Mechanical and shrinkage behavior of magnesium oxide based concrete with recycled aggregates

ABSTRACT

This study investigates the combined effect of reactive Magnesium Oxide (MgO) and Recycled Aggregates (RA) on the shrinkage and mechanical behavior of High-Performance Concrete (HPC), with a focus on sustainable and eco-concrete development. Unlike previous studies that typically examined MgO or RA in isolation, this work uniquely explores their interaction across varying RA levels (0%, 25%, 100%) and two curing periods (7 days and 6 months). Ten concrete mixes were prepared, with 10% Ordinary Portland Cement (OPC) replaced by MgO in half of them. The findings reveal that MgO reduced autogenous shrinkage by up to 93% due to Mg(OH)² formation, while overall shrinkage was lowered by 22–39% depending on RA content. However, MgO also increased drying shrinkage due to greater water demand, especially at higher RA levels. A key novelty is the identification of threshold RA contents—35% at early age and 42% at maturity—beyond which the shrinkage contribution from RA offset MgO’s benefits. Mechanically, MgO caused a modest 5–8% reduction in strength properties. However, at 100% RA, the combination with MgO led to up to 21% strength loss, attributed to poor aggregate–matrix interaction. This study provides new insights into the synergistic and competing effects of MgO and RA in HPC, helping define optimal material combinations for shrinkage control without severe compromise in strength. The results contribute valuable data for sustainable concrete design involving both recycled materials and alternative cementitious components.

Keywords:
Recycled aggregates; Magnesium Oxide; Mechanical Properties; Shrinkage; ANOVA; Sustainability; Eco-Concrete

1. INTRODUCTION

Concrete remains the most widely used construction material globally due to its versatility, strength, and durability. However, the increasing demand for high-performance applications, coupled with growing environmental concerns, has led to the exploration of more sustainable practices. These include utilizing recycled aggregates (RA) to reduce natural resource consumption and incorporating supplementary materials like reactive magnesium oxide (MgO) to address issues such as shrinkage and cracking. This study investigates the combined impact of RA and MgO on the performance and durability of high-performance concrete (HPC), with a particular focus on shrinkage behavior and material interactions.

1.1. Shrinkage in high-performance concrete

The shrinkage phenomenon refers to the inherent volumetric reduction of concrete, especially during its early stages, occurring independently of external forces [¹]. This reduction can lead to cracking, which significantly compromises durability by allowing aggressive agents to penetrate and accelerate deterioration and reinforcement corrosion [²]. Shrinkage is a complex phenomenon triggered by several mechanisms. Plastic shrinkage results from water loss as concrete transitions from a fresh to hardened state and typically concludes around 24 hours after casting, known as “time zero” [³]. Autogenous shrinkage follows, caused by capillary water absorption during delayed hydration, while drying shrinkage results from water evaporation [⁴]. Autogenous shrinkage is typically measured under sealed conditions, whereas total shrinkage is observed under environmental exposure [⁵]. High-Performance Concrete (HPC), characterized by its dense microstructure and high strength due to elevated cement content, is especially prone to shrinkage. The intense heat generated during hydration and reduced porosity increase susceptibility to cracking [⁶]. Therefore, every modification to HPC composition must consider its impact on shrinkage.

1.2. Recycled aggregates in HPC and the role of RA maturity

In recent years, Recycled Aggregates (RA) have been widely studied as substitutes for Natural Aggregates (NA) in HPC. Despite the concerns of reduced stiffness and higher water absorption, studies indicate that HPC exhibits a relatively smaller reduction in mechanical performance when incorporating RA than conventional concrete [⁷]. For instance, a 100% replacement of coarse NA with RA can reduce strength by 15–20% in HPC, compared to 40% in normal concrete [⁸]. Shrinkage behavior, however, is more significantly impacted. RA can decrease autogenous shrinkage due to its internal curing ability—releasing absorbed water to sustain hydration without affecting workability [⁹]. Yet, this internal water release can also increase drying shrinkage through evaporation. An important yet less explored aspect is the maturity of RA, defined by the curing time of Parent Concrete (PC) before it is crushed. RA produced from early-age PC (e.g., 7 days) retains higher shrinkage potential than RA from older PC (e.g., 6 months), due to ongoing hydration and mechanical instability. This maturity influences both mechanical strength and shrinkage behavior in HPC mixes. Studies have demonstrated a near-linear relationship between the proportion of RA and shrinkage in HPC, regardless of RA maturity. However, the combined effects of RA maturity and its interaction with supplementary materials like MgO remain underexplored.

1.3. Reactive MgO: potential and environmental relevance

Magnesium Oxide (MgO) has emerged as a promising additive for shrinkage compensation in cementitious materials. When light-burned at relatively low temperatures (700–1000°C), reactive MgO rapidly hydrates to form brucite (Mg(OH)₂), which has a higher molar volume and thus induces expansion [¹⁰]. This expansion, if timed correctly, can counteract early-age shrinkage without damaging the matrix [¹¹]. Although traditionally limited to 5% in cement (EN 197-1, 2011) due to concerns about late expansion from hard-burned MgO [¹²,¹³], recent research supports the use of up to 10–20% reactive MgO for shrinkage control in mortars [¹⁴]. At 10% replacement, shrinkage reductions of 18–49% have been recorded depending on MgO fineness. The environmental advantage is significant: producing MgO emits roughly 30% less CO₂ than cement clinker production, making it a more sustainable alternative [¹⁵]. When used in mortars with RA, MgO can reduce total shrinkage by 30–60%, though some strength reduction (10–20%) may occur [¹⁶,¹⁷].

1.4. Research gap and study objective

Although studies have explored the use of RA and MgO individually in cementitious systems, comprehensive investigations into their combined effects—particularly in High-Performance Concrete—are scarce. This study addresses that gap. The novelty of this research lies in its systematic evaluation of 10% reactive MgO as a partial cement replacement in HPC incorporating RA of different maturities. It uniquely assesses mechanical properties and shrinkage behaviors (autogenous, drying, and total shrinkage) across ten HPC mixes with varying RA levels (0%, 25%, 100%) and two RA maturities (7 days and 6 months). Additionally, it explores the synergy between RA maturity and MgO reactivity, contributing new insights into sustainable HPC design. By integrating sustainability benefits and focusing on performance metrics, this work advances the use of RA and MgO in high-strength concrete applications while addressing critical durability challenges.

2. MATERIALS AND METHODS

2.1. Materials used

As mentioned in the introduction, MgO and RA were included in the mixtures. Ordinary Portland cement (OPC) and NA were used to finish the composition of HPC.

2.1.1. Water and binder admixtures

All mixtures were made using OPC and CEM I 42.5 R, which comply with European standards. These mixtures’ approximate imaginary and real densities were equal to 0.9 and 2.8 mg/m³, respectively. According to modern standards (EN 197-1), this type of cement contains approximately 98 % clinker. On the other hand, in half of the mixtures, reactive MgO produced by one of the Indian suppliers, in the amount of 10% of total OPC weight, was added. Using the X-ray fluorescence (XRF) analysis, the purity of reactive MgO was 96%, as shown in Table 1. Additionally, it demonstrated a high degree of fineness in its surface area and size gradation, enabling it to successfully hydrate and, thus, achieve its maximal strength and expansion when combined with water. Lastly, compared to OPCs, it has a lower bulk density and a greater actual density. The investigation was conducted in Coimbatore, India, and all mixtures included drinking water from the city’s main water supply. In addition, a polycarboxylate-based superplasticizer (Sika ViscoCrete) was used primarily to provide high-workability concrete.

Table 1
Chemical composition and properties of MgO.

CHEMICAL COMPOSITION OF MgO
TYPE OF COMPOSITION	MgO	CaO	SiO2	Fe2O3	Al2O3
OBSERVED VALUE (%)	96.74	1.42	1.18	0.16	0.27
PROPERTIES OF MgO
TYPE OF PROPERTY	Bulk density (mg/m3)	BET surface area (m2/g)	Real density (mg/m3)	Particle size < 75µm (%)	Particle size < 45µm (%)
OBSERVED VALUE	0.48	61.18	3.62	99.98	94.92

2.1.2. Recycled aggregates

Before manufacturing, the parent concrete’s (PC) composition and slump were determined (100% NA, Table 2). The classification of PC as belonging to the C30/37 class is based on the mechanical characteristics detailed in Table 3, as per Eurocode 2. A notable rapid growth in strength was observed in the PC, with its compressive strength at 7 days reaching 87% of the value recorded at 28 days. After its manufacturing, the HPC underwent air curing for two different durations: seven days, during which concrete shrinkage remained quite considerable, and six months, at which point the shrinkage could be disregarded. These two materials allowed for the investigation of how the reduction of RA influenced the contraction of HPC. Two varieties of RA were produced by crushing PC in a jaw crusher after each curing period: early-age (air-cured for seven days) and matured (air-cured for six months). Following crushing, the RA (0–31.5 mm) was sieved to separate the fine (0–4 mm) and coarse (4–22 mm) fractions, as shown in Figure 1(a). The findings obtained would not be affected if the fraction of coarse (>4 mm) and fine (<4 mm) aggregate was maintained consistent throughout all blends. The primary physical characteristics of RA are listed in Table 2. Based on the observations, the absorption of water rates was found to be higher in coarse and fine fractions than natural aggregates (NA), and the density was 2.38 mg/m³. Additionally, the continuous gradation depicted (Refer to Figure 1(b)) makes these materials appropriate for High-Performance Concrete (HPC) production.

