Alkali–Silica Reactivity Potential of Reactive and Non-Reactive Aggregates under Various Exposure Conditions for Sustainable Construction

Abstract

The alkali–silica reaction (ASR) is a primary cause for premature concrete degradation. An accelerated mortar bar test is often used to access the detrimental phenomena in concrete caused by the ASR of aggregates. However, this test requires a certain environmental conditioning as per ASTM C1260. The objective of this study is to explore the effects of the cement alkali content, exposure solution concentration, temperature, and test duration on mortar bar expansion. Factorial experimental design and analysis was conducted to delineate the effects of the individual factors as well as their interaction. Five different aggregates with various mineralogical properties were used, representing reactive and non-reactive aggregates. Various dosages of cement alkalis (0.40, 0.80, and 1.20 Na₂Oe), sodium hydroxide (NaOH) solution concentrations (0.5, 1.0, and 1.5 N), and temperature (40 °C, 80 °C, and 100 °C) were the studied variables. Mortar bar expansion was measured at 3, 7, 14, 21, 28, 56, and 90 days. Mortar bars incorporating Jhelum aggregates incurred expansion of 0.32% at 28 days, proving to be reactive aggregates as per ASTM C1260. Similarly, specimens incorporating Taxila aggregates showed expansion of 0.10% at 28 days, indicating non-reactive nature. It was observed that specimens with Sargodha aggregates showed expansion of 0.27% at 28 days for 0.50 N NaOH solution concentration compared to 0.31% expansion for identical specimens exposed to 1.5 N solution. Moreover, expansion increased with exposure duration for all the tested specimens. Experimental results showed that the cement alkali contents had relatively lesser effect on expansion for 1.0 N NaOH; while, in the case of 0.5 N and 1.5 N NaOH, the cement alkali had a significant effect. It was noted that expansion increased with an increase in the temperature. Jhelum aggregates showed 28-day expansion of 0.290% when exposed to 40 °C, but at a temperature of 100 °C, expansion increased to 0.339%. Factorial analysis revealed that the exposure solution had a major contribution towards the expansion of mortar bar specimens. This study highlights the contribution of various exposure conditions on the ASR expansion, which leads to a decisive role in selecting the aggregate sources for various applications and exposure conditions leading to sustainable construction.

1. Introduction

The alkali–silica reaction (ASR) is one of the most well-known detrimental processes in concrete. Aggregates that have reactive silicates react with Portland cement’s alkalis to form an alkali–silica gel that expands on interaction with moisture [1]. It results in concrete cracking, leading to degradation of the mechanical characteristics [2,3,4,5,6,7,8]. The presence of alkalis, humidity, temperature, and sensitive aggregates cause the ASR in concrete. This reaction starts from the surface of aggregates and leads to formation of the ASR gel. This gel formation will be increased with the passage of time. Aggregate’s mineralogy, physical properties, and atomic structures are the main factors of its reactivity [9,10]. Various literature review studies have been conducted in the past on the mechanisms, test methodologies, modeling, and factors affecting the ASR in concrete structures (

Table 1. Published manuscripts on literature review of alkali–silica reaction.

Sr No.	Year	Main Discussion and Findings	References
1	2022	Alkali silica reaction (ASR) mechanism was examined using multiscale and multimethod techniques and the associated concrete damage was investigated.	Weise et al. [11]
2	2021	The ASR was examined using aggregates from different countries using petrographic analysis and expansion measurements.	Kazemi et al. [12]
3	2020	Investigating the ASR in recycled aggregate concrete, it was concluded that the age, exposure circumstances, crushing technique, and the heterogeneity of source can affect the alkalis and reactive silica present in the aggregates and responsible for the ASR.	Santos et al. [13]
4	2019	The ASR mechanism, its mitigation, and test methodology were reviewed.	Figueira et al. [14]
5	2018	The ASR mechanism of reactive aggregates was reviewed. Moreover, the role of fly ash in controlling the ASR was discussed.	Saha et al. [15]
6	2018	The ASR mechanism, its consequences, factors affecting the ASR, and preventive measures were discussed.	Munir et al. [16]
7	2016	Various models were reviewed to assess the ASR damage in concrete.	Esposito and Hendriks [17]
8	2016	Critical review on alkali–silica reactivity of concrete structures in Pakistan was conducted.	Munir et al. [18]
9	2015	Existing knowledge gaps on the ASR mechanism (chemical and physical reactions), composition of pore solution, and aggregates and surrounding exposures on the acceleration of the ASR were discussed.	Rajabipour et al. [19]
10	2013	Mechanism and deterioration due to the ASR in concrete was reviewed. Mitigation measures using coal ash and steel fibers were also evaluated.	Abbas [20]
11	2013	Factors affecting the ASR were reviewed. Further, the behavior of concrete mixtures incorporating polymer with regards to the ASR was also discussed.	Feiteira et al. [21]
12	2012	The ASR modelling and chemical expansion in concrete were reviewed.	Pan et al. [22]
13	2012	Various factors that affect the ASR testing methodologies were reviewed.	Lindgard et al. [23]
14	2011	Review of Supplementary cementitious Materials on the ASR was conducted.	Thomas [24]
15	2005	Effect of aggregates on the ASR was discussed. A test method was developed to evaluate the ASR of aggregates.	Chatterji [25]
16	2000	Literature review on the existing condition of alkali aggregate reaction in various part of Canada was presented.	Fournier and Berube [26]
17	1988	Implications of proposed the ASR mechanism were discussed.	Chatterji et al. [27]
18	1987	Mechanism of ASR was discussed. Factor affecting the ASR expansion were reviewed.	Urhan [28]
19	1976	Role of aggregates in concrete on the expansion mechanism of the ASR was reviewed.	Diamond [29]
20	1975	Role of alkalis in concrete on the expansion mechanism of the ASR was reviewed.	Diamond [30]

).

