### Autogenous shrinkage of cement-based materials : from the fundamental role of self-desiccation to mitigation strategies based on alternative materials

Yang Lyu UGent (2017)
abstract
With the increasing application of (ultra-) high performance concrete (HPC/UHPC) all over the world, autogenous shrinkage has attracted ample interest as the large magnitude of autogenous shrinkage develops uniformly through the concrete rapidly especially at very early age when the cement paste has poorly developed mechanical properties, leading to high cracking risks. In order to fully benefit from the HPC/UHPC, it is necessary to control its early age cracking potential by mitigating the early age autogenous shrinkage. Attempting to do so, one needs to thoroughly understand the mechanisms of autogenous shrinkage. It is generally agreed that the autogenous shrinkage is closely related to self-desiccation (internal relative humidity changes) in the capillary pores of hardening cement paste. Among the proposed mechanisms explaining the relation between autogenous shrinkage and self-desiccation in a hardening cement paste, the capillary pressure induced by the self-desiccation in the capillary pore system is generally accepted as the main driving force for autogenous shrinkage. One main conflicting issue involved in the field of autogenous shrinkage lies in the precise determination of the onset of autogenous shrinkage, i.e. the time-zero. This is essential for accurate autogenous shrinkage evaluation, as autogenous shrinkage occurring before the time-zero is of little practical significance as it may have little practical consequence in residual tensile stress development. The time-zero represents the suspension-solid transition in a hydrating cement paste. However, it is practically impossible to determine the precise onset of this solidification transition. Thus, many time criteria have been suggested in the literature, causing great difficulty in the comparison study of results available in the existing literature. By looking into the mechanism of autogenous shrinkage, it is clear that the autogenous shrinkage is closely linked to the self-desiccation as mentioned above. Under sealed curing condition, the internal RH is mainly related to self-desiccation. So the precise determination of internal RH changes of sealed curing cement pastes starting from the pastes being still in fluid phase would supply important experimental data to evaluate the evolution of self-desiccation. It leads to a proposal for determining the time-zero by monitoring the internal RH development of a sealed hydrating cement paste. In order to evaluate this time-zero proposal, the fundamental experiments of autogenous shrinkage and internal RH measurements were conducted on Ordinary Portland cement pastes. Microstructure evolution of cement pastes during the early age hydration period from the suspension state to the solid state was also monitored by means of ultrasonic pulse velocity (UPV) method. The results indicate that the occurrence of self-desiccation indicated by the start of internal RH drop links to the self-supporting solid skeleton formation in cement pastes with low w/c ratio (0.25-0.30). This critical time represents the time-zero for autogenous shrinkage. Meanwhile, final setting time determined by the penetration method does not link to the self-supporting solid skeleton formation where selfdesiccation begins to occur; it leads to an earlier estimation of time-zero and overestimation of autogenous shrinkage. Among the mitigation strategies, it is of great importance to counteract the effect of self-desiccation and keep high internal RH by applying extra curing water. However, due to the very dense microstructure of HPC/UHPC, the penetration depth of the external curing water is limited. Instead, internal curing, referring to the use of water filled inclusions that can provide curing water throughout the cross section of the concrete, has been found to be an effective method for mitigating self-desiccation and autogenous shrinkage. Among the most used water inclusions (lightweight aggregates (LWAs) and Superabsorbent polymers (SAPs), etc,), SAPs used as internal curing agent show more promising properties in HPC/UHPC. It gives a motivation to reevaluate the efficiency of applying SAPs in self-desiccation compensation and autogenous shrinkage mitigation based on the newly defined time-zero. In the case of internal curing by using SAPs, choosing final setting time as time-zero would lead to underestimation of the internal curing efficiency of SAPs. If zeroing the autogenous shrinkage at final setting time, autogenous shrinkage is not fully successfully mitigated although the internal RH results suggest efficient self-desiccation mitigation in SAPs mixtures. Based on the newly defined time-zero, SAPs with proper size (for example, SAP-A with mean particle size of 100.0 ± 21.5 μm) and sufficient amount of water show efficient autogenous shrinkage mitigation, corresponding to their successful mitigation of self-desiccation. The behaviors of SAPs as internal