Challenges and recent progress on the application of rapid sand casting for part production: a review

Various activators, binders, and sand mechanisms are applied to fabricate 3D-printed sand cores and moulds needed for many types of alloy casting [76‐78]. The organic binder often applied for 3D printing can initiate harmful gases in casting, which are not ecologically friendly, although they help the industry better in respect to the casting. Likewise, it has been discovered that furan initiates a health danger to machinists because of its carcinogen. Hence, it is important to search for alternative binders. The available binder in the market for 3D sand core and mould printing is a silicate binder with low gas emission, but this binder requires further study. The features of sand cores and moulds fabricated via the 3D printers can be influenced by diverse conditions during the printing/curing cycle. These features can be affected by the sand properties, the type of binder, concentration of binder, type of printing speed and activator, building orientation of sample printed and job-box co-ordinate, and locker thickness [79‐82]. It has been discovered that reducing the layer thickness and improving saturation enhances the tensile strength and decreases the surface quality of moulds or cores [83‐87].

Huge differences in the physical features of 3DP sand moulds bonded with resin have been encountered by foundries. The features such as permeability and three-point bending (3 PB) strength of 3DP resin-bonded sand moulds rely on specific parameters like binder and moisture content. Generally, higher binder compositions in the sand moulds often produce enhanced mechanical strength, although it consequently evolves more gas during the metal casting. Likewise, a huge quantity of binder allows the 3DP sand mould highly rigid hence inhibiting efficient expansion and prompting hot tearing defects and elevated residual stresses. In comparison, lesser binder quantity minimizes the off-gassing, although it negatively affects the mechanical strength, which causes molten metal penetration into the interstices of the large inter-sand; thus, bloated rough surfaces are developed on the casting. Contrastingly, the mould filling techniques need some measures in air evacuation from the mould cavity because there is often gas compression induced the melting. The melt can capture gases and air in the mould cavity, aided by turbulent filling. Therefore, it is important to remove the gas effectively to achieve a complete casting component with minimal defects [84, 88‐90]. Owing to these facts, the achievement of this modern technique is highly accustomed via the establishment of enough permeable sand moulds with suitable mechanical strength for their operation. As a result of the immense demand for high mechanical strength and dimensional precision of the cast components, 3DP furan bonded resin sand is hugely applied in casting. The furan binder (FB) is comprised of furfuryl alcohol and acid catalyst (toluene sulfonic acid) that establishes a 3-D polymer chain system (furan resin bridges) via polymerization, acid-hardening reaction, and condensation. It has been discovered that the polymer bridges in furan-bonded resin sand mould give additional strength and cohesion to the silica sand particles, which is important to maintain the 3D-printed sand-shaped moulds when in connection with the melt. The reaction from the condensation of furan binder gives water (dehydration); this inclines to lower the curing rate and thus influences the permeability and strength of the moulds [34, 79, 91].

Snelling et al. [92] differentiated the binder burnout features and the strength of the ExOne™ furan resin with a sulphonic acid catalyst system (3D printed) with selected available binder and sand as well as their strength. In assessing the binder burnout, the materials were set to curing cycles of 105^–900 °C. It was discovered that the ExOne™ printed materials possessed an optimum strength of 1.3 MPa because they were well-distributed binder and sand. The printed samples also sustained their shape to greater temperature burnout of 450 °C. It was discovered that the 3D printed mechanism performed more excellently than the conventionally arranged sand mould using phenolic acid and furan resin system. The research revealed that the printed moulds are higher in quality than traditional fabricated moulds and possess lesser and exact well-regulated binder constituents, which brings about gas build-up minimization and thus enhanced casting.

Mita et al. [93] researched the impacts of binder composition on the mechanical strength and permeability of 3DP sand moulds at varying parameters. With regard to distinguishing between the pores, the sand particles, and the furan resin bridges, the SEM images of the samples (Fig. 5) depict the furan resin bridge’s formation within the 3D-printed sample cross-section. In determining the impacts of the inertia pressure, the local permeability of 3DP samples was assessed as a dependency on the injection flow rate. The printed moulds’ mechanical strengths were discovered to be highly reliant on the curing system and the amount of binder. For all the binder compositions, the three-point bending (3 PB) strength was detected to improve when cured at 100 °C and declined at 200 °C (Fig. 6). Hence, the optimum strength of 3 PB was achieved when cured at 100 °C for all the binder compositions. However, the amount of binder within the functional binder range mass fraction (1.02–1.98%) was observed not to significantly influence the samples’ initial permeability before curing. Optimum permeability was achieved at similar parameters as the 3 PB strength. Thus, indicating that the samples’ mechanical strength can be improved within the examined binder range constituents without producing any decline in permeability.

