzite and chert C + F P1 Silicified limestone C + NRF C = coarse aggregate; F = fine aggregate; NRF = non-reactive fine aggregate.


9

3.2 Tests carried out AAR-3 method Aggregate Origin Comb. BRE SP NORCEM CESI LABEIN DTI AIDICO IBRI VOZ TITAN B1 C+NRF X X X B1 C+F X X X D1 C+F X X D2 F+NRC X X FI C+NRF X X X F2 C+F X X X F3 C+F X X X G1 C+NRF X X IT1 C+F X X X IT2 C+F X X X N1 C+NRF X X X N2 C+NRF X N3 C+F X N4 C+F X X X N5 C+F X X N6 C+F X S1 C+F X X UK1 C+F X X UK2 C+F X X X P1 C+NRF X X AAR-4 method Aggregate Origin Comb. LCPC VDZ CRIC CESI IMBIGS HOLCIM B1 C+NRF X X X B1 C+F X X X D1 C+F X X X D2 F+NRC X X X D3 F FI C+NRF X X X X F2 C+F X X X F3 C+F X X X X G1 C+NRF X X X X IT1 C+F X X X IT2 C+F X X X X N1 C+NRF X X X N2 C+NRF N3 C+F X X N4 C+F X X X N5 C+F X X X N6 C+F S1 C+F X X X X UK1 C+F X X X UK2 C+F X X X X P1 C+NRF X X X X


10

AAR-4 alternative method Aggregate Origin Comb. BRE ISSEP CRIC B1 C+NRF X X B1 C+F X X D1 C+F X X D2 F+NRC X X D3 F FI C+NRF X X F2 C+F X F3 C+F X G1 C+NRF X X IT1 C+F X IT2 C+F X N1 C+NRF X UK1 C+F X X UK2 C+F X X P1 C+NRF X “GERMAN” and “NORWEGIAN” methods Aggregate GERMAN NORWEGIAN Origin Comb. VDZ SINTEF NORCEM B1 C+F X X D1 C+F X X FI C+NRF X F2 C+F X X G1 C+NRF X N1 C+NRF X X N2 C+NRF X X N3 C+F X N4 C+F X X X N5 C+F X X X N6 C+F X S1 C+F X X UK1 C+F X X 3.3 Fresh concrete properties The fresh concrete properties of the batches produced solely for the laboratory concrete specimens were not consistently reported. However, in the laboratories where concrete for the field site specimens was produced, the fresh concrete properties were better reported as the same concretes were used by these laboratories in making the laboratory specimens. This data is given in Annex 3.


11

4. Results All the laboratories reported the results electronically on the result sheets included in the notes (see Annex 1). In Table 4.1-4.20 the % expansion results from the final measurements for each aggregate combination in each of the test methods are given. (x) indicates no test carried out. In the corresponding figures the expansion results for each aggregate combination from each of the concrete methods used are given. Where there is a known problem with the data, the results are shown with a dashed line. For each aggregate type its known reactivity in field are summarised in the table heading (see also Annex 2). The suggested critical limits for each of the testing methods are given in Table 2.1. Table 4.1: Aggregate B1 (Silicified limestone) coarse and non-reactive fines.

Reactivity: Damage to several structures.

Laboratory/country BRE IBRI TITAN CRIC CESI HOLCIM

Test method GB IS EL B I B

RILEM AAR-3 (12 Months) 0.362 0.146 0.002 x x x

RILEM AAR-4 (20 weeks) x x x 0.156 0.084 0.170

RILEM AAR-4 Alternative (20 weeks)

0.178 x x 0.120 x x

German Method (9 months) x x x x x x

Norwegian method (12 months) x x x x x x

B1 C+NRF all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400

Storage period in days

Expa

nsio

n of

the

pris

ms

% M4alt BRE

M4alt CRIC

M3 BRE

M3 IBRI

M3 TITAN

M4 CRIC

M4 CESI

M4 HOLCIM


12

Table 4.2: Aggregate B1 (Silicified limestone) coarse and fines. Reactivity: Damage to several structures

Laboratory/country

NORCEM AIDICO IBRI VDZ CRIC BRE HOLCIM Test method N ES IS D B GB B

RILEM AAR-3 (12 Months) 0.274 0.29 0.17 x x x x

RILEM AAR-4 (20 weeks) x x x 0.15 0.12 x 0.129


x x x x 0.11 0.17 x

German Method (9 months) x x x 0.12 x x x

Norwegian method (12 months) 0.234 x x x x x x

B1 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZM4alt BREM4alt CRICM3 NORCEMM3 IBRIM3 AIDICOM4 VDZM4 CRICM4 HOLCIMNorwegian NORCEM


13

Table 4.3: Aggregate D1 gravel with opaline flint coarse and fines. Reactivity: Severe and rapid damage to structures, but pessimum behaviour when coarse and fines combined.

Laboratory/country

BRE DTI LCPC VDZ CRIC ISSEP Test method GB DK F D B B

RILEM AAR-3 (12 Months) 0.233 0.026

(182 d) x x x x

RILEM AAR-4 (20 weeks) x x 0.019 0.024 0.025 x


x x x x 0.036 0.031

German Method (9 months) x x x 0.032 x x


D1 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZ

M4alt ISSEP

M4alt CRIC

M3 DTI

M3 BRE

M4 LCPC

M4 VDZ

M4 CRIC


14

Table 4.4: Aggregate D2 sea dredged semi dense flint fines and non-reactive coarse. Reactivity: Severe but slow damage to many structures

Laboratory/country

BRE DTI LCPC VDZ CRIC Test method GB DK F D B


(182d) x x x

RILEM AAR-4 (20 weeks) x x 0.120 0.095 0.042


0.034 x x x 0.042

German Method (9 months) x x x x x

Norwegian method (12 months) x x x x x

D2 F+NRC all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

M4alt BREM4alt CRICM3 DTIM3 BREM4 LCPCM4 VDZM4 CRIC


15

Table 4.5: Aggregate F1 gravel with flint coarse and non-reactive fines. Reactivity: No known damage to structures.

Laboratory/country

BRE CESI LABEIN LCPC VDZ CRIC Test method GB I ES F D B

RILEM AAR-3 (12 Months) 0.002 -0.015 -0.005 x x x

RILEM AAR-4 (20 weeks) x 0.009 x 0.010 0.007 0.014


0.028 x x x x 0.019

German Method (9 months) x x x x 0.028 x


F1 C+NRF all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZM4alt BREM4alt CRICM3 BREM3 CESIM3 LABEINM4 LCPCM4 VDZM4 CRICM4 CESI


16

Table 4.6: Aggregate F2 non-reactive limestone coarse and fines. Reactivity: Non-reactive

Laboratory/country BRE NORCEM VOZ LCPC VDZ CESI Test method GB N A F D I


RILEM AAR-4 (20 weeks) x x x 0.009 0.012 -0.004


0.016 x x x x x

German Method (9 months) x x x x 0.021 x

Norwegian method (12 months) x 0.005 x x x x

F2 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZ

M4alt BRE

M3 BRE

M3 NORCEM

M3 VOZ

M4 LCPC

M4 VDZ

M4 CESI

NorwegianNORCEM


17

Table 4.7: Aggregate F3 siliceous gravel coarse and fines. Reactivity: No known damage to structures

Laboratory/country

LABEIN AIDICO TITAN LCPC VDZ CRIC CESI Test method ES ES EL F D B I

RILEM AAR-3 (12 Months) -0.001 0.030 0.000 x x x x

RILEM AAR-4 (20 weeks) x x x 0.018 0.016 0.023 0.007


x x x x x 0.023 x

German Method (9 months) x x x x x x x

Norwegian method (12 months) x x x x x x x

F3 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

M4alt CRIC

M3 LABEIN

M3 AIDICO

M3 TITAN

M4 LCPC

M4 VDZ

M4 CRIC

M4 CESI


18

Table 4.8: Aggregate G1 gravel with siliceous limestone and chert coarse and non-reactive fines. Reactivity: Long term damage to concrete pavements

Laboratory/country

BRE VOZ LCPC VDZ CRIC IMBIGS Test method GB A F D B PL

RILEM AAR-3 (12 Months) 0.963 0.164 x x x x

RILEM AAR-4 (20 weeks) x x 0.141 0.168 0.140 0.114

(16 w) RILEM AAR-4 Alternative (20 weeks)

0.162 x x x 0.121 x

German Method (9 months) x x x 0.166 x x


G1 C+NRF all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZM4alt BREM4alt CRICM3 BREM3 VOZM4 LCPCM4 IMBIGSM4 VDZ M4 CRIC


19

Table 4.9: Aggregate It1 gravel with limestone chert and flint coarse and fines. Reactivity: Rapid damage to structures

Laboratory/country

CESI LABEIN AIDICO LCPC CRIC ISSEP Test method I ES ES F B B


RILEM AAR-4 (20 weeks) 0.186 x x 0.039 0.232 x


x x x x x 0.083



IT1 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

% M4alt ISSEP

M3 CESI

M3 LABEIN

M3 AIDICO

M4 LCPC

M4 CRIC

M4 CESI


20

Table 4.10: Aggregate It2 gravel with quartzite and gneiss coarse and fines. Reactivity: Slow (+50 years) damage to one structure

Laboratory/country

CESI VOZ TITAN VDZ HOLCIM ISSEP Test method I A EL D B B


RILEM AAR-4 (20 weeks) 0.045 x x 0.097 0.061 x


x x x x x 0.071



IT2 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

M4alt ISSEP

M3 CESI

M3 VOZ

M3 TITAN

M4 VDZ

M4 CESI

M4 HOLCIM


21

Table 4.11: Aggregate N1 cataclasite coarse and non-reactive fines. Reactivity: Severe damage to structures

Laboratory/country

SP DTI NORCEM VDZ CRIC HOLCIM ISSEP Test method S DK N D B B B


(182 days) 0.435 x x x x

RILEM AAR-4 (20 weeks) x x x 0.241 0.176 0.175 x


x x x x x x 0.139


Norwegian method (12 months) x x 0.316 x x x x

N1 C+NRF all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZ

M4alt ISSEP

M3 SP

M3 DTI

M3 NORCEM

M4 VDZ

M4 CRIC

M4 HOLCIM

NorwegianNORCEM


22

Table 4.12: Aggregate N2 sandstone coarse and non-reactive fines. Reactivity: Slow damage to structures (+15-20 years)

Laboratory/country

NORCEM SINTEF Test method N N

RILEM AAR-3 (12 Months) 0.209 x

RILEM AAR-4 (20 weeks) x x


x x

German Method (9 months) x x

Norwegian method (12 months) 0.265 0.256

N2 C+NRF all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

M3 NORCEM

Norwegian SINTEF

Norwegian NORCEM


23

Table 4.13: Aggregate N3 non-reactive granitic sand coarse and fines. Reactivity: No damage reported

Laboratory/country

NORCEM IMBIGS HOLCIM Test method N PL B

RILEM AAR-3 (12 Months) 0.012 x x

RILEM AAR-4 (20 weeks) x 0.026

(16w) 0.039


x x x

German Method (9 months) x x x

Norwegian method (12 months) 0.012 x x

N3 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

M3 NORCEMM4 IMBIGSM4 HOLCIMNorwegian NORCEM


24

Table 4.14: Aggregate N4 gravel with sandstone and cataclasite rocks coarse and fines. Reactivity: Slow damage to structures (+20 years)

Laboratory/country

SP NORCEM IBRI VDZ CRIC HOLCIM SINTEF Test method S N IS D B B N




x x x x x x x


Norwegian method (12 months) x 0.068 x x x x 0.061

N4 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZ

M3 SP

M3 NORCEM

M3 IBRI

M4 VDZ

M4 CRIC

M4 HOLCIM

Norwegian SINTEF

Norwegian NORCEM


25

Table 4.15: Aggregate N5 gravel with rhyolite and quartzite coarse and fines. Reactivity: Slow damage to structures (+20 years)

Laboratory/country

SP NORCEM IBRI VDZ CRIC HOLCIM SINTEF Test method S N IS D B B N




x x x x x x x


Norwegian method (12 months) x 0.087 x x x x 0.083

N5 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZ

M3 SP

M3 NORCEM

M3 IBRI

M4 VDZ

M4 CRIC

M4 HOLCIM

Norwegian SINTEF

Norwegian NORCEM


26

Table 4.16: Aggregate N6 gravel with sedimentary rocks coarse and fines. Reactivity: Slow damage to structures (+20 years)

Laboratory/country

NORCEM Test method N

RILEM AAR-3 (12 Months) 0.059

RILEM AAR-4 (20 weeks) x


x

German Method (9 months) x

Norwegian method (12 months) 0.076

N6 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

M3 NORCEM

Norwegian NORCEM


27

Table 4.17: Aggregate S1 gravel with porphyritic rhyolite coarse and fines. Reactivity: Moderate damage to structures, but reactivity of source is variable.

