Additive Manufacturing 40 (2021) 101889

Contents lists available at ScienceDirect

Additive Manufacturing
journal homepage: www.elsevier.com/locate/addma

Review

A review on the progress and challenges of binder jet 3D printing of sand

moulds for advanced casting
Tharmalingam Sivarupan a, *, Nagasivamuni Balasubramani a, Prateek Saxena b,
Devarajan Nagarajan c, Mohamed El Mansori d, e, Konstantinos Salonitis b, Mark Jolly b,
Matthew S. Dargusch a, f
a
Centre for Advanced Material Processing and Manufacturing (AMPAM), School of Mechanical and Mining Engineering, The University of Queensland, Brisbane,
Australia
b
Sustainable Manufacturing Systems Centre, School of Aerospace, Transport and Manufacturing, Cranfield University, Bedfordshire MK430AL, United Kingdom
c
Department of Metallurgical and Materials Engineering, National Institute of Technology, Tiruchirappalli 620015, Tamil Nadu, India
d
Laboratoire MSMP-EA7350 Arts et Metiers ParisTech, Aix en Provence, France
e
Texas A&M Engineering Experiment Station, Institute for Manufacturing Systems, Texas A&M University, College Station, TX 77843, USA
f
DMTC Ltd., Victoria, Australia

A B S T R A C T

3D sand mould printing using binder jet technology can enable many technical improvements in casting practice, including part consolidation, design of parts to
optimise the consumption of materials and hazardous chemicals, and on-demand and flexible size part manufacturing near the customer. Incorporating artificial
intelligence in optimising the design of moulds, printing process parameters, and solidification processes may help automate a production facility and reduce labour
time. Elimination of hazardous chemicals from industrial use may be a challenge. Therefore, an alternative technology to fuse the sand particles during printing or an
environmentally friendlier alternative option for the binders and other consumables should be utilised with the 3D sand printing process. Properties of parts produced
using 3D printed sand moulds have the potential to be better than the properties of parts produced using traditional casting due to this technology’s benefits. This
technology is an enabling technology for traditional casting processes rather than a competing technology. It is causing a paradigm shift in casting design because of
the mould geometries achievable by using additive manufacturing to produce the sand mould. This paper reviews the first twenty years of research and challenges in
developing 3D sand printing processes as an innovation for sustainable manufacturing.

1. Introduction
Traditional sand casting mould making processes involve the mixing
of sand and binder materials, and then the mixture is poured into a wood
pattern to shape pieces for a sand mould assembly. Molten metal or an
alloy is allowed to solidify inside the desired mould cavity within a sand
mould assembly. Sand moulds can now be 3D (three dimensional)
printed by binder jet or laser sintering processes.
Selective Laser Sintering (SLS) can also be applied on resin-coated
foundry sand particles to induce the phenolic resin’s thermo-
polymerisation process. A relatively low-power laser (200 W CO2
laser) is used, to sinter the silica sand without any binder [1]. Using a
similar approach, the binder jet sand mould printing involves the
spreading of acid-coated sand particles onto a job-box and spraying of
resin layer-by-layer in order to induce the polymerisation and bond the
sand particles to form a mould or a section of a mould—a review by Néel
et al. [2] provides further insights into the selective laser sintering
process but this review will focus on addressing the recent developments
and challenges in binder jet processes for sand mould printing. An
example of a printed sand mould and the casting obtained using such a
mould is shown in Fig. 1.
Almost all of the additive manufacturing (AM) techniques require
specific feedstock material with a well designed shape for 3D printing
any components [3,4] but metal part production through 3D sand mould
printing does not rely on the shape of metal feedstock (ie. powder size
and shape). However, it may be beneficial to have a specific sand par­
ticle type and size to produce moulds and ultimately castings with
acceptable surface finish. Different stages and processing variables
involved in 3D sand mould printing, casting and characterisation for
quality control are shown in Fig. 2. Firstly, designing a model using a
computer software package, before choosing the sand particle size and
printing process parameters, choice of binder and activator. Finally, the
mould sections are printed. These can be assembled into a mould of any
size if the design is castable with an acceptable defect limit. Therefore, a

* Corresponding author.
E-mail address: t.sivarupan@uq.edu.au (T. Sivarupan).

