Microstructure and Mechanical Properties of Sustainable Concrete Incorporating Used Foundry Sand and Coal Bottom Ash

Abstract

This study investigates the potential for sustainable concrete production using industrial by-products: used foundry sand (UFS) and coal bottom ash (CBA). These materials were partially substituted for natural aggregates to reduce environmental impact and promote circular economy practices. UFS was used as a replacement for fine aggregate, while both fine and coarse CBA were tested as substitutes for sand and gravel, respectively. The materials were characterized using X-ray diffraction (XRD), scanning electron microscopy (SEM), and energy-dispersive X-ray spectroscopy (EDS) to evaluate their mineralogical and microstructural properties. Six concrete mixtures were prepared with varying replacement levels (up to 70% total aggregate substitution) at a constant water-to-cement ratio of 0.50. Compressive strength tests were conducted at 28 days, supported by microstructural observations. Results showed that high levels of UFS and CBA led to reduced strength, mainly due to weak interfacial bonding and porous ash particles. However, moderate replacement levels (e.g., 20% fine CBA) maintained high strength with good structural integrity. The study concludes that both UFS and CBA can be used effectively in concrete when carefully dosed. The findings support the use of industrial waste in construction, provided that material properties are well understood and replacement levels are optimized.

1. Introduction

The disposal of industrial by-products in construction materials aligns with the goals of sustainability and circular economy. Foundry sands and coal combustion ashes are two main solid wastes with potential for beneficial reuse in concrete. Used foundry sand (UFS) is a by-product of metal casting industries generated after repetitive use of silica sand in mold production [1]. Foundries around the world produce millions of tons of UFS annually, most of which has traditionally been landfilled or discarded [2]. In the United States alone, 6–10 million tons of waste foundry sand are estimated to be generated annually [3]. This indicates a significant potential for reuse in construction materials. Similarly, coal-fired power plants generate large amounts of coal ash residues; while the finer fraction (fly ash) is often reused as a pozzolan, the coarser fraction known as bottom ash (or boiler slag) is often underutilized and disposed of as waste. The use of these materials in concrete can reduce environmental burdens and preserve natural aggregate resources.

UFS is typically a subangular silica sand that may be bonded with clay (e.g., bentonite) and carbonaceous additives during casting. After use, the sand particles become coated with residual binders (burnt clay, carbon/coal dust, and other additives), and this spent foundry sand is classified as non-hazardous industrial waste [4]. Prior research has shown that foundry sand can be partially substituted for fine aggregate in concrete without significantly compromising strength. The potential to reuse UFS in concrete has been investigated for several decades. One of the first studies by Naik et al. [5] demonstrated the feasibility of incorporating UFS in concrete, showing that, with proper handling, UFS could serve as a partial replacement for fine aggregate. Siddique et al. [1] observed a marginal increase (4–10%) in compressive strength when replacing 10–30% of natural sand with UFS in concrete. The compressive strength of mixes at 28 days with 0%, 10%, 20%, 30% UFS was reported as 28.5 MPa, 29.7 MPa, 30.0 MPa, and 31.3 MPa, respectively, indicating that UFS can be used effectively as a partial fine aggregate without detrimental effect on strength, and even with slight improvements. Similarly, Güney et al. [6] demonstrated that waste foundry sand can be effectively reused in high-strength concrete, achieving satisfactory mechanical properties at replacement levels of up to 30%. Son et al. [7] reported that the incorporation of WFS can enhance certain aspects of durability, depending on the mix design and the quality of the sand. These and other studies have concluded that foundry sand can be used successfully in concrete production [8,9], improving sustainability by consuming a waste material instead of virgin sand. However, the actual impact of UFS on concrete properties may depend on the type of foundry sand and binder residue. Many foundry sands are silica-based, but their impurity content (such as leftover binders or metals) can vary. Therefore, a thorough characterization of the mineralogy and microstructure of UFS is important to understand its behavior in cementitious systems. Recent studies have also explored the use of UFS in ultra-high performance concrete, confirming its potential for high-strength applications [10].

Coal bottom ash (CBA) is the coarse granular material that falls to the bottom of coal furnaces, distinct from the lighter fly ash carried out with flue gases. It typically has a porous, sand-like, or gravel-like texture, with a particle size ranging from fine sand up to a few centimeters, depending on the coal and boiler type. CBA particles are often partially vitrified (glassy) with angular shapes and a porous or vesicular structure. They can retain unburned carbon and may contain crystalline mineral phases formed during combustion [11]. Chemically, CBA usually comprises predominantly silica and alumina (originating from mineral matter in coal) with smaller amounts of iron oxides, lime, magnesia, sulphates, etc. [11,12]. For example, the CBA examined in this study (from a bituminous coal power plant) has about 51% of SiO₂, 27% of Al₂O₃, 5.4% of Fe₂O₃, 2.1% of CaO, 1.5% of MgO and 0.85% of SO₃, along with ~7% loss on ignition (unburnt carbon) [11,12]. Such a composition indicates a material that is largely siliceous but not reactive, such as fly ash (due to its low amorphous content and higher carbon). Earlier research on the use of CBA as aggregate has shown that it can partially replace natural sand or gravel in concrete, although often with some reduction in workability and strength depending on the replacement level. According to ASTM C33/C33M-18 [13], alternative aggregates, such as CBA, must meet specific grading, durability, and soundness criteria in order to be used in structural concrete. Aggarwal et al. [14], for example, reported that concrete with up to 30% to 50% CBA (by fine aggregate volume) had slightly lower early strength but was still acceptable for structural use, especially at 20–30% replacement. Similarly, the early work of Ghafoori and Cai [15] found that properly graded CBA could be successfully used as a fine aggregate in concrete, with only moderate reductions in compressive strength, depending on the quality of the ash and the design of the mix. Singh and Siddique [12] also reported that concrete mixes with high volumes of CBA as fine aggregates exhibited acceptable strength and durability when properly designed, confirming the potential of this material for sustainable concrete applications. Recent reviews [8] have summarized the potential benefits and challenges of incorporating both UFS and CBA into sustainable concrete, particularly for pavement applications. Previous studies [16] also examined the combined use of coal cinder and waste foundry sand in high-performance concrete, showing promising mechanical and durability properties. Some studies note that long-term curing can mitigate strength loss—concrete with CBA can develop 90-day strength comparable to that of control concrete despite lower strength at 28-day [17]. However, the variability of CBA (porosity, residual carbon, and particle size) means that its effects on concrete can be inconsistent, and microstructural analysis is needed to explain the observed mechanical performance. In particular, CBA aggregates tend to be weaker than natural stone aggregates, which can become the weak link in concrete under load.

