Advanced search
1 file | 10.76 MB Add to list

Development of standard testing methods to evaluate the self-healing efficiency of concrete

Tim Van Mullem (UGent)
(2021)
Author
Promoter
(UGent) and (UGent)
Organization
Abstract
Concrete is one of the most widely used construction materials, because it is economical, it can be used in a variety of different forms, and it is able to carry high compressive loads. However, it is prone to cracking due to its limited tensile capacity. These cracks allow for a fast ingress of water and harmful substances (e.g. chlorides and carbon dioxide) into the concrete, where they degrade the passive layer protecting the reinforcement from corroding. The steel reinforcement is placed inside the concrete to carry the tensile stresses, which the concrete is unable to carry. As the steel reinforcement corrodes, it expands, thereby pushing off the concrete cover which is supposed to protect it from the environment. Once this has occurred, degradation continues at an increased rate, until eventually this might even endanger the safety of the structure and its users. To prevent this from occurring, regular maintenance and repair works have to be carried out. These are costly, as they require skilled labour and specialised products, in addition to posing several safety risks due to accessibility problems. It was estimated in 􀍲􀍰􀍰􀍶 that the cost for repair, rehabilitation and protection of concrete structures in the USA would amount to $18 to $21 billion a year. For Europe, it was estimated that half of the entire construction budget is spent on maintaining and repairing existing structures. In many countries, like e.g. Belgium, these budgets and further investments are under constant pressure. Aside from the direct costs, indirect costs which result from traffic delays and loss of productivity cannot be overlooked, as they can be 10 times higher than the direct costs. Self-healing concrete is a promising way to deal with these necessary repairs. By enhancing the mix design and incorporating healing agents, this type of concrete is able to repair its own cracks with no external interventions. This requires an increased investment at the start but results in a lower lifetime cost, due to a lower maintenance cost and an extension of the service life. The latter reduces the environmental impact of structures, and delays the emission of greenhouse gasses related to the construction of new structures. As such, it can reduce the emission of e.g. carbon dioxide (CO2, for which it is estimated that the construction sector is responsible for approximately 23 % of the global emission. Many different self-healing methods have already been developed. Yet, it remains difficult to compare different studies due to a lack of normative testing methods. This also hinders further development and commercialisation, because the construction sector uses a strictly regulated concrete production. The goal of this thesis is to pave the way towards the standardization of testing methods to examine the efficiency of different self-healing cementitious materials. To test the self-healing efficiency of mortar and concrete, the material first needs to be cracked, prior to evaluating parameters such as the water permeability or resistance to chloride ingress or carbonation. Methods described in literature are often only applicable to mortar. Specifically to evaluate and develop the test methods for self-healing concrete, this thesis also investigates the performance of a bacterial concrete which has been used in the first large-scale self-healing demonstration project in Belgium, as part of the preparatory works of the Oosterweel infrastructure project in Antwerp. Testing water permeability and capillary water absorption of selfhealing concrete and mortar Different water permeability methods have already been used in literature to study the regain in liquid tightness, i.e. sealing efficiency. In these methods the water is recorded which comes from an upper reservoir and passes through the crack. In this thesis two different permeability methods are studied to assess the sealing efficiency of mortar. In the first method, cylindrical specimens were split and the halves were tied back together with silicone sheets in-between. Water coming from the top reservoir, leaked through the crack into a lower reservoir. It was found that this lower reservoir severely affected the accuracy of the results when specimens with a narrow crack (100 μm) were studied. Therefore, it was decided to omit this lower reservoir when investigating the sealing efficiency of high volume fly ash mortar enhanced with superabsorbent polymers (SAPs). SAPs are highly hydrophilic polymers (in powder form) which absorb water during mixing, which they release back to the cementitious matrix as it sets. SAP particles can again absorb water when the cementitious matrix cracks and the crack comes into contact with water. The SAP swelling can block (part of) the crack and the SAPs release the water during dry periods, which is favorable for ongoing hydration and calcium carbonate precipitation to close the crack. The rather high variation on the crack width of the studied samples made it challenging to accurately determine the sealing efficiency directly from the measured flow rates. In order to determine the sealing efficiency without a dependency on the crack width, the trend lines calculated from the graphs obtained from plotting the flow rates versus the crack width were used, resulting in a sealing efficiency of 81 % for the specimens with SAPs compared to 39 % for the reference