https://doi.org/10.1016/j.addma.2021.101889
Received 15 September 2020; Received in revised form 31 December 2020; Accepted 31 January 2021
Available online 10 February 2021
2214-8604/© 2021 Elsevier B.V. All rights reserved.
T. Sivarupan et al. Additive Manufacturing 40 (2021) 101889

final casting is obtained with minimal post-processing. Binder jet 3D

Fig. 1. Parts of the 3D printed sand mould, semi-assembled mould sections and
the intake manifold casting obtained using such mould parts (courtesy
of VoxelJet).
printing offers distinct advantages compared to other additive
manufacturing methods because alloys can be cast into moulds manu­
factured by jetting monomer liquid droplets onto a powder-bed, e.g. acid
mixed sand, to bind them through a polymerisation process. The sand
moulds can even be used for cold casting materials such as concrete [5].
Although it is reported that the sand moulds produced through this
process exhibit anisotropic properties, e.g., three-point bending strength
is weak in the z-direction [6–8], there is no significant influence on the
isotropic properties of the final casting as this depends on the solidifi­
cation behaviour of the molten alloy. Any complex geometry can be
manufactured rapidly using 3D printed moulds, even those that are
difficult to be produced with computer numerical control (CNC); 5-axis
machining methods [9] with undercuts, enclosures, sharp internal cor­
ners, and other features. Consequently, 3D printing of sand moulds has
been revolutionising the casting industry by providing energy-saving
options [10,11], defect optimisation [12], as well as the ability to
combine multiple components, previously separated due to the limita­
tions of traditional manufacturing. Production facilities can potentially
shift closer to the customer base, e.g. on a naval ship or very close to a
large city, but the design can be prepared from home.
Research in the area of 3D sand mould printing technology has been
trending up, Fig. 3. Over the past ten years, the authors have published
several papers [6–8,10,13–18] related to the 3D printing of sand
moulds, including a review article which summarises several important
attributes and processing variables with a focus on further improve­
ments needed in this field [18]. Most of the research has focused on the
optimisation of process parameters [6–8,19] while other proposals

Fig. 2. Stages and processing variables involved in 3D sand mould printing, casting and characterisation for quality control.

2
T. Sivarupan et al. Additive Manufacturing 40 (2021) 101889

include an alternative or improved method of design [20–24] for casting possibility of printing multi-material moulds that can provide geometry
using 3D printed sand moulds and testing methods using acoustic and dependent properties has been proposed and studied [48,49].
optical techniques [25] for inorganically bound granular materials, with Fig. 3 indicates a growing trend in sand mould printing research
conclusions for the benefits of this technology [10]. compared to the other popular aspects of additive manufacturing.
It is well established that many engineering applications involving Although the “3D sand mould printing” research is much smaller
these high temperatures and/or complex stress states can benefit from compared to overall “Additive Manufacturing” or “3D Printing”
cast products due to the isotropic nature of material properties that are research, a significant increase in reseatrch with 3D sand printing is
produced during casting. Sand casting is one of the major industrially evident. Also, it emphasises the paradigm shift in casting from Design for
employed mass-casting techniques used to manufacture high melting Manufacture to Manufacture for Design. Research on this layer by layer
point alloys and thin-wall sectioned engineering components [26]. manufacturing technology is progressing rapidly.
Nevertheless, sand casting suffers from considerable limitations on ge­ Some of the commercially available 3D printers are shown in Table 1
ometry and higher lead times due to the involvement of pattern making shows that the largest such printer is having a job-box volume of 8000 L
in the process [27]. In recent times, the increasing complexity in the with a build speed greater than 135 L/h. A new sand printer model
design of many engineering components requires complex pattern released by Voxeljet company overcomes this volume restriction due to
making, which in turn limits the usage of the sand-casting process. In the inclined printing method; the continuous sand printing method has
order to overcome this issue, 3D printing of sand moulds started only two-dimensional limitations. This is further discussed in the later
replacing the conventional pattern making process and has been suc­ paragraphs.
cessfully utilised for the production of complex-shaped sand cast prod­
ucts [27–29]. 2. Progress in 3D sand moulds printing
The main advantage of using 3D sand printing (3DSP) for making
sand casting moulds is that this technology allows the direct fabrication 2.1. Computer design and simulations
of complex-shaped moulds and cores without the necessity of using any
hard tool [28,30]. The final properties and quality of the 3D sand printed The computer simulation of mould filling and solidification has been
(3DSP) moulds depend on the process parameters used for the 3DSP widely used for the design of traditional sand casting since the mid-
process. The most critical parameters are the sand characteristics 1980s. However, the real advantages of using computer simulations
[31–36], types and concentrations of resin, binder, and activator [35, are better realised in the design of 3D printed sand moulds. This is where
37–40], the printing or powder recoating speed [35], layer thickness
[37,41], and curing temperature and time [42–44].
3D printing sand moulds can address challenges in centralised Table 1
manufacturing, especially the fabrication of custom-made, single pro­ 3D sand printing machines. Manufacturers and scale of sand printing machines.
duction products. The hard tooling requirement is significantly reduced Manufacturer Size (mm) and volume (L) Layer Build
if the initial investment on a large 3D sand mould printer is absorbed. thickness speed (L/
(mm) h)
Operational costs are comparable to conventional casting, with addi­
tional energy and material savings, from the ability to manufacture non- ExOne 1800 × 1000 × 700 (1260) 0.26 to 0.38 135
1800 × 1000 × 700 (1260) 0.26 to 0.38 105
traditional shapes. It is expected that the initial capital cost of the
800 × 500 × 400 (160) 0.26 to 0.38 36
printing machine will be affordable as the market pull increases, though, Voxeljet 300 × 200 × 150 (9) 0.15 or 0.3 Not
in the current market, the total cost of such printers and related equip­ available
ment are on the scale of a million US dollars. Nonetheless, the network of 500 × 400 × 300 (60) 0.15 or 0.3 Not
on-demand production facilities around the world means custom part available
1000 × 600 × 500 (300) 0.15 or 0.3 Not
production must be at the minimum affordable price. This is the only available
indirect 3D printing technology used to construct a significantly large 2000 × 1000 × 1000 (2000) 0.15 or 0.3 Not
casting in a production facility within a room. Therefore, it can be easily available
accessed by a large-scale city considerably in a small-time frame. The 4000 × 2000 × 1000 (8000) 0.15 or 0.3 Not
available
technique is not just for parts with high dimensional tolerance (layer
3D systems Discontinued sand printers/ Not Not
thickness in 280 µm -500 µm), but it can also be used for line forming (previously using different material available available
methods for large scale (volume) parts to speed up the printing process ZCorp)
by increasing the layer thickness up to 2 mm [45–47]. Another

