Chapter 7
Accelerated Expansion Test: Switzerland
Andreas Leemann, Christine Merz, and Stéphane Cuchet
7.1 Scope
This test method covers the laboratory determination of the expansion rate of concrete cores (designated as “residual expansion potential” in the following context)
extracted from structures affected by alkali silica reactions.
This test differs considerably from methods where aggregates are extracted from
concrete extracted from structures and reused in accelerated mortar tests with added
alkali. In the concrete structures the decreasing alkali level in the pore solution seems
to be the factor limiting expansion and not the availability of reactive minerals as in
the accelerated mortar tests.
This method does not encompass an assessment of the expansion taken place
before the extraction of the cores.
A. Leemann (B)
Empa, Swiss Federal Laboratories for Materials Testing and Research, Laboratory for Concrete
and Construction Chemistry, Überlandstrasse 129, 8600 DÜbendorf, Switzerland
e-mail: andreas.leemann@empa.ch
C. Merz
Merz Ingenieurberatung, Mőriken, Switzerland
e-mail: merz@merz-ingenieurberatung.ch
S. Cuchet
Stéphane Cuchet, Holcim Suisse, Laboratoire des matériaux, Eclépens, Switzerland
e-mail: stephane.cuchet@lafargeholcim.com
© RILEM 2021
V. E. Saouma (ed.), Diagnosis & Prognosis of AAR Affected Structures,
RILEM State-of-the-Art Reports 31,
https://doi.org/10.1007/978-3-030-44014-5_7
175
176
A. Leemann et al.
7.2 Principles and Methodology
The method described below is based on the method LCPC 44 Fasseu [1]. It has been
used for various structures including the ones investigated in a project in Switzerland
financed by the Federal road authorities [2]. Some of the results of this project are
published in [3].
The residual expansion potential is determined on cores taken from structures.
They are stored at 38 ◦ C and 100% RH for several months up to more than one year.
The measured parameters are mass change, diametral and longitudinal expansions
Fig. 7.1, measurements every 4 weeks at 20 ◦ C). The longitudinal expansion is used
to assess the residual expansion potential.
The principle of the method is described in the following paragraphs. It is not a
manual for conducting the test, but describes its major characteristics.
Advantages:
1.
2.
3.
4.
no alkalis added.
connection between length and mass change.
classification of the residual expansion potential in three categories.
concrete test.
Disdvantages:
1. some alkali leaching, although minimized by protective measures (wrapping)
2. how can the measured expansion be used in models?
The method is a test for concrete extracted from structures. It is not an aggregate
test and differs considerably from methods where aggregates are extracted from
concrete and reused in accelerated mortar tests with added alkali.
In the concrete structures the decreasing alkali level in the pore solution seems to
be the factor limiting expansion.
Fig. 7.1 Core (diameter =
100 mm, length = 200 mm)
used for the expansion test.
The longitudinal expansion
is measured on the length
axis using two stainless steel
pins fixed with glue in small
boreholes. The diametral
expansion is measured at two
positions perpendicular to
the length axis
7 Accelerated Expansion Test: Switzerland
177
7.3 Sampling and Sample Preparation
Immediately after being taken from the structure, the cores are sealed in plastic bags.
The cores are cut in the lab to appropriate size (typical diameter and length: 100 and
200 mm). Usually, three cores are tested per coring site. Small holes are drilled into
the cores fitting the pins. As soon as the drying state of surface permits, the pins are
glued (e.g. Araldite 2014-1) into the the small boreholes of the cores and aligned
with a jig. Stable glue (e.g. organic two-component glue) has to be used that shows
no volume changes at the conditions in the reactor.
7.4 Apparatus
The apparatus used for the test (based on AFNOR P18–454 and employed in Swiss
standard SIA MB 2042) is shown in Figs. 7.2. Three cores are stored in one container.
There is water at the bottom of the containers, but as the cores stand on a grid no
1 Inlet preventing condensed water to drop
on samples
2 Cover
3 Sealing
4 Handle with joint
5 Stainless steel grid keeping samples in
upright position
6 Stainless steel container
7 3 samples (Ø ~ 100mm, L ~200 mm)
8 Stainless steel grid with opening width
of ~ 10 mm
9 Water
38°C ± 2°C
Fig. 7.2 Stainless steel container for Empa test
1
2
3
4
5
6
7
8
Cover
Stainless steel
Thermal isolation
Container
Water
Immersion heater
Stainless steel grid
Thermocouple for temperature control
178
A. Leemann et al.
direct contact with water is possible. The containers themselves are placed into the
reactor with a climate of 38 ◦ C and 100% relative humidity.
