Sodium silicate bonded waste foundry sand - A substitute for fine aggregates in concrete and a potential material for cement phase synthesis

. Sodium silicate bonded Waste Foundry Sand (WFS) is being discarded by the foundries after single use. The studies so far are suggestive that the WFS after being discarded cannot be reclaimed by any physico-chemical methods and the silica transformation is within the crystal structure. Hence such sand utilisation in bulk quantity will be one of the options for managing the issue of sodium silicate bonded WFS. Recent research trend in construction materials involve utilization and blending of different industrial byproducts and waste materials to solve the environmental problems. In the present work, sodium silicate bonded WFS was used as a replacement for fine aggregates in concrete in varying percentages of 0%, 10%, 20%, 30%, 40% and 50% by weight. The 30% replacement showed better compressive strength compared to control samples. Further, these samples also passed the durability tests like rapid chloride penetration, water absorption, sorptivity, rebound hammer and ultrasonic pulse velocity. Both the strength and durability results can be attributed to the property of the sand in terms of its size, shape and its reactivity. To prove this hypothesis further fine sodium silicate bonded WFS (less than 45µm) was mixed with calcium carbonate in appropriate molar proportions. The mixture was calcined at 1400 o C. This mixture was analyzed using XRD and the results revealed that alite and belite phases were generated during the reaction. This gives new dimension to utilize sodium silicate bonded WFS in concrete or as source of silica in


Abbreviations:
WFS-Waste Foundry Sand; XRD-X-Ray Diffraction; JCPDS-Joint Committee for Powder Diffraction Standards; OPC-Ordinary Portland Cement, M-Sand-Manufactured Sand, ASTM-American Society for Testing and Materials, IS-Indian Standards

Introduction
High quality silica sand is conventionally used in the ferrous and non-ferrous metal casting industries in the mould making process.So far silica sand has not found its replacement in metal casting due to its high thermal conductivity and stability.Due to increased stress on the ecosystem, the availability of fresh silica sand poses new challenge to the foundry sector.Before disposal foundries successfully reuse this sand multiple times till it loses its binding property.Such sand is termed as Waste Foundry Sand (WFS) [1].Approximately 6-10 million tonnes of WFS is originating in US.Very less quantity, approximately 15% is recycled back and remaining sand does not find any application apart from landfilling [2].About 5000 foundries are operational in India, which will produce 17,10,000 tonnes WFS per year [3].Attempts to use WFS in different sectors have been reported by earlier researchers which include road construction, replacement in concrete for fine aggregates, manufacturing of bricks, blocks and pavers etc. [4,5].Research also shows the use of WFS in asphalt concrete [6].On the other hand many studies reported the drawbacks of leachate produced from usage of such WFS in various sectors [3].However still the bulk utilisation of sand is not up to the expectations of these various sectors.The fines generated and cost of reclamation are the factors that negatively influence the reuse of such sands.
The classification of foundry sands depends upon the type of binder systems used in metal casting.Generally, two types of binder systems are used, based on that foundry sands are classified as: Clay bonded sands (green sand) and Chemically bonded sands.Green sand is composed of naturally occurring materials which are blended together; high quality silica sand (85-95%), bentonite clay (4-10%) as a binder, a carbonaceous additive (2-10%) to improve the casting surface finish, and water (2-5%) [7].The addition of carbon makes the molding mixture very harmful to the natural environment.It also creates harmful working conditions [8].The foundries use variety of chemical binder systems namely, phenol-urethanes, epoxyresins, furfuryl alcohol, and sodium silicates.Though binders like phenolic urethanes and furans provide good strength, casting quality, shake-out etc., they thermally decompose during the casting process because of high temperature and release gaseous hazardous pollutants [9].Sodium silicate is widely used because of the ease in mold making.CO2 is used as hardener in sodium silicate bonded sand.The CO2 process gives instant hardening by polymeric reactions between sodium silicate and sand.It proves to be advantageous process as the mold making process requires less time.It has gained popularity in last 25 years due to its insitu curing across the world [10].The present study utilizes sodium silicate bonded WFS.In sodium silicate as a binder system, the sodium silicate reacts with foundry sand and undergoes polymeric chain reaction in presence of CO2 giving rigidity to the casting moulds.The reaction of sodium silicate follows three step process, initially the sodium silicate reacts with CO2 followed by segregation of combined water from sodium silicate and ultimately leading to hydration of sodium silicate.[10] Such sand casting is used in metal casting process during which the sand is exposed to the molten metal and high temperature.After demoulding such sand loses its binding property.It is reported that the binders added during the casting gives a thin film on the sand surface.[11,12].Several attempts have been made to reclaim and reuse the WFS in the foundry itself, via wet reclamation, dry reclamation, mechanical agitation, water wash and chemical wash etc., These methods are found to be inefficient and does not give satisfactory binding strength to cast the mold.[11,[13][14][15][16]. Also, when the sand is exposed to elevated temperatures in the foundry during metal casting process, the phase transformation of silica is reported at different temperatures by various authors.The quartz undergoes phase transformation from α-quartz to β-Quartz at 573°C and further to tridymite and cristobalite at 870°C, and 1470°C, respectively [17][18][19].When sand gets exposed to these temperatures, the structure of the sand is likely to change.Hence rebonding of such sand with fresh sodium silicate is not possible.
The reasons for the lost binding property of such sand remains unexplored.Hence it is worth to investigate the role of such sands when they are physically replaced for fine aggregates in cement concrete.Additionally, if such sands can be utilized as a source of silica in cement industry when their size is significantly reduced.Considering above discussion, the main objective of the present work is to utilize sodium silicate bonded WFS as a partial replacement for fine aggregates in cement concrete in varying percentages of 10, 20, 30, 40 and 50% and check for its strength and durability.Also, to utilize the powdered WFS (less than 45 µm) as a source of silica to synthesize cement phases such as alite and belite by reacting with lime.The mineralogical characterization of these samples is done using XRD to identify the minerals responsible for the strength.