Previous studies have shown that hydraulic structures had a more devastating effect due to the ASR compared to other structures [31]. For instance, in Pakistan, many structures have reported damaged owing to the devastating effect of the ASR. Izhar and Hassan [32] reported the first case related to the ASR in Tarbela dam. Similarly, Warsak dam and Jinnah barrage were also affected by the damaging effect of the ASR. Ahsan et al. [33] and WAPDA [34] reported that Himalayan rocks that have gneisses and quartzite and other metamorphic rocks considered as non-reactive aggregates were also affected by the ASR [35,36,37,38].

Table 2. Review of previous studies conducted on the ASR in last decade.

Sr No.	Studied Variables	Test Performed and Methodologies	Results and Findings	References
1	The ASR potential was studied by examining local aggregates	Expansion (ASTM C1260), compressive and flexural strength	The ASR was investigated in locally available unexplored aggregates following the ASTM C1260.	Abbas et al. [39]
2	Recycled untreated rubber waste (RW) was used for controlling the ASR in concrete mixtures.	Expansion test (ASTM C1260)	The use of untreated recycled rubber waste was effective for controlling the ASR. Microstructural and petrographic analysis confirmed the findings of expansion test as per ASTM C1260.	Abbas et al. [40]
3	Pre-treated glass fibers with Ca (OH)2, NaOH, and a combination of Ca (OH)2 and NaOH was investigated to mitigate the ASR.	Expansion (ASTM C1260) and compressive strength.	The ASR was mitigated for all types of pre-treatment methods. Compressive strength was improved. For instance, it improved by 6% when the glass aggregate was preheated in Ca (OH)2 solution at 80 °C for 7 days.	Sun et al. [41]
4	The ASR resistance of ground glass fiber (GGF) and fly ash based geopolymers was compared with OPC and dynamic modulus of elasticity was tested.	Expansion and the change in DME were tested (ASTM C1260 and ASTM E1876).	Both geopolymer mixtures (GGF and fly ash-based geopolymers) exhibited lesser expansion as compared to OPC mixture. However, these specimens exhibited significant reduction in dynamic modulus of elasticity (especially for the specimen prepared with siliceous aggregate and high alkali environment curing).	Dezfouli et al. [42]
5	The ASR of glass aggregate in alkali activated concrete (AAC) was explored using expansion tests.	Expansion test (ASTM C1260 and ASTM C1293).	One year expansion of OPC and ground granulated blast furnace slag (GGBFS) were above the ASTM limiting value; while the expansion of fly-ash based AAC and GGBFS were below the limiting value.	Khan et al. [43]
6	The ASR in alkali-activated copper slag (AACS) was examined with varying alkali content and silicate modulus.	Expansion test, shrinkage test, compressive strength test.	Expansion in AACS was lesser than OPC. The expansion due to the ASR in AACS with 4% and 6% alkali content was within standard limits at 14 and 28 days while the AACS specimen with 8% and 10% alkali content exhibited higher expansion.	Singh et al. [44]
7	Effect of potassium (K), sodium (Na) and calcium (Ca) on alkali silica formation and C-S-H gel.	X-ray diffraction analysis and thermodynamic modelling.	The formation of the ASR requires certain amount of K, or Na. Too low or too high concentrations did not favor the ASR. The most favorable concentration of K and Na were found to be ranging between 200 and 500 mM and Ca/Si ranged between 0.10 and 0.40.	Shi et al. [45]
8	Comparison between ASR products formed before and after cracking of aggregates.	Micro-structural analysis.	The ASR products in thin grain boundaries were found to be amorphous and may be defined as untextured ASR products. The platy products formed in veins were mostly crystalline. Although the morphology was different (untextured and platy products), the composition was found to be same.	Courjault et al. [46]
9	Effect of fly ash on controlling the ASR in concrete.	Compressive strength, expansion, density, water absorption, ultra-sonic pulse velocity test.	On adding fly ash, the compressive strength increases. The most optimum value causing an increase in compressive strength by 13% and 9% at 7 and 28 days, respectively, was found to be 35%. Expansion was reduced to 0.069% from 0.113% for mixture with 65% of fly ash. Other benefits included higher density and lower water absorption.	Nagrockiene et al. [47]
10	The role of temperature on the expansive behavior and microstructural changed in concrete due to the ASR.	Expansion test (Alkali-wrapped concrete prism test), Polarizing microscopy, SEM/EDS.	The effect of temperature was found to be more dominant during early stages while the trend was reversed at late stages. At lesser temperatures (20 °C), more cracks were found around aggregates during SEM analysis as compared to higher temperature (60 °C).	Kawabata et al. [48]
11	Role of calcium on the ASR.	X-ray diffraction analysis and thermodynamic modelling.	By increasing the Ca/Si ratio, the ASR formation initially increases and then decreases. The reduction of the ASR was accompanied by formation of CSH formation. This may hint at the conversion of the ASR to CSH at higher temperatures.	Shi et al. [49]
12	Ground glass fiber (GGF), ground glass powder (GLP) and Metakaolin (MK) in controlling of the ASR.	Compressive strength, thermogravimetric analysis, accelerated mortar bar test.	GGF, GLP, and MK were successful in reducing the ASR. GGF at 30% replacement was found to be most effective (all values were under code limitations for concrete and mortar specimen). The strength activity index of all mixtures was greater than 75% at 28 days where all the GGF and MK mixtures exhibited SAI of more than 100%.	Dezfouli et al. [50]
13	ASR in glass activated slag mortars incorporating fly ash and metakaolin.	Mortar bas expansion, alkalinity, SEM and EDS.	The most optimum replacement value of fly ash and metakaolin to reduce the ASR was found to be 30%. Moreover, 70% replacement with slag could suppress the ASR completely.	Shi et al. [51]
14	Role of calcium sulfo-aluminate (CSA) cement to mitigate the alkali silica reaction.	Expansion test, SEM, XRD.	The expansion of CSA mortar was found to be lesser than OPC mortar by 7% along with fewer deteriorations at micro level and fewer cracks. The amount of gel formed was also less in case of CSA mortar.	Kleib et al. [52]
15	Effect of alkali on the ASR.	Expansion test, alkali content.	Lower curing temperature caused more expansion as compared to higher temperatures (38 °C vs. 60 °C). No effect of Na content released from aggregate was found on the expansion. Similarly, enough evidence was not found to establish the impact of alkali released from aggregates on the ASR expansion.	Drolet et al. [53]
16	Role of activating solution on the ASR of concrete with fly ash.	Compressive strength, expansion test, SEM.	The alkali activated fly ash concrete (AAFAC) prepared with 8 M NaOH activator exhibited highest compressive strength and the expansion was also very low (0.006% after 24 months) as compared to 5% expansion found in OPC concrete. Mixture with 10 M NaOH showed 0.02% expansion at 24 months.	Williamson et al. [54]
17	Role of glass powder in controlling the ASR.	Expansion test, thermal analysis, XRD analysis, SEM and EDS.	The incorporation of soda-lime glass decreased the content of mono-sulphates. On using highly reactive aggregates, GPF specimen exhibited lesser expansion (0.1%) as compared to control mortar (0.6%) at 180 days.	Zheng et al. [55]
18	Role of rice husk ash (RHA) in mitigating the ASR.	Mortar bar expansion, petrographic analysis, chemical analysis.	A decrease in the ASR expansion can be possible with RHA due to its pozzolanic properties. Therefore, RHA replacement (up to 40% with cement) can be effective in controlling the ASR.	Munir et al. [56]