Fig. 3. From 2010 to 2020, the number of publications appeared on Scopus search for the keyword of (a) “3D sand mould printing” a total of around 130 publi­
cations, (b) “Binder jetting-362′′ and “3D sand mould printing-130′′ , (c) “additive manufacturing-29983′′ , “3D printing-31958′′ , “rapid proto typing-10551′′ , 3D sand
mould printing-130, and “Binder jetting-362′′ .

3
T. Sivarupan et al. Additive Manufacturing 40 (2021) 101889

the generative design or topology optimisation can be incorporated. towards defect elimination.
Modelling software packages have been incorporating additive
manufacturing-focused functionalities even though none of those
2.2. Flexible design methodologies
packages has a specialised module for 3D sand mould printing. It would
be an added advantage and reduce time on modelling and optimisation
Castings produced by conventional sand moulds require extensive
if these packages could incorporate functionalities such as optimum
preparation by skilled labour, including pattern and core making, that
mould sectioning for a dimension specified job-box of a sand mould
have inadequacies in facilitating new designs. Limitations of the con­
printer. This would assist the users to minimise sand waste and printing
ventional method include the pattern making process with expensive
time, thereby improving the economic performance. Traditional casting
machining works, storage of patterns, degradation of dimensional tol­
processes have been utilising computer technology in order to simulate
erances due to prolonged storage or usage, undercut in design requiring
the mould filling and solidification processes when a molten alloy is
additional cores, and other geometrical design considerations for gating,
poured into a sand mould assembly (gating system, casting cavity, and
feeding and runner systems. On the other hand, 3D printing facilitates
risers). There are several commercial software producers such as Quik­
rapid fabrication of moulds by reducing the processing time, tooling
CAST/ProCAST, MAGMA SOFT/ MAGMA5, FLOW 3D, and Altair
complexities associated with cores and undercuts [12,49,52]. Fig. 4
Inspire Cast. A comprehensive review on the simulation software
highlights some of the essential attributes of the design freedom ach­
packages has been published [50]. The benefit of casting simulation
ieved with 3D printing compared to conventional sand moulds. Sama
highly depends on the boundary conditions, material properties, and
et al. showed the potential of 3D printing and hybrid approaches
meshing of the computer-aided design (CAD) model at a high cost to
through three case studies of a vane impeller, a complex nested bracket,
benefit ratio [50]. The major problem of modelling traditional casting
and a centre bearing housing for a duplex stainless-steel casting. As
processes is providing the exact boundary conditions for the simulations.
opposed to the closed vane impeller’s conventional design, the 3D
The input of the correct data is crucial for the accurate simulation
printed sand mould consists of only three risers and one ingate that
outcome. Functional design and manufacturing of the 3D sand mould
eliminates shrinkage porosity with a design that has a vertical parting
will address these issues by providing geometry dependent gas perme­
line. The intricate bracket design using 3D printing enables the reduc­
ability and thermal information with minimal deviation. Improved input
tion of turbulence (velocity of 1.81 m/s in conventional moulds is
data on all boundary conditions not only simulates the casting process
reduced to 0.5 m/s in 3DSP moulds) and thermal losses, thereby
more realistically, but also leads to the saving of materials and energy,
decreasing the propensity for misruns. Misruns occur when the liquid
thus the quality and economic improvements. This issue has been
metal is too cold / too viscous to flow to the extremities of the mould
well-reviewed in a publication by one of the current authors [51] and
cavity before solidifying. Therefore, it would not be able to fill the mould
concluded with suggestions for the integration of simulation techniques
cavity completely. The unfilled portion of space in the mould is referred
on the atomic level, e.g. a Molecular Dynamics simulation, on a
to as a misrun. Providing additional sprues to the design enhances the
real-time scale which would result in better decision making to optimise
metal feeding at the locations of the misrun.
the energy consumption as well as most importantly the possibility
Provision for easy drain off for the loose sand and seamless

Fig. 4. Design flexibility and the advantages associated with 3DP [12,21–23,49,52–54].