In summary, incorporating UFS and CBA in concrete is feasible and can produce sustainable concrete if properly proportioned [1,5,6]. Choudhary et al. [8] provided a comprehensive review of recent research on the combined effects of UFS and CBA in sustainable concrete, summarizing their influence on mechanical and durability properties. However, more systematic studies are still needed to understand the combined use of these two wastes and their influence on concrete at both macroscopic and microscopic levels. This study aims to characterize the mineralogical composition and microstructure of a UFS and a CBA and to evaluate their effects on concrete compressive strength when used as partial replacements for fine and coarse aggregate, respectively. By using X-ray diffraction (XRD) and scanning electron microscopy/energy-dispersive X-ray spectroscopy (SEM/EDS) analyzes, we elucidated how the presence of these wastes alters the cement hydration products, the interfacial transition zone (ITZ), and the failure mechanisms of the concrete. The experimental program focuses on relatively high replacement levels (30% UFS and 20–40% CBA) to assess the worst-case performance, thus providing information on the limitations and proper use of these materials in structural concrete. The results are expected to help develop guidelines for the UFS and CBA in construction, promoting recycling and sustainable development in the concrete industry.

2. Materials and Methods

2.1. Materials

2.1.1. Cement

A commercial Portland cement CEM I 52.5 N–HSR/NA (CEMEX, Chełm, Poland) was used as the binder, which is a high-strength, sulfate-resistant type of Portland cement with low alkali content. This cement meets the requirements of PN-EN 197-1 [18] and is designed for enhanced durability (for example, in bridge or pavement concrete).

Table 1. Chemical composition of used foundry sand (UFS) and cement (in %).

Compound	Used Foundry Sand	Cement
SiO2	95.3	20.92
Al2O3	1.9	3.50
Fe2O3	0.7	4.38
CaO	0.35	64.69
MgO	—	1.20
SO3	—	3.07
K2O	—	0.38
Na2O	—	0.22
Cl	—	0.08
Loss on ignition	—	1.27
Insoluble matter	—	0.20

presents the chemical compositions of the cement and UFS.

2.1.2. Natural Aggregates

Crushed basalt gravel (maximum size 14–16 mm) and natural sand (0–4 mm) were used as coarse and fine aggregates in the control mix. The basalt had a specific gravity of ~2.66 and the sand of ~2.62. Both aggregates met the classification requirements for concrete aggregate; their particle size distribution was determined according to PN-EN 933-1 [19].

2.1.3. UFS

UFS was obtained from a local cast iron foundry (Lublin, Poland), and its appearance is shown in Figure 1a. The sand is a siliceous molding sand that had been used in ferrous metal casting with a glauconite (clay) binder. After being cast, the spent sand was collected as waste. The color of the UFS was visually assessed and described as dark gray. It was composed primarily of fine sand-sized particles. For the experiments, the UFS was dried and sieved to remove any debris, and only the fraction passing the 4 mm sieve (to match the fine aggregate range) was used.

To characterize the mineralogical composition of the UFS, an XRD analysis was conducted, as described in Section 2.3.1. The phase identification and discussion of results are presented in Section 3.1.1. The UFS had a bulk density of about 1.52 g/cm³ and an estimated porosity of ~20%, attributed to residual fines and internal voids. Chemically, foundry sands often contain silica (~85–95%) with clay and carbon binders making up the rest [4]. In this case, the clay (glauconite) binder and carbonaceous additives remained as a fine powder coating on the sand grains, as observed in SEM images (discussed later). The oxide composition of UFS determined by XRF analysis is summarized in Table 1.

For the purpose of this study, UFS was used as a mass-based replacement of natural sand. Although its bulk density was slightly lower due to adhered fines, this approach is justified by its comparable particle size distribution (see Figure 2) and aligns with practical batching procedures commonly reported in the literature [20,21].

Recent studies have reinforced the suitability of treated UFS as a fine aggregate replacement. Tangadagi and Ravichandran [20] showed that chemically treated UFS can replace 30–40 wt % of natural sand in self-compacting concrete (SCC) without any detrimental effects on flowability, strength, or durability. Similarly, Kumar and Singh [21] confirmed that high-purity silica-based UFS improved packing density and mechanical performance in sustainable concrete mixes.

2.1.4. CBA

The CBA slag was sourced from a local power plant burning bituminous coal (class II A) in steam boilers. Its macroscopic appearance is presented in Figure 1b. After combustion, the heavier ash particles that settled in the boiler were collected. This CBA has a heterogeneous appearance: the particles are mostly black or dark gray, with some lighter, reddish porous pieces (indicating more completely burned portions). The CBA was air-dried and then separated by sieving into two fractions. The coarse CBA had particles retained in the 4 mm sieve, with a size up to ~16 mm. These are generally larger clinkers or fused chunks and were designated for use as coarse aggregate substitute. The fine CBA had particles passing through a 4 mm sieve (and mostly retained on 0.25 mm), resembling coarse sand. These were considered as an alternative fine aggregate. Both fractions were used without further crushing or grinding. The specific gravity of the CBA particles ranged from 1.2 to 1.8 (lighter than natural aggregates), and their porosity was high (30–60% internal porosity reported for such ashes). The water absorption of CBA can be up to ~20% due to its porous nature.