19	Effect of fibers on residual strength of concrete affected by the ASR.	Flexural strength, toughness, expansion.	The incorporation of steel fibers inhibited the development of the expansion. Mostly cracks were oriented along the beam axis. However, fiber incorporation reduced the width and number of cracks. As the damage increased, the compressive strength and the modulus of elasticity decreased.	Giaccio et al. [57]
20	Role of fineness of ground recycled glass on the ASR.	Flow, compressive strength, expansion test, micro-structure analysis.	Glass powder of various fineness was used as aggregate and cement replacement was examined following ASTM C1260 and ASTM C1567. It was found that on incorporating glass powder, the flow of fresh concrete increases. The pozzolanic behavior of fiber glass particles was better (more compressive strength). The incorporation of glass powder reduced expansion due to the ASR; the 17-micron particles, especially, could reduce expansion at all ages.	Afshinnia et al. [58]
21	Role of silica fumes (SF) and rice husk ash (RHA) in controlling the ASR in high performance concrete.	Expansion test.	In comparison, SF was more efficient in decreasing expansion due to the ASR than RHA. It was found that larger particle size caused more expansion. Microscopic analysis revealed that cracks were also originated from RHA particles.	Le et al. [59]
22	Role of ternary blends containing slag and high-calcium fly ash on the ASR.	Expansion test (ASTM C1260), alkali leaching test.	It was found that using a ternary blend may not offer significant advantage over binary blend. As the ability to mitigate the ASR was found to be related to the capacity of a supplementary cementitious particle to retain alkalis, it was suggested that alkali leaching test was more realistic than accelerated mortar bar test to predict long term ASR potential.	Kandasamy et al. [60]
23	Effect of particle size on the ASR in recycled glass concrete.	Expansion test.	Various supplementary cementitious materials with varying particle sizes were utilized to study the impact on the ASR. It was found that smaller particles of green and brown glass (>2.36 mm) can be used to mitigate the ASR effectively. Similarly, the optimum values for incorporation of fly-ash, GGBS, silica fumes, steel fibers, and lithium compounds were found to be 10–50%, 45–60%, 12.5%, 1.5–2.0%, and 0.5–2.0%, respectively.	Du et al. [61]
24	Role of clay pozzolan as an admixture to mitigate ASR.	Expansion test, XRD.	Calcined clay reduced the expansion, between 42% and 108% at 14 days and between 9% and 16% at 84 days. Furthermore, 25% incorporation of pozzolan was found to be the most optimum. XRD analysis exhibited stable calcium silicates and the ASR gel reduced on addition of pozzolan.	Ansah et al. [62]
25	Impact of the ASR on fly ash based geopolymer concrete.	Expansion test and compression strength test.	It was found that the ASR potential of geopolymer concrete was lesser than OPC concrete. The values were found to be within ASTM limits.	Patil et al. [63]
26	Role of waste glass as sand in mortar to mitigate the ASR.	Expansion test and micro-structural study.	It was found that the replacement of sand with clear glass particles increased the ASR expansion due to micro-cracks in glass particles. The effects become more deleterious beyond 50% replacement of sand. However, the use of smaller particles causes lesser expansion due to better pozzolanic behavior. Furthermore, it was found that incorporation of fly ash and GGBS could suppress the ASR effectively.	Du et al. [64]
27	Twenty-year test on mitigation of the ASR using blast furnace slag and silica fume.	Expansion test on reinforced and unreinforced beams (0.6 m × 0.6 m × 2 m) and pavement slabs (0.2 m × 1.2 m × 4 m).	The content of alkalis in cement was proportional to ASR expansion where cracks in concrete appeared after 5 years in case of high alkali cement and 12 years in case of low alkali cement. However, it was found that the incorporation of 50% GGBFS or a blend of slag and silica fumes could eliminate the expansion.	Hooton et al. [65]
28	Role of aluminum on the dissolution of amorphous silica and its relation to the ASR.	Expansion test, pore solution analysis.	Aluminum was found to effectively suppress the ASR effects on aggregates. Even a small amount of aluminum (3.9 mm) was found to be quite effective. The main reason was the ability of aluminum to reduce the dissolution of amorphous silica inside aggregates. The damage concentrations were found to be comparable at various dosages of aluminum.	Chappex et al. [66]
29	Effects of aggregate size on the ASR.	Expansion test.	The expansion due to the ASR mainly depends upon the fracture of aggregates at early stages and fracture of paste at later stages. It was found that aggregate size between 4 to 8 mm expands at faster rate.	Dunant et al. [67]
30	Role of biomass fly ash for controlling the ASR.	Material characterization (TG/DTA, XRF, XRD), expansion test, compressive strength, flexural strength.	Two types of fly ashes were used in the study with different chemical characteristics. Amounts of 20–30% of these were used to replace cement. Although fly ash reduced the expansion at 28 days, the 14 days values were above ASTM limits while increasing both compressive and flexural strengths.	Esteves et al. [68]
31	Effect of blended cements produced with natural zeolite and industrial by-products on the ASR.	Compressive strength, expansion test, sulfate resistance.	It was found that blended cement incorporating GGBFS, zeolite, and FA reduced the ASR expansion. Blended cement with GGBFS was found to be more effective. Further, the incorporation of these materials improved the compressive strength.	Karakurt et al. [69]
32	Role of fly ash (FA) on the expansion of concrete due to ASR.	Expansion test on concrete blocks placed outside for 18 years.	It was found that FA was effective at 25% and 40% replacement levels in reducing expansion. The control specimen containing flint aggregate showed excessive expansion; however, on incorporation of FA, the expansion was found to be very little (within ASTM limits).	Thomas et al. [24]
33	Influence of superplasticizers (SP) on pore solution composition and on expansion of concrete due to the ASR.	Calorimetry, pore solution, micro bar test (AFNOR P 18-588), expansion test (AFNOR P 18-454).	On using naphthalene sulphonate-based SP, the expansion increased during the first two weeks due to increase in hydroxide concentration. The expansion was found to be more than the limiting values. While, on using polycarboxylate-based SP, the expansion did not increase.	Leemann et al. [70]
34	Effect of recycled glass aggregates on the ASR.	SEM analysis, expansion test.	It was found that larger glass particles cause a greater ASR and more expansion as compared to smaller particles. Main reason being the larger micro cracks in large glass particles. At the interface, glass particles show reactivity with cement causing production of non-expansive CSH. Smaller sized glass particles show no deleterious effects of ASR.	Rajabipour et al. [71]
35	Role of binary and ternary blends of metakaolin and fly ash for the ASR.	Expansion test.	On using metakaolin and fly ash as binary blend, the ASR expansion was reduced. The reduction was greater in the case of metakaolin as compared to fly ash. Metakaolin was effective mainly due to smaller particles size and higher reactivity. It was found that ternary blends have no significant improvement.	Moser et al. [72]