Table 2
Composition and mechanical properties of HPC in kg/m³.

COMPOSITION OF HPC
TYPE OF COMPOSITION	CEM I 52.5	Limestone filler	Water	CG	MG	CS	FS	Water – Binder ratio	Slump (cm)
OBSERVED VALUE	250	80	150	710	405	405	500	0.45	5.1 ± 0.3
MECHANICAL PROPERTIES OF HPC
TYPE OF PROPERTY	Compressive strength (7 days)			Compressive strength (28 days)		Split tensile strength (28 days)		Modulus of elasticity (28 days)
OBSERVED VALUE	40.28 ± 0.48			47.28 ± 0.38		2.82 ± 0.31		38.21 ± 0.24

Table 3
Aggregate density and water absorption as per EN 1097 – 6.

TYPE OF THE AGGREGATE	SIZE OF THE AGGREGATE IN FRACTION (mm)	DRY DENSITY – SATURATED (mg/m3)	DRY DENSITY – OVEN (mg/m3)	APPARENT DENSITY (mg/m3)	WATER ABSORPTION IN 24 HOURS (%)	WATER ABSORPTION IN 10 MINUTES (%)
CG	12–22	2.62	2.58	2.68	1.30	0.90
MG	4–12	2.64	2.61	2.69	1.15	0.88
CS	1–4	2.56	2.54	2.60	0.79	0.69
FS	0–1	2.61	2.60	2.64	0.59	0.42
Coarse RA	4–22	2.42	2.41	2.46	4.92	3.28
Fine RA	0–4	2.37	2.34	2.39	6.82	4.93

Figure 1
(a) Size distribution of aggregates and (b) global gradation of mixes.

2.2. Design of concrete mix

The reference mix RHPC was prepared for the beginning of the mix design process. This work aimed to design an HPC with a good workability slump of 170 ± 10 mm categorized under S4 as per EN 206 2013, and to ensure optimal performance in its fresh and hardened forms [¹⁸]. This guaranteed the concrete was strong and easy to pour, making it appropriate for structural application. The 400 kg/m³ density, a greater than typical OPC content, was used. To maintain clarity in assessing the effects of RA or MgO, mineral additions prevalent in HPC were omitted from consideration to prevent any potential interactions. The Eurocode 2 2010 standards described the ratio of cement to aggregate put into the mix [¹⁹]. Using a maximum aggregate size of 22 mm, the size fractions of the four NA specified in section 2.1.2 were adjusted using least squares to reference Faury’s curve to calculate their proportions [²⁰]. Therefore, the mix’s total aggregate volume was established to contain 45% fine aggregate, which included 20% crushed stone, 25% fly ash, and 55% coarse aggregate, comprised of 35% coarse gravel and 20% medium gravel. This allowed the aggregates to be packed well. The investigation concluded that the water-to-binder (w/b) ratio was low, precisely at 0.45. This conclusion was based on the water absorption recorded after a 10-minute mixing interval (Ref. Sec. 2.3) and the moisture of the natural aggregate (NA), which was maintained in the laboratory for the duration of the experiment. The water absorption capacity of the recycled aggregates was determined following ASTM C127 for coarse RA and ASTM C128 for fine RA, using a 10-minute partial immersion method to better represent in-mix absorption behavior. These values were used to calculate the necessary water adjustments for each mix. A high strength was achieved with this low w/b ratio, although workability was decreased. A two-stage mixing technique and a superplasticizer at a proportion of 2% cement mass were used to solve the final issue (section 2.3). Following the design of the reference mix, mixes, including RA of various maturities, were created. As a result, both developed and early-age RA were used to replace 25% and 100% of the coarse and fine NA, respectively. To mitigate the influence of water content on the results, the water content in each mix was carefully modified to preserve constant workability, aiming for a slump of 175 ± 15 mm. The quantity of extra water for each RA content was experimentally calculated based on the RA’s inherent moisture content and water absorption after 10 minutes. EA25 and EA100 (early-age RA, labelled EA) and MA25 and MA100 (matured RA, labelled MA) were the names given to the generated mixes. In the end, five more mixes were defined, each having the same composition as the others, except that 10% of the OPC by weight was substituted with MgO. This MgO level was selected since it has been shown to provide an ideal compromise between the reduction of shrinkage and the strength loss that it produces in earlier structural mortar studies. In addition to examining the impact of MgO, these mixtures allowed for a detailed investigation of its relationship with RA of various maturities. Around five times, the water content was increased with a density of 15 kg/m³ to achieve the desired slump of 175 ± 15 mm. The mixtures with MgO (MO) were designated RMO, MA25MO, MA100MO, EA25MO, and EA100MO. Figure 1(b) displays the joint gradation of the 10 HPC mixes that were created, and Table 4 compiles their composition. The aggregates are correctly packed. Figure 2 displays the time scale for each mix’s manufacturing moment.

Table 4
Mix design in kg/m³ with HPC.

TYPE OF MATERIAL	RA	MA25	EA25	MA100	EA100	RMO	MA25MO	EA25MO	MA100MO	EA100MO
Cement	400	400	400	400	400	360	360	360	360	360
MgO	0	0	0	0	0	40	40	40	40	40
Water	150	160	160	195	195	165	175	175	210	210
CG	690	520	520	0	0	690	520	520	0	0
MG	390	295	295	0	0	390	295	295	0	0
CS	380	285	285	0	0	380	285	285	0	0
FS	470	355	355	0	0	470	355	350	0	0
Coarse RA	0	245	245	980	980	0	245	245	980	980
Fine RA	0	190	190	760	760	0	190	190	760	760
Superplasticizer	8	8	8	8	8	8	8	8	8	8

Figure 2
Experimental plan.

2.3. Experimental plan

The mixing procedure was carried out in two steps to optimize the cement’s hydration and aggregates’ water absorption and, therefore, the HPC’s workability. Utilizing a vertical-axis concrete mixer, all aggregates were combined with 70% of the mixing water, and the mixture underwent stirring for four minutes. The remaining water containing the superplasticizer and the binders (OPC and MgO) was added after the mixer was turned off for two minutes. The fresh-state tests were conducted after four more minutes of mixing the HPC: According to the Abramscone test in EN 12350-2, 2020 [²¹], two slump measurements were made at one-minute intervals. Two recent density measurements were performed in compliance with EN 12350-6, 2020 [²²]. Following this, the specimens necessary for the various tests in the hardened state were assembled. Table 5 lists these tests, the type of material utilized in each instance, and the testing age. Since each test was conducted on two specimens, the arithmetic means of the data acquired for each specimen served as the final result. Two conditions were used to keep the generated specimens until the testing point: The specimens were generally maintained in a wet chamber (95 ± 5% humidity and 23 ± 5°C temperature). The specimens that assessed shrinkage and hardened density were dry-stored in a room with 40 ± 6% humidity and 23 ± 5°C temperature. The experimental strategy followed and the time each mix was created is detailed in the flowchart in Figure 2.

Table 5
Tests performed for hardened state.

S. No.	TYPE OF TEST	SIZE OF THE SPECIMEN (cm3)	SHAPE OF THE SPECIMEN	CURING PERIOD IN DAYS
1.	Harden density	10 × 10 × 10	Cubic	7, 28
2.	Compressive strength	15 × 15 × 15	Cubic	7, 28, 91*
3.	Ultrasonic pulse velocity	15 × 15 × 15	Cubic	7, 28
4.	Modulus of elasticity	15 × 30	Cylindrical	28
5.	Split tensile strength	15 × 30	Cylindrical	28
6.	Autogenous shrinkage	10 × 10 × 50	Prismatic	From 1 to 91
7.	Total shrinkage	10 × 10 × 50	Prismatic	From 1 to 91

• ^*Only for EA mixes for early-age RA studies.

2.3.1. Shrinkage tests

Despite the specimens’ varying circumstances, the autogenous and total shrinkage was assessed in this study using a digital comparator and by LNEC-E398 [²³]. To ascertain the overall shrinkage, specimens that interacted with the controlled room temperature were exposed to evaporation. The current protocol was expanded to assess autogenous shrinkage in the near term (up to two or three days after mixing). The aluminium foil tape was used to cover the specimens to avoid the water evaporation. However, as the introduction outlines, “time zero” represents the instant when drying and autogenous shrinkage begin while plastic shrinkage ends. This “time zero” occurs in high-performance cement pastes about 24 ± 0.5 hours after the mixing procedure is finished. As a result, measurements of both kinds of shrinkage began at that point. The number of weekly measurements was consistent with the amount of shrinkage anticipated for each concrete age. Until the concrete had aged 91 days, the shrinkage was tracked daily with three observations, and finally, the testing stage was stopped. The termination age for the shrinkage tests was determined according to the recommendations of previous analogous studies.