specimens. The fact that the sealing efficiency had to be determined from trend lines was a drawback, because the trend lines overestimated the flow rates of specimens with smaller cracks. The second permeability method which was studied in detail was a water flow test. Mortar prisms were manufactured with a cast-in tube, which was connected to an upper reservoir via a small hose. The water, which leaked through the crack, crossing the cast-in tube, leaked directly on a scale. The specimens were supplied with a carbon fibre reinforced polymer (CFRP) strip prior to cracking. The CFRP strip allowed the two halves of the prisms to remain connected when the specimens were cracked. Subsequently, the crack width, which was slightly too big after cracking, was actively controlled by pushing the two halves back together with screw jacks. This active crack width control was executed under a microscope, which significantly reduced the variation on the crack width within a series of specimens (to as low as 2.3 %). This is an important aspect in water permeability testing, since the flow rate is influenced by the crack width to the third power, meaning that a high variation on the crack width of a series will make the flow rates incomparable. For series with a low variation on the crack width (± 2 % - 3 %), the variation on the flow rate was found to be a magnitude higher (10 % - 20 %), indicating the important contribution of internal crack geometry to the flow rate. The latter becomes dominant at low variations on the crack width. Using probability theory it was estimated that, when the coefficient of variation on the crack width is smaller than 4 % for a series with 6 specimens, the variation on the mean flow rate will generally not be larger than 10 %. The combination of the water flow test and the active crack width control technique was also studied in an inter-laboratory study with six participating labs. Both reference specimens and self-healing specimens containing macrocapsules filled with liquid polyurethane (PU) were shipped to the participating labs. When self-healing specimens cracked, the macrocapsules cracked simultaneously, allowing the PU to fill the crack. The crack widths of the individual specimens of the different labs, were nearly all within a narrow predefined range (290-310 μm). As a consequence, the mean crack widths of the series of each lab were statistically equal. This resulted in all labs obtaining comparable flow rates and sealing efficiencies. This confirmed the potential of the investigated water flow test, in combination with an active crack width control technique, as a standardized test method. The inter-laboratory study also examined reinforced concrete specimens which were cracked in a displacement-controlled bending test. This passive crack control resulted in larger variations on the crack width than for the active technique, which was attributed to partial crack closure due to elastic regain of the reinforcement upon load removal. Nevertheless, labs which opened the cracks to a value as high as 485 μm were able to obtain the target crack width of 300 μm with good accuracy. After drying of the cracked specimens, two capillary water absorption tests were performed, once with aluminium tape waterproofing and once with coated waterproofing. This highlighted the importance of the quality of the waterproofing when executing a capillary water absorption test, which is more vital than having exactly the same area of concrete exposed to water. The capillary water absorption test can be easily affected by the operator sensitivity, e.g. the frequency and amount of water addition in the reservoir to compensate for the absorbed water. To limit the operator sensitivity as much as possible, detailed guidelines and an instruction video were developed. Laboratory control tests of self-healing bacterial concrete of Oosterweel demonstration project For the first large-scale application of self-healing concrete in Belgium, a bacterial healing agent, consisting of a Mixed Ureolytic Culture and anaerobic granular bacteria, together with nutrients, was successfully added to a concrete mix in an industrial concrete mixer. This bacterial concrete was used to cast a roof slab for an inspection pit of water drainage pipes, as well as accompanying lab specimens. When a crack in bacterial concrete comes into contact with water, the bacteria leave their dormant stage and use the nutrients to produce healing products to fill up the crack. A visual assessment on part of the laboratory specimens indicated the need for liquid water to start calcium carbonate production by the bacteria, with the best crack closures being obtained in cyclic wet-dry conditions. The orientation of the cracks during healing was found to be important, as cracks at the bottom of specimens healed more consistently than cracks at the top or at the side of specimens. A water permeability test, which was similar to the water flow test for mortar, indicated a nearly perfect regain in liquid tightness (>90%, for 3 out of 5 specimens >98.5%). This test also confirmed the influence of the internal crack geometry on water permeability results; for two specimens with an equal mean crack width, the initial flow rate (before healing) of the first specimen was six times larger than the flow rate of the second specimen. An inspection on the construction site, more than 1 year after casting, showed favourable conditions for healing but no signs of cracking. Most likely, the acting loads had remained below the design cracking load. Testing chloride ingress of self-healing mortar When concrete