summarizes previous studies conducted on the ASR in last decade.

According to the literature analysis, it was found that mega structures in developing countries have faced challenges related to the ASR. Therefore, proper testing is required for evaluating the aggregates in large mega scale projects. Before using the possibly reactive aggregates, it is important to fully comprehend all the elements present in them and how they are related to the ASR. Accelerated mortar bar test is the commonly used method for determining the aggregate’s reactivity with respect to the ASR. Specimens are placed in an alkali solution throughout the duration of the test, using concentration of alkaline mixture equivalent to 1.0 N NaOH at 80 ℃. Other characteristics, including aggregate size and their proportions, cement alkali, and water to cement ratio, need to be considered following the ASTM C1260 specifications. Due to its accelerated nature, other exposure conditions also need to be investigated for addressing the actual field conditions and variety of various materials used. Therefore, this study aimed to explore the cement alkalis (0.40%, 0.80%, and 1.20% Na₂Oe), exposure solution concentrations (0.5, 1.0, and 1.5 N), exposure temperatures (40 ℃, 80 °C and 100 °C), and testing ages (14, 28, 56, and 90 days) on the ASR expansion behavior. Furthermore, aggregates from five different sources were procured to investigate their reactivity potential. The percentage contribution to mortar bar expansion for individual and combined effects of solution concentration, cement alkali, and test duration on expansion was also investigated using full-factorial analysis. The outcome of this study will facilitate for researchers and construction related stakeholders an improved understanding of the individual and combined effects of various factors on the ASR expansion behavior.

2. Materials and Methods

Aggregates were collected from different local crusher plants (Sakhi Sarwar, Sargodha, Taxila, Jhelum, and Taunsa) (Figure 1). Afterwards, aggregates were washed with water to remove any dust or other materials. Cleaned aggregates were placed in the laboratory oven to become dried. Aggregates were then crushed and sieved to various size fractions and proportions in accordance with ASTM C1260 [73]. Sodium hydroxide (NaOH) was obtained from the local chemical industry. Commercially available OPC was used. The concrete mixture design was prepared in a ratio of 1:2.25 (cement-to-aggregate ratio by weight), and the water to cement ratio (w/c) was kept at 0.47 for the preparation of the test specimens, as per ASTM C1260 [73].

Table 3. Mixture design for casting specimens.

Cement	Aggregate	Water
646 kg/m3	1615 kg/m3	304 kg/m3

shows the mixture proportions for the prepared test specimens.