Laboratory/country

SP NORCEM VDZ CRIC IMBIGS HOLCIM Test method S N D B PL B

RILEM AAR-3 (12 Months) 0.005 0.014 x x x x


(16w) 0.021


x x x x x x

German Method (9 months) x x 0.03 x x x

Norwegian method (12 months) x 0.049 x x x x

S1 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZ

M3 SP

M3 NORCEM

M4 VDZ

M4 IMBIGS

M4 CRIC

M4 HOLCIM

Norwegian SINTEF


28

Table 4.18: Aggregate UK1 greywacke coarse and fines. Reactivity: Slow (+20 years) damage to structures

Laboratory/country

BRE NORCEM LCPC CRIC HOLCIM ISSEP VDZ Test method GB N F B B B D

RILEM AAR-3 (12 Months) 0.207 0.527 x x x x x

RILEM AAR-4 (20 weeks) x x 0.191 0.175 0.180 x x


0.071 x x x x 0.156 x

German Method (9 months) x x x x x x 0.196

Norwegian method (12 months) x 0.255 x x x x x

UK1 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

German VDZ

M4alt BRE

M4alt ISSEP

M3 BRE

M3 NORCEM

M4 LCPC

M4 CRIC

M4 HOLCIM

N i


29

Table 4.19: Aggregate UK2 gravel with quartzite and chert coarse and fines. Reactivity: Some damage to structures

Laboratory/country

BRE DTI VOZ LCPC CRIC IMBIGS HOLCIM ISSEP Test method GB DK A F B PL B B


(182d) 0.118 x x x x x


(16w) 0.094 x


0.051 x x x x x 0.058



UK2 C+F all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

M4alt BREM4alt ISSEP M3 BREM3 DTIM3 VOZ M4 LCPCM4 IMBIGSM4 CRIC M4 HOLCIM


30

Table 4.20: Aggregate P1 silicified limestone coarse and non-reactive fines. Reactivity: Similar material has caused damage to structures, but material tested probably has lower silica content

Laboratory/country

CESI LABEIN AIDICO LCPC CRIC HOLCIM ISSEP Test method I ES ES F B B B


RILEM AAR-4 (20 weeks) x x x 0.026 0.03

4 0.028 0.003


x x x x x x 0.036



P1 C+NRF all methods

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 50 100 150 200 250 300 350 400


Expa

nsio

n of

the

pris

ms

%

M4alt ISSEP

M3 CESI

M3 LABEIN

M3 AIDICO

M4 LCPC

M4 CRIC

M4 HOLCIM

M4 CESI


31

5. Discussion / interpretation 5.1 Identification of reactive aggregate combinations In the overwhelming majority of cases all the concrete methods correctly identified those aggregate combinations that had been shown by field experience to be involved in cases of damaging alkali reactivity. 5.1.1. Aggregates that react in “normal” timescales The methods were particularly effective in identifying aggregate combinations that caused damage in “normal” time scales; 10 to 20 years. In the case of D1 all the laboratories except one identified this aggregate combination as non-reactive, whereas it is known from field experience to cause rapid and severe damage to structures in Denmark. It is also known, however, that this aggregate type has a marked pessimum behaviour and it is probable that the coarse and fine combination tested is well above the pessimum content of reactive silica.

0

0.05

0.1

0.15

0.2

0.25

0.3

B1 C+Nr SilLim

B1 C+F SilLim

UK1 C+FGreywacke

G1 C+Nr SilLim Gr

IT1 C+F Lim& Fl Gr

N1 C+NrCatacl

UK2 C+F Qz Chert Gr

D1 C+F

Reactive aggregate combinations: % expansion at the end of the test period – all concrete methods[Order in chart is AAR-3 (light blue), AAR-4 (medium blue), AAR-4alt (dark blue), German (red), Norwegian (green)]


32

5.1.2 Non-reactive aggregate combinations Similarly all of the methods correctly identified those aggregate combinations that were established from field experience as being non-reactive. In the case of S1(C+F) all of the methods showed this combination to be on the margins of reactivity. Although this aggregate has been involved in known failures; its composition and reactivity is known to be variable.

-0,020

0,020,040,060,080,1

0,120,140,16

F1 C+NrFlint Grav

F2 C+FLim

F3 C+FSil Grav

N3 C+FGransand

S1 C+FRhy Grav

P1 C+Nr reactive

Non-reactive aggregate combinations: % expansion at the end of the test period – all concrete methods [Order in chart is AAR-3 (light blue), AAR-4 (medium blue), AAR-4alt (dark blue), German (red), Norwegian (green). One “reactive” result for scale] 5.1.3 Slowly reacting aggregate combinations The one class of aggregates where some uncertainties showed up were those where the damaging reactions were known to be slow. This was found in the following cases: D2F+NRC - one of the AAR-3 tests and the AAR-4 (alt) did not show the slow reactivity of this combination. However, the AAR-4 reactor tests did correctly identify its slow reactivity. IT2C+F - the AAR-3 tests again did not identify the long term reactivity of this aggregate. The AAR-4 tests did identify its reactivity although in one case only after the normal final test date. N4C+F - again, the AAR-4 tests more clearly identified the long term reactivity of this combination. Three of the four AAR-3 results identified this combination as reactive, but only by the smallest of margins.


33

0

0,05

0,1

0,15

0,2

0,25

0,3

D2 F+Nr Sea DrFlint

IT2 C+F Qrtz + GnGrav

N2 C+Nr Sand st N4 C+F Cata cl Gr N5 C+F Rhy& QrtzGrav

N6 C+F Sed Grav

Slowly-reactive aggregate combinations: % expansion at the end of the test period – all concrete methods [Order in chart is AAR-3 (light blue), AAR-4 (medium blue), AAR-4alt (dark blue), German (red), Norwegian (green)]

6. Comments on the methods 6.1 Comparison of test methods As stated above, in most cases all of the tests gave results that agreed. The main exceptions were the differences between the AAR-3 and AAR-4 results when very slowly reacting aggregates were tested. In these cases the AAR-4 method more clearly identified the reactivity. There is some limited evidence that the AAR-4 reactor method is better in this respect than the alternative wrapping method. Where the Norwegian and German methods were compared with the RILEM methods the agreement is good. The actual expansion values recorded were very variable for some aggregates, but quite consistent for others. The variability was greatest for the most expansive aggregates and least for the least expansive. The expansions revealed in the AAR-3 method seemed particularly variable. In general, amongst the RILEM methods, the results from the AAR-4 methods and in particular the AAR-4 reactor method, were the most consistent and most clearly identified the reactivity of also the slowly reactive aggregate combinations. There is also evidence that the experience of the laboratories with a particular method has a significant effect on the variability of results. Where laboratories were carrying out procedures with which they were very familiar, for example the Norwegian method carried out by NORCEM and SINTEF, the expansion values were very close. 6.2 Problems with the test methods. The most common problems experienced were in ensuring adequate humidity around the specimens. One laboratory failed to get any expansion at all, and this was traced to problems with the humidity.


34

The fresh concrete results in Annex 4 show that, where reported, the properties of the concretes were fairly consistent from one laboratory to another. Some concerns were expressed by some laboratories about the procedures for working with pre-wetted aggregates, allowing for the different densities of the aggregates, the resulting mix design and the mixing procedure. It was also felt that the proportion of fine aggregate in the mix (30%) was not typical of real concretes and that too low a proportion of fines could mean that some reactive fine aggregates would not be identified. There is, however, no evidence from these results that apart from humidity control, such issues have lead to variations in the results. 6.3 “Anomalous” results Although the vast majority of the results were in agreement, some laboratories recorded results that were out of line with field experience and/or with the majority of other laboratories. In some cases there was a clear explanation for these; e.g. difficulties with humidity control at TITAN labs. (AAR-3 results for B1C+F), in others there was no clear reason. Examples: LCPC AAR-4 results for IT1C+F: Little expansion where other labs. recorded considerable expansions. BRE AAR-3 results for D1C+F: Large expansion where other labs. recorded small expansions. DTI AAR-3 results for four aggregates: Strange expansion curves deviating from the other labs. In the case of the LCPC results, humidity loss could be a cause. The BRE results are unexplained, but as described in 5.1, this aggregate has a marked pessimum behaviour which may have contributed to the difference. 6.4 Recommendations for changes to RILEM methods Humidity: On the basis of the experience with these tests it is recommended that there is greater emphasis on monitoring the weight change of the specimens to identify any weight loss and hence low humidity. Aggregates: It is recommended that the proportion of fines should be increased to 40% and the total aggregate grading should conform as far as possible to an ideal aggregate grading curve which will ensure that the concrete matrix is optimised. Concrete mix design: Examples for calculating the concrete mix design, using the revised aggregate proportions and ideal aggregate grading curve should be given. Mixing procedure: A mixing procedure should be given. 6.5 Overall recommendation on quality of testing. More information on the precision of these tests will come from WP4 of the PARTNER project. However, it is clear from this work that experienced laboratories can obtain much more consistent and reliable results with these tests than laboratories new to the methods. This points to the value of continuing inter-laboratory “round robin” tests as part of a quality control procedure.


35

7 Conclusion Five concrete prism tests have been evaluated against 20 combinations of aggregates:

- RILEM AAR-3 concrete prism method (storage at 38 degrees C) - RILEM AAR-4 concrete prism method (storage at 60 degrees C) - RILEM AAAR-4 alternative method (storage at 60 degrees C) - German concrete test method (storage at 40 degrees C) - Norwegian concrete prism method (storage at 38 degrees C)

In the overwhelming majority of cases all the concrete methods correctly identified those aggregate combinations that had been shown by field experience to be involved in cases of damaging alkali reactivity. Similarly all of the methods correctly identified those aggregate combinations that where established from field experience as being non-reactive. However, some uncertainties showed up where the damaging reactions were known to be particularly slow. In general, amongst the RILEM methods, the results from the AAR-4 methods and in particular the AAR-4 reactor method most clearly identified the reactivity of also the slowly reactive aggregate combinations. Where the Norwegian and German methods were compared with the RILEM methods the agreement is good. The most common problems experienced were in ensuring adequate humidity around the specimens. Other concerns were the procedures for working with dry aggregates, the proportion of the fine aggregates in the mix, allowing for the different densities of the aggregates, the resulting mix design and the mixing procedure. Recommendations for improving the methods in the light of these tests have been made.