4
T. Sivarupan et al.
integration with conventional mould design is also possible to eliminate
defects in the casting process. Often the component geometry decides
the orientation, location of the riser, and size, whereas 3D printing en­
ables the incorporation of suitable orientation for increased metallic
yield and reduced defects. A parabolic and conical-helix sprue offers a
smooth flow of liquid melt, reducing the turbulence at the in-gate
compared to the conventional straight sprue. Significant reductions of
(i) void spaces or internal flaws in the casting (56 % and 99 %), (ii)
inclusions, mainly oxides or sand particles (21 % and 35 % area) and (iii)
centreline shrinkage defect (1.5 % and 0.6 % area) have been measured
for parabolic and conical-helix design respectively (for the values indi­
cated in the brackets) A decrease in the melt flow also reduces the sand
inclusions and increases flexural strength in the conical-helix design,
which could be used as a novel design for hybrid moulding for Al and
steel, or cast-iron melts [49].
Several improvements are also achieved with novel designs of the
pouring basin, runner and riser. Increasing the number of sprues while
reducing the width allows for less air entrapment, and the multiple sprue
design reduces the flow velocity due to frictional forces. The higher-
pressure head with the pouring basin shown in Fig. 4, avoids common
casting defects observed in thin-section castings and eliminates metal
splashing. A complex runner network shown in Fig. 4 avoids gas
entrapment at the entrance of the runner and in the gating system.
Additional vents act as a reservoir to reduce entrapped air and regulate
the flow velocity by maintaining backpressure during mould filling.
Round corners, instead of the rectangular cross-section, further decrease
the turbulence issues. Generally, cylindrical feeders or risers are stan­
dard in traditional casting. Modifications to the existing design by the
addition of air vents or a novel spherical design by 3DP delays the so­
lidification time and enhances feeding. For instance, a spherical design
showed a total solidification time of 7 min compared to the cylindrical
counterpart that solidified in 4 min for a particular casting design [55].
In contrast to the traditional thick sand mould fabricated by the
conventional process, lattice or shell type of moulds with varying
thickness and adequate strength can be produced in the 3DSP process. A
few models of designs for these moulds are shown in Fig. 4 [20–22,24].
A rib-reinforced design on the outer side of the casting shell forms a
network-like structure that enables increased cooling rates to reach the
shakeout temperature quickly and a significant reduction in sand con­
sumption (nearly 90 %) [22]. Kang et al. proposed several lattice frames
(skeleton mould design) [24], including a basic cross-frame, internal
and surface enforced cross frame, diamond, cellular and radial types of
frames. These frames are associated with higher cooling power, reduced
sand consumption, and reduced weight of the mould. Forced cooling can
be effectively introduced into this type of mould structure to reduce
production time and increase production efficiency without defects in
the casings.
A case study reported for the A-356 Al alloy wheel hub casting (0.3 m
diameter) with a mould that has no parting line showed that the lattice
design saved 50 % of the sand used while achieving higher heat transfer
for adequate cooling, castings devoid of shrinkage, and better surface
finish compared to traditional castings [21]. Distinct cooling behaviour
was observed for the shell mould and the dense traditional mould.
Initially, the dense mould extracts heat from the melt, whereas the shell
mould cools at a relatively lower rate. The shift in the cooling rate occurs
at temperatures approximately below 300 ◦ C (for A356 Al alloy), where
the shell mould cools faster to reach the shakeout temperature of 200 ◦ C
than the solid mould [20].
Deng et al. [23] investigated the effect of hollow structures around
the riser and evaluated the solidification and heat transfer characteris­
tics. A series of riser structures having one to three hollow cavities are
designed with a wall thickness of 3 mm and an air cavity of 12 mm when
an aluminium alloy (A-356) solidified in a riser with a single cavity
layer, the solidification time was delayed by 11.7 %, and it was further
extended to 30 % for a riser with three layers of air cavities when
compared to the conventional riser without hollow structure. The
5
Additive Manufacturing 40 (2021) 101889
shrinkage area between the riser and casting was increased by 6 % and
18 % for the risers with single and three layers of cavities respectively. In
addition, no shrinkage porosity was found on the risers with air cavities
indicating that the solidification was delayed with enhanced feeding,
and it also reduced residual stresses in the casting [23].
The design flexibility with moulds can also be extended to core
design, where several small components of the core can be integrated
into one part. An example is shown in Fig. 4 for a train’s air brake in
which eight different cores are consolidated into one 3DSP core that
reduces the misalignment errors and undercut difficulties in design [53,
56].
Snelling et al. [54] fabricated a cellular structure of the A-356 Al
alloy using a 3D printed sand mould using an ExOne (leading 3D sand
printer manufacturer by market capital) S-print machine without a
curing cycle. The model was created using CAD specialised for creating
cellular materials. Initially, a single unit cell was designed, which was
then patterned to create the whole structure (using Netfabb SSS soft­
ware). The printed mould was used as a core, which creates a cavity in
the resultant casting, and an outer mould with gates and down sprue
arrangement was fabricated, and this setup brings liquid metal into the
cavity. Castings made from the printed mould exhibited no sand burn or
metal penetration. Compression tests revealed that the castings could
withstand a maximum load of 120 kN to 130 kN. Fig. 4 shows the
manufacturing process of the cellular structure.
A case study on the comparison of cost estimation between con­
ventional and 3D printing of a train air brake (8 cores) and turbocharger
(3 cores) reveals that the total cost (tooling and fabrication) for a con­
ventional process increases as a function of part complexity (estimated
by a complexity factor). 3D printing becomes the cost-effective approach
for high volume production, and time spent on the tool design or pattern
and core making process can be significantly reduced [28].
3D sand printing offers a new paradigm of Manufacture for Design.
3DSP offers a flexible intermediate stage, not creating the part directly,
but fabricating the disposable mould tool without other hard tooling and
with a creative and functional mould design such as one described in the
next section.
2.3. A new avenue for manufacturing
3D printed sand models can be coated with, for example, carbon fibre
sheets with resin and hardener, a flexible material that can set into a
rigid shape; then the sand mould can be washed out to create a large-
scale, complex and lightweight hollow structure (composite structure),
e.g., aircraft ducting, pressure tanks, struts, and mandrels [57]. Nor­
mally, making pattern is challenging with other polymer 3D printing
techniques due to their non-recyclability and the scale of the job-box of
those printers as well as their dimensional stability. In this case, suitable
bonding material can be chosen, instead of furan resin. The mould
making can be implemented at room temperature using a water-soluble
substance that can bond the sand particles and keep the structure until
washout. There are also other possible alternative applications of 3DSP
such as the cold casting of concrete; these moulds can be used to produce
a complex structure in concrete.
3. Challenges with sand handling and processing variables
3.1. Binders
The use of furan binders or any other polymer materials are not new
to the casting industry, but these are considered as hazardous materials
and need to be replaced with substances that can provide better prop­
erties to the printed mould with improved environmental properties
[56], including energy-efficient production [10]. A list of currently used
sand particles and binders are tabulated in Table 2 and Table 3,
respectively. However, foundries have already started using epoxy/SO2
based curing system to minimise the carbon emission [58]. Koltygin
T. Sivarupan et al. Additive Manufacturing 40 (2021) 101889