7.5 Experimental Procedures
The expansion of the cores can be divided into three different phases, Fig. 7.3:
Phase 1: The initial measurements of dimension and mass are performed before
sample conditioning. Sample conditioning is done at 20 ◦ C in containers with
a thin layer of water at its bottom permitting capillary suction of the cores. The
mass of the cores has to be measured daily. Conditioning ends when constant mass
is reached. This is usually achieved after one week. Immediately after reaching
constant mass, the cores have to be moved into the reactor. The water saturation
is accompanied with a first swelling of the cores (hygroscopic swelling).
Remark: Conditioning is mandatory, because otherwise the expansion related to
moisture uptake cannot be distinguished from expansion related to ASR, after the
samples have been placed into the reactor. This is the only way to ensure identical
conditions as a starting point for cores of different structures making possible
comparisons and classifications for the residual expansion.
Phases 2 & 3: The cores are stored in the reactor at 38 ◦ C and at 100 % RH. After
the cores are wrapped in a plastic film (minimizes mass loss and alkali leaching),
they are moved into the containers. The containers are placed into the reactor. For
the length measurements the containers are removed from the reactor and stored
at 20 ◦ C for 24 h before the cores are measured. Afterwards the containers are put
back into the reactor.
The same apparatus as for the French performance-test AFNOR P18–454 is used.
At the beginning of the storage at 38 ◦ C and 100% RH a strong but quickly levelling
off expansion during 30–60 days is usually observed (phase 2). In the following
phase 3 the expansion rate is constant. However, it can also level off in certain
cases.
Remark: Mass control is very important, but not required in all published methods
[e.g. 1]. An preparation of the length change is difficult without having the data
for the mass change available. The mass change is used as a control parameter in
phases 2 and 3. It allows detecting unintentional changes of RH in the reactor or
containers and assessing the reasons for irregularities in the expansion behavior
of cores.
Phase 4: The cores are removed from the reactor after 160 days. Afterwards, they
are dried at 60–70% RH until they reach the identical mass as before conditioning
started (phase 1). The difference in length change is defined as “irreversible”
expansion and enables to distinguish between hygroscopic swelling due to the
formation of newly formed ASR products. This step requires 2–8 weeks.
So far, the longitudinal expansions during phases 2 and 3 are used to assess the
residual expansion potential of the cores. However, a first assessment is usually
7 Accelerated Expansion Test: Switzerland
179
Fig. 7.3 Different phases of expansion during the residual expansion measurements in longitudinal
and diametral direction of the cores
already possible at the end of the non-linear expansion (phase 2). The determination of the irreversible expansion in phase 4 is used as a control parameter. It gives
a rough indication about the magnitude of hygroscopic swelling and ASR-induced
expansion.
7.6 Preparation
Calculate the difference between the initial length of the specimen and the length at
the end of each period. The mean of the longitudinal length change at the beginning
of phase 2 (non linear expansion rate) and the mean of following linear length change
(phase 3, linear expansion rate) are used to assess the residual expansion potential
Fig. 7.4. The diametral length changes are used as a control parameter. They can
indicate an inhomogeneity of expansion over the length of the cores.
Constant relationship between longitudinal and diametral expansion has been
observed during expansions test by several authors as Smaoui, Bérubé, Fournier, and
Bissonnette [4], Larive [5] Multon, Leclainche, Bourdarot, and Toutlemonde [6] and
ourselves. For instance we use the diametral expansion as control parameter and also
to detect an eventual warping of the cores.
The method has been used to assess the residual expansion potential of numerous
structures in Switzerland, Fig. 7.4. The cores can be classified into three different
levels.
180
A. Leemann et al.
Fig. 7.4 Results of residual expansion measurement of cores extracted from concrete structures in
Switzerland
The differently colored domains are defined empirically, based on the experience
from Swiss concrete structures.
Blue domain:
Green domain:
Red domain:
concrete with a residual expansion potential zero or very low.
concrete with a moderate residual expansion potential.
concrete with a high residual expansion potential.
The domains shown in Fig. 7.4 were defined based on the experience gained from
Swiss concrete structures They take into account the degree of damage of the structure at the time of core extraction, age and exposition of the structure and concrete
composition (alkali content, cement type, aggregates). As these additional characteristics are considered and influence the assessment of the structure, the domains
overlap. The measured values give an indication about range of expansion rates to
be expected in a structure. These values are influenced by the expansion that already
took place and by the exposition of the studied structure or component. As an example, cores taken from a part of the structure with little ASR-induced damage may
show a higher expansion compared to cores taken from a strongly damaged part. This
emphasizes that the degree of ASR can vary widely within a structure. It is advantageous if structures are monitored and the measured expansions can be compared
with the one determined in the laboratory test. This facilitates the assessment of the
structure.