3. RESULTS AND DISCUSSIONS

3.1. Properties of fresh concrete mix

3.1.1. Workability

Table 6 demonstrates that every mix had a slump within the first set of values, 175 ± 15 mm. As a result, according to EN 206, 2013, all combinations might be categorized as S4 slump class. It may be concluded from the data that workability was consistent across all combinations. This was made feasible by adding RA and MgO, which increased the effective w/b ratio. To counterbalance the more excellent water absorption exhibited by Recycled Aggregate (RA) relative to Natural Aggregate (NA), the water-to-binder ratio was elevated. This adjustment also accounts for the increased friction between the components of the mix due to the irregular geometry of RA²⁴. Early-age RA’s less angular form, which results from PC’s decreased crushing strength, caused the droop to rise by around 1%. Table 1 represents the data indicating that MgO possesses a higher specific surface area than Ordinary Portland Cement (OPC), as per EN 197 – 1, 2011 standard, even though both binders exhibit a comparable size distribution. Consequently, incorporating MgO into 100%-OPC mixtures increased 0.035 units in the water-to-binder (w/b) ratio. Moreover, the somewhat sharper form of MgO can be harmful. Finally, when RA was incorporated into a blend containing MgO as a replacement for cement, it showed a decrease in workability (Table 6). The usage of 100% RA made this behaviour more noticeable.

Table 6
Concrete mixes w/b ratio and slump workability.

TYPE OF MIX	SLUMP (cm)	w/b RATIO	ΔRA*	ΔRA# MATURITY	ΔMgO$
RA	18.28 ± 0.31	0.37	–	–	–
MA25	17.82 ± 0.57	0.40	-2.38	–	–
EA25	18.08 ± 0.18	0.40	-1.18	+1.28	–
MA100	17.52 ± 0.15	0.48	-6.67	–	–
EA100	17.82 ± 0.28	0.48	-5.48	+0.89	–
RMO	17.67 ± 0.19	0.40	–	–	-3.72
MA25MO	17.28 ± 0.32	0.43	-2.37	–	-3.86
EA25MO	17.81 ± 0.23	0.43	-2.92	-0.59	-5.57
MA100MO	16.82 ± 0.17	0.52	- 9.16	–	-6.58
EA100MO	16.38 ± 0.09	0.52	-8.37	+0.89	-6.62

• *% loss in a slump due to RA addition
• ^#% variation due to early age RA and
• ^$% decrease due to MgO.

3.1.2. Density of fresh concrete

Figure 3 shows the fresh density of the HPC mix. A linear decline in fresh density is noted with the rising RA content due to the comparatively lower density of RA than that of NA, in addition to the increase in the water-to-binder ratio and air content. However, developed RA had a longer curing duration than early-age RA, so they suffered more considerable water evaporation during air curing. However, the resultant fresh density was unaffected by this mass loss because it was negligible compared to the combined mass of all HPC components [²⁴]. It is essential to mention that, despite the actual density of MgO being more excellent, the initial density of the mixtures with MgO was less than that of the mixtures composed entirely of 100% OPC. This is explained by adding this binder, which requires a more extensive water content to maintain consistent workability. Additionally, the MgO has many irregular shapes and surface areas, which may increase the air content of HPC [²⁵]. About the blends consisting exclusively of 100% Ordinary Portland Cement (OPC), all mixtures experienced a density reduction of approximately 0.02 Mg/m³. Regardless of the mix’s RA concentration or RA maturity, MgO always had the same effect.

Figure 3
High-performance concrete and its fresh density.

3.2. Hardened properties

3.2.1. Density of concrete

Figure 4 displays the mixtures’ hardened densities of 7 and 28 days. The 2.15 to 2.45 Mg/m³ range was typical for recycled aggregate concrete. As the fresh density shows, the hardened density was reduced when 10% OPC was substituted with MgO. This situation is attributed to workability, which is less dense than the other concrete components and requires a higher water content to be effectively maintained. However, adding MgO also makes concrete more porous, which counteracts the rise in microstructural density brought on by MgO’s greater actual density.

Figure 4
High-performance concrete and its hardening density (a) 7 days and (b) 28 days.

3.2.2. Compressive strength

All mixtures’ compressive strengths at 7 and 28 days are displayed in Table 7 and Figure 5. RA maturity, RA content, and MgO addition were the three elements that were examined, and they all impacted the compressive strength achieved and how it changed over time. Figure 5(a), (b) shows the compressive strength exhibited a linear decrease with the increasing concentration of Recycled Aggregates (RA). Due to the formation of substandard Interfacial Transition Zones (ITZ) between the cementitious matrix and the larger RA particles, as well as the presence of mortar particles within the finer fraction, this reduction may be attributed. A more significant loss of compressive strength was caused by early-age RA’s decreased strength and stiffness [²⁶]. When MgO was added at 10% by weight, HPC’s compressive strength decreased from 8% to 20% (Table 7). The occurrence of this event can be attributed to the hydration of reactive MgO, which produced a magnesium hydroxide (Mg(OH)₂) that was weaker than the calcium-silicate-hydrates (C-S-H) formed through the hydration of Ordinary Portland Cement [²⁷]. Additionally, the cement became more diluted and lost strength due to the increased water content caused by this binder [²⁸]. The strength decrease observed in HPC with 100% NA was comparable to that found in related studies. Table 7 readings show a more significant decrease in compressive strength with adding MgO. Adding MgO into the RHPC mix caused a reduction in compressive strength by 8.2%, resulting in the formation of the RMO mix. In comparison, adding MgO to the MA100 and EA100 mixes resulted in a compressive strength reduction of 15% to 18%, culminating in the MA100MO and EA100MO mixes, respectively. This interaction was clarified by the heightened porosity produced through the concurrent use of both materials, a result that has been validated with different alternative binders [²⁹]. It was comparable to the interaction seen with a slump (section 3.1.1). Finally, the strength’s temporal progression varied because each HPC mix had a distinct composition. The compressive strength of High-Performance Concrete (HPC) containing early-age recycled aggregates (RA) at 91 days was comparable to that of matured-RA HPC at 28 days. The internal curing caused by the RA, due to their excellent water absorption, improved the delayed hydration of the binders, resulting in a significant strength increase from 7 to 28 days. Also, at an early age, RA’s strength and stiffness increased over time. Adding Magnesium Oxide (MgO) led to a more significant temporal improvement in strength, as shown by compressive strength development over time. This occurrence can be linked to the gradual strength growth of MgO compared to Ordinary Portland Cement (OPC), akin to the patterns seen with other substitute binders like fly ash.

Table 7
Compressive strength (CS) and split tensile strength (ST) of HPC and its factors.

TYPE OF MIX	COMPRESSIVE STRENGTH (MPa)							SPLIT TENSILE STRENGTH (MPa)
	CS AT 7 DAYS	CS AT 28 DAYS	ΔRA* %	ΔRA# MATURITY	ΔMgO$ %	Δ7–28@ DAYS %	CS 91 DAYS	ST AT 28 DAYS	ΔRA* %	ΔRA# MATURITY	ΔMgO$ %
RA	69.2 ± 0.21	73.2 ± 1.92	–	–	–	96.2	–	3.48 ± 0.14	–	–	–
MA25	65.8 ± 0.17	69.8 ± 0.45	-4.98	–	–	95.4	–	3.08 ± 0.09	-11.9	–	–
EA25	62.8 ± 2.84	66.7 ± 1.12	-8.81	-3.97	–	95.1	70.8 ± 1.1	3.01 ± 0.10	-14.4	-2.72	–
MA100	56.7 ± 3.08	60.2 ± 0.31	-16.37	–	–	91.4	–	2.59 ± 0.07	-26.5	–	–
EA100	50.3 ± 2.28	56.1 ± 0.61	-21.26	-6.47	–	90.6	60.5 ± 0.7	2.48 ± 0.11	-29.8	-3.61	–
RMO	61.2 ± 1.29	67.2 ± 0.83	–	–	-8.14	91.2	–	3.27 ± 0.08	–	–	-4.9
MA25MO	56.7 ± 1.72	62.4 ± 1.21	-7.38	–	-10.62	89.7	–	2.76 ± 0.09	-15.8	–	-8.4
EA25MO	51.8 ± 1.27	59.8 ± 0.71	-11.41	-3.96	-10.71	88.2	61.2 ± 1.8	2.49 ± 0.14	-22.9	-8.9	-15.2
MA100MO	43.9 ± 0.81	51.2 ± 0.96	-22.89	–	-15.28	87.6	–	2.09 ± 0.15	-36.7	–	-18.7
EA100MO	41.08 ± 1.11	45.6 ± 1.12	-30.45	-10.34	-18.93	87.1	49.8 ± 0.5	2.01 ± 0.06	-40.8	-6.8	-19.6

• *Decrease in CS due to RA use
• #Decrease in % due to early age RA use
• $% decrease due to MgO and
• @28 days CS developed from 7 days in %.

Figure 5
High-performance concrete and its compressive strength (a) 7 days and (b) 28 days.

3.2.3. Splitting tensile strength

Figure 6 illustrates the various blends’ split tensile strength after 28 days. Table 7 shows all lowered strength values of three different analyses. i.e., RA, maturity RA, and MgO addition. RA’s introduction weakened the bond between the cementitious matrix and the aggregates, causing a decline in the splitting tensile strength of High-Performance Concrete (HPC). Figure 6 shows the constant reduction of early-age RA’s split tensile strength. An analysis of the linear trend indicates that the decrease in splitting tensile strength associated with 25% recycled aggregate was greater than anticipated, irrespective of the age of the recycled aggregate. This suggests that the negative impact of RA on splitting tensile strength is often evident. The MgO was replaced with 10% OPC, resulting in a 5–21% reduction in splitting tensile strength. These reductions were quite comparable to those seen for compressive strength. The decline in splitting tensile strength is influenced by multiple factors, including the relative strength of Mg(OH)², greater than that of C-S-H, and the increased porosity due to adding MgO. The dilution effect on Ordinary Portland Cement (OPC) results from the elevated effective water-to-binder ratio when MgO is incorporated [³⁰]. The tensile strength of OPC is typically in the range of 2 to 5 MPa, while MgO has a tensile strength of about 1 MPa. Table 7 illustrates that incorporating MgO negatively influenced and intensified the adverse effects on the splitting tensile strength associated with recycled aggregate content and its age. Although this phenomenon is considered less significant, it was also observed in compressive strength, likely due to the increased reliance on splitting tensile strength on concrete porosity, which was elevated by the interaction of MgO and RA [³¹].