XRD was used to characterize the mineralogical composition of the CBA, as described in Section 2.3.1. The identified phases and interpretation are presented in Section 3.1.2.

Table 2. Example chemical compositions of coal bottom ash (CBA) aggregates.

Components	Content [% by Weight]
	CBA Aggregate (Coal Combustion), CHP Lublin [This Study]	EPO Aggregate, Opole Power Plant [Unavailable Web Source]	CBA Aggregate (Polish Coals) [22]
SiO2	51.1	50 ± 5	30–55
Al2O3	26.7	22.5 ± 4.5	12–32
Fe2O3	5.44	10.7 ± 1.9	4–14
CaO	2.13	4.2 ± 0.6	2–10
MgO	1.46	3.0 ± 0.7	1–6
SO3	0.85	≤1.0	0.5–4
Na2O	0.45	–	–
K2O	2.27	–	–
TiO2	1.22	–	–
P2O5	0.64	–	–
MnO	0.04	–	–
Cl	0.13	–	–
Loss on ignition (LOI)	7.19	≤5.5	1–35

(from chemical analysis) summarizes the composition of the CBA. The first column presents the results obtained in this study. The second column (EPO Aggregate) is based on data formerly available on the website of the manufacturer (EPO Aggregate, Opole Power Plant), which is now unavailable. The third column contains reference values from published data on Polish bituminous coal combustion products [22].

Loss on ignition (LOI) was approximately 7.2% which is considered relatively high, confirming the presence of unburnt carbon and possibly moisture. Because the CBA can contain residual sulfides and sulfates, its use in concrete requires the cement to be resistant to sulfates (which is the case for the chosen cement) to avoid sulfate attack issues. In this study, the fine CBA was intended to replace the sand in some mixes and the coarse CBA to replace the basalt gravel in others, allowing for an evaluation of its use in both aggregate functions.

Figure 2 presents the particle size distributions of the raw materials used in the study. The results highlight significant differences in gradation between the fine powders (cement and UFS) and the coarser aggregates (CBA, natural sand, and basalt). These differences influence packing density and, potentially, the fresh properties of the concrete mixtures.

Recent reviews [23] have summarized the physical and chemical characteristics of CBA and its applications in structural concrete, highlighting porosity, chemical stability, and partial hydraulic reactivity. More recent experimental work showed that CBA can be effectively used as fine aggregate (≤75 µm) up to 6 wt% in mortars, significantly improving workability and without strength loss after 45 days [24]. Comprehensive reviews also indicate optimal CBA replacement levels between 10 and 30 wt%, balancing sustainability goals and mechanical performance [25].

2.1.5. Admixture

A polycarboxylate-based high-range water reducer (superplasticizer) was used in the mixes containing waste materials to improve the workability. Based on the trial mixes, a dose of 1.6% (by cement weight) was added in mixes with UFS and/or CBA. The control mix (without waste) did not require admixture in the 0.50 w/c ratio to achieve a similar slump. However, for consistency, a small dose (~0.5%) was added to the control and to maintain the same water content and comparable workability.

2.2. Mix Proportions and Specimen Preparation

A total of six concrete mixtures were prepared: one control (reference) mix with all-natural aggregates and five mixes incorporating UFS and/or CBA (see

Table 3. Mix proportions of concrete with UFS and CBA.

Concrete Mix Component	Unit	M-1	M-2	M-3	M-4	M-5	M-6
CEM I 52.5 cement	(kg/m3)	372	372	372	372	372	372
Used foundry sand	(kg/m3)	–	162	–	162	–	162
	(%)	–	30	–	30	–	30
Natural sand	(kg/m3)	539	377	539	377	539	377
Basalt aggregate 2–16 mm	(kg/m3)	1229	1229	737	737	983	983
Coal bottom ash	(kg/m3)	–	–	492	492	246	246
	(%)	–	–	40	40	20	20
Superplasticizer	(L/m3)	5.9	5.9	5.9	5.9	5.9	5.9
Water	(L/m3)	186	186	186	186	186	186
Water-to-cement ratio (w/c)	–	0.5	0.5	0.5	0.5	0.5	0.5
Waste replacement total ratio	(%)	–	30	40	70	20	50
Fresh concrete density [kg/m3]	(kg/m3)	2332	2332	2332	2332	2332	2332

). All mixes were designed with the same cement content and water-to-cement ratio (w/c = 0.50) to isolate the effect of aggregate replacement.

These replacement levels (30% for sand and 20–40% for basalt) were chosen based on the upper limits reported in the literature for feasible use without severe strength loss [26,27]. In particular, the combined replacements in M-4 and M-6 represent a very high total replacement of natural aggregate (reflecting a scenario to push the limits of waste utilization). All mixes had the same cement and water content (since w/c = 0.5) to target a comparable 28-day strength in the control of around 70–80 MPa (which is high for w/c = 0.5, but achievable with the cement and aggregate quality given). The superplasticizer was added to M-2 through M-6 at 1.6% cement weight to achieve a workable consistency (slump of ~50–100 mm). The mixing procedure involved first dry mixing of cement and aggregates (including waste) for 2 min, then adding approximately 80% of the mixing water premixed with the superplasticizer, followed by the remaining water, and mixing for another 3–4 min until a homogeneous mixture was obtained. Mixes with a high UFS content were observed to require slightly longer mixing to break up any clumps of foundry sand (due to residual clay making them clumpy) and mixes with CBA absorbed water quickly, requiring careful mixing to ensure an even distribution of moisture.

Standard 100 mm cube specimens were cast from each mix for compressive strength testing. For each concrete mixture, six cubes were cast. Fresh concrete was placed in two-layer oiled molds and vibrated on a table vibrator to eliminate air voids.