8. References 1. RILEM TC 106-AAR: Alkali-aggregate reaction - Recommendations - B -TC106-3 (now AAR-3) - Detection of potential alkali-reactivity of aggregates - Method for aggregate combinations using concrete prisms, Materials & Structures, 33, No 229, 290-293, 2000. 2. RILEM Recommended Test Method TC-106-4 (now AAR-4): Detection of potential alkali-reactivity of aggregates: Accelerated (60oC) concrete prism test, RILEM TC191-ARP Committee Draft to be published in Materials and Structures, 2006. 3. Deutscher Ausschuss für Stahlbeton, DAfStb (Ed.): Vorbeugende Maßnahmen gegen schädigende Alkalireaktion im Beton : Alkali-Richtlinie. Beuth, Berlin, 2001(DAfStb-Richtlinie). 4. Norwegian Concrete Association, NB (2005): Alkali-aggregate reactions in concrete. Test methods and requirements to test laboratories. NB Publication No. 32:2005 (in Norwegian).

PARTNER Copyright © PARTNER-project-GRD1-CT-2001-40103

All rights reserved.

Annex 1

Test methods and notes on testing * Note: RILEM AAR-3 Concrete Prism Method * Note: RILEM AAR-4 Accelerated Concrete Prism Method * Note: RILEM AAR-4 Accelerated Concrete Prism Alternative Method * Note: German Concrete Test Method * Note: Norwegian Concrete Prism Method

WP3.3-BRE-SINTEF-031028-AAR3NOTE-FINAL VERSION.DOC 28 Oct. 2003 Page 1 (7) Final version



Note: RILEM AAR-3 Concrete Prism Method. This note contains some of the key-elements, which have to be observed when performing this test. For further details consult the test-method. Outline The concrete prism method is a longer term (12months) test method used to assess the reactivity of a particular aggregate or aggregate combination within a real concrete mix, although in an accelerated manner. Cement Reference cement supplied by the project (Norcem). Total alkali content of cement is 1.26% by mass of sodium equivalent, and the density is 3.12 kg/m3. Cement content shall be kept close to 440kg/m3. No additions of sodium hydroxide are required in this test program. Aggregate The aggregates used shall consist of one of the following: 1. The fine and coarse test aggregates combined (C+F) 2. The fine test aggregate combined with a non-reactive coarse (F+NR) 3. The coarse test aggregate combined with a non-reactive fine (C+NR)

The aggregates shall always be combined where ever possible on a dry basis in the mass proportions 30% fine aggregate (4-0mm), 30% (10-4mm), and 40% (20-10mm). If this is not the case the fact should be reported in the results sheet at the end of testing, and the actual total grading curve shall be enclosed. The coarse to fine fraction ratio will be nominally 70:30. Combinations of other size fractions should equate as closely to these masses as possible. However, where such gradings are not available or the workability characteristics do not allow this combination to work, grading combinations in accordance with local usage shall be employed. If a fine aggregate of a particular material is not available, then crushed 10-4mm material suitably sieved and recombined should be employed. If concerns still remain regarding the aggregate combinations to be employed contact the task leader of the concrete methods. The aggregates should be tested according to the following methods (which will be performed by the laboratory responsible for the supply of the actual aggregate and distribute it to the other laboratories): Water absorption and density: EN 1097-6: Tests for mechanical and physical properties of aggregates – Part 6. “Determination of particle density and water absorption” (September 2000). Grading: EN 933-1: Tests for geometrical properties of aggregates – Part 1: “Determination of particle size distribution – Sieving method” (August 1997). Final aggregate preparation In the RILEM method document, after the sample has been processed the aggregate is to be washed and dried at 100-110°C for 16 hours and allowed to cool. For this testing it is proposed not to undertake this stage as this will potentially effect the size gradings and thus water absorption characteristics of the various fine aggregates used.




For the production of each concrete mix shall be used: - fine aggregate with a total moisture content (= absorbed + surface water) of 6 ± 1% (by weight of

oven dry aggregate), obtained during a moisture conditioning period of no less than 16 hours - coarse aggregate with the total moisture content as it is (no artificial drying or wetting necessary) The total moisture content of each aggregate used in the mix, shall be documented. Concrete mix design Prior to producing the concrete mix the calculated aggregate water absorption and relative densities should be used in the initial mix design development. The methods of obtaining the water absorption and density figures are explained in the associated document (WP3-SINTEF-030818- Note-final aggregate preparation-Rev1). The aggregate density is employed to calculate an assumed wet density. The most important criteria is to keep the cement content equivalent to 440kg/m3 thus ensuring sufficient alkalis (440kg/m3 x 1.26% Na2Oeq =5.54kg/m3 Na2Oeq) in the mix. The slump figures obtained should be in the range 20 to 120mm. To obtain this a free (effective) water to cement ratio 0.5 is required. Examples of two extreme mix designs are shown below in Table 1. It is anticipated that all the aggregates tested in this program will lie somewhere between these two figures. Table 1: Concrete Mix design extreme range examples Thames Valley sand and gravel Bally Barnes Greywacke Dry mass for

1m cube Mass Percentage (%) [dry aggregate only]*

Dry mass for 1m cube

Mass Percentage (%) [dry aggregate only]*

RILEM reference Cement

440kg/m3 19% 440kg/m3 18%

20-10mm agg 654kg/m3 28.5% [40%] 706kg/m3 29% [40%] 10-5 mm agg 490.5kg/m3 21.4% [30%] 530kg/m3 21.8% [30%] 5-0 mm agg 490.5kg/m3 21.4% [30%] 530kg/m3 21.8% [30%] Free (effective) Water

220kg/m3 9.6% 220kg/m3 9.1%

Slump ~120mm 20mm TOTAL mass (kg)

2295kg/m3 Aggregate to cement ratio = 3.7

2426kg/m3

Aggregate to cement ratio = 4.0

* These are dry mass for the aggregates. To these mixes suitable additional water must be added at the calculated rate to take account of the water absorption of the aggregates to saturated surface dry status (SSD) as calculated by the method shown above. Note: Aggregate water absorption values: Thames Valley mean for entire aggregate = 2.13% which equated to an extra water addition of 34kg Bally Barnes mean for entire aggregate = 0.24% which equated to an extra water addition of 4kg The method(s) to be employed to calculate the water absorption of the aggregate shall be given in a separate document. Note: The total cement content and thus the total alkali content of the concrete mix should not change. On the PARTNER web page an Excel work sheet for concrete mix design is available, taking the water absorption and the water content of the aggregates into account. If the workability is less than the aimed one, some laboratories shall add extra water and some shall add a superplasticizer (but not one containing air-entraining agent) to obtain the aimed workability. Rambøll will supply the superplasticizer. What method your laboratory eventually shall use is given in the note for mix design placed on the PARTNER web page. Mix method and curing Concrete shall be mixed and cured in accordance with the RILEM standard and normal laboratory procedure to produce three concrete prisms as also referred to EN480-1. During the production of the




concrete prisms, an air content determination, fresh concrete density, a slump test and / or flow test should be undertaken (all according to EN 12350) wherever possible and reported on the Excel result sheet available at the PARTNER web page. Concrete prisms of dimensions ranging between 250+50mm and 75+5 mm are acceptable for the RILEM method. The relative availability of prism moulds indicates that 280 x 70mm2 would be a preferred size for this testing if the partner is proposing to purchase new moulds for this testing. However, as long as the partner present mould conform to the range above they may be used in this test program. Make sure these dimensions are recorded on your record sheets, and that suitable reference stud locators are present and reference studs are available. If the testing request you have received indicates that you should not wrap your concrete prisms in twill weave cotton cloth as recommended in the RILEM procedure, please confirm that this has not been done on your measurement results sheet. Where the particular laboratory is undertaking a series of tests; AAR-3, AAR-4, AAR-4 alternative and field exposure trials, it is recommended that where possible the same large concrete batch is used for all the test specimens.

Initial curing After mixing the wet concrete within the moulds is left to hydrate at 20 ± 2 °C in Relative Humidity ≥ 90 % under moist cover for 24h ± 0.5 h before demoulding.

Subsequent curing and measuring procedure 1. After demoulding (1 day old), the prisms are measured to the nearest 1 mm using a steel rule

(l (*)) and weighed (W0) to the nearest ± 1 g. THERE IS NO IMMERSION STEP.

2. Wrap the prism in saturated cotton cloth and polythene bag by the method described in the Rilem TC 106-3 document and then stored for 24h ± 0.5h at 20 ± 2°C

3. At the end of this 24h ± 0.5h period ( 2 days old), take the initial measurement of length (C0+ / comparator ) and weight (W0+)

4. Store 5 days ± 2h at 20 ± 2°C. 5. Measure C7 and W7 (7 days old = 7 days after mixing) 6. Store at 38 ± 2°C 7. Measure Ct and Wt after 2, 4, 13, 26 and 52 weeks 8. Report the results in the form attached at page 5

(*) l is used for calculating the expansion percentage General comment on measuring/storage procedure

• All the measurements have to be performed at 20 ± 2°C: that requires a 24h ± 2h cooling period for the 38°C storage

• Demould and measure the prisms in set of three for avoiding drying which could affect expansion

• For performing the measurement, remove the prism from its polythene bag, leaving the saturated cloth and polythene wrapping undisturbed. Clean the reference studs before taking the length reading.

• Check the comparator using the invar rod before performing the readings • After each measurement, replace prism in its polythene bag and pour 5 ml of

deionised water over the upper end of the prism before sealing the bag • Check the water level in the storage container: it must be at least 20 mm depth.


• When moistened, the cotton cloth is about 260 mm width. It must be 10 mm wider than the prism. For prism > 250 mm, wrap with cotton cloth in 2 parts with a central recovery of 50 mm. (see figure below)

The addresses of two possible sources of the towelling, cut or as whole rolls, are listed below. Remember white material is specified.

• Joli Triste (UK) Professional Sourcing & Marketing: Cite: cotton roller towelling, white, uncut Contact: Mr Grant J M St Clair Armstrong 48 Ansgar Road Saffron Walden Essex, CB11 3EJ, United Kingdom Tel/Fax: +44 1799 503947 Mobile: +44 7973 253700 E-mail: [email protected]

• Ragbags

Cite: cotton roller towelling, white, uncut Contact Mr Terry or David Smith St Denys Nurseries Dappers Lane Angmering West Sussex, BN16 4EN, United Kingdom Tel: 01903 774230

COMMENT: SP in Sweden has bought 120 kg of the actual type of cotton cloth. You may order small quantities from them. If difficulties still remain in obtaining this material, the small volume required for this test procedure can be obtained from BRE Limited. Estimation of materials required to carry out one AAR-3 test on one aggregate. 3 concrete prisms of maximum size 300 x 80mm2 + 10% = 0.0065m3 =6.5litres Slump test (5.5litres), Wet density (10litres) can be used in the prisms, but the air content analyses material (6.5 litres) cannot be reused. This equates approximately to 22 litres of wet concrete required per mix +10%. 440 kg/m3 Cement = 11kg 706kg/m3 20-10mm aggregate = 17.6kg 530kg/m3 10-4mm aggregate = 13.2kg



530kg/m3 4-0mm aggregate = 13.2kg




RILEM AAR-3 Concrete Prism Method: Results sheet Laboratory performing test: Type of aggregate(s): Coarse: Fine: ________________________ Water absorption figure for aggregate(s): Coarse: Fine: ___________ Size of mixing equipment used and capacity: _____________________________________________ Size of batch mix: _________________________________________________________________ Size of prisms: _____________________________________________________________________ Wrapping: ________________________________________________________________________ Comments and discussions regarding procedure:

Composition of Concrete Constituent name or reference

and petrographical assessment of reactivity, if known.