Kafara et al. [94] examined the effect of binder quality on resilience and dimensional accuracy in 3D printing. Five different categories of samples were printed with binder contents equivalent to 40%, 55%, 70%, 85%, and 100%, respectively. Depending on the binder content, the printed samples’ bending strength is depicted in Fig. 7a. It was discovered that the printed samples’ bending strength improves almost linearly in a researched area with increasing binder quantities. The amount of binder effect on the compressive strength is revealed in Fig. 7b. The samples’ compressive strength rises almost linearly with an increasing amount of binder (40%, 55%, and 70%). In contrast, a lower compressive strength was achieved for the samples with 85% and 100% amount of binders. Similarly, the results of examined samples with the two amounts of binders show a huge deviation among the samples with a similar quantity of binder. This is because with the increasing amount of binder, the samples’ surface loaded during the experiment decreases (Fig. 8). Thus, this revealed that with an increasing amount of binders in a sample, there is declination in the dimensional accuracy [85, 95]. It is consequently projected that with the evaluated outcomes for the binder saturation, there is no reflection of 85% and 100% of the binder. Most likely, binder saturation has the same effect on the compressive strength as similarly detected on the bending strength.

Khandelwal and Ravi [96] investigated the impact of binder quantities on the features and shrinkage of moulds and cores fabricated via a three-part binder system (resin, catalyst, and crosslinking agent). It was found that the rise in the quantity of resin (of approximately 2.4% of the sand weight) produces an increased core shrinkage of about 0.15% in length. A rise in the quantity of catalyst (with the quantity of resin unchanged) depicted a similar total shrinkage degree; however, the initial shrinkage rate increased (during the first 3 h after curing). Additional assessment via hardness test, weight measurement, and SEM provides details on the shrinkage system. The bridges which were developed during the resin polymerization reaction occur between compacted sand particles in the mould (Fig. 9). The bridge’s shortening, which consequently led to core shrinkage, was prompted by the gradual evaporation of the solvent existing in the resin bridges. This was vindicated via weight reduction measurements (Fig. 10) comprising only a chemical mixture (deficient of sand). Another important assessment is that the catalyst with a higher percentage yields rapid curing and thus a greater core shrinkage rate. This observation was buttressed by hardness measurement (Fig. 11) and weight reduction tests.

Ramakrishnan et al. [97] examined the mechanism of an organic binder for sand mould printing by Voxeljet 3D printer, which was first established for developing polymer powders. The binder-dehydrate sodium silicate powder was mixed with quartz sand, and the fluid for printing contained thickened water, which is capable of liquefying the sodium silicate (deposited via the printing head). Changing two factors, such as heat input and the concentration of binder ratio to fluid printing, can all be used to examine the effect of strength and fluid migration of the printed samples. Hence, by heating the sample while printing or by decreasing the input fluid to the sand binder mechanism, the fluid migration can be decreased. The average peak strength of 5.65 N/mm² was attained when the fluid migration was increased. However, it declined to 4.14 N/mm² during printing when the heat was used. It is being observed that the printed components that can resist higher mechanical weight can produce smaller dimensional tolerances between thin walls because of sand adhesion.

Curing is a chemical reaction technique (such as polymerization) or physical action (like evaporation), producing a tougher, harder, or more firm substance (like concrete) or linkage (like an adhesive bond). In some curing techniques, there may be the need to sustain the procedure for specific humidity and temperature level; others may be specific time and pressure [99‐101]. Cure monitoring techniques provide an important understanding of the chemical system and describe procedural actions to attain a certain quality of the cured component. Curing may entail any technique where heat application is engaged to initiate or catalyse molecular-level and chemical structural modification in a polymeric substance like silicones, epoxies, polyesters, and phenolics. These substances are used in various routes to many products for insulation, sealing, protective coating, bonding, and other applications. Also, curing with efficient parameters provides enhanced mechanical properties, ensures material durability, inhibits cracking in material, and minimizes permeability [96, 102, 103].