Table 2 MgCO3, magnetite Fe3O4, Olivine (Mg, Fe)2(SiO4), forsterite Mg2SiO4

List of a few currently used (traditional casting) sand materials. and other compounds. A sand-gypsum mixture with Iron (III) sulphate as
Material Properties Limitations Advantages mass fraction of 0.2 % as hardening accelerator and dolomite as mass
fraction of 2 % as moisture absorber were found to contain adequate
Silica (SiO2) Low volumetric Low heat Good for gating
heat capacity conductivity system strength to that of ZCast 501 powder. The details of the alternative
Cheap/most binder used instead of Zb56 was not provided. A matched tensile
abundant strength and a good dimensional tolerance were obtained for the new
Zirconia (ZrO2), High volumetric High heat Expensive binder compared to that of the ZCast powder and binder. The mould
heat capacity conductivity
Olivine ((Mg2+/ Good for mould
obtained using the new sand mixture had sufficient strength.
Fe2+)SO4), cavity surface Ramakrishnan et al. [60] investigated the organic binder system for
Chromite printing sand moulds using a Voxeljet VX-500 3D printer, which was
(FeCr2O4), initially designed for processing polymer powders. The quartz sand was
Zircon (ZrSO4),
mixed with dry sodium silicate powder (acting as a binder), and the
Chamotte
printing fluid comprised of thickened water that was able to dissolve the
sodium silicate (dispensed by the printing head). The influence of fluid
migration and strength of the printed parts were evaluated by varying
Table 3
two parameters, namely, the concentration ratio of binder to printing
Summary of the commercially available binders and activators and their prop­
fluid and heat input. Fluid migration can be reduced by reducing the
erties [79].
fluid input to the sand binder system or heating the specimen while
Binder Activator Properties printing. An increase in fluid migration has the highest mean strength of
Furan (chemically cured Sulphonic Can be used with/without heat 5.65 N/mm2, which is reduced to 4.14 N/mm2 when heating is applied
furfuryl alcohol-based) acid treatment to remove any moisture. during printing. It should be noted that the printed parts that can
CHP binders (ester-cured Additional post-curing at elevated
–
withstand high mechanical load can lead to low dimensional tolerances
alkaline phenolic resole) temperatures.
HHP binders (acid-cured – Additional post-curing at elevated between thin walls due to sand adhesion.
phenolic resole) temperatures.
Inorganic binders (water- – Additional post-curing at elevated
based, alkali-silicate temperatures. 3.2. Processing
binder)