Methods
The raw materials used in the present study for manufacturing of concrete cubes, were tested as per the standard procedures.The OPC was tested for specific gravity and initial and final setting time as per IS 4031 (Part-11):1988 and IS 4031 (Part-5):1988 respectively [20,21].The fineness modulus and grading of M-sand, WFS and coarse aggregate was carried out as per IS: 383-1970 [22].Also the specific gravity and water absorption were determined as per IS: 2386 (Part-3)-1963 [23].The concrete mix design was done as per IS:10262-2019 [24].The mix design adopted for the present study after trial mix is 1:1.67:3.1 (Cement: Fine Aggregates: Coarse Aggregates) with water-cement ratio of 0.5.The details of the concrete mix used for different combination of cubes casted in the present work is shown in Table 1.The concrete blocks casted with only M-Sand was termed as control mix and designated as 0%.The other sets of cubes were casted by replacing M-sand with WFS by 10, 20, 30, 40 and 50%.All the sets were casted in triplicates and were checked for strength and durability.Slump and compaction factor tests were carried out on fresh concrete as per IS: 1199-1959 [25].Compressive strength test was carried out after 7 and 28 days of curing as per IS:516-1959 [26].Also, ultrasonic pulse velocity and rebound hammer tests were carried out on hardened concrete as per IS:13311-1992 [27,28].Durability properties, such as water absorption, sorptivity and rapid chloride permeability tests were done after 28 days of curing in accordance with ASTM C 642-06 and ASTM C-1202 [29,30].
To utilize the WFS in cement manufacturing process, WFS was crushed to powdered form (less than 45 m) and was mixed with calcium carbonate and then calcined at 1400 o C in the high temperature furnace.Excess amount of calcium carbonate was taken initially as calcium carbonate decomposes to give CaO and this CaO was considered in molar proportions to form Belite (C2S) and Alite (C3S) after reaction with WFS.
In order to form C3S, 3:1 molar proportion of lime (CaO) and silica (SiO2), was calcined at 1400°C for 4 hours.The mixture was allowed to cool and then ground.The calcination procedure was then again repeated to ensure uniform mixing and complete reaction of CaO and SiO2.Similar procedure was used to get C2S, 2:1 molar proportion of lime and silica, calcined at 1400°C for 4 hours.[31]

