Advanced search
1 file | 16.45 MB Add to list

Effect of secondary copper slag as supplementary cementitious material and aggregate replacement on strength, hydration, microstructure and durability of ultra-high performance concrete

(2018)
Author
Promoter
(UGent) and (UGent)
Organization
Abstract
Since the ultra-high performance concrete (UHPC) has been developed under laboratory conditions in the 1960s, amazing progress has been made by researchers all over the world to improve the performance and durability of this material. The high mechanical strength and durability performance of UHPC allow for its use for structural elements in buildings and long span bridges. However, the production of UHPC requires a lot of Portland cement, which implies a high cost. Besides this, the production of Portland cement consumes a considerable amount of non-renewable energy for sintering the clinker. Moreover, the CO2 emissions from the cement industry contributes to global warming. Because of the rise in global copper demand for the construction and manufacturing industry, the copper production keeps increasing. The main environmental issue associated with this industry is the production of copper slag as a waste material. This slag cannot be recycled and needs a large area for storage, of which the availability is insufficient. Moreover, the impact on water quality when heavy metals and other harmful elements leach from this slag can be severe. For these reasons, a possible breakthrough can be sought in exploiting this waste material within cement and concrete production. To apply this idea, secondary copper slag from a Belgian recycling plant is used as supplementary cementitious material and aggregate replacement in UHPC. One of the goals of this PhD project is to study the development of pozzolanic activity of copper slag. By implementing different grinding methods, different fineness of copper slag will be achieved. To assess the pozzolanic activity of copper slag, several methods were applied. The copper slag has low pozzolanic activity determined by the Chapelle test. A similar result is also obtained when assessing the pozzolanic activity of this slag using the Frattini test after 15 days of curing. This result gives an explanation for the later obtained values of compressive strength and the reduced hydration heats when replacing cement by copper slag. The present results indicate that some mixtures at 90 days may show pozzolanic activity since they have a higher strength activity index (SAI) at that age. On the contrary, a promising result is obtained for clean slag which shows pozzolanic activity according to the Frattini test within 28 days of curing after applying a higher energy for grinding the slag. The latter result is achieved when the fineness of slag is comparable to or slightly higher than cement fineness. Effect of secondary copper slag as cementitious material in ultra-high performance mortar In this study, both quickly cooled granulated copper slag (QCS) and slowly cooled broken copper slag (SCS) were intensively ground using a planetary ball mill (dry method) to achieve two levels of fineness. Mortars were made with copper slag as partial cement replacment. The replacement ratios varied between 0 and 20 wt% in steps of 5 wt%. A very low water-to-binder ratio (w/b = 0.15) was chosen in order to produce ultra-high performance mortar (UHPM). In order to explain the effect of copper slag as a cement replacement, the compressive strength and heat production with isothermal calorimetry were investigated. Pozzolanic activity and strength activity index (SAI) were also determined to assess the reactivity of slag in cement paste and concrete. It was seen that a positive effect on the compressive strength of UHPM was achieved by increasing the fineness of SCS, while the increasing QCS fineness had no significant effect on strength development of UHPM. Using isothermal calorimetry, it was found that the binder reactions in UHPM can be enhanced by replacement of a small part (5%) of Portland cement by copper slag. Larger replacement levels rather delay the hydration reactions. This can be due to the dilution of the clinker content in the paste, to the limited pozzolanic activity of the copper slag, and to heavy metal compounds such as Zn in copper slag. Nevertheless, the use of finely ground slowly cooled copper slag in replacement levels of up to 20%, allows to reach similar heat production at 7 days as in a control mixture with Portland cement as only