Processing stages involved in 3D printing of sand moulds before

et al. [59] investigated an alternate mixture for ZCast 501 powder and
Zb56 inorganic binder (Z Corporation, the US, known as 3D Systems
since 2012) using a sand-gypsum mixture. Using XRD analysis, the au­
thors found that the ZCast mixture was composed of gypsum
CaSO4⋅2H2O, calcium silicate Ca2(SiO4), calcite CaCO3, magnesite
reaching the pouring stage are schematically shown in Fig. 5. Cleaning
or removal of unbound sand particles from the job-box is a time
consuming and labour-intensive process and hence reduces the pro­
duction speed ultimately. Voxel jet, a manufacturer of sand printers,
produced a design (model VXC 800) to partly overcome these issues. It
allows the user to continuously print the part of the mould with any

Fig. 5. Processing stages involved in 3D printing of sand moulds before reaching the pouring stage.

6
T. Sivarupan et al.
length (width and height are limited by the size of the job-box); inclined
printing on a conveyer belt enables part removal without stopping the
printing process. This design is based on the principle of the critical
angle of repose of a granular material. That is, the print head is inclined
on an angle less than the critical angle of repose of the sand particles.
There is a scarcity of literature available highlighting the dimensional
tolerance of each design. The design of moulds can be slightly improved
with the help of proposed lattice-shell and rib enforced shell mould
designs by Kang et al. [24] if the unbound sand particles within the
mould cavity can be removed with ease.
Although it has been shown that traditional foundry sand can be
utilised to print sand moulds, the dimensional tolerance is affected due
to the poor flowability (prone to clogging) of the non-engineered sand
particles [61]. Therefore, recycling of acid-wetted or unbound sand
particles after printing remains a chalenge requiring extreme processing
before they can be used for printing processes. The price of engineered
sand particles is relatively cheap, but the cleaning process and
re-transporting to the manufacturer may significantly increase the
casting price. A case was presented to re-use waste foundry sand in the
building industry; Mitterpach et al. [62] proposed that it would be
necessary to thoroughly verify the eco-toxicological properties of not
only the created waste foundry sand and other foundry waste, but
mainly the building products for which this waste could be used. The
authors [62] further suggested that handling foundry sand significantly
increases particulate emission, which penetrates the work environment
and the atmosphere due to its dustiness. Therefore, the subsequent
environmental analysis of waste foundry sand must be monitored for the
concentrations of particulate matter PM10 and PM2.5 entering the
environment in all stages of the process that create waste foundry sand
and must be thoroughly assessed the impact upon human health. Lack of
knowledge on how to safely transport the 3D printed moulds and cores
[63] is challenging due to their fragile nature.
Many research studies have focused on optimising the process pa­
rameters due to the vast range of process parameters used in the printing
process, such as sand particles size, layer thickness, print resolution in
all directions and ink (polymer) droplet volume, job-box dimension, re-
coater speed (sand spreading speed). Increasing the binder content can
increase the strength; however this results in low dimensional tolerance
[19,64] that will produce a significant amount of gases during casting. A
comparative study was performed by Gill et al. [65] on dimensional
tolerances, surface roughness, gas holes, and shrinkage defects of A356
Al alloy castings fabricated using moulds made from the investment
casting process and 3D printed sand moulds with varying thickness (2
mm to 12 mm). Though the investment casting process using
plaster-based powder produced consistent results in surface finish and
defect control, the ZCast process (ZCast 501 powder) also produced
promising results for 6 mm shell thickness. A further improvement is
needed to increase the surface finish of the 3D printed moulds. The
authors also investigated the comparative qualities of the castings and
produced using a 3D printing sand mould of a die-cast Zn alloy (ZA12
alloy) with an A356 alloy [66]. It was found that the shell thickness for
the A356 and ZA12 can be as small as 1 mm and 3 mm respectively, thus
the cost and material consumption can be reduced to maximise the
production efficiency. On the other hand, microstructure studies
revealed that the investment castings (0.125 mm) produced uniform
grain structures compared to the ZCast process, in which large dendrites
were observed in the A356 alloy. This study also emphasised that the
casting process alloy compositions’ thermal gradient during solidifica­
tion (with respect to mould thickness), pouring temperature, cooling
rate, and the effect of constitutional undercooling need to be considered
for the ZCast process [65,66].
Singamneni et al. [67] investigated the effect of curing time and
temperature on the compressive strength and surface roughness of the
A356 Al alloy castings. A prolonged heating time or increase in tem­
perature resulted in reduced permeability due to the fusion of low
melting phases during baking. A similar observation was also noted for
7
Additive Manufacturing 40 (2021) 101889
the compressive strength as a result of phase transformation in the
plaster binder regions. The surface roughness of the Al casting produced