Analytical Methods
The concrete samples which gave the best (30% WFS) and least (50% WFS) compressive strength after 28 days of curing were considered for XRD analysis and the results were compared with control sample (0% WFS).The representative samples from the concrete blocks were collected, finely ground and acetone was passed to seize the hydration reactions.Such samples were oven dried and then were taken for XRD analysis.XRD analysis was carried out on the samples to check the formation of cement phases generated by the reaction of finely ground silica with lime at 1400 o C to check the presence of Alite(C3S) and Belite(C2S).The XRD analysis was performed with 2Ɵ values ranging from 10 to 90 o using Rigaku Smart Lab diffractometer at constant scanning speed of 1 o /minute with CuKα radiation (λ= 1.540 nm).The resultant peaks were then compared with Joint Committee on Powder Diffraction Standards (JCPDS) cards to identify the constituent minerals.

Results and Discussion
The workability, strength and durability test results of the different combinations of concrete samples are discussed in this section.

Strength test results of different concrete samples
Compressive strength was checked after 7 and 28 days of curing for both control and samples.Other strength tests like ultrasonic pulse velocity and rebound hammer was conducted after 28 days of curing.Compressive strength of concrete samples replaced with WFS increased upto 30% replacement after 7 and 28 days of curing compared to control.However beyond 30% replacement with WFS, compressive strength showed a decreasing trend.The marginal increase in the compressive strength of concrete samples replaced with SWFS was seen up to 30%.However, this strength was found to reduce beyond 30% replacement for both samples cured for 7 and 28 days compared to the control.The maximum compressive strength at the end of 28 days was seen in the 30% replacement sample (28.70 N/mm 2 ) compared to control sample (26.33 N/mm 2 ).The results were supported by ultrasonic pulse velocity and rebound hammer tests which categorises concrete samples under excellent and satisfactory respectively.

Durability test results of different concrete samples
From the Figure 2, 30% replacement of WFS gave best compressive strength results among other samples.From Figure 3, when the same samples were subjected to water absorption and water sorptivity, it was seen that the water absorption and sorptivity were highest for control samples followed by 10% and 20% SWFS replacement.However, the least water absorption (3.8%) and sorptivity (6.78%) was seen for 30% replacement compared to the control (5.19% and 7.01%).Marginal increase in water absorption and sorptivity was seen for 40% and 50% replacement.It can be postulated that 30% SWFS replacement optimally gives a concrete cube in which the sand particles are densely packed leading to less water absorption and sorptivity compared to other samples.This could also possibly explain marginally higher compressive strength due to fewer voids compared to other samples.All the samples showed good resistance to chloride ions.It was expected that when the fine aggregates are replaced with finer aggregates such as SWFS, the concrete is expected to be dense which will show resistance to the anionic attacks such as chloride which is evident from the results.