Mix design

Mass for batch

NORCEM reference cement Cement kg kg/m3

Water (added) kg kg/m3

Fraction

Water content

Aggregate1: mm % kg1) kg/m3




Aggregate 5: mm % kg1) kg/m3

Superplasticizer (added) kg kg/m3

TOTAL theoretical wet mix mass

kg/m3

Measured wet mix mass kg/m3

Slump value mm Flow value mm Air content %

1) Included moisture




Measurement of prisms (RILEM AAR-3 method) Participant Name: Participant Country: Date and time of casting:

Stored in moulds for 24 hours and >90% RH, after which the prisms are stored wrapped in or unwrapped in saturated cloth and polythene tubing under damp sacking for a further 24 hours before the initial readings 1.* Test Specimen: Invar 1 2 3 Mean Any Specimen Defects:

Initial Mass, g: - after demoulding, W0 - after wrapping, W0+ Initial Length, mm: - steel rule (l) - after wrapping, C0+

Stored at 38°C within containers ~100% RH. Extracted to room temperature 24 hours prior to each testing time. Test Mass Readings: Mass Increase, X, % - after 1 week (W7) - after 2 weeks (W14) - after 4 weeks (W28) - after 13 weeks (W91) - after 26 weeks (W182) - after 52 weeks (W365) - after - after Test Length Readings: Length Increase, E, % - after 1 week (C7) - after 2 weeks (C14) - after 4 weeks(C28) - after 13 weeks(C91) - after 26 weeks (C182) - after 52 weeks (C365) - after - after Any Specimen Features:

* - Selected laboratories will be asked to undertake the tests with and without the cloth coverings. Calculated expansion verses time data and associated graphs should also be plotted in the final presented documentation. The results should be reported to BRE on the Excel sheet available at the PARTNER web page.


Plan of proposed testing to the RILEM AAR-3 method. X -indicate aggregate combination to be tested by that laboratory.

BRE SP NORCEM CESI LABEIN DTI AIDICO IBRI VOZ TITAN No. of Sample no. Details Comb. GB S N I ES DK ES IS A EL tests

B1 Silicified limestone C+NR X X X 3 B1 Silicified limestone C+F X+p X X 3 D1 Gravel with opaline flint C+F X X+p 2D2 Sea gravel semi-dense flint F+NR X X+p X 3D3 Non reactive siliceous sand F 0 F1 Gravel with flint C+NR X X X 3 F2 Non reactive limestone C+F X X X 3 F3 Siliceous gravel C+F X X X 3 G1 Gravel with sil lst and chert C+NR X+p X 2It1 Gravel with lst, chert and flint C+F X X X 3 It2 Gravel with quartzite and gneiss C+F X X X 3 N1 Cataclasite C+NR X X X+p 3 N2 Sandstone C+NR X+p 1 N3 Non reactive granitic sand C+F X 1 N4 Gravel with sst and catacl. rocks C+F X+p X X 3 N5 Gravel with rhyolite and quartzite C+F X+p X X 3 N6 Gravel with sedimentary rocks C+F X 1 S1 Gravel with porphyritic rhyolite C+F X+p X 2

UK1 Greywacke C+F X+p X 2UK2 Gravel with quartizite and chert C+F X X+p X 3 P1 Silicified limestone C+NR X X X 3

8 4 10 4 4 4 4 4 4 4 50

RILEM AAR-3 method = samples for field testing will be produced from the actual concrete mixX+p = one extra prism will be produced (to be transported to Ramboll in DK)

Total no. of tests:

Laboratory / country

Type of aggregate



WP3.3-BRE-SINTEF-031028-AAR4NOTE-final version.doc 28 Oct 2003 Page 1 (6) Final version



Note: RILEM AAR-4 Accelerated Concrete Prism Method. This note contains some of the key-elements, which has to be observed when performing this test. For further details consult the test-method. Outline The accelerated concrete prism method is a shorter-term (20 weeks) test method than RILEM ARP-3 concrete prism method. It is used to assess a particular aggregate or aggregate combinations reactivity within a real concrete mix, although in an accelerated manner. Cement Reference cement supplied by the project (Norcem). Total alkali content of cement is 1.26% by mass of sodium equivalent, and the density is 3,12 kg/m3. Cement content shall be kept close to 440kg/m3. No additions of sodium hydroxide are required in this test program. Aggregate The aggregates used shall consist of one of the following: 1. The fine and coarse test aggregates combined (C+F) 2. The fine test aggregate combined with a non-reactive coarse (F+NR) 3. The coarse test aggregate combined with a non-reactive fine (C+NR)

The aggregates shall always be combined where ever possible on a dry basis in the mass proportions 30% fine aggregate (4-0mm), 30% 10-4mm, and 40% 20-10mm. If this is not the case the fact should be reported in the results sheet at the end of the testing and the actual grading curve shall be enclosed. The coarse to fine fraction ratio will be nominally 70:30. Combinations of other size fractions should equate as closely to these masses as possible. However, where such gradings are not available or the workability characteristics do not allow this combination to work grading combinations in accordance with local usage shall be employed. If a fine aggregate of a particular material is not available, then crushed 10-4mm material suitable sieved and recombined should be employed. If concerns still remain regarding the aggregate combinations to be employed contact the task leader of the concrete methods. The aggregates should be tested according to the following methods (which will be performed by the laboratory responsible for the supply of the actual aggregate and distribute it to the other laboratories): Water absorption and density: EN 1097-6: Tests for mechanical and physical properties of aggregates – Part 6. “Determination of particle density and water absorption” (September 2000). Grading: EN 933-1: Tests for geometrical properties of aggregates – Part 1: “Determination of particle size distribution – Sieving method” (August 1997). Final aggregate preparation. In the RILEM method document, after the sample has been processed the aggregate is to be washed and dried at 100-110°C for 16 hours and allowed to cool. For this testing it is proposed not to undertake this stage as this will potentially effect the size gradings and thus water absorption characteristics of the various fine aggregates used. For the production of each concrete mix shall be used: - fine aggregate with a total moisture content (= absorbed + surface water) of 6 ± 1% (by weight of

oven dry aggregate), obtained during a moisture conditioning period of no less than 16 hours - coarse aggregate with the total moisture content as it is (no artificial drying or wetting necessary)




The total moisture content of each aggregate used in the mix, shall be documented. Concrete mix design Prior to producing the concrete mix the calculated aggregate water absorption and relative densities should be used in the initial mix design development. The methods of obtaining the water absorption and density figures are explained in the associated document (WP3-SINTEF-030818-Note-Final aggregate preparation). The aggregate density being employed to calculate an assumed wet density. The most important criteria is to keep the cement content equivalent to 440kg/m3 thus providing the test sufficient alkalis (440kg/m3 x 1.26% Na2Oeq =5.54kg/m3 Na2Oeq) in the mix. The slump figures obtained should be in the range 20 to 120mm. To obtain this it is felt a free (effective) water to cement ratio 0.5 is required. Examples of two extreme mix designs are shown below in Table 1. It is anticipated that all the aggregates tested in this program will lie somewhere between these two figures. Table 1: Concrete Mix design extreme range examples Thames Valley sand and gravel Bally Barnes Greywacke Dry mass for





440kg/m3 19% 440kg/m3 18%


220kg/m3 9.6% 220kg/m3 9.1%

TOTAL mass (kg)


2426kg/m3


* These are dry mass for the aggregates. To these mixes suitable additional water must be added at the calculated rate to take account of the water absorption of the aggregates to saturated surface dry status (SSD) as calculated by the method shown above. Note: Aggregate water absorption values: Thames Valley mean for entire aggregate = 2.13% which equated to an extra water addition of 34kg Bally Barnes mean for entire aggregate = 0.24% which equated to an extra water addition of 4kg The method(s) to be employed to calculate the water absorption of the aggregate shall be given in a separate document. Note: The total cement content and thus the total alkali content of the concrete mix should not change. On the PARTNER web page an Excel work sheet for concrete mix design is available, taking the water absorption and the water content of the aggregates into account. If the workability is less than the aimed one, some laboratories shall add extra water and some shall add a superplasticizer (but not one containing air-entraining agent) to obtain the aimed workability. Rambøll will supply the superplasticizer. What method your laboratory eventually shall use is given in the note for mix design placed on the PARTNER web page. Mix method and curing Concrete shall be mixed and cured in accordance with the RILEM standard and normal laboratory procedure to produce three concrete prisms as also referred to EN 480-1. During the production of the concrete prisms wherever possible, an air content determination, fresh concrete density, a slump test and / or flow test should be undertaken (all according to EN 12350) and reported on the Excel result sheet available at the PARTNER web page.




Concrete prisms of dimensions ranging between 250+50mm and 75+5 mm are acceptable for the RILEM method. The relative availability of prism moulds indicates that 280 x 70mm2 would be a preferred size for this testing if the partner is proposing to purchase new moulds for this testing. However, as long as the partners present moulds conform to the range stipulated above they may be used in this test program. Make sure these dimensions are recorded on your record sheets, and that suitable reference stud locators are present and reference studs are available. Where the particular laboratory is undertaking a series of tests AAR-3, AAR-4, AAR-4 alternative and field exposure trials it is recommended where possible the same large concrete batch is used for all the test specimens.


Subsequent curing, storage and measurement procedure

1. Mark the top/bottom of the prism before demoulding. 2. Directly after demoulding (1 day old), measure initial length (l (*)) to the nearest 1 mm using

a steel rule, C0 with comparator and weight (W0) to the nearest ± 1g at 20 ± 2 °C. 3. Immerse the prisms in water at 20 ± 2°C for 30 ± 5 min. 4. Remove from water and wipe excess water. 5. Perform a second measurement of weight (W0 slight increase possible ⇒ W0+ ) and length

(C0 slight change possible ⇒ C0+) 6. Install prisms in sealable container with 35 ± 5 mm water in the bottom (the prism is located

15 mm above the water)- Max capacity of 1 container = 6 prisms 7. Store the containers in a reactor / 60 °C, 100 % RH / water level from the bottom 200 mm –

Max capacity of the reactor = 12 containers 8. Measure, at 20 ± 2°C, Ct and Wt after 4, 8, 10, 12, 16 and 20 weeks 9. Report the results in the form attached at the next page

(*) l is used for calculating the expansion percentage

General comment on storage/measuring procedure • All the measurements have to be performed at 20 ± 2°C: that requires a 24h ± 2h

cooling period, preferentially in moist room. Keep the container cover in place with adhesive tape. DO NOT OPEN the container.

• Check the comparator using the invar rod before performing the readings • Check the water level in the storage containers and reactor.

Estimation of materials required to carry out one AAR-4 test on one aggregate. 3 concrete prisms of maximum size 300 x 80mm2 + 10% = 0.0065m3 =6.5litres Slump test (5.5litres), Wet density (10litres) can be used in the prisms, but the air content analyses material (6.5 litres) cannot. This equates approximately to 22 litres of wet concrete required per mix. 440 kg/m3 Cement = 11kg 706kg/m3 20-10mm aggregate = 17.6kg 530kg/m3 10-4mm aggregate = 13.2kg





RILEM AAR-4 Concrete Prism Method: Results sheet Laboratory performing test: ___ Type of aggregate(s): Coarse: _ ____ Fine: ________________________ Water absorption figure for aggregate(s): ________________________________________________ Size of mixing equipment used and capacity :____________________________________________ Size of batch mix: ___________________________________________________________________ Size of prisms: ______________________________________________ Comments and discussions regarding procedure:



Mix design

Mass for batch



Fraction

Water content








kg/m3



1) Included moisture




Measurement of prisms (RILEM AAR-4 method)

Participant Name: Participant Country: Mix Designation: Date and time of casting

Initial measurements of length and weight (plus comparator) are taken immediately after demoulding also within the 20+2oC environment. After measuring immerse the prisms in water at 20+2oC for 30 min +5 min, remove from the water and wipe away excess water with a damp cloth. Make a second measurement of weight (W0+) and the comparator (C0+). Test Specimen: Invar 1 2 3 Mean Any Specimen Defects:

Initial Mass, g: - after demoulding, Wo - after immersion, Wo + Initial Length, mm: - steel rule - comparator, Co - after immersion, Co+ Place the measured prisms in the containers. Then place all the containers in the 60+2oC reactor checking all the water levels are correct. Test Mass Readings: Mass Increase, X, % - after 4 weeks (W28) - after 8 weeks (W56) - after 10 weeks (W70) - after 12 weeks (W84) - after 16 weeks (W112) - after 20 weeks (W140) Test Length Readings: Length Increase, E, % - after 4 weeks (C28) - after 8 weeks (C56) - after 10 weeks (C70) - after 12 weeks (C84) - after 16 weeks (C112) - after 20 weeks (C140) Any Specimen Features:

Calculated expansion verses time data and associated graphs should also be plotted in the final presented documentation. The results should be reported to BRE on the Excel sheet available at the PARTNER web page.