Figure 6
High-performance concrete and its split tensile strength at 28 days.

3.2.4. Elasticity modulus

Because RA is less rigid than NA, it lowers the elasticity modulus of HPC. Furthermore, this decline is more pronounced if RA is derived from concrete of worse quality (lower strength and stiffness). The modulus of elasticity values for all combinations at 28 days are displayed in Figure 7. As anticipated, the outcomes appeared. Consequently, when the RA concentration increased, the modulus of elasticity dropped. It was also decreased by the employment of early-age RA, which was acquired when PC was less rigid, albeit this impact was only seen when its content was 100%. The modulus of elasticity of Mg(OH)₂ is comparable to that of C-S-H. Accordingly, recent research suggests that, provided no additional changes to the mix’s composition are made, replacing OPC with MgO can preserve the cement-based materials’ modulus of elasticity. However, adding MgO necessitated a larger water content to preserve workability, increasing HPC’s porosity. Thus, when MgO was added, the modulus of elasticity of the mixes in this investigation dropped by around 4–9%, as Table 8 illustrates. Nevertheless, this decline was less than the splitting tensile and compressive strength results.

Figure 7
High-performance concrete and its elasticity modulus at 28 days.

Table 8
HPC mixes modulus of elasticity, UPV and their factors.

TYPE OF MIX	MODULUS OF ELASTICITY (GPa)				ULTRASONIC PULSE VELOCITY (m/s)
	MODULUS OF ELASTICITY	ΔRA* %	ΔRA# MATURITY	ΔMgO$ %	UPV 7 DAYS	UPV 28 DAYS	ΔRA* %	ΔRA# MATURITY	ΔMgO$ %	Δ7–28@ DAYS %
RA	43.2 ± 0.48	–	–	–	5067 ± 28.6	5318 ± 31.2	–	–	–	98.2
MA25	41.5 ± 0.35	-3.27	–	–	5089 ± 29.2	5192 ± 30.8	-1.18	–	–	98.7
EA25	41.8 ± 0.35	-4.21	-0.86	–	4876 ± 66.4	5176 ± 11.2	-3.62	-2.36	–	97.2
MA100	31.8 ± 1.12	-30.81	–	–	4286 ± 48.7	4618 ± 71.5	-15.23	–	–	96.4
EA100	28.5 ± 0.18	-31.45	-1.27	–	4198 ± 51.1	4509 ± 49.6	-17.31	-2.18	–	95.6
RMO	40.2 ± 0.41	–	–	-4.91	4915 ± 53.9	5208 ± 58.3	–	–	-5.12	96.7
MA25MO	36.7 ± 0.19	-9.82	–	-10.89	4839 ± 49.8	5113 ± 62.7	-1.52	–	-5.57	94.5
EA25MO	38.2 ± 0.57	-7.11	+3.62	-7.52	4871 ± 41.9	5081 ± 71.4	-2.37	+0.78	-3.94	95.6
MA100MO	28.6 ± 0.32	-29.83	–	-3.98	4118 ± 38.9	4329 ± 28.4	-16.42	–	-6.37	92.8
EA100MO	26.9 ± 0.72	-32.27	-3.72	-6.34	4089 ± 27.4	4246 ± 28.2	-17.21	-1.27	-5.51	92.1

• ^*Decrease in % due to RA use
• ^#Decrease in % due to early age RA use and
• ^$% decrease due to MgO.

Additionally, in contrast to those strengths, Table 8 relative reductions reveal that independent of RA concentration, the impact of MgO was roughly the same in all combinations. Since microstructural density significantly impacts HPC’s modulus of elasticity, a lower and more consistent decrease in this mechanical characteristic was the outcome of its rise owing to MgO’s greater actual density than OPC’s [³²].

3.2.5. Ultrasonic pulse velocity (UPV)

In Figure 8(a), (b), the UPV of the mixes generated at 7 and 28 days is measured in 15 × 15 × 15 cm cubic specimens subjected to compressive strength testing. A high-stiffness concrete like HPC is representative of all the values measured, which range from 4000 to 5400 m/s [³³]. UPV and concrete’s modulus of elasticity are strongly associated. Thus, there are several parallels between HPC’s behaviour regarding this characteristic and the modulus of elasticity [³⁴]. Regardless of their age, UPV dropped linearly with RA content because adding this waste increased porosity and made low-quality ITZ emerge. Additionally, UPV somewhat decreased because of the reduced stiffness of early-age RA compared to developed RA, which became more noticeable when 100% RA was added (Table 8). The elasticity modulus and ultrasonic pulse velocity were in a similar relationship, with the incorporation of Magnesium Oxide (MgO) resulting in a UPV reduction of about 4–6%. This effect can be explained by the dilution of the binder as an increase in the water-to-binder (w/b) ratio, considering that magnesium hydroxide and calcium silicate hydrate have nearly identical stiffness [³⁵]. Furthermore, the slight rise in porosity attributed to MgO may also influence these outcomes. The percentage drop in UPV upon adding MgO was consistent, independent of the quantity and maturity of RA, much like in the elasticity modulus [³⁶]. Conversely, RA’s content and maturity caused a percentage decline that remained consistent irrespective of the binder. Table 8 clearly illustrates these elements. Finally, the proportion of UPV formed at 7 days was generally decreased by internal curing brought on by the RA’s delayed water release and the somewhat slower growth of MgO’s strength and stiffness in comparison to OPC (Table 8). A similar thing happened with early-age RA’s increasing stiffness over time.

Figure 8
High-performance concrete and its UPV (a) 7 days and (b) 28 days.

3.2.6. Statistical analysis

The addition of RA, RA maturity, and MgO were discussed in the previous studies. Additionally, research indicates that the influence of MgO was negative as the RA content increased, although this effect was not uniform in terms of compressive strength or splitting tensile strength. To substantiate these findings, a three-way Analysis of Variance (ANOVA) can be employed for each characteristic, considering the effects of all other elements by using minitab software. Table 9 displays the derived p-values. Given that this kind of study typically uses a significance threshold of 5%, it can be said that Increasing the RA concentration and substituting 10% OPC with MgO had a substantial impact on every property examined; that is, they consistently resulted in a discernible decline in the property under analysis. However, only compressive strength was a significant indicator of RA maturity, suggesting that RA strength appears more relevant to HPC mechanical behaviour than stiffness. The interaction among the elements was limited to a second-order manner concerning compressive and splitting tensile strength [³⁷]. The findings indicated that early-age Recycled Aggregate (RA) had a more damaging effect on compressive strength as the RA content in the mix increased. This may be due to the significant influence of RA maturity on its strength characteristics [³⁸]. However, regarding compressive and splitting tensile strength, the increase in RA quantity (RA content-MgO interaction) also made the effect of MgO more detrimental. Finally, maturity RA had a similar outcome on the MgO interaction regardless of the binder utilized.

Table 9
ANOVA for the properties of HPC and p values (3-way).

TYPE OF PROPERTY		RA*	RA* MATURITY	MgO*	RA – RA# MATURITY	RA – MgO#	RA MATURITY – MgO#
Slump		0.0042	0.3157	0.0036	0.6031	0.1501	0.4121
Density		0.0003	0.4168	0.0079	0.4900	0.4983	0.4121
Harden density	7 days	0.0003	0.9999	0.0015	0.4900	0.0581	1.0000
	28 days	0.0005	0.9001	0.0034	0.3952	0.1716	0.9215
Compressive strength	7 days	0.0001	0.0003	0.0001	0.0008	0.0035	0.2516
	28 days	0.0008	0.0115	0.0011	0.0381	0.0322	0.5842
Split tensile strength		0.0013	0.0592	0.0049	0.1927	0.0401	0.2911
Modulus of elasticity		0.0017	0.8711	0.0179	0.5002	0.1819	0.6628
UPV	7 days	0.0007	0.0572	0.0021	0.1972	0.7104	0.2444
	28 days	0.0005	0.1703	0.0049	0.0941	0.5101	0.2216

• ^*Each factor effect and
• ^#Second order iterations.