XRD graph of specimens containing different percentages of WFS
The XRD analysis was carried out for 0% replacement which served as a control sample, 30% replacement, which gave the best compressive strength and 50% replacement which gave the least compressive strength.All these samples were collected at the end of 28 days of curing to understand the cement hydration phases and the effect of SWFS replacement on strength parameters.From Fig. 4, the major peak of SiO2 at 26.6 o and 20.8 o was observed in 30% and 50% of SWFS replacement.This peak is originating from SWFS which matched with JCPDS card No. 872096.The peak intensity was highest in 50% SWFS compared to 30% SWFS.However, these peaks were absent in the control sample though they had silica in their composition because the silica in the control sample is of M-Sand origin.The peak position of 27.8 o in 0% replacement matched with JPCPDS card No. 830541 which is of SiO2.These peaks were with less intensity in 30% and 50% samples as M-Sand was replaced with SWFS in respective samples.
The peak position at 28.03 o was present in all three samples.The highest intensity of this peak was seen in 30% SWFS replacement followed by control and 50% SWFS replacement.This peak belongs to calcium aluminium silicate matching JCPDS card no.411486.The highest intensity peak 18.1 o of calcium aluminium oxide was seen in 50% SWFS followed by 0% and 30% SWFS.This peak matched with JCPDS card no.780910.Both these minerals could have originated from the cement phase and could contribute to the strength reactions leading to the formation of C-S-H.The calcium silicate hydrate peak was seen at 29.5 o which matched with JCPDS card no.150584.The intensity of this peak was highest in 30% SWFS followed by 0% control and the least with 50% SWFS.This finding justifies that C-S-H is a strength contributing mineral.These findings agree with compressive strength test results in

Synthesis of cement phases from WFS
To check if the SWFS could be a source of SiO2 in the cement industry, the experiments were carried out by choosing appropriate proportions of finely ground SWFS (<45m) and lime as mentioned in section 3).The post-calcination samples were chosen for XRD analysis to see the presence of cement phases.Fig. 5 shows the XRD analysis of the samples after the specified calcination of CaCO3 and SiO2 at 1400°C.Above 800°C the decomposition of CaCO3 takes place and CaO will get formed, which in turn reacts with silica (SiO2) to give Alite (C3S) and Belite (C2S) phases.With these XRD results it can be concluded that C2S and C3S were formed in the experiment.The experiments showed that the SWFS can be used as a potential source of SiO2 in the cement industry where it can find the bulk utilisation and replace other siliceous materials.
The reduced size of silica proves to react with lime which might find applications in various sectors involving high temperature.

Conclusion
The study demonstrated that optimum replacement of 30% WFS gives best compressive strength.Other strength and durability properties tested are also in alignment with the found results.The study brings out the important finding that the strength could be attributed to optimum physical replacement of WFS along with the strength contributing hydration minerals.
The findings also show that reduced size of WFS finds its application in cement industry as a source of silica due its reactivity with lime.
Slump and compaction factor of fresh concrete and the concrete samples replaced with WFS are shown in Fig 1.The control mix with 0% WFS showed slump value of 200 mm whereas the concrete samples with 10, 20, 30, 40 and 50% WFS replacement showed slump values of 180, 180, 170, 165 and 160 mm respectively.The slump values indicate that as the percentage replacement of SWFS increased the workability reduced.Also, the compaction factor values varied from 0.97 to 0.95 which is in agreement with slump values.The reduction in the workability with increasing percentage of SWFS is attributed to the fineness of SWFS.

Fig. 5 .
Fig. 5. XRD graph of synthesized cement powder The obtained 2Ɵ values were compared with a standard database of the JCPDS.It was found that 2 values at 15.7 o , 20.5 o , 21.8 o , 23.2 o , 26.3 o , 29.6 o , 30.8 o , 32.7 o , 47.6 o , 50.6 o and 52.02 o matched with JCPDS card number 871256.The card was of dicalcium silicate which was present in both samples.Although the molar composition in the mixing of CaCO3 and SiO2 was different, there is the possibility that during the formation of C3S some of C2S might also get formed.It can be seen that C2S peak was present in both syntheses (29.6 o ).However, the peak intensity was more in Fig. 5 (C2S-a) indicating C2S phase was dominating in the mixture.Similarly, the 2Ɵ values in Fig. 5 (C3S-b) peaks at 34.3 o , 38.7 o , and 62.3 o were of C3S which matched with JCPDS card number 730599.These peaks at 34.3 o were also seen in Fig. 5 (C2S-a) but their intensity was less.

Table 1 .
Details of concrete mix used for different combination of cubes.