Initially, cement and aggregates were mixed in a dry state using an electric mortar mixer. Afterwards, ordinary tap water was added to the mixture and mixing resumed until a uniform mixture was achieved. Mixture was poured into the expansion mold of size 25 mm × 25 mm × 285 mm with steel studs already placed in the mold. Mortar bar expansion specimens were cast in two layers on a vibratory table. Six specimens for each source and test exposure were cast. Specimens were covered with plastic sheets to stop any kind of moisture loss until they were demolded after 24 h. After demolding, specimens were placed in water for one day. The initial reading was taken using digital length comparator, and specimens were placed in desired NaOH solution (0.5 N, 1.0 N, and 1.5 N) and required temperature (40 °C, 80 °C, and 100 °C). Length change was monitored at 3, 7, 14, 21, 28, 56, and 90 days of exposure as per ASTM C490 [74]. Figure 2 shows the mortar preparation process. Before taking the length change reading, specimens were visually examined for any type of damage or surface cracking. Figure 3 shows testing process snapshots.

3. Results and Discussions

According to ASTM C1260, if expansion on mortar bar exposed to 1.0 N NaOH solution is higher than 0.10% at 14 days and 0.20% at 28 days, aggregates can be considered as reactive. Specimens incorporating Taxila aggregates showed expansion of 0.042% and 0.090% at 14 and 28 days, respectively. Expansion for aggregates of Taunsa source was 0.087 at 14 days and 0.190% at 28 days. Expansion of 0.112% at 14 days and 0.199% at 28 days was observed in the case of specimen having Sakhi Sarwar aggregates. Specimen prepared from Sargodha’s aggregates showed expansion of 0.230% at 14 days and 0.290% at 28 days. For specimens that have Jhelum aggregates, expansion of 0.250% at 14 days and 0.320% at 28 days was observed. This indicated that the Taxila, Taunsa, and Sakhi Sarwar were considered as non-reactive aggregates, while Sargodha and Jhelum were reactive aggregates as per ASTM C1260.

Table 4. Mortar expansion results (in %) for various exposure conditions for tested aggregates.

Sources	Days	Cement Alkali (%)			Solution Concentration (N)			Temperature (℃)
		0.4	0.8	1.2	0.5	1.0	1.5	40	80	100
Taxila	14	0.053	0.061	0.069	0.037	0.042	0.048	0.037	0.042	0.048
	28	0.097	0.104	0.111	0.079	0.090	0.096	0.075	0.090	0.199
	56	0.099	0.117	0.118	0.089	0.099	0.108	0.081	0.095	0.103
	90	0.123	0.129	0.136	0.104	0.112	0.118	0.106	0.110	0.115
Taunsa	14	0.096	0.103	0.115	0.078	0.087	0.095	0.078	0.042	0.048
	28	0.193	0.203	0.213	0.175	0.190	0.199	0.176	0.190	0.199
	56	0.219	0.224	0.234	0.206	0.220	0.233	0.205	0.220	0.229
	90	0.249	0.256	0.263	0.231	0.240	0.247	0.232	0.240	0.256
Sakhi Sarwar	14	0.112	0.119	0.129	0.101	0.112	0.119	0.102	0.107	0.117
	28	0.229	0.236	0.247	0.199	0.220	0.230	0.189	0.220	0.226
	56	0.247	0.254	0.264	0.241	0.250	0.260	0.236	0.250	0.259
	90	0.287	0.298	0.309	0.264	0.275	0.286	0.261	0.275	0.285
Sargodha	14	0.243	0.259	0.271	0.201	0.230	0.242	0.219	0.230	0.239
	28	0.308	0.318	0.331	0.267	0.290	0.311	0.271	0.290	0.304
	56	0.350	0.354	0.366	0.331	0.350	0.362	0.321	0.350	0.359
	90	0.379	0.387	0.398	0.357	0.370	0.381	0.345	0.370	0.38
Jhelum	14	0.256	0.269	0.279	0.221	0.250	0.263	0.231	0.250	0.259
	28	0.328	0.334	0.349	0.299	0.320	0.339	0.291	0.320	0.339
	56	0.379	0.383	0.399	0.359	0.379	0.394	0.339	0.370	0.383
	90	0.402	0.409	0.416	0.387	0.400	0.421	0.373	0.400	0.411

shows the expansion results for various tested exposure conditions. It was also observed that expansion increased with the passage of exposure duration for all the tested aggregates. For instance, specimens incorporating Jhelum aggregates showed expansion of 0.40% at 90 days compared to 0.32% at 28 days.

3.1. Effect of Solution Concentration

It was observed that the solution concentrations (0.5 N, 1.0 N, and 1.5 N) had a significant effect on the mortar bar expansion (Figure 4). Higher expansion was observed for specimens exposed to higher concentration solution. Tested non-reactive aggregates (i.e., Taxila) exposed to 0.5 N showed expansion of 0.037% and 0.079% at 14 and 28 days, respectively, which is lower than the expansion tested for 1.0 N. In case of reactive aggregates (i.e., Jhelum), expansion was noted as 0.078% and 0.299% at 14 days and 28 days, respectively, for 0.5 N. Similarly, specimens exposed to 1.5 N showed expansion higher than 1.0 N. For instance, expansion for specimens having non-reactive aggregates (i.e., Taunsa) was observed as 0.095% at 14 days and 0.199% at 28 days for 1.5 N. Specimens incorporating reactive aggregates (i.e., Jhelum) exposed to 1.5 N showed expansion of 0.263% and 0.339% at 14 and 28 days, respectively, which is higher than the expansion values for identical specimens tested for 1.0 N solution concentration.

It was noted that solution concentrations having higher hydroxyl ions (OH-) ions were responsible for large expansion due to alkali silica reactivity. The mortar expansion for 0.5 N solution concentration was lower as compared to 1.0 N and 1.5 N NaOH due to lesser availability of hydroxyl ions (OH-). During this study, surface cracks on mortar bars incorporating reactive aggregates exposed to higher concentration solution were observed at later ages.