After casting, the specimens were covered with plastic sheets and kept in the casting room (~23 °C) for 24 h. They were then demolded and cured in water at 20 ± 2 °C for 7 days. After the initial 7-day water cure, the cubes were stored in a standard laboratory environment (≈20 °C, ~60% RH) until the test age of 28 days. Thus, all specimens were air-drying at 28 days after 1 week of moist curing. No additional curing was provided beyond 7 days to simulate typical field curing practices and to accentuate any strength development differences between mixes (since extended curing could benefit mixes with latent hydraulic or pozzolanic activity).

2.3. Testing Methods

2.3.1. XRD

XRD analysis was performed on powdered samples of UFS, CBA, and all six concretes to identify mineralogical phases. A Philips X’Pert diffractometer (Philips, Almelo, The Netherlands) (Cu Kα radiation, 40 kV, and 30 mA) was used in Bragg–Brentano geometry (5–60° 2θ and step size of 0.02°). Samples were ground to <75 μm prior to testing. Phase identification was based on the PDF-2 database. For UFS, peak intensities were used to estimate relative phase abundance. Due to the amorphous content of CBA, only qualitative identification of crystalline phases was performed.

The X-ray diffraction patterns were processed using standard smoothing and background subtraction tools available in the Philips X’Pert software (X’Pert HighScore v2.2b) to enhance signal clarity. These corrections were applied uniformly across all samples and did not affect the peak positions or relative intensities relevant for phase identification.

2.3.2. SEM and EDS

Microstructural analysis was carried out using a FEI Quanta 250 FEG SEM (FEI, Hillsboro, OR, USA) equipped with an EDAX EDS detector. Imaging was performed in a high-vacuum mode, while a low-vacuum mode was used for nonconductive samples. UFS was examined in powder form on carbon tape; CBA samples were embedded in epoxy and polished. Concrete specimens were fractured after testing, and selected fragments were polished to observe ITZ regions. Samples were gold-coated prior to analysis. EDS spot analysis and mapping were performed on selected areas to assess elemental composition (e.g., Si, Al, Ca, and Fe). Spectra were collected from UFS surfaces, CBA cross-sections, and cement matrix/ITZ in selected concretes (M-2, M-3, and M-5). Due to material heterogeneity, EDS results are presented qualitatively. SEM/EDS analyzes were used to evaluate particle morphology, coating residues, porosity, and paste–aggregate interaction.

2.3.3. Compressive Strength Test

Compressive strength was tested at 28 days on 100 mm cubes in accordance with PN-EN 12390-3 [28], using a 3000 kN hydraulic press (Controls, Milan, Italy) with a loading rate of 0.6 MPa/s. Strength was calculated from the peak load and cross-sectional area. For each mix, the average of six specimens is reported. Failure modes were visually observed. Fragments from failed specimens were collected for SEM/EDS analysis. The control mix reached over 70 MPa, attributed to the high-quality basalt aggregate and cement hydration. Mixes with waste showed more abrupt or variable failure patterns.

3. Results and Discussion

3.1. Mineralogical Characterization of XRD

3.1.1. Characterization of UFS

XRD was used to identify the mineralogical phases present in the UFS. The mineralogical composition of this UFS, determined by XRD, is primarily crystalline silica quartz (SiO₂), ~98% by mass. Minor quantities (~1% each) of calcite (CaCO₃) and albite (NaAlSi₃O₈) were also detected. The presence of quartz as the dominant phase is expected since the foundry sand was originally a silica sand. Calcite may derive from additives or slight carbonation, and albite (sodium feldspar) could be an impurity from the sand source. Notably, typical high-temperature phases such as mullite were not observed in this sand, implying that the sand was not heavily altered mineralogically by the casting process (likely because the sand was used with a low-melting binder rather than being part of the metal).

The XRD diffractogram of the UFS is shown in Figure 3. The pattern is dominated by sharp peaks of quartz (SiO₂), confirming it as the predominant crystalline phase, comprising approximately 98% by weight of the material [29,30]. This is evidenced by intense diffraction peaks at 2θ ≈ 20.8°, 26.6°, and 50.1°, corresponding to the reflections (100), (101), and (112) of α-quartz, respectively. In addition to quartz, two minor phases were identified: calcite (CaCO₃) and albite (Na-feldspar and NaAlSi₃O₈), each representing approximately 1 wt%. Calcite produces a distinguishable peak at 2θ ≈ 29.4° ((104) reflection), while albite exhibits weaker peaks around 2θ ≈ 27.9° and 22.1°. The presence of calcite is probably attributable to molding additives such as limestone dust or carbonation of lime from the clay binder during metal casting (for example, reaction of CO₂ with calcium compounds produces calcite on sand grain surfaces). Albite may originate from the source of natural sand, which commonly contains trace amounts of feldspar. In particular, no peaks attributable to clay minerals (e.g., glauconite) are visible. This absence is expected, since the binder likely became amorphous during thermal exposure or is present in amounts below the detection limit. Similarly, iron oxides and other minor components did not produce distinct diffraction peaks. Importantly, the lack of mullite (3Al₂O₃·2SiO₂) suggests that the clay of the sand did not undergo significant ceramic transformation—in agreement with typical foundry temperatures which generally do not exceed ~1000 °C. Therefore, from a mineralogical point of view, the UFS is compositionally similar to natural quartz sand, with trace amounts of calcite and albite as benign impurities. However, XRD does not detect amorphous or organic content, such as residual carbon or degraded binder phases, which may still be present. The nearly complete quartz composition suggests that UFS will behave as an inert fine aggregate in concrete, with minimal pozzolanic or cementitious reactivity. Its high crystallinity also implies a relatively low surface area, except where it is coated with fine binder residues.