binder. The compressive strength at 7 days is even higher for the mixtures with copper slag than for the control with OPC. However, due to a slower strength development after 7 days, similar or slightly lower strengths are obtained for mixes with up to 20% copper slag, than for the OPC reference. Influence of intensive vacuum mixing and heat treatment on compressive strength and microstructure of reactive powder concrete incorporating secondary copper slag as supplementary cementitious material Copper slag has a hardness of 6-7 on the Mohs scale (hardness) and is mainly composed of iron silicate glass. Therefore, there will be a high energy need to grind this material. In the grinding process, the energy is determined by the time, speed, and number of balls charged. Based on the results obtained in the previous study, the specific surface area (SSA) of QCS reached a value of 2533 cm2/g with the Blaine permeability test, long duration of grinding (5 times during 12 minutes at 300 rpm) and 5 balls charged in the ball mill. Since this grinding process was time-consuming and not very productive, a short duration grinding process (6 times during 5 minutes at 390 rpm) and 7 balls charged was applied. This method reduced the grinding time with 30 minutes in comparison with that of the long duration method. With the increase in grinding speed and addition of two balls in this method it was expected to achieve a similar fineness as with the grinding method aforementioned (an SSA of 2277 cm²/g was realized). As mentioned before a significant compressive strength increase was not obtained when replacing the cement with the finer copper slag for UHPM. A possible way to enhance the pozzolanic activity of copper slag and compressive strength of the concrete was applying heat treatment combined with vacuum mixing to the concrete mixture. Therefore, these techniques were applied to RPC mixtures containing copper slag as a cement replacement, and a significant effect on the compressive strength of this concrete was expected. The porosity of RPC was determined by mercury intrusion porosimetry (MIP). The compressive strength obtained for RPC containing copper slag was comparable to or even better than for the reference mixture for all treatments. The effect of vacuum mixing to enhance the compressive strength of RPC is limited. This phenomenon is due to the fact that RPC is composed of fine particles, automatically reducing the porosity of this concrete and leaving only a limited amount of air voids. Therefore, when the vacuum mixing is applied, only remaining air bubbles can be removed. However, by applying heat treatment to the mixtures, a significant reduction in total porosity, macropores+entrained air voids, and capillary pores of RPC were obtained. The increase of the copper slag proportion up to 15% has only a limited effect on the heat production of cement paste. This shows that the filler effect of the slag is able to compensate for the reduced cement content, in spite of the fineness of the slag being lower than for the cement. The utilization of copper slag as cement replacement decreases the energy consumption and reduces the carbon footprint in the production of RPC. The small effect of copper slag used for UHPM and RPC is due to the insufficient fineness of this slag. Therefore, in the next steps of this study, the copper slag will be ground intensively to achieve a similar or higher fineness than cement. Influence of vacuum mixing and heat treatment on mechanical properties and microstructure of ultra-high performance concrete containing secondary copper slag as supplementary cementitious material (SCM) The compressive strength of RPC containing copper slag (QCS) under vacuum mixing and heat treatment was comparable or slightly increased compared to the reference mixture. From this result, it was realized that a higher compressive strength of concrete might be obtained if the copper slag fineness goes to nano-size. However, to achieve this size, higher energy for grinding is required since the copper slag has a hardness of 6-7 on the Mohs scale. Therefore, for UHPC the copper slag QCS was used as basalt aggregate replacement, and SCS slag was used as cementitious material, since the SCS seemed softer to grind than QCS. Since the heavy metal content in SCS and QCS used for UHPM and RPC is responsible for a retardation effect during the early hydration of cement paste, it is interesting to study a recently developed clean slag in concrete which contains less heavy