in the moulds was tested with and without various mould coatings.
Brush coated and spray-coated moulds produced smaller average
roughness (12 μm to 17 μm) with better tensile properties than the
uncoated moulds (12 μm to 17 μm).
3.3. Properties
3.3.1. Thermal properties/mechanical properties/gas permeability
Controlling thermophysical and heat transfer properties of the mould
materials in the sand-casting process is critical to understanding the
microstructure formation and the associated defects (due to gas evolu­
tion). Two types of bond mixtures based on phenolic and furan binder in
3D printed cores were comparatively reported by Toth et al. [39]. In­
verse Fourier analysis was conducted to extract the total absorbed heat
and the rate of heat absorption in liquid Al and iron. Results showed that
the furan-based binder using quartz sand absorbs 30 % more heat than
the phenolic binder. Higher water content in the phenolic binder with
high heat absorption capacity at the initial stages of the solidification
process can give rise to defects, while furan facilitates smooth decom­
position profiles. Although the 3D printed sand mould using furan
binder could be used to cast parts without heat treatment, Primkulov
et al. [68] found that curing at temperatures of nearly 60 ◦ C to 150 ◦ C for
an appropriate duration (30 h to 60 h) showed optimal unconfined
compressive strength (UCS) at 80 ◦ C. On the other hand, Mitra et al. [15]
studied the same for 3PB (three-point bending) strength and perme­
ability. The authors reported that only the permeability improved due to
natural ageing (evaporation of bound water) and there was no signifi­
cant change in 3PB strength even at 100 ◦ C. They further suggested that
the mould could be stored for a long term while roughly preserving the
strength. It appears that the tensile and compressive behaviour of the
printed sand mould is affected by the bound water/other fluid content;
therefore, it is apparent that properties of the sand mould are controlled
by the acid/furan ratio and pore connectivity within the sand mould.
There are challenges to implementing industrial hybrid technology
for rapid casting that combines 3D sand additive manufacturing of
moulds and traditional casting processes. The idea is to master the effect
of direct 3D powder binding to achieve high and sustainable produc­
tivity in the casting industry. Solutions to the research challenges bridge
the gap between fundamental powder metallurgy engineering and
manufacturing applications, between measurement methods and pro­
cess control of production diagnostics and data-based experimentation.
3.4. Factors influencing the quality of the 3D printed sand moulds
Many of the keyfactors affecting the quality of the mould have been
presented in the authors’ previous publication [69]. The effect of two
sand mould making processes, conventional and 3D printing, on the
weight savings, allowances provided to the moulds, surface finish and
fettling work required on the castings of a pump bowl have been
investigated [36]. The results [36] indicated that the employment of 3D
printing for making pump bowl moulds and cores has significant ad­
vantages over the conventional process. The amount of sand that was
used in the conventional pattern making was 301 kg, whereas the 3D
printing process consumed only 99 kg. A significantly higher weight
saving of 67.11 % was observed due to the usage of the additive
manufacturing process for the production of the sand mould and core for
the pump bowl. The weight saving was predominantly due to the
elimination of the draft, angle, machining, and finishing allowances
provided to the casting, which is not necessary for the 3D printing
process.
Additionally, the average surface roughness was improved using the
3D printing process (~200 µm) compared to the conventional process
(~ 500 µm), and the mechanical strength of the 3D printed sand moulds
was found to be superior to the conventional process due to better
T. Sivarupan et al.
bonding between the sand particles. Moreover, the fettling time was
reduced to 15 min from 60 min due to the very close tolerances of the 3D
printed sand moulds and subsequent little machining of the final pump
bowl castings [36]. The authors [36] also stated that the 3D sand mould
printing technique is effective, in terms of cost and lead-time savings
with minimum wastage of material, only when a small quantity of sand
mould is manufactured, whereas the conventional sand mould making
process is more favourable when large quantities are produced.
To the contrary, the applicability of Binder Jet technology to 3D print
sand moulds for the fabrication of complicated bespoke cast aluminium
space-frame structure was investigated [30], and the authors suggested
that rather than using conventional sand casting or direct 3D metal
printing for making non-standard customised metal joints, it is useful to
combine conventional sand casting with 3D printed sand moulds so that
large numbers of non-standard customised structures can be made in a
shorter time. Besides, 3D printing of sand moulds offered more design
freedom while fabricating such complicated structures. Neither con­
ventional sand casting nor direct 3D metal printing offered such geo­
metric freedom or fabrication flexibility for making complicated
structures [30].
The inherent benefit of design freedom of 3D printing and the ability
to make complicated sand moulds increased the complexity of the
casting process and led to more casting defects due to the inability to
predict heat dissipation within the complicated moulds [70]. An in­