curing with different powder size of SAPs and different amount of internal curing water in cement pastes with various w/c ratio have also been investigated. Due to the difficulty in proper dispersion of SAPs in concrete and the reported possible detrimental effect of SAPs on the mechanical property of concrete, it is of great interest to seek alternative materials with similar internal curing function but no or less possible negative influence on the mechanical properties of concrete. One natural clinoptilolite-Ca-heulandite-Ca type zeolite was used to evaluate its potential utilization as an internal curing agent for mitigation of self-desiccation and the subsequent autogenous shrinkage of cement paste. The results indicate that addition of zeolite could eliminate the self-desiccation partially and the autogenous shrinkage as well based on the time-zero defined by the onset of internal RH drop. Attempts have been done for the purpose of improving the internal curing property of zeolite by acid treatment and thermal treatment. However, nitric acid treatment tends to increase the Brunauer-Emmett-Teller (BET) surface area of zeolite by increasing the volume of 2-7 nm pores, which is undesirable. Thermal treatment tends to reduce the porosity rather than to enlarge the pore size, which is also undesirable. In the literature, it is shown that rice husk ash (RHA) presents effective internal curing function to produce UHPC with promoted hydration degree of cement and eliminated autogenous shrinkage. This leads to a way of seeking porous materials with finer particle size distribution in comparison with LWAs or SAPs and high amorphous silica content (corresponding to high pozzolanic properties) as internal curing agent. In this way, the volume of macropores (initially water filled) induced by the incorporation of LWAs or SAPs could be reduced thus the detrimental effect on the strength is reduced or even eliminated. To this end, three different biomass ashes were chosen to investigate their possible application as internal curing agents in low w/b ratio cement pastes. The experimental results reveal that only limited internal curing property can be expected from the industry generated biomass ash (Napier grass ash), because the NGA does not really show internal porous structure due to its highly crystallized structure. The lab-produced biomass ashes (MA and SGA) from combustion of two different energy crops (miscanthus and switchgrass) show high potassium content. After potassium extraction by acid leaching, both MA and SGA presented highly porous structure with BET surface area of 45.28 m2/g and 35.52 m2/g, respectively. Highly effective self-desiccation and autogenous shrinkage mitigation is observed in both MA and/or SGA incorporated cement pastes. The effect of particle size of MA/SGA on the self-desiccation and autogenous shrinkage mitigation efficiency has also been investigated. Take MA as example, with the decreasing mean particle size from 20.43 μm to 6.67 μm, lower self-desiccation mitigation efficiency was observed in the blended cement pastes. The grinding process may break MA particles and make more and more small pores (pores with size in diameter in the range of 2-4 nm and pores with size of around 20 nm) which are originally embedded inside MA particles exposed. Thus, more water is expected to be absorbed in those pores and is expected to be released later at lower RH in hydrating cement paste compared with the case of MA with D50 of 20.43 μm.
author
promoter
UGent and UGent
organization
year
type
dissertation
publication status
published
subject
defense location
Jozef Plateauzaal, Plateaustraat 22, 9000 Gent
defense date
2017-10-20 17:00
ISBN
978-94-6355-051-2
language
English
UGent publication?
yes
classification
D1
I have retained and own the full copyright for this publication
id
8535801
handle
http://hdl.handle.net/1854/LU-8535801
date created
2017-10-30 11:31:31
date last changed
2017-10-31 08:33:09
@phdthesis{8535801,
abstract     = {With the increasing application of (ultra-) high performance concrete (HPC/UHPC) all over the world,
autogenous shrinkage has attracted ample interest as the large magnitude of autogenous shrinkage
develops uniformly through the concrete rapidly especially at very early age when the cement paste
has poorly developed mechanical properties, leading to high cracking risks. In order to fully benefit
from the HPC/UHPC, it is necessary to control its early age cracking potential by mitigating the early
age autogenous shrinkage. Attempting to do so, one needs to thoroughly understand the
mechanisms of autogenous shrinkage. It is generally agreed that the autogenous shrinkage is closely
related to self-desiccation (internal relative humidity changes) in the capillary pores of hardening
cement paste. Among the proposed mechanisms explaining the relation between autogenous
shrinkage and self-desiccation in a hardening cement paste, the capillary pressure induced by the
self-desiccation in the capillary pore system is generally accepted as the main driving force for
autogenous shrinkage.