metal. The UHPC containing copper slag (QCS) aggregate, prepared under vacuum mixing or without vacuum mixing was used as a reference. The SCS or clean slag were then used as cement replacement, and vacuum mixing + heat treatment were applied to this UHPC. Similar to RPC and UHPM, the compressive strength of UHPC containing copper slag (QCS) as aggregate replacement for both vacuum mixing and non-vacuum mixing was comparable to or even better than the reference mixture. The same behaviour was also obtained for the compressive and flexural strength of UHPC containing SCS as cement replacement under vacuum mixing or without vacuum mixing combined with heat treatment. The use of clean slag seems beneficial as compressive and flexural strength are comparable or even better compared to reference mixtures, while a minimal effect on compressive strength of UHPC is obtained in the presence of SCS. This is related to the higher fineness of clean slag compared to SCS, and the reduced contents of Zn and Pb which may retard the hydration reactions. Replacement of the cement by SCS reduced the amount of chemically bound water. Quantitative analysis of porosity of reactive powder concrete based on backscattered-electron imaging and GUI-based Matlab When the porosity of RPC is quantified by using the Wong overflow method to determine the grey value threshold in SEM-BSE images, it is shown that this method overestimates the total porosity compared to mercury intrusion porosimetry. To quantify the porosity efficiently, accurately, and reliably, a new threshold selection method is proposed based on the overflow method. The image which is captured by scanning electron microscope-backscattered electron imaging (SEM-BSE) is quantified by using different types of thresholds. The proposed method furthermore divides the captured images in two groups, one of low and one of high brightness, respectively corresponding to low and high grey value thresholds. The new proposed threshold method for SEM-BSE image analysis provides a reliable result, and it can be a good alternative for investigating the porosity of reactive powder concrete. The procedure can be automated through combination with GUI-based Matlab. It was seen that the porosity determined with the proposed grey level thresholds (threshold 1 and 2) corresponded better with the porosity obtained by MIP than the Wong method. Still, a very good correspondence could not be obtained between the proposed threshold method and MIP even with the best threshold, due to the different sizes and characteristics of the pores that can be measured with BSE and MIP respectively (e.g. MIP can only reach open pores, while BSE visualizes also closed pores). Although the proposed threshold method performs well for the investigation of the porosity of RPC, by combining the result with other techniques such as computed tomography, air void analysis, or fluorescence microscopy a more complete result representing all pore sizes and types of pores may be obtained. Influence of vacuum mixing on the carbonation resistance and microstructure of reactive powder concrete containing secondary copper slag as supplementary cementitious material (SCM) When the CO2 penetrates into concrete, it will cause carbonation and the deterioration of reinforced concrete. In this study, the carbonation depth on the RPC samples was measured using phenolphthalein method and optical microscopy. The influence of vacuum mixing on the porosity reduction and the strength enhancement of RPC containing copper slag were also studied. The carbonation depth for both the samples mixed under vacuum condition (100 mbar) and atmospheric pressure (1013 mbar) which was measured on a freshly split RPC surface using phenolphthalein indicates no carbonation up to 48 weeks at CO2 concentration of 10%. The microscopic observation of the selected RPC mixture under vacuum and non-vacuum mixing shows no carbonation both near the surface and deeper down. These findings seem promising, however it is necessary to further investigate the carbonation for longer-term exposure (e.g. xviii Summary more than 96 weeks accelerated testing) in order to better assess the effect of copper slag and vacuum mixing on carbonation resistance. By applying vacuum mixing to the RPC mixture, the porosity decreased. The reduction is higher with increasing (15-20%) copper slag content. This result is in contrast with the compressive strength enhancement of RPC, which decreased for larger replacement levels of copper slag.