3.3. Shrinkage effect

3.3.1. Autogenous shrinkage

When cement is hydrated after a stable structure has developed in the concrete (sometimes referred to as “time zero”), a chemical process causes capillary water to be consumed, resulting in autogenous shrinkage [³⁹]. To avoid water absorption, 10 × 10 × 50 cm prismatic specimens were coated with the aluminium foil and used for this study. Whereas the impact of each condition is shown in Table 10, Figure 9 shows the results gathered. More water can be held in RA than in NA, as shown in Figure 10. It is evident from Figure 9(a) that the use of RA reduced the autogenous shrinkage. Consequently, their increased water release during the cement’s delayed hydration made it possible to replace the water used during this procedure more effectively [⁴⁰]. Moreover, because of their more excellent deformability, this event compensated for their decreased resistance to the cementitious matrix’s contraction. The rheological character of early-age RA causes them to shrink; hence, their usage accelerates autogenous shrinking. This circumstance could also benefit from their reduced rigidity [⁴¹]. Reactive magnesium oxide is expansive, as the introduction explains, since the volume of the resultant Mg(OH)₂ after hydration is greater than that of MgO. The data presented in Figure 9(b) reveals that the reference mix’s autogenous shrinkage was diminished by 93% upon substituting 10% OPC reactive with MgO. This confirms the agent for shrinkage reduction in the HPC. When combined with matured RA, the increased matured RA content resulted in a larger reduction in autogenous shrinkage when employing MgO as opposed to an HPC with the same amount of RA but 100% OPC (Table 10). This is why mixtures, including MgO, as opposed to 100% OPC, showed a greater shrinkage decrease when RA was increased. Initially, RA exhibited lower resistance to the growth of MgO due to its decreased stiffness compared to NA. A notable proportion of MgO particles that did not hydrate in the mixing phase were subsequently able to hydrate, owing to RA’s remarkable capacity to retain water and release it over time. This greater hydration of MgO particles resulted in increased expansion, which, in turn, strengthened the cementitious matrix’s resistance to contraction. Given these factors, the early-age RA mixes should have had a greater impact from MgO because of their reduced stiffness. However, the autogenous shrinkage increased when early-age RA was employed because their shrinkage was more significant than their stiffness [⁴²]. The abovementioned factors were consistent with the MA100MO formulation, demonstrating the most considerable length growth of 0.035 mm/m after 91 days. During the mixing period, the maximum amount of MgO was hydrated. Figure 10(b) further illustrates that MgO’s expansion generally occurs in the early days. On the other hand, no clear pattern of the duration of the HPC’s lengthening was found. Thus, the mixes that extended for two days were the RMO and MA100MO mixes, whereas the other mixes expanded for three days. Concrete composition does not affect this factor, which appears related to how well MgO hydrates during the mixing process [⁴³]. Therefore, it has been established that RA is important for the long-term hydration of MgO, as noted earlier, but it does not affect the initial hydration or the later expansion of MgO. Ultimately, Figure 10(b) shows that the time-dependent behaviour of autogenous shrinkage in High-Performance Concrete (HPC) with Magnesium Oxide (MgO) conforms to a logarithmic function similar to that observed in HPC with 100% Ordinary Portland Cement (OPC) in Figure 10(a) and conventional concrete. As long as the data from the initial period, particularly from day one to day four, is omitted because this is when most of the MgO growth occurs, it is indeed feasible to precisely fit this kind of function to mixtures with this composition.

Table 10
HPC mixes autogenous, drying and total shrinkage and its effects at 91 days.

TYPE OF MIX	AUTOGENOUS SHRINKAGE (mm/m)				DRYING SHRINKAGE (mm/m)				TOTAL SHRINKAGE (mm/m)
	AUTOGENOUS SHRINKAGE	ΔRA* %	ΔRA# MATURITY	ΔMgO$ %	DRYING SHRINKAGE	ΔRA* %	ΔRA# MATURITY	ΔMgO$ %	TOTAL SHRINKAGE	ΔRA* %	ΔRA# MATURITY	ΔMgO$ %
RA	-0.182 ± 0.012	–	-	–	-0.082 ± 0.021	–	–	–	-0.265 ± 0.018	–	–	–
MA25	-0.179 ± 0.022	-3.18	-	–	-0.131 ± 0.032	+56.8	–	–	-0.308 ± 0.006	+15.81	–	–
EA25	-0.176 ± 0.011	-2.08	+0.92	–	-0.188 ± 0.042	+117.6	+37.6	–	-0.357 ± 0.021	+36.92	+17.22	–
MA100	-0.162 ± 0.031	-15.9	–	–	-0.382 ± 0.031	+342.8	–	–	-0.538 ± 0.006	+97.65	–	–
EA100	-0.158 ± 0.028	-12.2	+5.72	–	-0.412 ± 0.005	+392.6	+11.9	–	-0.586 ± 0.008	+112.48	+8.15	–
RMO	-0.011 ± 0.014	–	–	-92.1	-0.161 ± 0.006	–	–	+73.2	-0.159 ± 0.022	–	–	-38.2
MA25MO	+0.008 ± 0.012	-158.6	–	-102.4	-0.231 ± 0.032	+58.2	–	+74.8	-0.218 ± 0.016	+36.83	–	-27.6
EA25MO	+0.003 ± 0.042	-128.2	+54.7	-99.2	-0.268 ± 0.019	+79.2	+12.4	+42.8	-0.244 ± 0.024	+57.26	+14.81	-28.4
MA100MO	+0.032 ± 0.008	-357.3	–	-112.4	-0.458 ± 0.016	+202.8	–	+20.3	-0.418 ± 0.021	+148.77	–	-21.6
EA100MO	+0.021 ± 0.011	-275.2	+33.2	-117.6	-0.489 ± 0.023	+231.7	+7.8	+17.6	-0.443 ± 0.032	+171.51	+8.48	-21.2

• ^*Decrease in % due to RA use
• ^#Decrease in % due to early age RA use and
• ^$% decrease due to MgO.

Figure 9
HPC and its autogenous shrinkage (a) without MgO and (b) with 10% MgO.

Figure 10
HPC and its drying shrinkage (a) without MgO and (b) with 10% MgO.

To quantitatively describe the time-dependent behavior of autogenous shrinkage, a logarithmic regression model was fitted to the experimental data. The shrinkage trend closely follows the function:

$${\text{ε}}_{\text{auto}}(\text{t})=\text{a}\times \text{ln(t)+b}$$

Where ε_auto is the autogenous shrinkage in mm/m, t is the age in days, and a and b are regression coefficients specific to each mix. For instance, the RMO mix (containing 10% MgO) showed a good fit with coefficients a = −0.012; b = 0.025, and a coefficient of determination ^R2 = 0.983. Similar strong correlations were observed across other MgO and RA mixes. These regression models strengthen the interpretation of shrinkage evolution and confirm the logarithmic nature of early-age behavior, particularly in HPC incorporating MgO.

3.3.2. Drying shrinkage

When water evaporates from concrete after “time zero,” drying shrinkage results. Since autogenous shrinkage and this phenomenon cannot be objectively distinguished, it is possible that this sort of shrinkage cannot be directly quantified [⁴⁴]. Thus, the best way to find it is to figure out how much the shrinkage of specimens covered with aluminium foil tape differs from that of specimens not wrapped. This method calculates drying shrinkage by knowing the difference between autogenous shrinkage and total shrinkage. While Table 10 details the impact of each element, Figure 10 displays the drying shrinkage of the mixtures. Increasing the RA concentration and utilizing MgO had the opposite effects on drying shrinkage compared to autogenous shrinkage, as seen in Figure 10, they enhanced drying shrinkage rather than decreased it. This was explained by the fact that to keep the workability constant, the w/b ratio had to be increased. The postponement of water flow from RA resulted in abundant water for evaporation, ultimately leading to a heightened drying shrinkage in HPC [⁴⁵]. Additionally, the reduced stiffness of RA promoted this process. To preserve workability when replacing OPC with MgO, it is essential to elevate the water content of HPC, as MgO possesses a larger specific surface area [⁴⁶]. Consequently, a greater amount of water was present for evaporation, leading to an increase in drying shrinkage in the mixes composed of 90% OPC and 10% MgO compared to those made up of 100% OPC with identical composition (Table 10). About Figure 10(b), it is evident that there was no increase in length, as the contraction resulting from water evaporation surpassed the expansion of MgO. The maturation of RA had the same impact on autogenous shrinkage because early-age RA shrank more and had less rigidity, which led to more drying shrinkage. The rise in drying shrinkage was mainly due to the higher water content in HPC, which resulted from the substantial addition of RA. In contrast, the shrinkage of early-age RA contributed additionally. The rise was especially pronounced with the addition of limited quantities of RA, as depicted in Table 10. To conclude, the escalation of RA content in HPC led to decreased drying shrinkage associated with incorporating MgO. This may be explained by the fact that the drying shrinkage increment brought on by MgO was consistent across all mixes. As a result, the drying shrinkage increase was less significant in mixes containing high levels of RA [⁴⁷]. In addition, although the drying-shrinkage curves did not reveal any visible MgO expansion, it indeed occurred and might have been slightly more significant in High-Performance Concrete (HPC) with Recycled Aggregates (RA), resembling autogenous shrinkage, which could potentially improve its effectiveness.