McKenna et al. [104] examined the impacts of curing time and temperature on the mechanical features, especially the sand moulds’ permeability and strength. These were fabricated using Zb56 resin binder mechanism with ZCast 501 powder. The temperature and curing time were altered between 150 and 250 °C and 4–8 h, respectively. These were applied to characterize the combined and individual influence of these conditions on the printed mould’s permeability and compressive strength. A linear regression model with a 99% confidence level was attained when the statistical design of the experiment was utilized. It was established that as a result of the volatile escape of components, the permeability rises with the curing time. Prolonged heating could have initiated embrittlement of gypsum plaster and low melting temperature fusion phases, causing decreased compressive strength and permeability, respectively. The research excluded the other features of the produced cores and moulds like cohesiveness, refractoriness, and collapsibility; the study ignited the curiosity to further establish the understanding for the correct uses of the technique to produce quality castings from the moulds and cores. Primkulov et al. [98] examined the impacts of curing temperature on strength samples of sand-based rock and furfuryl alcohol resin. It was discovered that the composite material’s compressive strength (equivalent to 19.0 MPa) got enhanced at the optimum temperature of 80 °C, and beyond this temperature, there was a depreciation in furfuryl alcohol resin and consequently reducing the 3D printed sand-stones strength.

Another research tried to optimize the temperature and time for optimum compressive strength with cylindrical ZCast samples [105]. In contrast to McKenna et al. [104], the temperature ranges from 423 to 523 K, and the compressive strength changed from 6.2 to 2.4 MPa. It was stated that the curing time was insignificant on the optimized temperature and strength. These conflicting outcomes necessitate more investigation in this area. Others have considered the casting flaws because of volatile off-gassing of the binder materials; this was attempted by optimizing the curing cycle. Owing to the high surface area to volume ratio of ZCast powder, it needs 8–9% higher quantity of binder in contrast to foundry sand (1.4%). Diverse curing cycles were done to adequately remove the binder and the casted component from the moulds, and a curing temperature of 316 °C for 1 h was discovered to be efficient with adequate mould strength and with no visible defects.

The 3DP powders, which are ZCast and ViriCastTM and no-bake foundry sand, were juxtaposed by Snelling et al. [49], depending on their properties and handleability of the fabricated cast metal. From A356 alloy, cylindrical samples of length 101.6 mm and 25.4 mm diameter were cast out. The samples were examined for hardness, surface roughness, density, microstructure, and porosity. It was detected that the casted parts’ strength and hardness from the 3D moulds were similar to conventional cast alloy (A356). Although from the microstructural evaluation (Fig. 12), ZCast and ViriCast mould reveal dendritic arm spacing. The prepared samples with no-bake moulds had considerably lesser spacing in the dendrite arm than the prepared 3DP samples. This indicated the 3DP and the no-bake’s heat treatment parameters differ, and thus, their strength and hardness are incomparable. The no-bake sand had a marginal higher density and hardenability in the casted parts, with much lower porosity. Contrastingly, ZCast moulds exhibited higher surface roughness. It was established that in producing cast parts with the same feature of no-bake sands, the powders applied for the 3D printing were appropriate.

The influence of shell mould wall thickness, casting volume, and pouring temperature on the surface roughness of fabricated castings was carried out by Chhabra and Singh [106]. Mould of varying shell thickness and volumes were produced and applied to cast brass, aluminium, and copper. The shell wall thickness and pouring temperature were discovered to have a respective 2% and 97.55% contribution to the surface roughness. This outcome is ascribed to the volatile substance decomposition at the metal-mould interface, and thus, the produced gas might influence the cast part’s property which is above the sand particle size effect.