One main conflicting issue involved in the field of autogenous shrinkage lies in the precise
determination of the onset of autogenous shrinkage, i.e. the time-zero. This is essential for accurate
autogenous shrinkage evaluation, as autogenous shrinkage occurring before the time-zero is of little
practical significance as it may have little practical consequence in residual tensile stress
development. The time-zero represents the suspension-solid transition in a hydrating cement paste.
However, it is practically impossible to determine the precise onset of this solidification transition.
Thus, many time criteria have been suggested in the literature, causing great difficulty in the
comparison study of results available in the existing literature.
By looking into the mechanism of autogenous shrinkage, it is clear that the autogenous shrinkage is
closely linked to the self-desiccation as mentioned above. Under sealed curing condition, the internal
RH is mainly related to self-desiccation. So the precise determination of internal RH changes of
sealed curing cement pastes starting from the pastes being still in fluid phase would supply important
experimental data to evaluate the evolution of self-desiccation. It leads to a proposal for determining
the time-zero by monitoring the internal RH development of a sealed hydrating cement paste. In
order to evaluate this time-zero proposal, the fundamental experiments of autogenous shrinkage
and internal RH measurements were conducted on Ordinary Portland cement pastes. Microstructure
evolution of cement pastes during the early age hydration period from the suspension state to the
solid state was also monitored by means of ultrasonic pulse velocity (UPV) method. The results
indicate that the occurrence of self-desiccation indicated by the start of internal RH drop links to the
self-supporting solid skeleton formation in cement pastes with low w/c ratio (0.25-0.30). This critical
time represents the time-zero for autogenous shrinkage. Meanwhile, final setting time determined
by the penetration method does not link to the self-supporting solid skeleton formation where selfdesiccation
begins to occur; it leads to an earlier estimation of time-zero and overestimation of
autogenous shrinkage.
Among the mitigation strategies, it is of great importance to counteract the effect of self-desiccation
and keep high internal RH by applying extra curing water. However, due to the very dense
microstructure of HPC/UHPC, the penetration depth of the external curing water is limited. Instead,
internal curing, referring to the use of water filled inclusions that can provide curing water
throughout the cross section of the concrete, has been found to be an effective method for
mitigating self-desiccation and autogenous shrinkage. Among the most used water inclusions
(lightweight aggregates (LWAs) and Superabsorbent polymers (SAPs), etc,), SAPs used as internal
curing agent show more promising properties in HPC/UHPC. It gives a motivation to reevaluate the
efficiency of applying SAPs in self-desiccation compensation and autogenous shrinkage mitigation
based on the newly defined time-zero.
In the case of internal curing by using SAPs, choosing final setting time as time-zero would lead to
underestimation of the internal curing efficiency of SAPs. If zeroing the autogenous shrinkage at final
setting time, autogenous shrinkage is not fully successfully mitigated although the internal RH results
suggest efficient self-desiccation mitigation in SAPs mixtures. Based on the newly defined time-zero,
SAPs with proper size (for example, SAP-A with mean particle size of 100.0 {\textpm} 21.5 \ensuremath{\mu}m) and sufficient
amount of water show efficient autogenous shrinkage mitigation, corresponding to their successful
mitigation of self-desiccation. The behaviors of SAPs as internal curing with different powder size of
SAPs and different amount of internal curing water in cement pastes with various w/c ratio have also
been investigated.