3.3.3. Total shrinkage

Total shrinkage is the easiest to ascertain since it is measured on unwrapped specimens that come into direct contact with the environment. The overall effect includes both drying and autogenous shrinkage, which occur concurrently when water evaporation is permitted and the elements that lead to autogenous shrinkage are activated at the same time [⁴⁸]. The total shrinkage variations of High-Performance Concrete (HPC) mixes are depicted in Figure 11, with Table 10 providing a breakdown of the percentage contributions from each component. Ultimately, autogenous and drying shrinkage determines the overall shrinkage resulting from modifications in HPC composition. Additionally, drying shrinkage is more significant; hence, it mostly determines overall shrinkage. Thus, drying shrinkage and total shrinkage typically show comparable patterns. However, only two of the three characteristics examined in this study met this requirement. The impact of RA content and early-age RA was identical to drying shrinkage. Because of their increased deformability and water absorption, the overall shrinkage increases with the quantity of RA. Moreover, the overall shrinking of HPC was exacerbated by the early-age RA’s reduced rigidity compared to matured RA and their shrinkage. To enhance clarity and assess the relative contributions of autogenous and drying shrinkage, a quantitative comparison was made based on the 91-day data in Table 10. In the reference HPC mix (RA), autogenous shrinkage was 0.182 mm/m while drying shrinkage was 0.082 mm/m, showing that autogenous shrinkage was initially dominant. However, with increased RA content, this trend reversed. For instance, in MA100, autogenous shrinkage was 0.162 mm/m versus 0.382 mm/m for drying shrinkage, indicating drying shrinkage contributed over 70% to the total. In MgO-containing mixes, drying shrinkage overwhelmingly dominated. In RMO, autogenous shrinkage dropped to 0.011 mm/m, while drying shrinkage increased to 0.161 mm/m meaning over 93% of shrinkage came from drying. In MA100MO, autogenous shrinkage turned expansive (+0.032 mm/m), further confirming that drying shrinkage was the sole contributor to net length change (−0.458 mm/m). Therefore, the ratio between drying and autogenous shrinkage shifts markedly with the use of MgO and higher RA content, underlining the need to consider each mechanism’s role explicitly when interpreting shrinkage behavior in HPC systems. Despite the noted increase in drying shrinkage associated with its use, similar to what was observed in mortars, the incorporation of MgO contributed to a decrease in the total shrinkage of high-performance concrete [⁴⁹]. The increase in drying shrinkage resulting from the greater water content required for workability was notably surpassed by the reduction of autogenous shrinkage. In line with this, the overall shrinkage curves (Figure 11(b)) demonstrated the increase in the length of the specimens over the initial days due to MgO expansion. However, unlike autogenous shrinkage, no discernible pattern relating to the expansion time could be found.

Figure 11
HPC and its total shrinkage (a) without MgO and (b) with 10% MgO.

Furthermore, to characterize the development of total shrinkage over time, logarithmic regression models of the form:

$${\text{ε}}_{\text{total}}(\text{t})={\text{a}}^{\prime }\times \text{ln(t)+}{\text{b}}^{\prime }$$

were applied. The MA100 mix, for example, yielded a regression fit with a^′ = −0.086, b^′ = 0.285, and R² = 0.976. This trend demonstrates that while drying shrinkage dominates in later stages, the overall time-dependent behavior of shrinkage in HPC with MgO still follows a consistent logarithmic progression. These models offer a clearer, quantitative comparison of shrinkage progression among the different mixes.

Consequently, it may be said that in contrast to OPC, the impact of MgO in HPC depends on autogenous shrinkage [⁵⁰]. Because RA and MgO had opposing effects on overall shrinkage, it was possible that when both were utilized, the proportion of total shrinkage that MgO reduced dropped as HPC’s RA concentration rose (Table 10). According to Table 10, the rise in drying shrinkage due to Recycled Aggregates (RA) was significantly greater than the decline in autogenous shrinkage resulting from Magnesium Oxide (MgO). As a result, MgO was determined to be a more effective total-shrinkage-reducing agent in High-Performance Concrete (HPC) produced with Natural Aggregates (NA). As seen in Figure 11, the addition of over 42% developed RA made up for the decrease in overall shrinkage brought on by MgO. Because of their shrinking and decreased stiffness, the content was lower for early-age RA (35%) than for older RA.

3.3.4. Analysis of statistics

Table 11 presents the p-values from the three-way ANOVA carried out for the measurements taken at 91 days for the diverse categories of shrinkage investigated. These results provide strong evidence at the micro level to support the arguments made in the previous sections about the influence of RA composition, how it has changed over time, and the utilization of MgO [⁵¹]. The following may be stated at a confidence level of 95% and a significance level of 5%. All shrinkage types were highly influenced by the three parameters that were investigated RA content, RA maturity, and substitution of MgO for 10% OPC; therefore, any alteration to these factors resulted in noticeable alterations to every kind of HPC shrinkage. RA maturity and RA content had a substantial second-order interaction in both drying and autogenous shrinkage. As previously mentioned, the amount of HPC shrinkage brought on by early-age RA varied according to its composition [⁵²]. Given that the significant shrinkage levels observed in High-Performance Concrete (HPC) were considerably greater than the slight increase in shrinkage induced by early-age Recycled Aggregate (RA) with increased RA content, this influence was deemed insignificant regarding total shrinkage (Figure 12) [⁵³]. The RA content of HPC determined how MgO affected autogenous and drying shrinkage. Therefore, RA’s higher delayed water release and lower stiffness relative to NA enhanced MgO’s expansion and decreased various kinds of shrinking. This interaction also did not significantly affect overall shrinkage. Lastly, MgO’s impact was independent of RA’s maturity. Because early-age RA is less stiff than developed RA, it expands slightly more, although the shrinkage of early-age RA counteracts this.

Table 11
ANOVA for the shrinkage types and p values (3-way).

SHRINKAGE TYPE	AUTOGENOUS	DRYING	TOTAL
RA*	0.0005	0.0001	0.0003
RA* maturity	0.0121	0.0101	0.0106
MgO*	0.0001	0.0013	0.0007
RA – RA maturity#	0.0242	0.0421	0.0545
RA – MgO#	0.0092	0.0263	0.1224
RA maturity – MgO#	0.1821	0.2124	0.2381

• ^*Each factor effect and
• ^#second order iterations.

Figure 12
Variations in shrinkage (a) RA matured and (b) RA early age.

4. CONCLUSION

In this study, the effect of using reactive Magnesium Oxide (MgO) as a partial substitution for Ordinary Portland Cement (OPC) on the mechanical and shrinkage behaviour of High-Performance Concrete (HPC) has been investigated. The study utilized 100% OPC to create half of the combinations, whereas the remaining half featured a substitution of 10% OPC with reactive MgO. The following observations were made based on the experimental results.

• To maintain workability, it is imperative to increase the water-to-binder (w/b) ratio in High-Performance Concrete (HPC), as Magnesium Oxide (MgO) has a high precise surface area and a more irregular configuration compared to Ordinary Portland Cement (OPC). Although MgO has a higher actual density than OPC, this increase in water content also decreased HPC’s fresh and hardened densities;

• The replacement of 10% ordinary Portland cement with magnesium oxide resulted in a 5–8% decrease in the modulus of elasticity, splitting tensile & compressive strength, and ultrasonic pulse velocity. The incorporation of MgO resulted in an exacerbation of the negative effects due to the interface among the RA content and MgO, which manifested as a 19–21% reduction in both splitting tensile strength and compressive strength in High-Performance Concrete (HPC) with 100% RA. With the increase in Recycled Aggregate (RA) content within High-Performance Concrete (HPC), the total shrinkage also intensified, reflecting a similar pattern to the heightened drying shrinkage values. In entirely matured RA, the drying and overall shrinkage increased by 36% and 99%, respectively, while autogenous shrinkage decreased by 17%. When 10% OPC was substituted with MgO, autogenous shrinkage was either negligible or nonexistent due to the 93% reduction in autogenous shrinkage brought about by the expansion brought on by the formation of Mg(OH)₂ throughout the hydration of responsive MgO. On the other hand, drying shrinkage was 75% larger in HPC with MgO because of the water content. Using magnesium oxide led to a 39% decrease in overall shrinkage, suggesting that the reduction in autogenous shrinkage outweighed the rise in drying shrinkage. Hence, in HPC with 25% RA, the overall shrinkage resulting from MgO was reduced by 28%, and when 100% RA was added, it was reduced by 22%. The overall conclusion is that MgO was a binder that significantly decreased HPC shrinkage when applied at a rate of 10% of the total binder content while causing a reasonable drop in mechanical characteristics. However, using it in conjunction with RA reduced its efficacy as a shrinkage-reducing treatment.