is subjected to a chloride solution, chloride ions diffuse into the concrete, where they cause pitting corrosion of the steel reinforcement. In the current preliminary study the effect of SAPs on the chloride ingress in cracked and uncracked mortar was studied. To test the chloride ingress, the bulk diffusion test was chosen as it is a cheap and straightforward method to determine the chloride ingress. Its main advantage is that it uses a natural transport process and allows the chlorides to naturally bind to the matrix, as opposed to migration tests. Mortar specimens were made with additional mixing water, to compensate for a loss in workability due to the addition of SAPs absorbing part of this mixing water. This caused the mortar with SAPs to have a higher porosity, and as a result the chloride ingress in uncracked specimens was higher than for specimens made from a reference mortar. Water permeability tests, which were executed on cracked mortar samples, showed that there was only a limited amount of immediate crack sealing due to the swelling of the SAP particles when they came in contact with water. This explained the poor behaviour of the cracked specimens with SAPs which were placed in a chloride solution, prior to crack healing. On the other hand, when specimens were allowed to heal, prior to placing them in contact with the chloride solution, the specimens with SAPs showed a promising behaviour. At early ages the chloride penetration in cracked and healed specimens with SAPs was lower than in companion reference specimens, and at later ages the influence of the crack on the chloride ingress was reduced. Testing carbonation of self-healing mortar The calcium bearing phases (in a first instance the portlandite) of a cementitious matrix react with the carbon dioxide (CO2) in the air, resulting in a pH drop. When the pH drop reaches the reinforcement, it will start to corrode. The natural CO2 concentration is rather low, making carbonation a slow process. To speed up the experiments, carbonation chambers with elevated CO2 concentrations were used and specimens were cast from a high volume fly ash mixture, which contains a lower amount of portlandite compared to regular mortar. Comparing specimens with realistic cracks to specimens with artificial cracks (by pulling out thin metal plates), showed a wider and deeper carbonation front for the specimens with artificial cracks. Additionally, the carbonation in realistic cracks showed a dependency on the crack width, which was absent for artificial cracks. These observations lead to the conclusion that artificial cracks are not representative of realistic cracks, because they lack rough crack walls. The studied artificial cracks also had a wall effect, resulting in a higher amount of pores next to the crack, which influenced the carbonation process. The healed cylindrical specimens with SAPs, which had already been tested in a water permeability test, were also subjected to a carbonation test. This carbonation study showed that at the location of the crack, there was still CO2 ingress as a result of the imperfect crack healing which was measured in the permeability test. Yet, the partial crack closure, due to the deposit of healing products, did slow down the carbonation ingress. Away from the crack, in the uncracked zone, the addition of SAPs also slowed the carbonation ingress, most likely due to a densification of the matrix and a reduction in pore radius. In the current test, specimens had to be dry due to the strictly regulated environment in the carbonation chamber. In realistic conditions, where specimens are subjected to cyclic wetting and drying, the addition of SAPs might result in a wet surface for an extended period of time. CO2 enters in the matrix through the pores, when these are filled with water, the rate of carbonation decreases. Also specimens containing macrocapsules filled with a PU-based healing agent showed very promising results with respect to preventing the CO2 ingress in cracks. The best performance was obtained for macrocapsules filled with PU and sealed with methylmetacrylate. The healing agent flowed out of the capsules when the specimens were cracked, and this crack filling resulted in a carbonation depth which was only half of that of reference specimens without macrocapsules. Assessment of long-term deformations of self-healing concrete Most of the studies on self-healing concrete have been focussing on the restoration of durability parameters. Yet, as these healing agents influence the cementitious matrix, it is also important to know the influence on the long-term behaviour. More specifically long-term deformations are an important aspect, since a wrong estimation of these deformations might result in serviceability issues. In this preliminary test, a high dosage of SAPs (1 m% with respect to the weight of cement, a typical dosage to obtain crack healing) was added to concrete to study the influence on the long-term deformations. The concrete with the SAPs showed a reduced autogenous shrinkage, measured from 28 days after casting, up to more than 18 months. With regard to creep, the total creep deformations were lower for the specimens with SAPs, which were loaded at the same stress-to-strength ratio as the reference specimens. The presence of SAPs might have resulted in a stimulated precipitation of hydration products in the (macro-)pores and a higher internal humidity. This higher humidity in the specimens with SAPs, could reduce the microprestress, which in combination with the lower cracking potential might have resulted in a lower amount of creep.