7 Accelerated Expansion Test: Switzerland
181
7.7 Examination at End of Test
After the final length measurements, the condition of the wrapping, the stability of
the pins and the characteristics of each core are examined.
This includes:
1. Presence, location, and pattern of cracking,
2. Appearance of surfaces, surface mottling, and
3. Surficial deposits or exudations, their nature, thickness and distribution.
7.8 Report
The report has to contain the following information:
1. Distribution and severity of damages in the structure, preferably including crackindices.
2. Location of the coring sites.
3. Microscopic investigation including petrography of aggregates, type of cement,
mineral additions, microcracks with crack-indices on the microscopic scale, alterated aggregates, presence of ASR products and information about other processes
like leaching, ettringite formation, carbonation etc.
4. Alkali content of the concrete as mass percent of potassium oxide (K2 O), sodium
oxide (Na2 O), and calculated sodium oxide (Na2 O) equivalent. Refer to used
method for extracting alkalis (water soluble and/or acid soluble alkalis).
5. If possible: source of aggregate/concrete.
6. If possible: Type and source of Portland cement/mineral additions with their
respective alkali contents.
7. Any relevant information concerning the preparation of concrete core.
8. Development of length change of each set of three cores during phases 1–4.
9. Any significant features revealed by examination of the specimens during and
after the test.
7.9 Limitations and Applicability to Analysis
The method characterizes the kinetics of expansion at the conditions in the reactor
(temperature of 38 ◦ C and RH of 100%). As such, it does not provide a value for
residual expansion of the tested concrete.
The expansion measured in this test shows a strong positive correlation to the
crack-index determined in strongly damaged parts of structures with an age in the
range of 40 years [3]. This correlation seems to justify using a parameter determined
in this test for finite element analysis. Based on the experience with the method up to
182
A. Leemann et al.
now, the nonlinear expansion in phase 2 is influenced by the degree of damage of
the concrete at the time of coring. The linear expansion in phase 3 does not show
such an influence and consequently better represents solely the kinetics of expansion.
Therefore, it is recommended to use the approximately constant expansion rate in
phase 3 for finite element analysis. It is likely that the expansion rate in the structure
is lower due to the lower temperatures. However, a correction for the temperature at
the location of the structure seems inappropriate, because the decrease of expansion
by decreasing the temperature from 38 to 20 ◦ C is concrete- and aggregate-specific,
as shown in Merz and Leemann [3]. As such, the use of the expansion rate in phase
3 enables the modelling of a worst case scenario and should allow identifying the
areas of the structure for which expansion is the most critical.
7.10 Precision and Bias
The dominating factor for the uncertainty of the expansion is the variation of the
concrete cores. The magnitude of error of the measurement itself is negligible
(±0.002 mm). Depending on the homogeneity of the concrete at the coring site,
significant variations in the expansion of single cores within a set of three cores are
possible. Therefore, all cores have to be considered, even if there is an outlier.
References
1. Fasseu, P.: Alcali-Reaction Du Beton-Essai D’Expansion Residuelle Sur Beton DurciProjet De Methode D’Essai. Technical report No. 44. Laboratoire Central des
Ponts et Chaussées (LPC) (1997). http://www.ifsttar.fr/fileadmin/user_upload/editions/lcpc/
MethodeDEssai/MethodeDEssai-LCPC-ME44.pdf
2. Merz, C., Leemann, A.: Validierung der AAR-Prüfungen für Neubau und Instandsetzung. Technical report 648. Bundesamt Fuer Strassenbau (ASTRA)/Office Federal Des Routes (OFROU)
(2011)
3. Merz, C., Leemann, A.: Assessment of the residual expansion potential of concrete from structures damaged by AAR. Cem. Concr. Res. 52, 182–189 (2013)
4. Smaoui, N., Bérubé, M., Fournier, B., Bissonnette, B.: Influence of specimen geometry, orientation of casting plane, and mode of concrete consolidation on expansion due to ASR. Cem.,
Concr. Aggreg. 26(2), 1–13 (2004)
5. Larive, C.: Apports Combinés de l’Experimentation et de la Modélisation à la Comprehension del’Alcali-Réaction et de ses Effets Mécaniques. Ph.D. thesis. Paris: Laboratoire Central
des Ponts et Chaussées (1998). https://hal.inria.fr/docs/00/52/06/76/PDF/1997TH_LARIVE_
C_NS20683.pdf
6. Multon, S., Leclainche, G., Bourdarot, E., Toutlemonde, F.: Alkali-Silica reaction in specimens
under multi-axial mechanical stresses. In: Proceedings of CONSEC 4 (Concrete Under Severe
Conditions), pp. 2004–2011. Seoul (2004)