An increased expansion rate was observed between 14 and 28 days. Normal expansion change was evident from 28 to 56 days of exposure. After 56 days, the expansion rate was slowly changed for 1.0 N and 1.5 N solution exposure. In case of 0.5 N solution concentration, expansion rate was slowly increased with the passage of time. It was observed that specimens incorporating non-reactive aggregates (Taxila, Sakhi Sarwar, and Taunsa) showed slow or moderate expansion with exposure ages. However, specimens with reactive aggregates (Sargodha and Jhelum) exhibited accelerated expansion rates with the passage of exposure duration. Therefore, it can be argued that the rate of expansion was also dependent on aggregate mineralogical properties along with solution concentration. Similar observations have been reported in previous research. For example, Islam et al. [75] studied the role of solution concentration on expansion of accelerated mortar bar test by considering three variables i.e., 0.25 N, 0.5 N, and 1.0 N solution concentration. In the case of 0.25, it was observed that mortar expansion was very low compared to identical specimens exposed to 0.5 N and 1.0 N. It was also observed that mortar bar submerged in 0.5 N solution expanded slowly with an increase in test duration; mortar bars placed in 1.0 N solution showed higher expansion during the early age of testing and after 28 days’ expansion rate was reduced. Deboodt et al. [76] conducted an experimental study and found that reducing the normality of solution concentration NaOH to 0.5 N resulted in a lower expansion compared to identical specimens at 0.75 N. Oberholster and Davies [77] tested mortars bars by submerging specimens in 1.0 N solution concentration. It was examined that specimens in 1.0 N solution caused more expansion than 0.5 N and 1.5 N solutions.

3.2. Effect of Cement Alkalis

Figure 5 shows mortar bar expansion exposed to 1.0 N solution for various tested aggregates. It was observed that the mortar expansion was increased with an increase in cement alkalis. Specimens incorporating Sargodha aggregates showed expansion of 0.291% for exposure to 0.4 Na₂O_e, 0.299% for 0.8 Na₂O_e, and this increased to 0.311% for 1.2 Na₂O_e at 28 days for identical specimens. It can be argued that the expansion of mortar bars increased with an increase in cement alkalis due to the presence of more hydroxyl ions compared to those specimens that have lower contents of cement alkalis. Similarly, specimens incorporating Jhelum aggregates showed expansion of 0.319% for 0.4% cement alkalis which increased to 0.336% for 1.2% cement alkalis at 28 days. Specimens incorporating Taxila aggregates showed an expansion of 0.089% for 0.4%, which increased to 0.207% for 1.2% at 28 days.

It was also observed that the cement alkali contents had a significant influence for specimens exposed to 0.5 N and 1.5 N solution concentration. For example, specimens incorporating Sargodha aggregates showed an expansion of 0.308% for 0.4% cement alkali, which increased to 0.331% for 1.2% cement alkali at 28 days for 0.5 N solution concentration. In the case of 1.5 N solution exposure, there was an expansion of 0.247% for 0.4% cement alkali for Sargodha aggregates, which increased to 0.289% for 1.2% cement alkali at 28 days. However, cement alkali had relatively lesser influence for specimens exposed to 1.0 N solution on mortar bar expansion. For instance, Sargodha’s aggregates showed an expansion of 0.29% and 0.31% for 0.4% and 1.2% cement alkali contents, respectively, exposed to 1.0 N solution. Similar findings have been reported in earlier studies. For instance, Tosun and Felekoglu [78] observed the effect of total alkali content of cement on expansion caused by the ASR in mortar bars containing admixtures. Islam et al. [75] investigated the effect of cement alkalis on mortar bar expansion dependent on solution concentration NaOH and test duration. A higher rate of expansion was noted for low concentration of solution NaOH and at early test duration, while the effect was minor for higher solution concentration NaOH and at a later stage of testing. Tariq et al. [79] performed an experiment to determine the effect of cement alkali content on mortar bar expansion. It was observed that the expansion of specimens increased with an increase in cement alkali content from 0.43% to 0.58%.

3.3. Effect of Temperature

Figure 6 shows the mortar bar expansion exposed to various temperatures tested at 1.0 N solution concentration and 0.40% alkali contents. It was observed that the mortar bar expansion increased with the increase in temperature. Specimen incorporating Taxila’s aggregates showed an expansion of 0.075% at 40 °C, 0.090% at 80 °C and 0.199% at 100 °C at 28 days for 0.40% alkali content and 1.0 N solution concentration. Specimens with Sakhi Sarwar’s aggregates showed an expansion of 0.189% at 40 °C and 0.226% at 100 °C at 28 days exposed to 1.0 N solution. Similarly, reactive aggregates (i.e., Sargodha) showed an expansion of 0.271 at 40 ℃ which increased to 0.304% at 100 °C for specimens exposed to 1.0 N. It was also evident that higher exposure solution (1.5 N), tested at various temperatures (i.e., 40 °C, 80 °C and 100 °C), showed higher expansion.

Deboodt et al. [76] investigated the effect of temperature on mortar bar expansion. It was observed that decreasing the temperature below 80 ℃, caused a decrease in expansion. It was observed that the specimens exposed to 60 °C resulted in a decrease in expansion by 35% and 19% at 14 and 28 days, respectively, compared to that of the identical specimen tested at 80 °C, respectively. Stanton [1] stated that test conditions caused significant effect on expansion. Oberholster and Davies [77] performed a study to examine the effect of temperature on expansion of mortar bars. They concluded that specimens incorporating reactive aggregates showed more expansion when tested at 80 °C compared to 70 °C and 90 °C.

3.4. Factorial Analysis

Full factorial analysis (3³, three factors at three levels) was conducted to calculate the individual and combined effects of solution concentration (S), cement alkali (CA), and testing age (T) on mortar bar expansion. For each factor of S, CA, and T and their combined effects, percent contributions (P) to mortar bar expansion were calculated.

Table 5. Details of factors and their levels used for factorial analysis.

Factors	Level 1	Level 2	Level 3
S (N)	0.5	1.0	1.5
CA (%)	0.4	0.8	1.2
T (days)	14	28	90

shows the various levels of tested factors.