Plan of proposed testing to the RILEM AAR-4 method. X -indicate aggregate combination to be tested by that laboratory.

LCPC VDZ CRIC CESI IMBiGS Holcim No. ofSample no. Details Comb. F D B I PL B tests

B1 Silicified limestone C+NR X X X 3B1 Silicified limestone C+F X X X 3D1 Gravel with opaline flint C+F X X X 3D2 Sea gravel semi-dense flint F+NR X X X 3D3 Non reactive siliceous sand F 0F1 Gravel with flint C+NR X X X X 4F2 Non reactive limestone C+F X X X 3F3 Siliceous gravel C+F X X X X 4G1 Gravel with sil lst and chert C+NR X X X X 4It1 Gravel with lst, chert and flint C+F X X X 3It2 Gravel with quartzite and gneiss C+F X X+p X X 3N1 Cataclasite C+NR X X X 3N2 Sandstone C+NR 0N3 Non reactive granitic sand C+F X X 2N4 Gravel with sst and catacl. rocks C+F X X X 3N5 Gravel with rhyolite and quartzite C+F X X X 3N6 Gravel with sedimentary rocks C+F 0S1 Gravel with porphyritic rhyolite C+F X X X X 4

UK1 Greywacke C+F X X X 3UK2 Gravel with quartizite and chert C+F X X X X 4P1 Silicified limestone C+NR X X X X 4

11 12 15 7 4 11 49

RILEM AAR-4 Normal method = samples for field testing will be produced from the actual concrete mixX+p

Type of agggregate

Total no. of tests:


= one extra prism will be produced (to be transported to Ramboll in DK)



WP3.3-BRE-SINTEF-031028-AAR4ALT-NOTE-final version.doc 28 Oct 2003 Page 1 (7) Final version



Note: RILEM AAR-4 Accelerated Concrete Prism Alternative Method. This note contains some of the key-elements, which have to be observed when performing this test. For further details consult the test-method. Outline The accelerated concrete prism alternative method is a shorter-term (20 weeks) test method than RILEM ARP-3 concrete prism method. It is used to assess the reactivity of a particular aggregate or aggregate combination within a real concrete mix, although in an accelerated manner. Cement Reference cement supplied by the project (Norcem). Total alkali content of cement is 1.26% by mass of sodium equivalent, and the density is 3,12 kg/m3. Cement content shall be kept close to 440kg/m3. No additions of sodium hydroxide are required in this test program. Aggregate The aggregates used shall consist of one of the following: 1. The fine and coarse test aggregates combined (C+F) 2. The fine test aggregate combined with a non-reactive coarse (F+NR) 3. The coarse test aggregate combined with a non-reactive fine (C+NR)

The aggregates shall always be combined where ever possible on a dry basis in the mass proportions 30% fine aggregate (4-0mm), 30% (10-4mm), and 40% (20-10mm). If this is not the case the fact should be reported in the results sheet at the end of testing and the actual total grading curve shall be enclosed. The coarse to fine fraction ratio will be nominally 70:30. Combinations of other size fractions should equate as closely to these masses as possible. However, where such gradings are not available or the workability characteristics do not allow this combination to work, grading combinations in accordance with local usage shall be employed. If a fine aggregate of a particular material is not available, then crushed 10-4mm material suitable sieved and recombined should be employed. If concerns still remain regarding the aggregate combinations to be employed contact the task leader of the concrete methods. The aggregates should be tested according to the following methods (which will be performed by the laboratory responsible for the supply of the actual aggregate and the distribution of it to the other laboratories): Water absorption and density: EN 1097-6: Tests for mechanical and physical properties of aggregates – Part 6. “Determination of particle density and water absorption” (September 2000). Grading: EN 933-1: Tests for geometrical properties of aggregates – Part 1: “Determination of particle size distribution – Sieving method” (August 1997). Final aggregate preparation. In the RILEM method document, after the sample has been processed the aggregate is to be washed and dried at 100-110°C for 16 hours and allowed to cool. For this testing it is proposed not to undertake this stage as this will potentially effect the size gradings and thus water absorption characteristics of the various fine aggregates used




For the production of each concrete mix shall be used: - fine aggregate with a total moisture content (= absorbed + surface water) of 6 ± 1% (by weight of

oven dry aggregate), obtained during a moisture conditioning period of no less than 16 hours - coarse aggregate with the total moisture content as it is (no artificial drying or wetting necessary). The total moisture content of each aggregate used in the mix, shall be documented. Concrete mix design Prior to producing the concrete mix the calculated aggregate water absorption and relative densities should be used in the initial mix design development. The methods of obtaining the water absorption and density figures are explained in the associated document (WP3-SINTEF-0308018-Note- Final aggregate preparation-Rev1). The aggregate density is employed to calculate an assumed wet density. The most important criteria is to keep the cement content equivalent to 440kg/m3 thus ensuring sufficient alkalis (440kg/m3 x 1.26% Na2Oeq =5.54kg/m3 Na2Oeq) in the mix. The slump figures obtained should be in the range 20 to 120mm. To obtain this a free (effective) water to cement ratio 0.5 is required. Examples of two extreme mix designs are shown below in Table 1. It is anticipated that all the aggregates tested in this program will lie somewhere between these two figures. Table 1: Concrete Mix design extreme range examples Thames Valley sand and gravel Bally Barnes Greywacke Dry mass for





440kg/m3 19% 440kg/m3 18%


220kg/m3 9.6% 220kg/m3 9.1%

Slump ~120mm 20mm TOTAL mass (kg)


2426kg/m3


* These are dry mass for the aggregates. To these mixes suitable additional water must be added at the calculated rate to take account of the water absorption of the aggregates to saturated surface dry status (SSD as calculated by the method shown above.. Note: Aggregate water absorption values: Thames Valley mean for entire aggregate = 2.13% which equated to an extra water addition of 34kg Bally Barnes mean for entire aggregate = 0.24% which equated to an extra water addition of 4kg The method(s) to be employed to calculate the water absorption of the aggregate shall be given in a separate document. Note: The total cement content and thus the total alkali content of the concrete mix should not change. On the PARTNER web page an Excel work sheet for concrete mix design is available, taking the water absorption and the water content of the aggregates into account. If the workability is less than the aimed one, some laboratories shall add extra water and some shall add a superplasticizer (but not one containing air-entraining agent) to obtain the aimed workability. Rambøll will supply the superplasticizer. What method your laboratory eventually shall use is given in the note for mix design placed on the PARTNER web page. Mix method and curing Concrete shall be mixed and cured in accordance with the RILEM standard and normal laboratory procedure to produce three concrete prisms as also referred to EN480-1. During the production of the concrete prisms, an air content determination, fresh concrete density, a slump test and / or flow test




should be undertaken (all according to EN 12350) wherever possible and reported on the Excel result sheet available at the PARTNER web page. Concrete prisms of dimensions ranging between 250+50mm and 75+5 mm are acceptable for the RILEM method. The relative availability of prism moulds indicates that 280 x 70mm2 would be a preferred size for this testing if the partner is proposing to purchase new moulds for this testing. However, as long as the partner present mould conform to the range above they may be used in this test program. Make sure these dimensions are recorded on your record sheets, and that suitable reference stud locators are present and reference studs are available. If the testing request you have received indicates that you should not wrap your prisms in twill weave cotton cloth as recommended in the RILEM procedure please confirm that this has not been done on your measurement results sheet. Where the particular laboratory is undertaking a series of tests; AAR-3, AAR-4, AAR-4 alternative and field exposure trials. It is recommended that where possible the same large concrete batch is used for all the test specimens. Unlike the Rilem AAR-4 alternative procedure (see A.9.1), the measurement of l, the length of the prism, is performed BEFORE wrapping.


Subsequent curing and measuring procedure 1. Immediately after demoulding (1 day old ), the prisms are measured to the nearest 1 mm using

a steel rule (l (*)) and weighed (W0) to the nearest ± 1g at 20 ± 2 °C . THERE IS NO IMMERSION STEP.

2. Wrap the prism as described in the Rilem procedure ARP-4 (see A9) 3. Take the initial measurements of length and weight (C0+ / comparator and W0+) 4. Store at 60 ± 2°C 5. Measure, at 20 ± 2°C, Ct and Wt after 4, 8, 10, 12, 16 and 20 weeks . 6. Report the results in the form attached at the next page

(*) l is used for calculating the expansion percentage General comments on measuring/storage procedure

• All the measurements have to be performed at 20 ± 2°C: that requires a 24h ± 2h cooling period because of the 60°C storage

• Demould and measure the prisms in set of three for avoiding drying which could affect expansion

• For performing the measuremen, remove the prism from its polythene bag, leaving the saturated cloth and polythene wrapping undisturbed. Clean the reference studs before taking the length reading.

• Check the comparator using the invar rod before performing the reading • After each measurement, replace prism in its polythene bag and pour 5 ml of

deionised water aver the upper end of the prism before sealing the bag • Check the water level in the storage container: it must be about 20 mm depth. • When moistened, the cotton cloth is about 260 mm width. It must be 10 mm wider

than the prism. For prism > 250 mm, wrap with cotton cloth in 2 parts with a central recovery of 50 mm. (see figure below).




Note that in a significant change from the Rilem AAR-3, the measurement of l (the length of the prism), is performed BEFORE wrapping. The addresses of two possible sources of the towelling cut or as whole rolls are listed below. Remember white material is specified.