3.1.2. Characterization of CBA

XRD analysis was used to identify the crystalline and amorphous phases in the CBA. Mineralogically, the bottom ash is partly vitreous. XRD analysis (supported by the literature [12]) indicates a significant amorphous phase (glassy silicate) along with crystalline phases such as quartz (from unburnt mineral inclusions or impurities of sand in coal), mullite (Al₆Si₂O₁₃, formed from clays in coal at high temperature), anorthite (CaAl₂Si₂O₈, a calcium feldspar), and minor iron oxide phases such as magnetite (Fe₃O₄) [11]. Sulfate compounds (e.g., anhydrite or gypsum) may also be present in fully burned and oxidizing conditions [8]. The bottom ash used in this study showed dense, glassy particles and porous, cinder-shaped particles; the latter tend to contain higher sulfate (from partially reacted sulfur in coal) and unburnt carbon.

The XRD patterns of the CBA samples are shown in Figure 4a (dark slag) and Figure 4b (light slag). Both patterns exhibit a broad amorphous hump centered between 2θ ≈ 25 and 30°, indicative of a significant glassy (non-crystalline) phase formed by the rapid cooling of molten mineral matter during combustion.

For dark CBA (Figure 4a), the crystalline phases superimposed on the amorphous background include quartz (SiO₂), identified by the sharp peak at 2θ ≈ 26.6°, and mullite (Al₆Si₂O₁₃), evidenced by peaks at approximately 16.4°, 25.9°, and 33.2° 2θ. These phases are commonly associated with residual sand and kaolinite-derived transformation products in coal ash. Additional minor peaks may correspond to magnetite (Fe₃O₄) or hematite (Fe₂O₃), as well as calcite (CaCO₃), observed near 2θ ≈ 29.5°, possibly carbonation or unburnt lime. A weak peak around 11.6° can suggest the presence of gypsum (CaSO₄·2H₂O), potentially formed by sulfide oxidation in a moist environment. The light CBA (Figure 4b) similarly shows quartz and mullite as dominant crystalline phases but with clearer mullite reflections, suggesting a slightly different combustion or mineral transformation pathway. The presence of anorthite (CaAl₂Si₂O₈) and gehlenite (Ca₂Al(AlSi)O₇) is also suspected based on peak positions and the prior literature on CBA mineralogy. The iron-containing phases were less intense in the light slag, possibly due to a lower content or greater dispersion in the glassy matrix.

In general, both ashes are mineralogically dominated by inert aluminosilicates with a high amorphous content (estimated 40–60% by difference). Because of the low crystallinity of many phases and overlapping signals, quantitative phase analysis was not performed. However, quartz appears to be the most abundant crystalline constituent in both ashes. From a practical standpoint, the CBA used in this study was incorporated as coarse aggregate and is expected to act primarily as an inert filler. The amorphous glass may exhibit limited pozzolanic reactivity if finely ground, but in aggregate form, its effect on hydration is minimal. A minor concern is the presence of sulfates, as inferred from XRD and supported by a measured SO₃ content of ~0.85%. However, the use of sulfate-resisting cement mitigates the risk of internal sulfate attack. Thus, any influence on concrete properties is likely to arise from physical characteristics (particle shape, surface texture, and porosity) rather than from mineralogical reactivity.

3.1.3. Concrete

XRD analysis was performed on all concrete mixtures (M-1 to M-6) to evaluate the mineral phases after 28 days of hydration and to assess the influence of UFS and CBA on the phase assemblage. The XRD patterns of all samples are presented in Figure 5. The diffraction results are consistent with the expected mineralogical compositions based on the proportions of the mixture and confirm the largely inert behavior of the added waste materials. Figure 5 highlights the influence of increasing the content of waste on the crystalline composition of the concrete matrix.

The reference concrete (M-1), which did not contain waste additives, exhibited prominent peaks of portlandite (Ca(OH)₂) and calcite (CaCO₃), along with a broad hump corresponding to C-S-H gel, the main hydration product. Weak reflections from the unhydrated phases of the cement, such as alite and belite, were also observed, which is typical of the hydrated Portland cement paste [31]. In M-2, where 30% UFS was used as a sand replacement, intense peaks of quartz (SiO₂) dominated the pattern, with minor contributions from albite (NaAlSi₃O₈) and calcite, suggesting the presence of natural feldspar and possible carbonation. These phases are characteristic of UFSs, which primarily consist of high-purity quartz with minor impurities [32,33]. For M-3, containing 40% CBA, the diffraction pattern showed weaker quartz peaks and the appearance of mullite (Al₆Si₂O₁₃) at ~16.4°, 25.9°, and 33.2° 2θ, a phase commonly formed from the thermal transformation of kaolinite in coal combustion. A broad amorphous hump centered around 25–30° 2θ indicates the presence of a glassy aluminosilicate matrix, which is typical of CBA [34]. Mixtures M-4 and M-5, containing combinations of UFS and CBA (70% total replacement in M-4 and 50% in M-5), displayed composite patterns with contributions from both types of waste. Peaks from quartz, mullite, and minor feldspar were present, along with hydration products such as portlandite and ettringite. The intensities of the mullite and the amorphous background were lower in M-5, consistent with its reduced ash content. Finally, M-6, a variant with 30% UFS and 20% CBA, exhibited a similar pattern to M-5, confirming that the substitution levels were below the threshold to significantly disrupt hydration or form new cementitious phases. The XRD results across all samples confirm that both UFS and CBA act primarily as inert fillers, with minimal pozzolanic reactivity detectable by XRD.

3.2. SEM/EDS Analysis

3.2.1. Morphology of UFS and CBA Particles

The SEM examination provides information on the texture and shape of the surface of the waste aggregates. Figure 6 (SEM image of UFS at 400× magnification) shows that the spent foundry sand grains are generally subangular to rounded silica particles, similar in size to regular sand (diameters ~0.2–0.5 mm).