3.4.1. Solution Exposure and Testing Age on Mortar Bar Expansion for 0.40%, 0.80%, and 1.20% Cement Alkalis

Equation (1) shows the percent contribution of solution concentration (S), testing age (T), and their combinations for cement alkalis of 0.40%, 0.80%, and 1.20% Na₂Oe on mortar bar expansion.

$${\left(P\right)}_{S/CA}+{\left(P\right)}_{T/CA}+{\left(P\right)}_{\left(S, T\right)/CA}=100$$

(1)

where (P)_S/CA, (P)_T/CA, and (P)_(S,T)/CA are the percentage contribution to mortar bar expansion due to solution concentration, test duration, and their combined interaction for a given cement alkali, respectively. Figure 7 presents the results for each aggregate source. It was observed that solution concentration played a major contribution on mortar bar expansion, pursued by test duration and the interactions of solution concentration and test duration.

With increased cement alkali contents from 0.40% to 0.80% and 1.20%, the percent contribution to mortar bar expansion due to solution concentration showed comparable behavior for the tested Sakhi Sarwar and Sargodha aggregates. Maximum contribution was shown by Jhelum aggregates (Figure 7a). On the other side, the contribution to mortar bar expansion due to test duration was increased with an increase in cement alkalis. Sakhi Sarwar, Sargodha, and Jhelum’s source aggregates showed lesser contribution to mortar bar expansion due to test duration in comparison with specimens made with Taxila and Taunsa aggregates. Jhelum’s source aggregates showed the lowest contribution to mortar expansion for various tested cement alkalis in comparison to other tested aggregates (Figure 7b). Lower contribution to mortar bar expansion (less than 20%) due to the interaction of solution concentration and test duration was observed.

3.4.2. Testing Age and Cement Alkalis on Mortar Bar Expansion for the 0.5, 1.0, and 1.5 N NaOH

Equation 2 shows percent contribution to mortar bar expansion due to testing duration (T), cement alkali (CA), and their interactions (T, CA) for 0.5, 1.0, and 1.5 N NaOH solution.

$${\left(P\right)}_{T/S}+{\left(P\right)}_{CA/S}+{\left(P\right)}_{\left(T, CA\right)/S}=100$$

(2)

where (P)_T/S, (P)_CA/S, and (P) _(T,CA)/S are the percent contribution for the test duration, cement alkalis, and their combined effects for solution concentration, respectively, for mortar bar expansion. Figure 8 shows the results for each aggregate source.

The contribution to mortar bar expansion due to test duration was lower for the 0.5 N concentrated solution, moderate for the 1.0 N solution concentration, and higher for the 1.5 N solution concentration, as shown in (Figure 8a).

For instance, aggregates incorporating the Taxila source showed a contribution to mortar bar expansion of 69% for solution concentration for 0.5 N, 73% was noted for 1.0 N solution concentration, and 90% was observed in the case of 1.5 N solution concentration at 14 days. Other tested sources of aggregates showed similar behavior.

The contribution to mortar bar expansion due to cement alkali for solution concentration of 0.5 N was higher, moderate for 1.0 N, and lower for 1.5 N solution concentration. For instance, aggregates incorporating Jhelum source showed PC due to cement alkali to mortar bar expansion of 14% for 0.5 N solution concentration. It was decreased to 2% for the case of 1.5 N solution concentration as shown in (Figure 8b). The PC to mortar bar expansion, due to the combined effect of test duration and cement alkali (Figure 8c), was much less for all the tested aggregates sources compared to that of the testing age and cement alkali individually.

3.4.3. Solution Exposure and Alkali Content of Cement on Mortar Bar Expansion at 14, 28, and 90 Days

Equation (3) shows the percent contribution to mortar bar expansion due to solution exposure (S), cement alkali (CA), and their combined effects (S, CA) at various days, i.e., 14, 28, and 90 days.

$${\left(P\right)}_{S/T}+{\left(P\right)}_{\mathit{CA}/T}+(P){}_{(S, \mathit{CA})/T}=100$$

(3)

where, (P)_S/T, (PC)_CA/T, and (P)_(S,CA)/T are the percent contributions of solution concentration (S), cement alkali (CA), and their combinations to bar expansion for the test duration (T), respectively.

It can be posited that the percent contribution of solution concentration to mortar bar expansion was more noticeable compared to that of the cement alkali and their interaction effects (Figure 9).

The percentage contribution to mortar bar expansion of solution concentration increased with an increase in test duration for tested Taxila, Sargodha, and Jhelum aggregates. For instance, Sargodha aggregates showed the percent contribution to mortar bar expansion due to solution concentration as 75% at 14 days, which increased to 82% at 28 days and 93% at 90 days. Specimens with Taunsa and Sakhi Sarwar aggregates showed a decreased contribution with an increase in test duration. Sakhi Sarwar aggregates showed a percent contribution to mortar bar expansion due to solution concentration of 78% at 14 days, which decreased to 69% at 90 days (Figure 9a).

The percentage contribution of cement alkali to mortar bar expansion decreased with an increase in test duration for tested Taxila, Taunsa, Sargodha, and Jhelum aggregates. For instance, specimens incorporating Taunsa aggregates showed percent contribution to mortar bar expansion of 22% at 14 days and 18% at 90 days. Sakhi Sarwar aggregates showed a different trend; the percent to mortar bar expansion of 26% was noted at 14 days, which increased 30% at 90 days for Sakhi Sarwar aggregates (Figure 9b).

No significant contribution for the interaction effects was observed for the tested aggregates except for Sargodha aggregates. The percentage contribution of specimen incorporating Sargodha aggregates to mortar bar expansion due to their interaction was observed as 5% at 28 days and 12% at 90 days (Figure 9c).