Due to the difficulty in proper dispersion of SAPs in concrete and the reported possible detrimental
effect of SAPs on the mechanical property of concrete, it is of great interest to seek alternative
materials with similar internal curing function but no or less possible negative influence on the
mechanical properties of concrete. One natural clinoptilolite-Ca-heulandite-Ca type zeolite was used
to evaluate its potential utilization as an internal curing agent for mitigation of self-desiccation and
the subsequent autogenous shrinkage of cement paste. The results indicate that addition of zeolite
could eliminate the self-desiccation partially and the autogenous shrinkage as well based on the
time-zero defined by the onset of internal RH drop. Attempts have been done for the purpose of
improving the internal curing property of zeolite by acid treatment and thermal treatment. However,
nitric acid treatment tends to increase the Brunauer-Emmett-Teller (BET) surface area of zeolite by
increasing the volume of 2-7 nm pores, which is undesirable. Thermal treatment tends to reduce the
porosity rather than to enlarge the pore size, which is also undesirable.
In the literature, it is shown that rice husk ash (RHA) presents effective internal curing function to
produce UHPC with promoted hydration degree of cement and eliminated autogenous shrinkage.
This leads to a way of seeking porous materials with finer particle size distribution in comparison
with LWAs or SAPs and high amorphous silica content (corresponding to high pozzolanic properties)
as internal curing agent. In this way, the volume of macropores (initially water filled) induced by the
incorporation of LWAs or SAPs could be reduced thus the detrimental effect on the strength is
reduced or even eliminated. To this end, three different biomass ashes were chosen to investigate
their possible application as internal curing agents in low w/b ratio cement pastes.
The experimental results reveal that only limited internal curing property can be expected from the
industry generated biomass ash (Napier grass ash), because the NGA does not really show internal
porous structure due to its highly crystallized structure. The lab-produced biomass ashes (MA and
SGA) from combustion of two different energy crops (miscanthus and switchgrass) show high
potassium content. After potassium extraction by acid leaching, both MA and SGA presented highly
porous structure with BET surface area of 45.28 m2/g and 35.52 m2/g, respectively. Highly effective
self-desiccation and autogenous shrinkage mitigation is observed in both MA and/or SGA
incorporated cement pastes. The effect of particle size of MA/SGA on the self-desiccation and
autogenous shrinkage mitigation efficiency has also been investigated. Take MA as example, with the
decreasing mean particle size from 20.43 \ensuremath{\mu}m to 6.67 \ensuremath{\mu}m, lower self-desiccation mitigation efficiency
was observed in the blended cement pastes. The grinding process may break MA particles and make
more and more small pores (pores with size in diameter in the range of 2-4 nm and pores with size of
around 20 nm) which are originally embedded inside MA particles exposed. Thus, more water is
expected to be absorbed in those pores and is expected to be released later at lower RH in hydrating
cement paste compared with the case of MA with D50 of 20.43 \ensuremath{\mu}m.},
author       = {Lyu, Yang},
isbn         = {978-94-6355-051-2},
language     = {eng},
school       = {Ghent University},
title        = {Autogenous shrinkage of cement-based materials : from the fundamental role of self-desiccation to mitigation strategies based on alternative materials},
year         = {2017},
}


Chicago
Lyu, Yang. 2017. “Autogenous Shrinkage of Cement-based Materials : from the Fundamental Role of Self-desiccation to Mitigation Strategies Based on Alternative Materials.”
APA
Lyu, Y. (2017). Autogenous shrinkage of cement-based materials : from the fundamental role of self-desiccation to mitigation strategies based on alternative materials.
Vancouver
1.
Lyu Y. Autogenous shrinkage of cement-based materials : from the fundamental role of self-desiccation to mitigation strategies based on alternative materials. 2017.
MLA
Lyu, Yang. “Autogenous Shrinkage of Cement-based Materials : from the Fundamental Role of Self-desiccation to Mitigation Strategies Based on Alternative Materials.” 2017 : n. pag. Print.