Gill and Kaplas [107] made a comparison analysis on the parts produced from split pattern shells via 3D printing and the components designed from investment mould casting. In assessing the thickness effect on the casting, the moulds whose shell thickness decreases were produced; hence, 3 mm and 6 mm were the highest values discovered for A356 aluminium alloy and ZA12 zinc alloy respectively. The microstructural evaluation revealed the impact of shell wall thickness on the ZCast technique. Non-uniform dispersal of dendritic structures were observed on the A356 aluminium alloy microstructures whose shell wall thickness are 12 mm (Fig. 13a) and 6 mm (Fig. 13b). The ‘white’ areas in the Fig. 13, which are defined by the eutectic silicon elements edges, are the dendritic cells. The ‘darker black’ areas are the silicon elements in the A356 aluminium alloy microstructures, in which they can collate with other silicon elements to develop bands. The 3D printed components have poor properties in hardness and surface roughness in contrast to investment casting. This is because the solidification process affects the microstructure and consequently relies on the mould capability to convey the heat [108]. Other factors which could contribute to these are uneven distribution and/or higher amount of the binder about the sand substances inside the 3D printed mould. As for the binder, the binder distribution relies on the binder viscosity and chemical reaction speed, and thus, if these parameters are low, the sprayed binder via the print head possibly will not wet the entire surface of the individual sand particles. The moulds’ thermal and mechanical features manufactured with ZCast mixture via ZPrinter 310 plus have been analysed in [109]. The features were applied to enhance the mould wall thickness and to pattern the casting procedure. By changing their thickness and calcination period, the parts’ tensile strength was measured. After calcination, a 7-mm thickness was discovered to be tougher than the 15-mm indicated manufacturer thickness, thus promoting efficiency in mould printing. The thermal transfer properties such as heat capacity, heat diffusivity, and thermal conductivity were assessed and applied to plot the AK81 cooling curves. The results depicted better convergence with experimental cooling curves. The research has connected the knowledge space concerning mould characteristics and their importance for procedural modeling. More mould characteristics, especially density, permeability, and their relationship with thermal properties, are needed to be studied.

Shangguan et al. [110] examine the cooling efficiency of a casted stress frame Al alloy (A356) sample by comparing the influence of a conventional dense mould with a 3D printed rib-enforced shell. The outcomes indicated that the casting cooling efficiency significantly improved when rib-enforced shell mould was applied, and hence, in natural conditions, about 40% of cooling time is preserved, and in air blowing conditions, 35% is further preserved for the stress frame samples prior to shakeout. It was discovered that this mould technique compared to traditional mould can feasibly allow the associated cooling condition area to be easily modified and also ensures uniform and rapid casting cooling. This consequently enhances manufacturing efficiency and minimizes the casting residual stresses and deformation. Likewise, the fabricated casting has the excellent surface finish and dimensional accuracy. Furthermore, the 3D printing rib-enforced moulds preserved per mould about nine-tenths of the sand. Some samples’ analysis based on the variation of temperature with time at various stages on the dense sand mould and the rib-enforced sand mould is depicted in Fig. 14. Figure 14 shows that the rib-enforced sand mould approached the optimum temperature after pouring at 800 s. Also, during solidification, the shell approached a temperature of 596 K. Throughout the overall cooling procedure, the temperature range for the shell is from 620 to 380 K, therefore indicating that the new sand mould has an excellent heat discharging capacity to the surrounding.

Another vital parameter affecting the value of the cast metal components is the thermal gradient present in the mould and the heat trapped at the mould metal interface. The heat features and binder depreciation properties of 3D printed (phenolic and furan mechanisms) spherical cores were studied by Fourier thermal [111]. The conditions that were assessed are the overall and fraction heat absorbed, as well as the heat rate absorption under casting factors of aluminium and iron. The disparity of the amount of water in the phenolic mechanism and the thermal degradation behaviour causes the furan absorbed 30% higher heat of 283 kJ/kg as against 221 kJ/kg and possessed a greater heat absorption rate. The furan process was discovered to behave better since it has a moderate and smoother degradation and better cooling capacity. This fact will be necessary to establish numerical simulations of the heat distribution and binder degradation. The research examination varies in the route that it regard the core properties that will assist in achieving better core design and their influence on the internal cast part properties. Excluding the mould properties, it is pertinent to research the mould influence on the fabricated castings. As mentioned earlier, this can be a result of the heat transfer features at the metal-mould boundary and because of the off-gassing in the course of the binder depletion. The thermal transfer features influence the solidification rate, and as a result, the density and strength of the cast component are affected. The release of volatile substances produces unwanted surface abnormalities.