Downloads

  • (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 10.76 MB

Citation

Please use this url to cite or link to this publication:

MLA
Van Mullem, Tim. Development of Standard Testing Methods to Evaluate the Self-Healing Efficiency of Concrete. Universiteit Gent. Faculteit Ingenieurswetenschappen en Architectuur, 2021.
APA
Van Mullem, T. (2021). Development of standard testing methods to evaluate the self-healing efficiency of concrete. Universiteit Gent. Faculteit Ingenieurswetenschappen en Architectuur.
Chicago author-date
Van Mullem, Tim. 2021. “Development of Standard Testing Methods to Evaluate the Self-Healing Efficiency of Concrete.” Universiteit Gent. Faculteit Ingenieurswetenschappen en Architectuur.
Chicago author-date (all authors)
Van Mullem, Tim. 2021. “Development of Standard Testing Methods to Evaluate the Self-Healing Efficiency of Concrete.” Universiteit Gent. Faculteit Ingenieurswetenschappen en Architectuur.
Vancouver
1.
Van Mullem T. Development of standard testing methods to evaluate the self-healing efficiency of concrete. Universiteit Gent. Faculteit Ingenieurswetenschappen en Architectuur; 2021.
IEEE
[1]
T. Van Mullem, “Development of standard testing methods to evaluate the self-healing efficiency of concrete,” Universiteit Gent. Faculteit Ingenieurswetenschappen en Architectuur, 2021.
@phdthesis{8694912,
  abstract     = {Concrete is one of the most widely used construction materials, because it
is economical, it can be used in a variety of different forms, and it is able to
carry high compressive loads. However, it is prone to cracking due to its
limited tensile capacity. These cracks allow for a fast ingress of water and
harmful substances (e.g. chlorides and carbon dioxide) into the concrete,
where they degrade the passive layer protecting the reinforcement from
corroding. The steel reinforcement is placed inside the concrete to carry the
tensile stresses, which the concrete is unable to carry. As the steel
reinforcement corrodes, it expands, thereby pushing off the concrete cover
which is supposed to protect it from the environment. Once this has
occurred, degradation continues at an increased rate, until eventually this
might even endanger the safety of the structure and its users.
To prevent this from occurring, regular maintenance and repair works have
to be carried out. These are costly, as they require skilled labour and
specialised products, in addition to posing several safety risks due to
accessibility problems. It was estimated in 􀍲􀍰􀍰􀍶 that the cost for repair,
rehabilitation and protection of concrete structures in the USA would
amount to $18 to $21 billion a year. For Europe, it was estimated that half of
the entire construction budget is spent on maintaining and repairing
existing structures. In many countries, like e.g. Belgium, these budgets and
further investments are under constant pressure. Aside from the direct
costs, indirect costs which result from traffic delays and loss of productivity
cannot be overlooked, as they can be 10 times higher than the direct costs.
Self-healing concrete is a promising way to deal with these necessary repairs.
By enhancing the mix design and incorporating healing agents, this type of
concrete is able to repair its own cracks with no external interventions. This
requires an increased investment at the start but results in a lower lifetime
cost, due to a lower maintenance cost and an extension of the service life.