However, the surfaces of these grains are coated with a layer of fine particulate matter. This fine material appears as dark patches and loose fines on the sand surface. EDS analysis of these areas (spectrum in Figure 6) detected elements such as carbon (C) and aluminum (Al), in addition to the predominant silicon (Si) and oxygen (O) of quartz, and a small amount of iron (Fe). Carbon likely originates from residual coal dust used in the sand binder of the foundry, and Al (with Si and some Fe) corresponds to the clay binder (glauconite or bentonite, which are aluminosilicates with possibly Fe oxides). The presence of this fine binder film on the UFS is significant. It creates a weaker and more chemically complex interface when these sand grains are embedded in cement paste. The binder is essentially a very fine powder (much finer than cement) that could inhibit direct contact between the cement paste and the sand surface and could even consume some mix water (if clay) or act as a release agent (if carbonaceous) at the interface. This observation aligns with the hypothesis that the residual binder of the foundry sand can loosen the paste–aggregate bond [29,30,35]. No distinct resins were observed (some foundry sands use organic resins). A green sand (clay/coal) binder was used, which leaves an inorganic residue. The UFS particles did not show obvious signs of microcracks or crushing from their previous use, so their intrinsic strength as sand is intact. It is the surface condition that probably affects performance.

The CBA particles exhibit a starkly different morphology. Figure 7 shows a SEM image of a representative CBA grain.

One can observe highly irregular shapes with porous microstructure. Some particles have a pumice-like appearance, full of voids and open pores, especially the lighter-colored pieces that are rich in oxidized compounds (these often have a reddish-brown tint due to iron oxidation). Other particles are denser and glassier (dark, vitrified pieces), and these tend to have smoother surfaces but often with sharp edges from fracture. Within the porous particles, unburnt carbon inclusions were noted as black unreacted spots—EDSs in those showed primarily carbon. In contrast, the matrix of the ash particles showed peaks of Si, Al, O, and also calcium (Ca) and iron, indicating that the glassy phase contains dispersed Ca and Fe (likely as tiny crystallites of anorthite, melilite, magnetite, etc.). A notable feature is that sulfate salts were seen at some pore surfaces. EDS detected sulfur (S) in some localized spots in porous ash, suggesting the presence of sulfate crystals (possibly gypsum that precipitated from weathering of ash). The internal porosity of CBA means that, when used as aggregate, these particles can absorb water from the mix and later create voids if that water leaves. In addition, the strength of these porous particles is expected to be low—some ash grains can be crushed under finger pressure. This has implications for concrete because the CBA, especially the porous fraction, will be the weak inclusions that may fracture under load.

3.2.2. ITZ and Hydration Products

After 28 days of curing, polished sections and fracture surfaces of selected concretes were analyzed. The mix M-2 (30% UFS), M-3 (40% coarse CBA), and M-5 (20% fine CBA), which represent the cases of each waste individually, are documented in this article. In the control concrete (M-1), a typical ITZ around natural sand and basalt aggregate was about 10–20 μm thick, with some calcium hydroxide crystals and ettringite observed (through EDS mapping: Ca and S enriched near the aggregate surfaces).

Microstructural observations of M-2 concrete that incorporates UFS reveal several distinct features that point to the impact of UFS on the ITZ and the bulk cement paste (see Figure 8).

As shown in Figure 8a, a UFS grain has been cleanly removed from the paste matrix during fracture, leaving a void with smooth edges. This morphology supports the hypothesis that the residual binder prevents adequate mechanical interlocking and chemical bonding, thereby weakening the ITZ. The presence of such pull-out zones with minimal residual paste suggests poor adhesion at the aggregate interface. Complementarily, Figure 8b presents a region of the UFS–paste interface, where small gaps and partial debonding are visible along a line scanned by EDS. These features likely result from a carbonaceous or clay coating present in the UFS grains, which interferes with cement hydration and the mechanical interlock between the sand and the paste. ITZ in these areas appears wider and less dense compared to conventional concrete, consistent with an interface compromised by residual foundry materials. Further information is provided by the elemental distribution across the UFS–paste boundary, captured in the EDS line scan in Figure 8c. The scan reveals a sharp decrease in calcium and a simultaneous increase in silicon at the grain surface, confirming the presence of silica sand. In particular, a narrow interfacial band is observed that is enriched in aluminum and iron, which is consistent with the presence of residual clay binder from the foundry process. These elements are not typical of cement hydration products, and their presence reinforces the idea that the ITZ is chemically and structurally altered by UFS contamination. Despite these interfacial effects, the bulk cement paste remained largely unaffected. Figure 8d shows a representative SEM image of the paste matrix, located away from the aggregate particles, where hydration appears normal and well developed at 28 days. Elemental analysis of this region gave an average composition of approximately Ca, 18%, Si, 17%, Al, 2%, Fe, 2%, and S, ~0.8%, closely resembling that of ordinary hydrated cement paste. The modest increase in Si and Al may indicate a pozzolanic contribution of the clay components of UFS. The sulfur detected likely originates from both gypsum in cement and minor sulfur-bearing phases in UFS, such as glauconite or coal residues. Carbon was intentionally excluded from EDS quantification in polished sections to avoid interference with the conductive coating; however, EDS conducted on fracture surfaces detected local carbon peaks, implying the presence of embedded fine carbonaceous fragments from UFS in the hardened matrix. In general, these microstructural and chemical characteristics confirm that UFS affects the integrity of ITZ by weakening the paste–aggregate bonding, although the bulk cement matrix remains similar to that of conventional concrete.

In M-3 concrete (40% coarse CBA), the microstructure was markedly different in the vicinity of the CBA aggregate. The ITZ was less critical than in conventional systems because the entire ash particle was often failing under stress. SEM images of M-3 after the compression test show CBA particles that fractured internally or cleaved along weak planes (Figure 9a). Fresh fracture surfaces were visible through the interior of the ash, along with areas where the ash had separated from the paste. In many instances, the cement paste still adhered to the irregular surface of the ash, suggesting that the paste–ash bond was not particularly weak. Rather, the ash particles themselves disintegrated, indicating a cohesive failure within the aggregate. This interpretation is supported by the observation of ash fragments still embedded in the paste on one side of the crack and a corresponding negative imprint on the opposite surface (Figure 9b). This contrasts with natural basalt aggregates in the same mix, which remained intact during fracture, with cracks propagating around them, indicating failure along the ITZ rather than through the aggregate.