5. BIBLIOGRAPHY

• ^[1] TUTKUN, B., BARLAY, E.S., YALÇINKAYA, Ç., et al, “Effect of internal curing on shrinkage and cracking potential under autogenous and drying conditions”, Construction & Building Materials, v. 409, pp. 134078, 2023. doi: http://doi.org/10.1016/j.conbuildmat.2023.134078.
  » https://doi.org/10.1016/j.conbuildmat.2023.134078
• ^[2] LIU, D., WANG, C., GONZALEZ-LIBREROS, J., et al, “A review of concrete properties under the combined effect of fatigue and corrosion from a material perspective”, Construction & Building Materials, v. 369, pp. 130489, 2023. doi: http://doi.org/10.1016/j.conbuildmat.2023.130489.
  » https://doi.org/10.1016/j.conbuildmat.2023.130489
• ^[3] FOLORUNSHO, A.B., KIM, S., PARK, C., “A review on the performance of fibers on restrained plastic shrinkage cracks”, Buildings, v. 14, n. 8, pp. 2477, 2024. doi: http://doi.org/10.3390/buildings14082477.
  » https://doi.org/10.3390/buildings14082477
• ^[4] TANG, S., HUANG, D., HE, Z., “A review of autogenous shrinkage models of concrete”, Journal of Building Engineering, v. 44, pp. 103412, 2021. doi: http://doi.org/10.1016/j.jobe.2021.103412.
  » https://doi.org/10.1016/j.jobe.2021.103412
• ^[5] KIPKEMBOI, B., MIYAZAWA, S., “Evaluating the influence of drying shrinkage on cracking resistance of massive self-compacting concrete with blast furnace slag using finite element method”, Infrastructures, v. 10, n. 1, pp. 11, 2025. doi: http://doi.org/10.3390/infrastructures10010011.
  » https://doi.org/10.3390/infrastructures10010011
• ^[6] SUN, M., VISINTIN, P., BENNETT, T., “The effect of specimen size on autogenous and total shrinkage of ultra-high performance concrete (UHPC)”, Construction & Building Materials, v. 327, pp. 126952, 2022. doi: http://doi.org/10.1016/j.conbuildmat.2022.126952.
  » https://doi.org/10.1016/j.conbuildmat.2022.126952
• ^[7] JAGADESH, P., KARTHIK, K., KALAIVANI, P., et al, “Examining the influence of recycled aggregates on the fresh and mechanical characteristics of high-strength concrete: a comprehensive review”, Sustainability (Basel), v. 16, n. 20, pp. 9052, 2024. doi: http://doi.org/10.3390/su16209052.
  » https://doi.org/10.3390/su16209052
• ^[8] REVILLA-CUESTA, V., EVANGELISTA, L., DE BRITO, J., et al, “Mechanical performance and autogenous and drying shrinkage of MgO-based recycled aggregate high-performance concrete”, Construction & Building Materials, v. 314, pp. 125726, 2022. doi: http://doi.org/10.1016/j.conbuildmat.2021.125726.
  » https://doi.org/10.1016/j.conbuildmat.2021.125726
• ^[9] ZHANG, C., LI, Y., SHI, X., et al, “Effect of internal curing by alkali-resistant superabsorbent polymers on workability, mechanical property, hydration, pore structure and shrinkages of high-performance cementitious materials”, Case Studies in Construction Materials, v. 22, pp. e04309, 2025. doi: http://doi.org/10.1016/j.cscm.2025.e04309.
  » https://doi.org/10.1016/j.cscm.2025.e04309
• ^[10] MI, T., YANG, E.-H., UNLUER, C., “Investigation of the properties of reactive MgO-based cements and their effect on performance”, Cement and Concrete Composites, v. 138, pp. 104984, 2023. doi: http://doi.org/10.1016/j.cemconcomp.2023.104984.
  » https://doi.org/10.1016/j.cemconcomp.2023.104984
• ^[11] MOHAJERANI, A., VAJNA, J., CHEUNG, T.H., et al, “Practical recycling applications of crushed waste glass in construction materials: a review”, Construction & Building Materials, v. 156, pp. 443–467, 2017. doi: http://doi.org/10.1016/j.conbuildmat.2017.09.005.
  » https://doi.org/10.1016/j.conbuildmat.2017.09.005
• ^[12] BRITISH STANDARDS INSTITUTION, EN 197-1 Cement – Part 1: Composition, specifications and conformity criteria for common cements, Reino Unido, British Standards Institution, 2011. http://www.puntofocal.gob.ar/notific_otros_miembros/-mwi40_t.pdf, accessed in July 2025.
  » http://www.puntofocal.gob.ar/notific_otros_miembros/-mwi40_t.pdf
• ^[13] AZEVEDO, A.G.S., MOLANO, J.C.A., PARENTE, I., et al, “Effect of the rehydration method on the physical–mechanical properties of CO2-cured magnesium-based fiber cement boards”, CivilEng, v. 5, n. 1, pp. 247–264, 2024. doi: http://doi.org/10.3390/civileng5010013.
  » https://doi.org/10.3390/civileng5010013
• ^[14] BALOCH H, GRÜNEWALD S, MATTHYS S., “Assessment of shrinkage and bond behaviour of high performance cement-based composites as a repair mortar”, Developments in the Built Environment, v. 15, pp. 100203, 2023. doi: http://doi.org/10.1016/j.dibe.2023.100203.
  » https://doi.org/10.1016/j.dibe.2023.100203
• ^[15] BRAVO, M., FORERO, J.A., NOBRE, J., et al, “Performance of Mortars with commercially-available reactive magnesium oxide as alternative binder”, Materials (Basel), v. 14, n. 4, pp. 938, 2021. doi: http://doi.org/10.3390/ma14040938. PubMed PMID: 33669426.
  » https://doi.org/10.3390/ma14040938
• ^[16] ZHANG, N., YANG, Q., LI, D., et al, “Autogenous and drying shrinkage properties of precast recycled aggregate concrete”, Case Studies in Construction Materials, v. 22, pp. e04355, 2025. doi: http://doi.org/10.1016/j.cscm.2025.e04355.
  » https://doi.org/10.1016/j.cscm.2025.e04355
• ^[17] FORERO, J.A., BRAVO, M., PACHECO, J., et al, “Thermal performance of concrete with reactive magnesium oxide as an alternative binder”, Sustainability (Basel), v. 14, n. 10, pp. 5885, 2022. doi: http://doi.org/10.3390/su14105885.
  » https://doi.org/10.3390/su14105885
• ^[18] BRITISH STANDARDS INSTITUTION, Eurocode 2 Design of concrete structures. Part 1-1: General rules and rules for buildings, Reino Unido, British Standards Institution, 2010, https://www.phd.eng.br/wp-content/uploads/2015/12/en.1992.1.1.2004.pdf, accessed in July 2025.
  » https://www.phd.eng.br/wp-content/uploads/2015/12/en.1992.1.1.2004.pdf
• ^[19] BRITISH STANDARDS INSTITUTION, EN 206 Concrete – Specification, performance, production, and conformity, Reino Unido, British Standards Institution, 2013.
• ^[20] CARACOL, C., KRAVCHANKA, L., BRAVO, M., et al, “Recycled aggregate concrete using seawater: optimizing concrete’s sustainability”, Journal of Building Engineering, v. 97, pp. 110841, 2024. doi: http://doi.org/10.1016/j.jobe.2024.110841.
  » https://doi.org/10.1016/j.jobe.2024.110841
• ^[21] BRITISH STANDARDS INSTITUTION, EN 12350-2 Testing fresh concrete – Part 2: Slump test, Reino Unido, British Standards Institution, 2020. https://cdn.standards.iteh.ai/samples/64380/-68173c5f924b4a5083cdb5fa710f39c0/SIST-EN-12350-2-2019.pdf, accessed in July 2025.
  » https://cdn.standards.iteh.ai/samples/64380/-68173c5f924b4a5083cdb5fa710f39c0/SIST-EN-12350-2-2019.pdf
• ^[22] BRITISH STANDARDS INSTITUTION, EN 12350-6. Testing fresh concrete – Part 6: Density, Reino Unido, British Standards Institution, 2020. https://cdn.standards.iteh.ai/samples/64407/-d6e4d96fe65c4643b6369b92b72b4273/SIST-EN-12350-6-2019.pdf, accessed in July 2025.
  » https://cdn.standards.iteh.ai/samples/64407/-d6e4d96fe65c4643b6369b92b72b4273/SIST-EN-12350-6-2019.pdf
• ^[23] LABORATÓRIO NACIONAL DE ENGENHARIA CIVIL, LNEC-E398 Determination of Drying Shrinkage and Expansion, Portugal, Laboratório Nacional de Engenharia Civil, 1993.
• ^[24] ELSHAHAWI, M., HÜCKLER, A., SCHLAICH, M., “Infra lightweight concrete: a decade of investigation (a review)”, Structural Concrete, v. 22, n. S1, pp. E152–E168, 2021. doi: http://doi.org/10.1002/suco.202000206.
  » https://doi.org/10.1002/suco.202000206
• ^[25] WANG, L., ZENG, X., LI, Y., et al, “Influences of MgO and PVA fiber on the abrasion and cracking resistance, pore structure and fractal features of hydraulic concrete”, Fractal and Fractional, v. 6, n. 11, pp. 674, 2022. doi: http://doi.org/10.3390/fractalfract6110674.
  » https://doi.org/10.3390/fractalfract6110674
• ^[26] SKOCEK, J., OUZIA, A., VARGAS SERRANO, E., et al, “Recycled sand and aggregates for structural concrete toward the industrial production of high-quality recycled materials with low water absorption”, Sustainability (Basel), v. 16, n. 2, pp. 814, 2024. doi: http://doi.org/10.3390/su16020814.
  » https://doi.org/10.3390/su16020814
• ^[27] SREENIVASAN, H., BERNARD, E., SANTOS, H.S., et al, “A critical review of magnesium silicate hydrate (MSH) phases for binder applications”, Cement and Concrete Research, v. 178, pp. 107462, 2024. doi: http://doi.org/10.1016/j.cemconres.2024.107462.