The latter reduces the environmental impact of structures, and delays the
emission of greenhouse gasses related to the construction of new structures.
As such, it can reduce the emission of e.g. carbon dioxide (CO2, for which
it is estimated that the construction sector is responsible for approximately
23 % of the global emission.
Many different self-healing methods have already been developed. Yet, it
remains difficult to compare different studies due to a lack of normative
testing methods. This also hinders further development and
commercialisation, because the construction sector uses a strictly regulated
concrete production.
The goal of this thesis is to pave the way towards the standardization of
testing methods to examine the efficiency of different self-healing
cementitious materials. To test the self-healing efficiency of mortar and
concrete, the material first needs to be cracked, prior to evaluating
parameters such as the water permeability or resistance to chloride ingress
or carbonation. Methods described in literature are often only applicable to
mortar. Specifically to evaluate and develop the test methods for self-healing
concrete, this thesis also investigates the performance of a bacterial concrete
which has been used in the first large-scale self-healing demonstration
project in Belgium, as part of the preparatory works of the Oosterweel
infrastructure project in Antwerp.
Testing water permeability and capillary water absorption of selfhealing
concrete and mortar
Different water permeability methods have already been used in literature
to study the regain in liquid tightness, i.e. sealing efficiency. In these
methods the water is recorded which comes from an upper reservoir and
passes through the crack. In this thesis two different permeability methods
are studied to assess the sealing efficiency of mortar.
In the first method, cylindrical specimens were split and the halves were tied
back together with silicone sheets in-between. Water coming from the top
reservoir, leaked through the crack into a lower reservoir. It was found that
this lower reservoir severely affected the accuracy of the results when
specimens with a narrow crack (100 μm) were studied. Therefore, it was
decided to omit this lower reservoir when investigating the sealing efficiency
of high volume fly ash mortar enhanced with superabsorbent polymers
(SAPs). SAPs are highly hydrophilic polymers (in powder form) which
absorb water during mixing, which they release back to the cementitious
matrix as it sets. SAP particles can again absorb water when the
cementitious matrix cracks and the crack comes into contact with water.
The SAP swelling can block (part of) the crack and the SAPs release the
water during dry periods, which is favorable for ongoing hydration and
calcium carbonate precipitation to close the crack. The rather high variation
on the crack width of the studied samples made it challenging to accurately
determine the sealing efficiency directly from the measured flow rates. In
order to determine the sealing efficiency without a dependency on the crack
width, the trend lines calculated from the graphs obtained from plotting the
flow rates versus the crack width were used, resulting in a sealing efficiency
of 81 % for the specimens with SAPs compared to 39 % for the reference
specimens. The fact that the sealing efficiency had to be determined from
trend lines was a drawback, because the trend lines overestimated the flow
rates of specimens with smaller cracks.
The second permeability method which was studied in detail was a water
flow test. Mortar prisms were manufactured with a cast-in tube, which was
connected to an upper reservoir via a small hose. The water, which leaked
through the crack, crossing the cast-in tube, leaked directly on a scale. The
specimens were supplied with a carbon fibre reinforced polymer (CFRP)
strip prior to cracking. The CFRP strip allowed the two halves of the prisms
to remain connected when the specimens were cracked. Subsequently, the
crack width, which was slightly too big after cracking, was actively
controlled by pushing the two halves back together with screw jacks. This
active crack width control was executed under a microscope, which
significantly reduced the variation on the crack width within a series of
specimens (to as low as 2.3 %). This is an important aspect in water
permeability testing, since the flow rate is influenced by the crack width to
the third power, meaning that a high variation on the crack width of a series
will make the flow rates incomparable. For series with a low variation on the
crack width (±  2 % -  3 %), the variation on the flow rate was found to be a
magnitude higher (10 % - 20 %), indicating the important contribution of
internal crack geometry to the flow rate. The latter becomes dominant at
low variations on the crack width. Using probability theory it was estimated
that, when the coefficient of variation on the crack width is smaller than 4 %
for a series with 6 specimens, the variation on the mean flow rate will
generally not be larger than 10 %.