This change in failure mode, from interfacial debonding to aggregate crushing, confirms that the CBA acted as the strength-limiting phase in M-3. EDS analysis of the paste adjacent to the CBA particles showed slightly elevated levels of aluminum and silicon, possibly due to limited pozzolanic activity from fine ash debris embedded in the paste. However, no distinct secondary C–S–H or new reaction products were detected at 28 days. The sulfur introduced by the CBA did not appear to form deleterious phases: ettringite was present in typical quantities, and no signs of an excessive sulfate reaction (e.g., thaumasite) were found. This is likely attributed to the use of sulfate-resistant cement in the mix.

An additional observation in M-3 was the presence of microcracks that run through the paste near ash particles, even before mechanical loading. These cracks may have originated from differential drying shrinkage or thermal stress due to the porous nature of ash, which could have absorbed more water and induced local stress gradients. This early age microcracking may contribute to the reduced compressive strength observed in M-3.

For M-5 (20% fine CBA), the microstructure was more similar to a control. Since only fine ash (under 4 mm) was added and at a moderate replacement level, the majority of the fine aggregate was still natural sand. The fine CBA particles in M-5 behaved like a porous sand. SEM images show that most of these particles were embedded in the paste without undergoing a full fracture (Figure 10a)—probably because their smaller size made them less critical in crack propagation. Some ash particles showed signs of cracking, but overall the impact on strength was limited. The ITZ around fine ash grains can be slightly more porous due to the particle’s porosity; this could allow the paste to penetrate the pores, creating a degree of mechanical interlock, although it could also introduce local stress concentrations. The SEM analysis of the general matrix (Figure 10b) showed a stable microstructure without signs of deleterious products or phase separation, consistent with the chemical stability confirmed by EDS (no abnormal concentrations apart from minor variations in Fe and Al). In general, the microstructural study confirms that, at 20% fine ash substitution, the concrete structure remains relatively sound—consistent with the relatively high compressive strength measured (65.8 MPa, which is close to the control).

3.2.3. EDS Spectra

Representative EDS spectra for UFS and CBA are given in Figure 6 and Figure 7. Figure 6 (for the UFS grain surface) shows strong Si and O peaks, a moderate Al peak, and a small Fe peak. The Si/O ratio ~1:2 is consistent with silica (quartz or clay). The Al and Fe signals confirm the clay (glauconite contains iron). A tiny Mg or K peak might also be present (glauconite is a mineral containing K and Mg). Importantly, a noticeable Au peak (~2.1 keV) is seen due to the gold coating; this was ignored in chemical analysis. Figure 7 (EDS of a CBA particle cross-section) would show Si and O as major (from glass), a significant Al, and notable Ca and Fe peaks. In an analysis of a dense ash part, we found ~Si, 25%, Al, 8%, Ca, 3%, Fe, 5%, and O, ~55% (wt.%), which aligns with a mix of silicate and oxide phases. Another spectrum in a porous part of the ash showed S, ~1–2%, confirming sulfate in that area. Tables in Figure 8, Figure 9 and Figure 10 compile EDS elemental percentages for different points in concrete. The data basically support that the chemical makeup of the cement paste is not drastically changed by the inclusion of UFS or CBA; the main effects are physical (as discussed).

The microstructural analysis reveals two key factors influencing the mechanical performance of the investigated concretes. First, the presence of UFS introduces interfacial weaknesses due to the residual clay and carbonaceous binder films coating the sand particles. These coatings hinder the effective bonding with the surrounding cement paste, resulting in the formation of partially debonded interfacial zones. Such zones can serve as preferential sites for crack initiation and propagation under mechanical loading, contributing to the observed reduction in compressive strength. Second, the coarse CBA used as aggregate exhibits inherently low strength and high porosity. Under stress, these particles tend to fracture internally rather than transfer load through the matrix. In concretes with a high CBA content, this leads to a change in failure mode, from cracking through the paste or interfacial zones to crushing of the aggregate itself, ultimately resulting in a significant decrease in compressive strength. These microstructural findings correlate well with the mechanical results discussed next.

3.3. Compressive Strength

The 28-day compressive strength results for all concrete mixes (M-1 to M-6) are summarized in Figure 11 and

Table 4. Statistical summary of 28-day compressive strength results for concrete mixes M-1 to M-6 (n = 6 specimens per mix).

Mix ID	Mean (MPa)	Std Dev (MPa)	CV (%)
M-1	75.8	1.40	1.85
M-2	32.7	3.13	9.58
M-3	20.5	3.50	17.09
M-4	13.8	3.85	27.83
M-5	65.7	2.33	3.54
M-6	30.2	3.31	10.97

. Each value represents the mean of six tested specimens, with the corresponding standard deviation and coefficient of variation (CV) reported. The control mix M-1 exhibited the highest compressive strength of 75.80 MPa, with excellent repeatability (CV = 1.85%). The addition of waste materials led to varying degrees of strength reduction, strongly dependent on the type and amount of replacement. Mix M-2 (30% UFS) showed a significant reduction to 32.70 MPa (CV = 9.58%), while mix M-3 (40% coarse CBA) reached 20.50 MPa with higher variability (CV = 17.09%). The combination of UFS and fine CBA in the M-4 mix resulted in the lowest strength of 13.80 MPa and the highest variability (CV = 27.83%), reflecting the heterogeneity and poor workability of this mixture. In contrast, the M-5 mix (20% fine CBA) retained a high compressive strength of 65.70 MPa with very good repeatability (CV = 3.54%). The M-6 mix exhibited a moderate strength reduction of 30.20 MPa, with a CV of 7.23%, confirming the combined influence of UFS and coarse CBA. In general, statistical analysis confirms that the experimental results are consistent and reliable, with variability levels typical for concretes containing recycled or waste-derived aggregates. The observed trends in the reduction in compressive strength align well with the expected influence of waste type and replacement level, providing a robust basis for further discussion and interpretation. Although the compressive strength of mix M-6 was numerically lower than that of M-2, the difference was not statistically significant at the 95% confidence level. This can be attributed to the relatively high standard deviations in both groups, leading to overlapping confidence intervals despite differences in mean values. Such cases are not uncommon in studies involving recycled materials, where material heterogeneity can increase result dispersion and obscure statistically meaningful differences.