  » https://doi.org/10.1016/j.cemconres.2024.107462
• ^[28] HANAFI, M., JAVED, I., EKINCI, A., “Evaluating the strength, durability and porosity characteristics of alluvial clay stabilized with marble dust as a sustainable binder”, Results in Engineering, v. 25, pp. 103978, 2025. doi: http://doi.org/10.1016/j.rineng.2025.103978.
  » https://doi.org/10.1016/j.rineng.2025.103978
• ^[29] BALDOVINO, J.A., NUÑEZ DE LA ROSA, Y.E., PALMA CALABOKIS, O., “Analysis of novel porosity–water-to-binder index for prediction of strength, stiffness and durability for cemented soils”, Materials (Basel), v. 16, n. 19, pp. 6354, 2023. doi: http://doi.org/10.3390/ma16196354. PubMed PMID: 37834490.
  » https://doi.org/10.3390/ma16196354
• ^[30] MOHAMED, A.M., TAYEH, B.A., MAJEED, S.S., et al, “Fresh, hardened, durability and microstructure properties of seawater concrete: a systematic review”, Journal of CO2 Utilization, v. 83, pp. 102815, 2024. doi: http://doi.org/10.1016/j.jcou.2024.102815.
  » https://doi.org/10.1016/j.jcou.2024.102815
• ^[31] FAROOQ, U., RIZWAN, M., KHALIQ, W., et al, “Towards sustainable construction: Harnessing potential of pumice powder for eco-friendly concrete, augmented by hybrid fiber integration to elevate concrete performance”, Case Studies in Construction Materials, v. 21, pp. e03815, 2024. doi: http://doi.org/10.1016/j.cscm.2024.e03815.
  » https://doi.org/10.1016/j.cscm.2024.e03815
• ^[32] FORERO, J.A., BRAVO, M., PACHECO, J., et al,“Fracture behaviour of concrete with reactive magnesium oxide as alternative binder”, Applied Sciences (Basel, Switzerland), v. 11, n. 7, pp. 2891, 2021. doi: http://doi.org/10.3390/app11072891.
  » https://doi.org/10.3390/app11072891
• ^[33] LYU, C., YU, C., LU, C., et al, “Long-term performance and microstructural characterization of dam concrete in the three gorges project”, Engineering (Beijing), v. 33, pp. 237–262, 2024. doi: http://doi.org/10.1016/j.eng.2023.04.017.
  » https://doi.org/10.1016/j.eng.2023.04.017
• ^[34] CARRILLO, J., RAMIREZ, J., LIZARAZO-MARRIAGA, J., “Modulus of elasticity and Poisson’s ratio of fiber-reinforced concrete in Colombia from ultrasonic pulse velocities”, Journal of Building Engineering, v. 23, pp. 18–26, 2019. doi: http://doi.org/10.1016/j.jobe.2019.01.016.
  » https://doi.org/10.1016/j.jobe.2019.01.016
• ^[35] JOHN, E., LOTHENBACH, B., “Cement hydration mechanisms through time–a review”, Journal of Materials Science, v. 58, n. 24, pp. 9805–9833, 2023. doi: http://doi.org/10.1007/s10853-023-08651-9.
  » https://doi.org/10.1007/s10853-023-08651-9
• ^[36] PANTIĆ, V., ŠUPIĆ, S., VUČINIĆ-VASIĆ, M., et al, “Effects of grinding methods and water-to-binder ratio on the properties of cement mortars blended with biomass ash and ceramic powder”, Materials (Basel), v. 16, n. 6, pp. 2443, 2023. doi: http://doi.org/10.3390/ma16062443. PubMed PMID: 36984322.
  » https://doi.org/10.3390/ma16062443
• ^[37] KIM, Y., PARK, W., HANIF, A., “Steam-cured recycled aggregate concrete incorporating moderately high early strength cement: effect of binder content and curing conditions”, SN Applied Sciences, v. 1, n. 5, pp. 445, 2019. doi: http://doi.org/10.1007/s42452-019-0472-x.
  » https://doi.org/10.1007/s42452-019-0472-x
• ^[38] AL AJMANI, H., SULEIMAN, F., ABUZAYED, I., et al, “Evaluation of concrete strength made with recycled aggregate”, Buildings, v. 9, n. 3, pp. 56, 2019. doi: http://doi.org/10.3390/buildings9030056.
  » https://doi.org/10.3390/buildings9030056
• ^[39] WANG, H., LIU, Y., HU, Z., et al, “Assessment of determination methods for time-zero of autogenous shrinkage in cement mortars”, Cement and Concrete Composites, v. 152, pp. 105607, 2024. doi: http://doi.org/10.1016/j.cemconcomp.2024.105607.
  » https://doi.org/10.1016/j.cemconcomp.2024.105607
• ^[40] AN, M., LIU, Y., ZHANG, G., et al, “Properties of cement-based materials with low water–binder ratios and evaluation mechanism under further hydration effect”, Applied Sciences (Basel, Switzerland), v. 13, n. 17, pp. 9946, 2023. doi: http://doi.org/10.3390/app13179946.
  » https://doi.org/10.3390/app13179946
• ^[41] LI, Z., CHEN, Y., PROVIS, J.L., et al, “Autogenous shrinkage of alkali-activated slag: A critical review”, Cement and Concrete Research, v. 172, pp. 107244, 2023. doi: http://doi.org/10.1016/j.cemconres.2023.107244.
  » https://doi.org/10.1016/j.cemconres.2023.107244
• ^[42] LIM, J.L.G., RAMAN, S.N., SAFIUDDIN, M., et al, “Autogenous shrinkage, microstructure, and strength of ultra-high performance concrete incorporating carbon nanofibers”, Materials (Basel), v. 12, n. 2, pp. 320, 2019. doi: http://doi.org/10.3390/ma12020320. PubMed PMID: 30669570.
  » https://doi.org/10.3390/ma12020320
• ^[43] MAHMOOD, A., KAISH, A.B.M.A., GULAM, N.F.B.A., et al, “Effects of MgO-based expansive agent on the characteristics of expansive concrete”, Engineering Proceedings, v. 11, n. 1, pp. 14, 2021. doi: http://doi.org/10.3390/ASEC2021-11165.
  » https://doi.org/10.3390/ASEC2021-11165
• ^[44] ZHU, L., ZHENG, M., ZHANG, W., et al, “Evolution and mechanism on shrinkage of high- performance concrete with full aeolian sand cured by superabsorbent polymer”, Construction & Building Materials, v. 400, pp. 132846, 2023. doi: http://doi.org/10.1016/j.conbuildmat.2023.132846.
  » https://doi.org/10.1016/j.conbuildmat.2023.132846
• ^[45] GONZALEZ-COROMINAS, A., ETXEBERRIA, M., “Effects of using recycled concrete aggregates on the shrinkage of high performance concrete”, Construction & Building Materials, v. 115, pp. 32–41, 2016. doi: http://doi.org/10.1016/j.conbuildmat.2016.04.031.
  » https://doi.org/10.1016/j.conbuildmat.2016.04.031
• ^[46] LI, S., MO, L., DENG, M., et al, “Mitigation on the autogenous shrinkage of ultra-high performance concrete via using MgO expansive agent”, Construction & Building Materials, v. 312, pp. 125422, 2021. doi: http://doi.org/10.1016/j.conbuildmat.2021.125422.
  » https://doi.org/10.1016/j.conbuildmat.2021.125422
• ^[47] LIN, X., CASTEL, A., DENG, Z., et al, “Effect of crystalline admixtures on shrinkage and alkali-silica reaction of biochar-cementitious composites”, Developments in the Built Environment, v. 18, pp. 100456, 2024. doi: http://doi.org/10.1016/j.dibe.2024.100456.
  » https://doi.org/10.1016/j.dibe.2024.100456
• ^[48] AMRAN, M., ONAIZI, A.M., MAKUL, N., et al, “Shrinkage mitigation in alkali-activated composites: a comprehensive insight into the potential applications for sustainable construction”, Results in Engineering, v. 20, pp. 101452, 2023. doi: http://doi.org/10.1016/j.rineng.2023.101452.
  » https://doi.org/10.1016/j.rineng.2023.101452
• ^[49] XI, Y.F., LEE, J., CHEN, B.L., et al, “Impact of high-performance expansion and shrinkage-reducing agents on the mechanical properties and shrinkage compensation of high-strength concrete”, Buildings, v. 13, n. 3, pp. 717, 2023. doi: https://doi.org/-10.3390/buildings13030717.
  » https://doi.org/10.3390/buildings13030717
• ^[50] RODRÍGUEZ-ÁLVARO, R., GONZÁLEZ-FONTEBOA, B., SEARA-PAZ, S., et al, “Internally cured high performance concrete with magnesium based expansive agent using coal bottom ash particles as water reservoirs”, Construction & Building Materials, v. 251, pp. 118977, 2020. doi: http://doi.org/10.1016/j.conbuildmat.2020.118977.
  » https://doi.org/10.1016/j.conbuildmat.2020.118977
• ^[51] GATOU, M.-A., SKYLLA, E., DOUROU, P., et al, “Magnesium oxide (MgO) nanoparticles: synthetic strategies and biomedical applications”, Crystals, v. 14, n. 3, pp. 215, 2024. doi: http://doi.org/10.3390/cryst14030215.
  » https://doi.org/10.3390/cryst14030215
• ^[52] GHANEM, H., RAMADAN, R., KHATIB, J., et al, “A review on chemical and autogenous shrinkage of cementitious systems”, Materials (Basel), v. 17, n. 2, pp. 283, 2024. doi: http://doi.org/10.3390/ma17020283. PubMed PMID: 38255451.
  » https://doi.org/10.3390/ma17020283
• ^[53] THANG, N.C., TUAN, N.V., YANG, K.H., et al, “Effect of zeolite on shrinkage and crack resistance of high-performance cement-based concrete”, Materials (Basel), v. 13, n. 17, pp. 3773, 2020. doi: http://doi.org/10.3390/ma13173773. PubMed PMID: 32859074.
  » https://doi.org/10.3390/ma13173773

Publication Dates

• Publication in this collection
  08 Aug 2025
• Date of issue
  2025

History

• Received
  02 May 2025
• Accepted
  11 June 2025