The combination of the water flow test and the active crack width control
technique was also studied in an inter-laboratory study with six
participating labs. Both reference specimens and self-healing specimens
containing macrocapsules filled with liquid polyurethane (PU) were shipped
to the participating labs. When self-healing specimens cracked, the
macrocapsules cracked simultaneously, allowing the PU to fill the crack. The
crack widths of the individual specimens of the different labs, were nearly
all within a narrow predefined range (290-310 μm). As a consequence, the
mean crack widths of the series of each lab were statistically equal. This
resulted in all labs obtaining comparable flow rates and sealing efficiencies.
This confirmed the potential of the investigated water flow test, in
combination with an active crack width control technique, as a standardized
test method.
The inter-laboratory study also examined reinforced concrete specimens
which were cracked in a displacement-controlled bending test. This passive
crack control resulted in larger variations on the crack width than for the
active technique, which was attributed to partial crack closure due to elastic
regain of the reinforcement upon load removal. Nevertheless, labs which
opened the cracks to a value as high as 485 μm were able to obtain the target
crack width of 300 μm with good accuracy. After drying of the cracked
specimens, two capillary water absorption tests were performed, once with
aluminium tape waterproofing and once with coated waterproofing. This
highlighted the importance of the quality of the waterproofing when
executing a capillary water absorption test, which is more vital than having
exactly the same area of concrete exposed to water. The capillary water
absorption test can be easily affected by the operator sensitivity, e.g. the
frequency and amount of water addition in the reservoir to compensate for
the absorbed water. To limit the operator sensitivity as much as possible,
detailed guidelines and an instruction video were developed.
Laboratory control tests of self-healing bacterial concrete of
Oosterweel demonstration project
For the first large-scale application of self-healing concrete in Belgium, a
bacterial healing agent, consisting of a Mixed Ureolytic Culture and
anaerobic granular bacteria, together with nutrients, was successfully added
to a concrete mix in an industrial concrete mixer. This bacterial concrete
was used to cast a roof slab for an inspection pit of water drainage pipes, as
well as accompanying lab specimens. When a crack in bacterial concrete
comes into contact with water, the bacteria leave their dormant stage and
use the nutrients to produce healing products to fill up the crack. A visual
assessment on part of the laboratory specimens indicated the need for liquid
water to start calcium carbonate production by the bacteria, with the best
crack closures being obtained in cyclic wet-dry conditions. The orientation
of the cracks during healing was found to be important, as cracks at the
bottom of specimens healed more consistently than cracks at the top or at
the side of specimens. A water permeability test, which was similar to the
water flow test for mortar, indicated a nearly perfect regain in liquid
tightness (>90%, for 3 out of 5 specimens >98.5%). This test also confirmed
the influence of the internal crack geometry on water permeability results;
for two specimens with an equal mean crack width, the initial flow rate
(before healing) of the first specimen was six times larger than the flow rate
of the second specimen.
An inspection on the construction site, more than 1 year after casting,
showed favourable conditions for healing but no signs of cracking. Most
likely, the acting loads had remained below the design cracking load.
Testing chloride ingress of self-healing mortar
When concrete is subjected to a chloride solution, chloride ions diffuse into
the concrete, where they cause pitting corrosion of the steel reinforcement.
In the current preliminary study the effect of SAPs on the chloride ingress
in cracked and uncracked mortar was studied. To test the chloride ingress,
the bulk diffusion test was chosen as it is a cheap and straightforward
method to determine the chloride ingress. Its main advantage is that it uses
a natural transport process and allows the chlorides to naturally bind to the
matrix, as opposed to migration tests. Mortar specimens were made with
additional mixing water, to compensate for a loss in workability due to the
addition of SAPs absorbing part of this mixing water. This caused the mortar
with SAPs to have a higher porosity, and as a result the chloride ingress in
uncracked specimens was higher than for specimens made from a reference
mortar. Water permeability tests, which were executed on cracked mortar
samples, showed that there was only a limited amount of immediate crack
sealing due to the swelling of the SAP particles when they came in contact
with water. This explained the poor behaviour of the cracked specimens
with SAPs which were placed in a chloride solution, prior to crack healing.