Downloads

  • PhD RomySuryaningratEdwin.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 16.45 MB

Citation

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

MLA
Edwin, Romy Suryaningrat. Effect of Secondary Copper Slag as Supplementary Cementitious Material and Aggregate Replacement on Strength, Hydration, Microstructure and Durability of Ultra-High Performance Concrete. Ghent University, 2018.
APA
Edwin, R. S. (2018). Effect of secondary copper slag as supplementary cementitious material and aggregate replacement on strength, hydration, microstructure and durability of ultra-high performance concrete. Ghent University, Ghent.
Chicago author-date
Edwin, Romy Suryaningrat. 2018. “Effect of Secondary Copper Slag as Supplementary Cementitious Material and Aggregate Replacement on Strength, Hydration, Microstructure and Durability of Ultra-High Performance Concrete.” Ghent: Ghent University.
Chicago author-date (all authors)
Edwin, Romy Suryaningrat. 2018. “Effect of Secondary Copper Slag as Supplementary Cementitious Material and Aggregate Replacement on Strength, Hydration, Microstructure and Durability of Ultra-High Performance Concrete.” Ghent: Ghent University.
Vancouver
1.
Edwin RS. Effect of secondary copper slag as supplementary cementitious material and aggregate replacement on strength, hydration, microstructure and durability of ultra-high performance concrete. [Ghent]: Ghent University; 2018.
IEEE
[1]
R. S. Edwin, “Effect of secondary copper slag as supplementary cementitious material and aggregate replacement on strength, hydration, microstructure and durability of ultra-high performance concrete,” Ghent University, Ghent, 2018.
@phdthesis{8581729,
  abstract     = {Since the ultra-high performance concrete (UHPC) has been developed under laboratory
conditions in the 1960s, amazing progress has been made by researchers all over the world to
improve the performance and durability of this material. The high mechanical strength and
durability performance of UHPC allow for its use for structural elements in buildings and long
span bridges. However, the production of UHPC requires a lot of Portland cement, which
implies a high cost. Besides this, the production of Portland cement consumes a considerable
amount of non-renewable energy for sintering the clinker. Moreover, the CO2 emissions from
the cement industry contributes to global warming.
Because of the rise in global copper demand for the construction and manufacturing industry,
the copper production keeps increasing. The main environmental issue associated with this
industry is the production of copper slag as a waste material. This slag cannot be recycled and
needs a large area for storage, of which the availability is insufficient. Moreover, the impact on
water quality when heavy metals and other harmful elements leach from this slag can be severe.
For these reasons, a possible breakthrough can be sought in exploiting this waste material within
cement and concrete production. To apply this idea, secondary copper slag from a Belgian
recycling plant is used as supplementary cementitious material and aggregate replacement in
UHPC.
One of the goals of this PhD project is to study the development of pozzolanic activity of copper
slag. By implementing different grinding methods, different fineness of copper slag will be
achieved. To assess the pozzolanic activity of copper slag, several methods were applied. The
copper slag has low pozzolanic activity determined by the Chapelle test. A similar result is also
obtained when assessing the pozzolanic activity of this slag using the Frattini test after 15 days
of curing. This result gives an explanation for the later obtained values of compressive strength
and the reduced hydration heats when replacing cement by copper slag. The present results
indicate that some mixtures at 90 days may show pozzolanic activity since they have a higher
strength activity index (SAI) at that age. On the contrary, a promising result is obtained for
clean slag which shows pozzolanic activity according to the Frattini test within 28 days of
curing after applying a higher energy for grinding the slag. The latter result is achieved when
the fineness of slag is comparable to or slightly higher than cement fineness.

Effect of secondary copper slag as cementitious material in ultra-high
performance mortar
In this study, both quickly cooled granulated copper slag (QCS) and slowly cooled broken
copper slag (SCS) were intensively ground using a planetary ball mill (dry method) to achieve
two levels of fineness. Mortars were made with copper slag as partial cement replacment. The
replacement ratios varied between 0 and 20 wt% in steps of 5 wt%. A very low water-to-binder
ratio (w/b = 0.15) was chosen in order to produce ultra-high performance mortar (UHPM). In
order to explain the effect of copper slag as a cement replacement, the compressive strength
and heat production with isothermal calorimetry were investigated. Pozzolanic activity and
strength activity index (SAI) were also determined to assess the reactivity of slag in cement
paste and concrete.
It was seen that a positive effect on the compressive strength of UHPM was achieved by
increasing the fineness of SCS, while the increasing QCS fineness had no significant effect on
strength development of UHPM. Using isothermal calorimetry, it was found that the binder
reactions in UHPM can be enhanced by replacement of a small part (5%) of Portland cement
by copper slag. Larger replacement levels rather delay the hydration reactions. This can be due
to the dilution of the clinker content in the paste, to the limited pozzolanic activity of the copper
slag, and to heavy metal compounds such as Zn in copper slag. Nevertheless, the use of finely
ground slowly cooled copper slag in replacement levels of up to 20%, allows to reach similar
heat production at 7 days as in a control mixture with Portland cement as only binder. The
compressive strength at 7 days is even higher for the mixtures with copper slag than for the
control with OPC. However, due to a slower strength development after 7 days, similar or
slightly lower strengths are obtained for mixes with up to 20% copper slag, than for the OPC
reference.