• Joli Triste (UK) Professional Sourcing & Marketing: Cite: cotton roller towelling, white, uncut Contact: Mr Grant J M St Clair Armstrong 48 Ansgar Road Saffron Walden Essex, CB11 3EJ, United Kingdom Tel/Fax: +44 1799 503947 Mobile: +44 7973 253700 E-mail: [email protected]

• Ragbags

Cite: cotton roller towelling, white, uncut Contact Mr Terry or David Smith St Denys Nurseries Dappers Lane Angmering West Sussex, BN16 4EN, United Kingdom Tel: 01903 774230

COMMENT: SP in Sweden has bought 120 kg of the actual type of cotton cloth. You may order small quantities from them. If difficulties still remain in obtaining this material, the small volume required for this test procedure can be obtained from BRE Limited. Estimation of materials required to carry out one ARP-3, 4 or 4alt test on one aggregate. 3 concrete prisms of maximum size 300 x 80mm2 + 10% = 0.0065m3 =6.5litres Slump test (5.5litres), Wet density (10litres) can be used in the prisms, but the air content analyses material (6.5 litres) cannot. This equates approximately to 22 litres of wet concrete required per mix. 440 kg/m3 Cement = 11kg 706kg/m3 20-10mm aggregate = 17.6kg 530kg/m3 10-4mm aggregate = 13.2kg





RILEM AAR-4 Alternative Concrete Prism Method: Results sheet Laboratory performing test: Type of aggregate(s): Coarse: _ ____ Fine: ________________________ Water absorption figure for aggregate(s): ________________________________________________ Size of batch mix : ___________________________________ ___ ______ Size of mixing equipment used and capacity: _____________________________________________ Size of prisms: ______________________________________________ Wrapping of prisms: ____ ___________ Comments and discussions regarding procedure:



Mix design

Mass for batch



Fraction

Water content








kg/m3



1) Moisture included




Measurement of prisms (RILEM AAR-4 alternative method)

Participant Name: Participant Country: Mix Designation: Date and time of casting Initial measurements of length (l) and weight (W0), plus comparator (C0) are taken immediately after demoulding and also after wrapping of the prisms in saturated cloth coverings. All this being undertaken within the 20+2oC environment. Test Specimen: Invar 1 2 3 Mean Any Specimen Defects:

Initial Mass, g: - after demoulding, Wo - after wrapping, Wo + Initial Length, mm: - steel rule (l) - comparator, Co - after wrapping, Co + After measuring place the wrapped concrete prisms in their polythene sleeves as described in the method document with an further addition of deionised water before placing them in their individual containers with adequate water at the bottom. Seal the lid and place the containers in a 60+ 2oC test environment. Test Mass Readings: Mass Increase, X, % - after 4 weeks (W28) - after 8 weeks (W56) - after 10 weeks (W70) - after 12 weeks (W84) - after 16 weeks (W112) - after 20 weeks (W140) Test Length Readings: Length Increase, E, % - after 4 weeks (C28) - after 8 weeks (C56) - after 10 weeks (C70) - after 12 weeks (C84) - after 16 weeks (C112) - after 20 weeks (C140) Any Specimen Features:

*- Selected laboratories will be asked to undertake the test with and without the cloth coverings. Calculated expansion verses time data and associated graphs should also be plotted in the final presented documentation. The results should be reported to BRE on the Excel sheet available at the PARTNER web page.


Plan of proposed testing to the RILEM AAR-4 alternative method. X -indicate aggregate combination to be tested by that laboratory.

BRE ISSEP CRIC No. ofSample no. Details Comb. GB B B tests

B1 Silicified limestone C+NR X X 2B1 Silicified limestone C+F X X 2D1 Gravel with opaline flint C+F X X 2D2 Sea gravel semi-dense flint F+NR X X 2D3 Non reactive siliceous sand F 0F1 Gravel with flint C+NR X X 2F2 Non reactive limestone C+F X 1F3 Siliceous gravel C+F X 1G1 Gravel with sil lst and chert C+NR X X 2It1 Gravel with lst, chert and flint C+F X 1It2 Gravel with quartzite and gneiss C+F X 1N1 Cataclasite C+NR X 1N2 Sandstone C+NR 0N3 Non reactive granitic sand C+F 0N4 Gravel with sst and catacl. rocks C+F 0N5 Gravel with rhyolite and quartzite C+F 0N6 Gravel with sedimentary rocks C+F 0S1 Gravel with porphyritic rhyolite C+F 0

UK1 Greywacke C+F X X 2UK2 Gravel with quartizite and chert C+F X X 2P1 Silicified limestone C+NR X 1

8 7 7 28

RILEM AAR-4 Alternative method = samples for field testing will be produced from the actual concrete mix

Type of agggregate

Total no. of tests:




WP3.3-VDZ-031030-NOTE GERMANCONC-REV6.DOC 30 Oct 2003 Page 1 (4) Rev. 6



Note: German Concrete Test Method This note contains some of the key-elements, which has to be observed when performing this test. For further details consult the test-method (see www.partner.eu.com/index.htm). Outline The German concrete test method is a test method used to assess a particular aggregate or aggregate combinations reactivity within a real concrete mix. The hardened concrete specimens are stored in a 40 °C fog chamber. The test method is intended to be suitable for dense aggregates (minimum saturated surface dry density equal to 2.60 Mg/m³). Cement. Reference cement supplied by the project (Norcem). Total alkali content of cement is 1.26% by mass of sodium equivalent, the density is 3.12 g/cm³. Cement content shall be kept as 400 kg/m3 or equivalent for the size of mix being produced. The concrete test with the low effective alkali cement has not to be performed in the PARTNER project. Grading of aggregate. The aggregate used shall be composed as described in the table below.

Portion of aggregate in % by mass (on an air-dried basis) Grain size

in mm Natural non reactive sand (supplied by the project) Aggregate to be tested

0-4 (0-5) 30 --- 4-8 (5-10) --- 40

8-16 (10-20) --- 30 Combinations of other size fractions should equate as closely to these quantities as possible. However, where such gradings are not available or the workability characteristics do not allow this combination to work grading combinations in accordance with local usage shall be employed. If concerns still remain regarding the aggregate combinations to be employed contact the task leader of the concrete methods. The aggregates can be used on an air dried basis (ocularly no moisture on the aggregate surface), because the influence of water absorption is not significant for aggregates with densities greater than 2.60 Mg/dm³. Artificial drying or wetting is consequently not necessary. The density of the aggregates on a saturated surface dry basis shall be used for the calculation of the concrete composition. Concrete mix design The most important criteria is to keep the cement content equivalent to 400 kg/m3 thus providing the test sufficient alkalis (400kg/m3 x 1.26% Na2Oeq =5.04kg/m3 Na2Oeq). An water to cement ratio of 0.45 is required (400 kg/m³ x 0.45 = 180 kg/m³). Additional water to compensate water absorption of the aggregate has not to be considered. Concerning the workability there are no demands but a flow test or a compaction test according to EN 12350 is recommended. The figures should be reported on the result sheet. Note: The total cement content and thus the total alkali content of the concrete mix should not change.




Mix method, curing and initial measurement Mixing and casting After adding all constituents the mixer shall run for two minutes. Cast the concrete into the moulds and compact it in the moulds or on a vibrating table. The vibrating time is dependent on the workability of the concrete. The laboratory technician should be of such experience to guarantee good compaction. For control purposes the air content and the density of the fresh concrete should be determined according to EN 12350. The figures should be reported on the result sheet. Curing and storing Once the concrete specimen are produced they have to be stored in the moulds, covered with moist fabric, at 20 °C for 24±1 hours. After demoulding measure the length and the sectional dimensions of the beams with a steel rule. Weigh them to nearest full gramme afterwards. The beams have to be prepared for measuring length changes, if there is no possibility to fix the reference studs into the moulds before casting. Then the specimens have to be stored in a fog chamber at 40 °C. In doing so the beams should be stored in horizontal direction and be allowed to expand unrestrained. Good experience has been made with storing the beams on a stainless steel grating. Concerning the cubes it should be possible to check cracking apparently on every side face of the cube, if monthly moving of the cube (specific weight is around 60 kg) out of the fog chamber shall be avoided. But it is also allowed to remove the cube from the fog chamber shortly. Initial measuring and standard measuring procedure After demoulding the initial measurements of length and weight of the beams have to be taken in a room with 20 °C. If the reference studs have been placed at the front sides of the beams already, please mark the top in order to make sure further measurements in the same position. If the reference studs shall be glued to two opposite side faces of the beams after demoulding the initial measurements of length and weight have to be taken after hardening of the glue. Thereby the length is the mean value of two measurements taken at the two opposite side faces of the beams. In both cases the length should be measured in a horizontal direction. At the end of the procedure the beams and the 30-cm-cube shall be brought into the 40 °C fog chamber. For the standard measuring procedure take the three beams out of the fog chamber and measure the length between the reference studs immediately. The room temperature should be not lower than 20 °C. Depending on the reference studs the length is defined as prescribed for the initial measuring procedure. The beams have to be weight to the nearest full gramme and finally put back into the fog chamber. The storing place of the beams inside the fog chamber should not be changed during the test. Avoiding evaporation and unnecessary cooling the whole procedure shall take not more than 5 minutes. The cube has to be checked visually after measuring the length change of the beams. If the cube has to be removed from store, the cube should be not longer than 5 minutes outside the fog chamber. If cracking is visible the crack width has to be determined with a crack width gauge. Write down all figures at the result sheet.




Estimation of materials required to carry out one German concrete test on one aggregate. 3 concrete prisms 100 x 100 x 500 mm³ = 0.015 m³ 1 concrete cube 300 x 300 x 300 mm³ = 0.027 m³ => 0.042 m³ density and air content of fresh concrete => 0.008 m³ + 20 % for safety => 0.060 m³ This equates approximately to 400 kg/m³ Cement = 24.00 (25.00 kg) 744 kg/m³ 4-8 (5-10) mm aggregate = 44.62 (46.00 kg) 558 kg/m³ 8-16 (10-20) mm aggregate = 33.48 (35.00 kg) 558 kg/m³ 0-4 (0-5) mm aggregate (NR) = 33.48 (35.00 kg) or equivalent. Annotation: The concrete necessary for the determination of consistency shall be reused for the preparation of the test specimens.

WP3.3-VDZ-031030-NOTE GERMANCONC-REV6.DOC 30 Oct 2003 Page 4 (4) Rev. 6 Test plan:



Lab./countryVDZ

Sample no. Details Comb. DB1 Silicified limestone C+NRB1 Silicified limestone C+F XD1 Gravel with opaline flint C+F XD2 Sea gravel semi-dense flint F+NRD3 Non reactive siliceous sand FF1 Gravel with flint C+NR XF2 Non reactive limestone C+F XF3 Siliceous gravel C+FG1 Gravel with sil lst and chert C+NR XIt1 Gravel with lst, chert and flint C+F It2 Gravel with quartzite and gneiss C+FN1 Cataclasite C+NR XN2 Sandstone C+NR N3 Non reactive granitic sand C+FN4 Gravel with sst and catacl. rocks C+F XN5 Gravel with rhyolite and quartzite C+F XN6 Gravel with sedimentary rocks C+FS1 Gravel with porphyritic rhyolite C+F X

UK1 Greywacke C+F XUK2 Gravel with quartizite and chert C+FP1 Silicified limestone C+NR

10

"German method"

Type of agggregate

Total no. of tests:

2007-03-28

German Concrete Test Method: Results sheet

Laboratory performing test:Type of aggregate(s)Coarse (4-16 mm R):Fine(0-4 mm NR):Water absorption figure for aggregate(s):Size of mixing equipment used and capacity: Size of prisms:

NORCEM reference cement kg kg/m³kg kg/m³

Fraction Water contentAggregate 1: mm % kg1) kg/m³Aggregate 2: mm % kg1) kg/m³Aggregate 3: mm % kg1) kg/m³Aggregate 4: mm % kg1) kg/m³Aggregate 5: mm % kg1) kg/m³

Flow test cmCompaction test (optional) -Air content %Density (fresh concrete) kg/m³

1) dried on air (for the calculation of the concrete composition the ssd density shall be used)

Basic Data

Properties of fresh concrete

The flow/compaction test should be performed directly after the mixing procedure. Air content and density of fresh concrete shall be measured 10 minutes after adding the water to the cement.

CementWater (added)

TOTAL

Constituent name or reference and petrographical assessment of reactivity, if known.

Comments and discussions regarding procedure :

Mix design

Composition of Concrete

Mass for batch

WP3.3-VDZ-031030-German Concrete Method Results sheet.xls / Basic data

2007-03-28

Measurement of prisms and cube (German concrete method)

Mean

#DIV/0!#DIV/0!