On the other hand, when specimens were allowed to heal, prior to placing
them in contact with the chloride solution, the specimens with SAPs showed
a promising behaviour. At early ages the chloride penetration in cracked and
healed specimens with SAPs was lower than in companion reference
specimens, and at later ages the influence of the crack on the chloride
ingress was reduced.
Testing carbonation of self-healing mortar
The calcium bearing phases (in a first instance the portlandite) of a
cementitious matrix react with the carbon dioxide (CO2) in the air, resulting
in a pH drop. When the pH drop reaches the reinforcement, it will start to
corrode. The natural CO2 concentration is rather low, making carbonation
a slow process. To speed up the experiments, carbonation chambers with
elevated CO2 concentrations were used and specimens were cast from a high
volume fly ash mixture, which contains a lower amount of portlandite
compared to regular mortar. Comparing specimens with realistic cracks to
specimens with artificial cracks (by pulling out thin metal plates), showed a
wider and deeper carbonation front for the specimens with artificial cracks.
Additionally, the carbonation in realistic cracks showed a dependency on
the crack width, which was absent for artificial cracks. These observations
lead to the conclusion that artificial cracks are not representative of realistic
cracks, because they lack rough crack walls. The studied artificial cracks also
had a wall effect, resulting in a higher amount of pores next to the crack,
which influenced the carbonation process.
The healed cylindrical specimens with SAPs, which had already been tested
in a water permeability test, were also subjected to a carbonation test. This
carbonation study showed that at the location of the crack, there was still
CO2 ingress as a result of the imperfect crack healing which was measured
in the permeability test. Yet, the partial crack closure, due to the deposit of
healing products, did slow down the carbonation ingress. Away from the
crack, in the uncracked zone, the addition of SAPs also slowed the
carbonation ingress, most likely due to a densification of the matrix and a
reduction in pore radius. In the current test, specimens had to be dry due to
the strictly regulated environment in the carbonation chamber. In realistic
conditions, where specimens are subjected to cyclic wetting and drying, the
addition of SAPs might result in a wet surface for an extended period of time.
CO2 enters in the matrix through the pores, when these are filled with water,
the rate of carbonation decreases.
Also specimens containing macrocapsules filled with a PU-based healing
agent showed very promising results with respect to preventing the CO2
ingress in cracks. The best performance was obtained for macrocapsules
filled with PU and sealed with methylmetacrylate. The healing agent flowed
out of the capsules when the specimens were cracked, and this crack filling
resulted in a carbonation depth which was only half of that of reference
specimens without macrocapsules.
Assessment of long-term deformations of self-healing concrete
Most of the studies on self-healing concrete have been focussing on the
restoration of durability parameters. Yet, as these healing agents influence
the cementitious matrix, it is also important to know the influence on the
long-term behaviour. More specifically long-term deformations are an
important aspect, since a wrong estimation of these deformations might
result in serviceability issues.
In this preliminary test, a high dosage of SAPs (1 m% with respect to the
weight of cement, a typical dosage to obtain crack healing) was added to
concrete to study the influence on the long-term deformations. The
concrete with the SAPs showed a reduced autogenous shrinkage, measured
from 28 days after casting, up to more than 18 months. With regard to creep,
the total creep deformations were lower for the specimens with SAPs, which
were loaded at the same stress-to-strength ratio as the reference specimens.
The presence of SAPs might have resulted in a stimulated precipitation of
hydration products in the (macro-)pores and a higher internal humidity.
This higher humidity in the specimens with SAPs, could reduce the
microprestress, which in combination with the lower cracking potential
might have resulted in a lower amount of creep.},
  author       = {Van Mullem, Tim},
  isbn         = {9789463554558},
  language     = {eng},
  pages        = {XXXI, 299},
  publisher    = {Universiteit Gent. Faculteit Ingenieurswetenschappen en Architectuur},
  school       = {Ghent University},
  title        = {Development of standard testing methods to evaluate the self-healing efficiency of concrete},
  year         = {2021},
}