3.4.4. Solution Exposure, Cement Alkali and Testing Age on Mortar Bar Expansion

Equation (4) shows the percent contribution to mortar bar expansion due to solution concentration, cement alkali, testing duration, and their two and three factor combinations.

$$\left(P\right){}_S+\left(P\right){}_T+\left(P\right){}_{\mathit{CA}}+\left(P\right){}_{S, T}+\left(P\right){}_{T, \mathit{CA}}+\left(P\right){}_{S, \mathit{CA}}+\left(P\right){}_{S, T, \mathit{CA}}=100$$

(4)

where (P)_S, (P)_T, (P)_CA, and (P)_S, _T, _CA are the percent contributions of solution concentration, test duration, cement alkali, and their combination to mortar bar expansion, respectively. (P)_{S, T}, (P)_{T, CA}, and (P)_{S, CA} are the percent contributions of the combined effects of solution concentration and testing age, testing age and cement alkali, and solution concentration and cement alkali to bar expansion, respectively.

It can be observed that for the tested aggregates sources, the solution concentration had a major contribution to mortar bar expansion (from 40% to 78% for various tested aggregates), followed by test duration (from 14% to 37%). It was noted that the percent contribution of solution concentration for the tested Jhelum aggregates was maximum and minimum for Taxila aggregates (Figure 10). It can be argued that the effect of solution concentration on mortar expansion was less for non-reactive aggregates (i.e., Taxila) compared to reactive aggregates (Sargodha and Jhelum). The contribution for the interaction of solution concentration and the test duration ranged from 1% to 17% for various tested aggregates (Figure 10). The two-factor combinations, i.e., (S, CA) and (T, CA), and the three factor combinations (S, T, CA) showed an insignificant contribution (<1%) to mortar bar expansion for each aggregate sources.

It can be argued that despite the aggregates’ mineralogy, solution concentration exposure had a major effect on mortar bar expansion, followed by the testing age. The percentage contributions to mortar bar expansion due to cement alkali and the combined effects of solution concentration and testing age depend on the type of aggregates used. The added value of this research is that is investigates the effect of various factors (such as different concentrations of solutions, dosages of cement alkali, and different temperatures) on the ASR potential of reactive and non-reactive aggregates. In past studies, the individual effects of each factor were studied rather than their combined effects. For instance, Pathirage et al. [80] investigated the effect of alkali–silica reaction on the mechanical properties of mortar bars by considering two dosages of cement alkali contents. Ghafoori et al. [81] studied the effect of the ASR reactivity of mortar bars submerged in three different dosages of solution concentrations. Bavasso et al. [82] evaluated the potential of alkali–silica reactivity of aggregates by using solution concentrations and temperatures. This present study was mainly focused on their combined effects.

Factorial analysis was also conducted to evaluate the percentage contribution of the individual test parameters as well as their combined effects. In this study, it was observed that exposure solution had a major contribution towards the expansion of mortar bars. The main concern of this research is to highlight the contribution of various exposure conditions on the ASR expansion, which has a significant role in selecting the aggregate sources for different applications.

4. Conclusions

This study explored the effects of cement alkalis (0.40%, 0.80%, and 1.20% Na₂Oe), exposure solutions (0.5, 1.0, and 1.5 N), exposure temperatures (40 °C, 80 °C and 100 °C), and duration (14, 28, 56 and 90 days) on the expansion behavior of various tested aggregates. The following conclusions can be drawn from the present study.

The solution concentration played a major role in the mortar bar expansion. Solution concentration was responsible for an increase in the expansion of mortar bars. For example, specimens incorporating Taxila aggregates exhibited an expansion of 0.079% at 28 days for 0.5 N solution, which increased to 0.090% for identical specimens tested at 1 N solution and reached 0.096% for 1.5 N solution. Similarly, Sargodha aggregates showed an expansion of 0.267% at 28 days for the 0.5 N solution, which increased to 0.290% for similar specimens tested at 1.0 N solution and increased to 0.311% for the 1.5 N solution. This increased expansion due to higher solution concentration was attributed to the higher level of OH⁻ ions leading to increased ASR gel formation. It was also observed that the expansion rate was not regular (i.e., it did not follow a similar trend) for all the tested aggregate sources during the testing period. Three chosen aggregates exhibited the reactive nature (i.e., Sargodha, Jhelum, and Taunsa) as indicated by their expansion (more than 0.10% and 0.20% at 14 and 28 days, respectively, in accordance with ASTM C 1260). The remaining two sources of aggregates were considered non-reactive aggregates.

Specimens containing higher cement alkali contents caused higher expansion compared to specimens having low cement alkali contents. For instance, specimens incorporating Jhelum aggregates showed an expansion of 0.319% at 28 days for 0.40% cement alkali at 1.0 N solution, which increased to 0.324% for 0.80% cement alkali, and then further increased to 0.336% for 1.20% cement alkali for similar specimens tested at 28 days. This increase in expansion due to higher cement alkali was caused due to increased OH⁻ ions entering through the pores, which leads to increased formation of the ASR gel. It was observed that increasing the temperature from 40 °C to 100 °C increased the expansion for all the tested aggregates. For instance, specimens incorporating Taunsa aggregates incurred 28-days expansion of 0.176% at 40 °C, which increased to 0.190% at 80 °C, and further increased to 0.199% for identical specimens tested at 100 °C in 1.0 N solution. It can be argued that increasing the temperature of exposure promotes the production of the ASR gel and the associated expansion.

In this research, factorial analysis was conducted to evaluate the percentage contribution of the individual test parameters as well as their combined effects. It was observed that among the tested variables, the solution concentration was the primary governing factor that contributed towards mortar bar expansion for all tested aggregates, followed by the testing age and cement alkali content. No significant contributions of the combined parameters and their interaction effects were observed.