Influence of intensive vacuum mixing and heat treatment on compressive
strength and microstructure of reactive powder concrete incorporating
secondary copper slag as supplementary cementitious material
Copper slag has a hardness of 6-7 on the Mohs scale (hardness) and is mainly composed of iron
silicate glass. Therefore, there will be a high energy need to grind this material. In the grinding
process, the energy is determined by the time, speed, and number of balls charged. Based on
the results obtained in the previous study, the specific surface area (SSA) of QCS reached a
value of 2533 cm2/g with the Blaine permeability test, long duration of grinding (5 times during
12 minutes at 300 rpm) and 5 balls charged in the ball mill. Since this grinding process was
time-consuming and not very productive, a short duration grinding process (6 times during 5
minutes at 390 rpm) and 7 balls charged was applied. This method reduced the grinding time
with 30 minutes in comparison with that of the long duration method. With the increase in
grinding speed and addition of two balls in this method it was expected to achieve a similar
fineness as with the grinding method aforementioned (an SSA of 2277 cm²/g was realized). As
mentioned before a significant compressive strength increase was not obtained when replacing
the cement with the finer copper slag for UHPM. A possible way to enhance the pozzolanic
activity of copper slag and compressive strength of the concrete was applying heat treatment
combined with vacuum mixing to the concrete mixture. Therefore, these techniques were
applied to RPC mixtures containing copper slag as a cement replacement, and a significant
effect on the compressive strength of this concrete was expected. The porosity of RPC was
determined by mercury intrusion porosimetry (MIP).
The compressive strength obtained for RPC containing copper slag was comparable to or even
better than for the reference mixture for all treatments. The effect of vacuum mixing to enhance
the compressive strength of RPC is limited. This phenomenon is due to the fact that RPC is
composed of fine particles, automatically reducing the porosity of this concrete and leaving
only a limited amount of air voids. Therefore, when the vacuum mixing is applied, only
remaining air bubbles can be removed. However, by applying heat treatment to the mixtures, a
significant reduction in total porosity, macropores+entrained air voids, and capillary pores of
RPC were obtained.
The increase of the copper slag proportion up to 15% has only a limited effect on the heat
production of cement paste. This shows that the filler effect of the slag is able to compensate
for the reduced cement content, in spite of the fineness of the slag being lower than for the
cement. The utilization of copper slag as cement replacement decreases the energy consumption
and reduces the carbon footprint in the production of RPC.
The small effect of copper slag used for UHPM and RPC is due to the insufficient fineness of
this slag. Therefore, in the next steps of this study, the copper slag will be ground intensively
to achieve a similar or higher fineness than cement.

Influence of vacuum mixing and heat treatment on mechanical properties
and microstructure of ultra-high performance concrete containing
secondary copper slag as supplementary cementitious material (SCM)
The compressive strength of RPC containing copper slag (QCS) under vacuum mixing and heat
treatment was comparable or slightly increased compared to the reference mixture. From this
result, it was realized that a higher compressive strength of concrete might be obtained if the
copper slag fineness goes to nano-size. However, to achieve this size, higher energy for
grinding is required since the copper slag has a hardness of 6-7 on the Mohs scale. Therefore,
for UHPC the copper slag QCS was used as basalt aggregate replacement, and SCS slag was
used as cementitious material, since the SCS seemed softer to grind than QCS. Since the heavy
metal content in SCS and QCS used for UHPM and RPC is responsible for a retardation effect
during the early hydration of cement paste, it is interesting to study a recently developed clean
slag in concrete which contains less heavy metal.
The UHPC containing copper slag (QCS) aggregate, prepared under vacuum mixing or without
vacuum mixing was used as a reference. The SCS or clean slag were then used as cement
replacement, and vacuum mixing + heat treatment were applied to this UHPC.
Similar to RPC and UHPM, the compressive strength of UHPC containing copper slag (QCS)
as aggregate replacement for both vacuum mixing and non-vacuum mixing was comparable to
or even better than the reference mixture. The same behaviour was also obtained for the
compressive and flexural strength of UHPC containing SCS as cement replacement under
vacuum mixing or without vacuum mixing combined with heat treatment. The use of clean slag
seems beneficial as compressive and flexural strength are comparable or even better compared
to reference mixtures, while a minimal effect on compressive strength of UHPC is obtained in
the presence of SCS. This is related to the higher fineness of clean slag compared to SCS, and
the reduced contents of Zn and Pb which may retard the hydration reactions. Replacement of
the cement by SCS reduced the amount of chemically bound water.