Storage time Reading Increase Reading Increase Reading Increase Mean increasedays g % g % g % % Number width [mm]

- after 0 days storage, E0 (initial value 0 #N/A #N/A #N/A #N/A #N/A- after 1 day storage, W1 1 #N/A #N/A #N/A #N/A #N/A- after 7 days storage, W7 7 #N/A #N/A #N/A #N/A #N/A- after 28 days storage, W28 28 #N/A #N/A #N/A #N/A #N/A- after 60 days storage, W60 60 #N/A #N/A #N/A #N/A #N/A- after 91 days storage, W91 91 #N/A #N/A #N/A #N/A #N/A- after 121 days storage, W121 121 #N/A #N/A #N/A #N/A #N/A- after 151 days storage, W151 151 #N/A #N/A #N/A #N/A #N/A- after 182 days storage, W182 182 #N/A #N/A #N/A #N/A #N/A- after 212 days storage, W212 212 #N/A #N/A #N/A #N/A #N/A- after 242 days storage, W242 242 #N/A #N/A #N/A #N/A #N/A- after 273 days storage, W273 273 #N/A #N/A #N/A #N/A #N/A

Storage time Reading Reading Increase Reading Increase Reading Increase Mean increasedays mm mm % mm % mm % %

- after 0 days storage, E0 (initial value 0 #N/A #N/A #N/A #N/A- after 1 day storage, E1 1 #N/A #N/A #N/A #N/A- after 7 days storage, E7 7 #N/A #N/A #N/A #N/A- after 28 days storage, E28 28 #N/A #N/A #N/A #N/A- after 60 days storage, E60 60 #N/A #N/A #N/A #N/A- after 91 days storage, E91 91 #N/A #N/A #N/A #N/A- after 121 days storage, E121 121 #N/A #N/A #N/A #N/A- after 151 days storage, E151 151 #N/A #N/A #N/A #N/A- after 182 days storage, E182 182 #N/A #N/A #N/A #N/A- after 212 days storage, E212 212 #N/A #N/A #N/A #N/A- after 242 days storage, E242 242 #N/A #N/A #N/A #N/A- after 273 days storage, E273 273 #N/A #N/A #N/A #N/A

300 mm-cube2 3Invar

Any specimen defects:

100 x 100 x 500 mm³ beams

Cracks (cube)

Specimen dimensions, mm x mm x mm:Initial mass W0 after demoulding, g:Initial length E0 after demoulding, mm:

Test Mass Readings:

Test Length Readings

Initial measurements of length and weight are taken immediately after demoulding or gluing the reference studs in a room with a temperature of 20 °C.

Place the measured prisms and the cube in the fog chamber afterwards. Remove them again only for measuring purposes. Any measurement shall not take longer than 5 minutes.

Participant name:Participant country:Mix designation:Date and time of casting

Test specimen: 1

WP3.3-VDZ-031030-German Concrete Method Results sheet.xls / test results

WP3.3-SINTEF-031028-NOTE-NORWEGIAN METHOD-FINAL VERSION.DOC Page 1 (6)

Test plan: NORWEGIAN Concrete Prism Method. This note contains some of the key-elements, which have to be observed when performing this test. For further details, consult the test-method. Outline. In this test the Norwegian concrete prism method is used to provide information within 12 months on the potential of the particular aggregates or aggregate combinations to produce deleterious expansions in concrete as a consequence of AAR. The method is accelerated by the use of elevated temperature (38°C), humidity (100 % RH) and content of alkalis (5 kg Na2O eq/m3). Cement. Reference cement supplied by the project (Norcem). Total alkali content of cement is 1.26 % by mass of sodium equivalent, and the density is 3,12 kg/m3. Cement content shall be kept at 400 ± 10 kg/m3. No additions of sodium hydroxide are required in this test program. Aggregate. The aggregates used shall consist of one of the following: 1. The coarse (5 - 20 mm) and fine (0 - 5 mm) test aggregates combined (C+F) 2. The coarse (5 - 20 mm) test aggregate combined with a non-reactive fine (0 - 5 mm) (C+NR)

The aggregates shall be combined on a water saturated and surface dry basis in mass proportions corresponding to the volume proportions 45 % fine aggregate (0 - 5 mm) and approx 18.3 % of each of the three coarse fractions 5 - 10 mm, 10 -14 mm and 14 - 20 mm. To calculate the correct mass proportions, the water absorption and density properties of the aggregates have to be known. For aggregates to be used, these properties should be determined according to EN 1097-6 by the laboratory responsible for the supply of the actual aggregate. The above aggregate combination will normally correspond to a 0 - 4 mm : 4 - 20 mm combination of approx. 40 : 60 (% by volume). Final aggregate preparation. For the production of each concrete mix shall be used: - fine aggregate with a total moisture content (= absorbed + surface water) of 6 ± 1% (by weight of

oven dry aggregate), obtained during a moisture conditioning period of no less than 16 hours - coarse aggregate with the total moisture content as it is (no artificial drying or wetting neccesary) The total moisture content of each aggregate used in the mix, shall be documented. Concrete mix design On the PARTNER web page an Excel work sheet for concrete mix design is available, taking the water absorption and the water content of the aggregates into account. The concrete mix design shall be based on: - the aggregate water absorption and relative densities

PARTNER Copyright PARTNER-project-GRD1-CT-2001-40103


- the density of the cement

WP3.3-SINTEF-031028-NOTE-NORWEGIAN METHOD-FINAL VERSION.DOC Page 2 (6) - an assumed demand for free water corresponding to approx. 180 kg/m3 when gravel is used as fine

aggregate and approx. 200 kg/m3 when crushed sand is used - an assumed air content of approx. 2 % The most important criteria is to keep the cement content equivalent to 400 kg/m3 thus ensuring sufficient alkalis (400kg/m3 x 1.26% Na2Oeq ≈ 5.0 kg/m3 Na2Oeq) in the mix. Mixing and testing of fresh concrete The slump shall be in the range 60 -100 mm. If the slump is less than 60 mm, some laboratories shall add extra water and some shall add a superplasticizer (but not one containing air-entraining agent) to obtain the aimed slump. Rambøll will supply the superplasticizer. What method your laboratory eventually shall use is given in the note for mix design placed on the PARNTER web page. Concrete shall be mixed in an efficient forced action mixer according to the following procedure: - 1 min dry mixing of cement and aggregate - 2 min wet mixing (with water addition during the first min) - 5 min rest (slump shall be determined within this period) - 2 min wet mixing with addition of extra water or a superplasticizer if necessary to obtain aimed

slump Immediately after completed mixing, slump, density and air content shall be determined according to EN 12350. Reference test of compressive strength for QA purposes From each mix, 2 cubes (100 mm) shall be produced, water cured and tested for compressive strength (as described in EN 12390) at age 28 days. If the compressive strength results from different laboratories performing the same test differ significantly, significant differences in concrete composition and / or concrete production are indicated. Production, storage and testing of prisms Three prisms (100x100x450 mm, with measuring studs in the ends) shall constitute one test. The procedures for production, storage and testing of the prisms shall be in accordance with the technical specifications for Class P: "Methods for Testing of Aggregates-Part 2" given by the Norwegian Control Council for Concrete Products. Casting Prism moulds, with the measuring studs in place, shall be filled with concrete in two layers. Each layer shall be consolidated with a straight edged trowel (towards the mould edges) and a tamper in a manner ensuring no large air voids to occur and proper compaction to be achieved. Any excess concrete shall be struck off, and the surface shall be brought to a smooth finish by use of the straight edged trowel. Storage After casting, and till demoulding at age 23 ± 1 hour, the prisms shall be cured within the moulds (under cover of polythene sheet) at 20 ± 2°C, RH ≥ 98 %. After being demoulded, the prisms shall be immersed in water at 20 ± 1°C for 1/2 hour.



Further storage shall be at RH ≥ 98 % in sealed containers (of volume approx. 60 dm3) with a water reservoir (about 20 mm depth, refilling whenever neccesary) at bottom and a wick of efficient adsorbent material placed around the inside wall from the top, so that the bottom extends into the water. One container shall contain the 3 individual prisms from one mix, only. The (unwrapped) prisms shall stand in a vertical position on perforated rack above the water reservoir. The container

WP3.3-SINTEF-031028-NOTE-NORWEGIAN METHOD-FINAL VERSION.DOC Page 3 (6) shall be stored for 12 months in a room or cabinet where the temperature and ventilating conditions are sufficient to secure a permanent temperature within the container at 38 ± 2°C. During this 12 months storage period, the container holding the prisms shall be removed from the 38°C storage and maintained at 20 ± 2°C for 20 ± 4 hours prior to each length (and weight) measurement period. The most important criterion linked to the Norwegian method is to keep the humidity within the containers high enough during the 12 months 38°C storage period. At an earlier stage, the method allowed use of a less efficient absorbent material and much larger containers "filled up" with far more- and horizontally placed prisms. The expansions obtained then, was in general far less than those obtained after the 38°C storage specifications was changed to those valid today. Further, most sandstone aggregate concrete stopped expanding after less than 6 months when tested according to the earlier, unsatisfactory storage specifications. When tested according to the specifications valid today, sandstone aggregate concrete develop major parts of their 12 months expansion during the period between 6 and 12 months. Testing Weighing of prisms shall be performed to the nearest 1 gram, length reading (deviation to the length of an invar rod reference bar) to the nearest 0.001 mm. The reference bar should be measured both prior to, and after the measurement of the 3 prisms within one test. In significant deviation between the two reference bar measurements, the measurement procedure shall be repeated. During the measurement periods, the prisms shall be protected towards loss of moisture. Prior to each length measurement, the studs (prisms and invar rod) must be cleaned. After the 1/2-hour immersion period at age 1 day, each prism shall be measured for initial weight (w0) and initial length reading (l0). Subsequent measurements of weight (wn) and length reading (ln) of each prism shall be performed after (± 2 days) 1, 8, 16, 26 and 52 weeks of the 12 months 38°C storage period. Calculate the weight change (∆wn) of each prism (wn - w0) in grams per prism, and as the mean value of the three individual prisms within each test (∆Wn) in kg/m3 (to the nearest 0.1 kg) at all test ages (n = 1, 8, 16, 26 and 52 weeks, respectively). Calculate the expansion (∆ln) of each prism in mm (ln - l0) and in % (100 % • (ln - l0) / 450) per prism, and as the mean value of the three individual prisms within each test (∆Ln) in % at all test ages (n = 1, 8, 16, 26 and 52 weeks, respectively). The % calculations shall be to the nearest 0.001 %. Reporting Report material compositions and test results in the form attached at the next page. To carry out one Norwegian AAR-3 test on one aggregate / aggregate combination a mix of min 20 dm3 should be carried out. If normal density (≈ 2.7 kg/m3) aggregates are used, this volume equates approximately to a mix containing: 400 kg/m3 cement = 8 kg 810 kg/m3 0 - 5 mm aggregate = 16.2 kg 330kg/m3 5 - 10 mm aggregate = 6.6 kg

330kg/m3 10 - 14 mm aggregate = 6.6 kg

330kg/m3 14 - 20 mm aggregate = 6.6 kg



In addition, neccesary amounts of aggregate for the determination of moisture content will be needed.


NORWEGIAN Concrete Prism Method: Results sheet Laboratory performing test: Type of aggregate(s): Coarse: ____________________________ Fine:_________________________ Water absorption figure for aggregate(s): ________________________________________________ Size of mixing equipment used and capacity: _____________________________________________ Size of prisms: _____________________________________________________________________ Comments and discussions regarding procedure:



Mix design

Mass for batch



Fraction

Water content

Aggregate1: 20-14 mm % kg1) kg/m3


Aggregate3: 10- 5 mm % kg1) kg/m3




kg/m3


Slump value mm

Flow value mm

Air content %

28-days compressive strength MPa

1) Moisture included




Calculated results from measurement of prisms from Mix No:

Participant Name: Participant Country: Date and time of casting:

Stored in moulds for 23 ± 1 hour and ≥ 98 % RH, after which the prisms are immersed in water for a further 1/2 hour before the initial readings. Test Specimen: 1 2 3 Mean Any Specimen Defects:

g (w0) Initial mass kg/m3 (W0) - - -

Stored at 38°C within containers ~≥ 98 % RH. Extracted to room temperature 20 ± 4 hours prior to each testing time.