Interestingly, despite the 20% addition of coarse CBA in M-6, its compressive strength remained comparable to that of M-2. This may be attributed to the partial compensatory effects of the porous CBA particles, such as internal curing and enhanced paste–aggregate interaction. Although CBA typically reduces mechanical strength due to its lower stiffness and higher porosity, its high water absorption capacity can contribute to better hydration and microstructural refinement in the surrounding cement paste. Similar observations have been reported in studies involving recycled or low-density aggregates [36].

A one-way analysis of variance (ANOVA) confirmed that the type of concrete mix had a statistically significant effect on the compressive strength at 28 days (F = 411.76, p < 0.0001; see

Table 5. Summary of one-way ANOVA for compressive strength of concretes M-1 to M-6.

Source	Sum of Squares	Degrees of Freedom	F	p-Value
C(Mix)	18,943.50	5	411.76	1.10 × 10−26
Residual	276.04	30	—	—

). In the ANOVA model, the designation of the concrete mixture (M-1 to M-6) was treated as a categorical factor C(Mix).

To further identify significant differences between individual mixes, a Tukey HSD post hoc test was performed. The results (

Table 6. Results of the Tukey HSD post hoc test for 28-day compressive strength of concrete mixes M-1 to M-6 (α = 0.05).

Comparison	Mean Difference (MPa)	p-Value	Lower CI	Upper CI	Significant (Yes/No)
M-1 vs. M-2	−43.13	0	−48.46	−37.81	yes
M-1 vs. M-3	−55.3	0	−60.63	−49.97	yes
M-1 vs. M-4	−61.97	0	−67.29	−56.64	yes
M-1 vs. M-5	−10.1	0	−15.43	−4.77	yes
M-1 vs. M-6	−45.63	0	−50.96	−40.31	yes
M-2 vs. M-3	−12.17	0	−17.49	−6.84	yes
M-2 vs. M-4	−18.83	0	−24.16	−13.51	yes
M-2 vs. M-5	33.03	0	27.71	38.36	yes
M-2 vs. M-6	−2.5	0.71	−7.83	2.83	no
M-3 vs. M-4	−6.67	0.01	−11.99	−1.34	yes
M-3 vs. M-5	45.2	0	39.87	50.53	yes
M-3 vs. M-6	9.67	0	4.34	14.99	yes
M-4 vs. M-5	51.87	0	46.54	57.19	yes
M-4 vs. M-6	16.33	0	11.01	21.66	yes
M-5 vs. M-6	−35.53	0	−40.86	−30.21	yes

and Figure 12) revealed that almost all pairwise comparisons were statistically significant (p < 0.001), except for M-2 vs. M-6, which showed no significant difference (p = 0.202).

These findings confirm that both the type and the amount of waste used in concrete mixes significantly influenced the compressive strength. The particularly close performance of M-2 and M-6 highlights the compounded effects of UFS and CBA at certain replacement levels.

4. Conclusions

This study investigated the UFS and CBA as partial replacements for fine and coarse natural aggregates in concrete, with a focus on their effects on mineralogy, microstructure, and mechanical performance. The following key conclusions can be drawn:

• XRD confirmed that UFS is predominantly composed of quartz, while CBA exhibits a largely amorphous matrix with crystalline inclusions of quartz, mullite, and aluminosilicates. Both materials act mainly as inert fillers without significant pozzolanic reactivity.
• SEM/EDS revealed that the UFS grains are coated with residual clay and carbonaceous binder films, which weaken the ITZ between sand and cement paste. In contrast, CBA particles are highly porous and mechanically fragile, often fracturing under loads and thus becoming the weak phase in the composite.
• Compressive strength tests demonstrated that high replacement levels of UFS and CBA substantially reduce the early age strength of concrete. The severity of this reduction depends on both the type and the amount of waste used. UFS primarily affects strength by degrading ITZ quality, while CBA contributes to loss of strength through aggregate fracture and increased porosity.
• Moderate substitutions, particularly 20% replacement of fine aggregate with CBA, were found to be feasible, resulting in only modest strength reduction and maintaining performance within the range suitable for structural applications. On the contrary, high levels of UFS (≥30%) or coarse CBA (≥40%) caused significant strength penalties and are not recommended without prior treatment.
• Practical implications include the potential for sustainable utilization of UFS and CBA in concrete, provided replacement levels are carefully controlled. Based on the experimental results, replacement levels up to 10–15% for UFS and up to 20% for fine CBA are recommended, as these combinations maintained acceptable mechanical properties. These thresholds reflect the performance boundaries observed in mixes M-5 and M-6, while higher dosages (e.g., 30% UFS or 40% coarse CBA) resulted in excessive strength reduction. Processing techniques such as binder removal (for UFS) and sieving or beneficiation (for CBA) are also encouraged to improve material quality and ensure better integration into the cement matrix.
• Future research should focus on long-term performance (≥90 days), durability aspects (e.g., permeability, freeze-thaw resistance, and sulfate attack), and optimized treatment methods for these industrial by-products. Properly engineered mixes could allow the safe and effective use of UFS and CBA in sustainable concrete applications.

In conclusion, the combined mineralogical, microstructural, and mechanical insights gained from this study provide a scientific foundation for the responsible incorporation of UFS and CBA into concrete, contributing to resource conservation and waste valorization in the construction industry.