Quantitative analysis of porosity of reactive powder concrete based on backscattered-electron imaging and GUI-based Matlab
When the porosity of RPC is quantified by using the Wong overflow method to determine the
grey value threshold in SEM-BSE images, it is shown that this method overestimates the total
porosity compared to mercury intrusion porosimetry. To quantify the porosity efficiently,
accurately, and reliably, a new threshold selection method is proposed based on the overflow
method. The image which is captured by scanning electron microscope-backscattered electron
imaging (SEM-BSE) is quantified by using different types of thresholds. The proposed method
furthermore divides the captured images in two groups, one of low and one of high brightness,
respectively corresponding to low and high grey value thresholds.
The new proposed threshold method for SEM-BSE image analysis provides a reliable result,
and it can be a good alternative for investigating the porosity of reactive powder concrete. The
procedure can be automated through combination with GUI-based Matlab. It was seen that the
porosity determined with the proposed grey level thresholds (threshold 1 and 2) corresponded
better with the porosity obtained by MIP than the Wong method. Still, a very good
correspondence could not be obtained between the proposed threshold method and MIP even
with the best threshold, due to the different sizes and characteristics of the pores that can be
measured with BSE and MIP respectively (e.g. MIP can only reach open pores, while BSE
visualizes also closed pores). Although the proposed threshold method performs well for the
investigation of the porosity of RPC, by combining the result with other techniques such as
computed tomography, air void analysis, or fluorescence microscopy a more complete result
representing all pore sizes and types of pores may be obtained.

Influence of vacuum mixing on the carbonation resistance and
microstructure of reactive powder concrete containing secondary copper
slag as supplementary cementitious material (SCM)
When the CO2 penetrates into concrete, it will cause carbonation and the deterioration of
reinforced concrete. In this study, the carbonation depth on the RPC samples was measured
using phenolphthalein method and optical microscopy. The influence of vacuum mixing on the
porosity reduction and the strength enhancement of RPC containing copper slag were also
studied.
The carbonation depth for both the samples mixed under vacuum condition (100 mbar) and
atmospheric pressure (1013 mbar) which was measured on a freshly split RPC surface using
phenolphthalein indicates no carbonation up to 48 weeks at CO2 concentration of 10%. The
microscopic observation of the selected RPC mixture under vacuum and non-vacuum mixing
shows no carbonation both near the surface and deeper down. These findings seem promising,
however it is necessary to further investigate the carbonation for longer-term exposure (e.g.
xviii
Summary
more than 96 weeks accelerated testing) in order to better assess the effect of copper slag and
vacuum mixing on carbonation resistance.
By applying vacuum mixing to the RPC mixture, the porosity decreased. The reduction is higher
with increasing (15-20%) copper slag content. This result is in contrast with the compressive
strength enhancement of RPC, which decreased for larger replacement levels of copper slag.},
  author       = {Edwin, Romy Suryaningrat},
  isbn         = {978-94-6355-093-2},
  language     = {eng},
  pages        = {234},
  publisher    = {Ghent University},
  school       = {Ghent University},
  title        = {Effect of secondary copper slag as supplementary cementitious material and aggregate replacement on strength, hydration, microstructure and durability of ultra-high performance concrete},
  year         = {2018},
}