Test mass readings: Individual mass increase (∆w), g Mean mass increase (∆W), kg/m3

- after 1 week - after 8 weeks - after 16 weeks - after 26 weeks - after 52 weeks

Test length readings: Individual lenght increase (∆l), % Mean lenght increase (∆L), %

- after 1 week - after 8 weeks - after 16 weeks - after 26 weeks - after 52 weeks Any specimen features:



Calculated expansion verses time data and associated graphs should also be plotted in the final presented documentation. The results should be reported to BRE on the Excel sheet available at the PARTNER web page.

WP3.3-SINTEF-031028-NOTE-NORWEGIAN METHOD-FINAL VERSION.DOC Page 6 (6) Page 6 (6)

Plan of proposed testing to the Norwegian Concrete Prism Method. X -indicate aggregate combination to be tested by that laboratory.

Laboratory /

country

Type of agggregate SINTEF NORCEM No. of

Sample no. Details Comb. N N testsB1 Silicified limestone C+NR 0 B1 Silicified limestone C+F X 1D1 Gravel with opaline flint C+F 0 D2 Sea gravel semi-dense flint F+NR 0 D3 Non reactive siliceous sand F 0 F1 Gravel with flint C+NR 0 F2 Non reactive limestone C+F X 1F3 Siliceous gravel C+F 0 G1 Gravel with sil lst and chert C+NR 0 It1 Gravel with lst, chert and flint C+F 0 It2 Gravel with quartzite and gneiss C+F 0 N1 Cataclasite C+NR X 1 N2 Sandstone C+NR X X 2N3 Non reactive granitic sand C+F X 1N4 Gravel with sst and catacl. rocks C+F X X 2

N5 Gravel with rhyolite and quartzite C+F X X 2

N6 Gravel with sedimentary rocks C+F X 1S1 Gravel with porphyritic rhyolite C+F X 1

UK1 Greywacke C+F X 1UK2 Gravel with quartizite and chert C+F 0 P1 Silicified limestone C+NR 0

Total no. of tests: 3 10 13

"Norwegian method"



Annex 2 – page 1

Petrographic composition and field performance of the aggregates tested Aggregate

combination Origin Brief petrographic description Reported alkali reactivity

B1 (C+F) Western Belgium

Crushed silicified, dark-grey argillaceous limestone with fossil debris; reactive mineral is crypto-microcrystalline quartz with sometimes fibrous habit.

Aggregate has caused damage in several concrete structures such as bridges and water structures.

D1 (C+F) Denmark Glaciofluvial gravel containing white to creamy white opaline flint; reactive mineral is opal.

Aggregate has produced severe deterioration in all types of concrete structures (can be very quick under severe conditions, clear pessimum effect).

D2 (C+F) Denmark

Sea dredged, polymictic gravel originally derived from glaciofluvial sediments, main component of interest is partly porous dense chalcedonic flint, in smaller amounts pure porous chalcedonic flint is included; reactive mineral is chalcedony.

Dense porous flint is considered to be non reactive, porous flint to be reactive. Aggregate has produced severe deterioration in all types of concrete structures (normally after 10-15 years). Not as severe as D1.

D3 (F) Denmark No description available yet. No deterioration reported.

F1 (C) France (Seine Valley)

Polymictic river gravel, mainly composed of flint/cherts; reactive mineral is micro-cryptocrystalline quartz.

Is in France considered to be potentially reactive but with clear pessimum effect. No evidence of damage in structures.

F2 (C+F) France Fine grained limestone with some fossils; no reactive minerals. Non-reactive.

F3 (C+F) France (Rhine Valley)

Polymictic river gravel (partly crushed), mainly composed of quartzite, alkali reactive constituents are flint, greywacke and granitoids; reactive minerals are micro-cryptocrystalline quartz and strained, highly metamorphically sutured quartz.

No deterioration reported.

G1 (C)

Germany (Upper Rhine

Valley)

Partly crushed polymictic river gravel, considerable variation in constituent lithologies, aggregates of interest are silicified limestone and chert; reactive minerals are micro- to cryptocrystalline quartz and chalcedony.

Considered to be reactive, concrete pavements with this aggregate have been deteriorated due to ASR. (Damage observed after 10 years under very severe conditions).

G2 (C) Northern Germany

Polymictic gravel from glaciofluvial deposit, alkali reactive due to opaline sandstone (with tridymite/christobalite) and flint (with crypto-crystalline quartz and chalcedony).

Has produced severe deterioration, very quickly, in concrete structures. No damage observed after introduction of national regulations.

It 1 (C+F)

Italy (Marche region in central Italy)

Polymictic river gravel, containing mainly micritic limestone, but also silicified limestone, flint, chert and strained quartz; reactive minerals are micro- to cryptocrystalline quartz and strained, high metamorphically sutured quartz).

Quick reaction (5-10 years) observed in all types of concrete structures.

Annex 2 – page 2

Aggregate


It2 (C+F) Italy

(Piemont region)

Polymictic river gravel, aggregate of interest is fine grained quartzite with strained quartz; reactive mineral is strained highly metamorphically sutured quartz.

Considered to be reactive, slowly (one example is 50 years old water constructions).

N1 (C) Norway (middle)

Crushed cataclasite, homogeneous and fine-grained, feldspar particles lie scattered within a matrix of about 0.02 mm grain size; reactive mineral is crypto- to microcrystalline quartz.

Used as concrete material. Has caused severe damage in local areas (e.g. 11 years old airport pavement).

N2 (C) Norway (south east)

Crushed sandstone, homogeneous and fine grained, with a sediment grain size ranging between 0.05 and 0.5 mm. These variously sized particles are embedded in a fine grained matrix; reactive mineral is crypto- to microcrystalline quartz.

Used as concrete material. Has caused severe damage in local areas (damage observed in bridges and dams after 15 to 20 years).

N3 (C+F) Norway (south

western)

Natural gravel/sand from a glaciofluvial deposit, originally composed of Precambrian crystalline rocks, consists of granites and gneisses; no reactive constituents.

No damage reported.

N4 (C+F) Norway (south east)

Natural gravel/sand from a moraine deposit.; Sandstones, siltstones and cataclastic rocks are reactive rocks; reactive mineral is crypto- to microcrystalline quartz.

Used as concrete material. The coarse fraction has caused moderate damage, if the humidity and the alkali content are high (e.g. 20-25 years old constructions, mainly bridges).

N5 (C+F) Norway (south)

Sand and coarse gravel from a glaciofluvial deposit. Rhyolite and fine grained quartzite are reactive rocks; reactive mineral is microcrystalline-fine grained quartz.

Used as concrete material. The coarse fraction has caused moderate damage, if the humidity and the alkali content is high (e.g. 20-25 years old constructions, mainly bridges).

N6 (C+F) Norway (south)

Sand and coarse gravel from a glaciofluvial deposit. The reactive rocks are mainly argillaceous rocks and sandstones in addition to small amounts of hornfels, rhyolite and mylonite. Reactive mineral is crypto- to microcrystalline quartz.

Used as concrete material. The coarse fraction has caused moderate damage, if the humidity and the alkali content are high (e.g. 20-25 years old constructions, mainly bridges).

P1 (C) Portugal Crushed, poorly silicified limestone; reactive minerals could be micro- or cryptocrystalline quartz.

Similar limestone, probably with higher content of silica, has caused damage in several concrete structures like bridges and dams.

S1 (C+F) Sweden

Polymictic glaciofluvial gravel and sand, primarily composed of meta-rhyolite and granite, aggregates of interest are meta-rhyolite and greywacke; reactive minerals are micro- or cryptocrystalline quartz or chalcedonic quartz.

Similar aggregate used as concrete material has caused moderate damage. The source is variable in composition.

UK1 (C+F) United Kingdom

Crushed greywacke, poorly sorted; reactive minerals are micro- or cryptocrystalline quartz, possibly volcanic glass.

Concrete with this aggregate has demonstrated high damage at moderate to high alkali levels in many real structures (more than 20 years until observed damage).

Annex 2 – page 3

Aggregate


UK2 (C+F) United Kingdom

Polymictic mature river gravel and sand, composed primarily of metaquartzite, ortho-quartzite, quartz (vein) and chert, which is the reactive portion in the aggregate; reactive minerals are micro- or cryptocrystalline or chalcedonic quartz.

Both, fine and coarse constituents have demonstrated high reactivity at moderately high alkali levels in many real structures, mainly bridges (damage after 10 to 15 years).

E1 (F) Spain

Dolostone with prismatic dolomite chrystals cemented with calcite, also opal and clay are apparent in considerable amounts, reactive mineral is opal.

Serious damage (due to ASR?) reported in 30 years old precast element (big water pipe).

C = coarse aggregate (> 4 mm); F = fine aggregate (≤ 4 mm)



Annex 3

Fresh concrete properties The fresh concrete properties of each batch used in the production of the blocks used in the field trials are compiled in table 1. The same concrete was used in the production of the laboratory concrete specimens. Additionally, if used, the quantity of superplasticizer is mentioned. Table 1: Fresh concrete properties

Aggregate combination/concrete Property [Unit] B1

C+F B1

C+NR D2

F+NR F1

C+NR F2

C+F G1

C+NR It2

C+F N1

C+NR N2

C+NR N4

C+F S1

C+F UK1 C+F

P1 C+NR

No. of batches 2 10? 5 2 2 2 4 2 4 2 2 4 4

Slump [cm]

5 5

6 7 7 6

n.d. n.d. 5

n.d. n.d. n.d.

12 12 12 12 n.d.

12 12

10 (s) 13 (s)

5 12 (s)

16 n.d. 12

n.d.

9 10

6 6

n.d. n.d.

13 (s) 16 (s)

13 15

5 4 4 5

n.d.

10 n.d. 10

n.d.

Air content [%]

0.9 0.8

0.4 0.4 0.4 0.4 n.d. n.d. 0.4 n.d. n.d. n.d.

0.3 0.3 0.3 0.3 n.d.

0.4 0.4

1.5 1.5

1.2 1.2

1.0 n.d. 1.0 n.d.

0.7 0.8

2.0 1.9 n.d. n.d.

1.5 1.5

1.0 0.9

0.5 0.4 0.5 0.5

n.d.?

0.9 n.d. 0.9 n.d.

Density [t/m³]

2.35

2.34

2.39 2.39 2.38 2.38 n.d. n.d. 2.39 n.d. n.d. n.d.

2.41 2.41 2.41 2.41 n.d.

n.d.

2.30 2.31

2.34 2.34

2.35 2.33

2.39 n.d. 2.38 n.d.

2.39 2.38

n.d. n.d. n.d. n.d.

2.34 2.35

2.34 2.33

2.42 2.42 2.40 2.40 n.d.

2.36 n.d. 2.36 n.d.

Super- plasticizer

[kg/m³]

0 0

0 0 0 0

n.d. n.d 0

n.d. n.d. n.d.

0 0 0 0

n.d.

n.d.

0 0

0 0

0 0

0 0 0 0

0 0

0.62 0.62 0.62 0.62

0 0

0 0

0 0 0 0

n.d.

0 0 0 0

(s) = sheared off n.d. = not determined/reported to field test laboratory

· aggregates according to several concrete prism ... aggregate and geological types found across Europe and calibrated the results of ... Aggregate Origin Comb.

Documents