strument methodology comprising of wired and wireless sensors was
developed to improve the understanding of such complex castings [70].
These sensors were housed at different strategic locations within the
mould to collect diversified data such as temperature, pressure, mois­
ture, gas chemistry, the motion of moulds and magnetic field, from the
moulds. The design freedom offered by 3D printing leveraged the usage
of different sensors and helped in understanding the thermodynamics
and physics of the process.
A lightweight structure in a smaller length scale instead of the
macroscopic design was implemented to reduce the manufacturing cost
and processing time of 3D printed sand moulds [71]. Two types of
lightweight structures or geometrical unit patterns, a box with a square
hole (Type-I) and a lattice or mesh structure with top and bottom pads
(Type-II), were designed and mechanically analysed. It was observed
that the Type-II pattern was more flexible in taking out sand particles
than the Type-I pattern. Similarly, both lightweight structure types were
suitable only for rigid single component materials such as metals or
plastics and not suitable for ceramic or polymer composites, which have
different mechanical behaviours.
The effect of a traditional dense mould compared to a rib-enforced
shell mould constructed with 3D sand printing technique on the cool­
ing efficiency of the Al alloy A356 casting for stress-frame applications
has been investigated [72]. The authors observed that the cooling effi­
ciency of the rib-enforced 3D sand printed shell mould castings was
significantly larger with cooling time savings of approximately 40 % in
natural conditions and further 35 % in air blown conditions before
shakeout. Furthermore, the rib-enforced 3D sand printed shell mould
enabled the adjustment of cooling conditions at locations of interest and
produced good dimensional tolerance and surface quality in the A-356
aluminium alloy stress frame. Moreover, the 3D sand printing of the
rib-enforced shell moulds saved almost nine-tenths of sand per mould
[72].
The effectiveness of the application of new technologies, additive
manufacturing (ink-jet bonding 3D printing) and subtractive
manufacturing (sand mould milling), in terms of resource consumption,
environmental impact and production efficiency in a traditional foundry
industry were investigated [73]. Material, energy, and human resources
were considered evaluation parameters for estimating resource con­
sumption, carbon emission as an evaluation index for environmental
impact, and actual casting output to evaluate production efficiency. The
results indicated that the 3D sand mould printing process combined with
traditional casting has better resource utilisation, reducing carbon
8
Additive Manufacturing 40 (2021) 101889
emission by over 20 %. Also, the production efficiency of the 3D printing
process was about 93 %, which has more significant economic benefits.
The authors provided a summary of whether and which new technolo­
gies could be combined in a traditional manufacturing process to realise
sustainability, which has practical significance.
The mass transport properties, i.e., permeability and microstructural
features of the 3D printed sand moulds were characterised [74] using
X-ray micro-computed tomography to achieve reproducibility and for
controlling gas defects in the final castings. The results showed that the
3D printed sand moulds’ mass transport properties could be precisely
predicted using the in-situ non-destructive methods like X-ray
micro-computed tomography, and the data can be used to improve the
quality of the castings and create optimised 3D printed sand mould
structures for foundry applications. Fig. 6 illustrates the
processing-oriented advantages and the research areas to be focused on
improving the capabilities of 3DSP.
Properties such as gas permeability and strength of the mould are not
always required to be at their maximum for good casting. Therefore,
flexibility in controlling these parameters is the strength of this tech­
nology. To summarise, the following would give a user a better under­
standing of the process.
1. The printing process parameter such as layer thickness which is
determined by the sand particle size.
2. Furan droplet size and distribution influence the mass distribution of
bonding material which can determine the density of the 3DSP
mould, hence the gas permeability.
3. Viscosity of the binder is a function of temperature. Therefore, the
printer should be maintained within a controlled environment.
3.5. Printhead and re-coater
The most important part of the binder jetting machine is the print­
head. The resolution, both in x- and y- directions, is determined by the
space between the nozzles and the speed of the printhead as well as the
volume (controlled by the applied electrical potential difference,
voltage, on the printhead) of the ink droplet; the volume is typically
around 78 pL to 104 pL (average diameter is between 50 µm and
60 µm; not perfect spheres) [75], and this will occupy around 2 % to 3 %
of the interstitial region between the sand particles. The volume of each
droplet may affect the dimensional tolerance even though the sand
particle (140 µm to 250 µm) is much greater in size than these droplets;
therefore, the effect of the layer thickness of sand particles (usually two
times the particle size) will determine the final tolerance. Regardless of
this, the binder’s distribution within the mould is determined by the size
of the droplet and therefore, it should be kept to a minimum. The
printing process’s speed is mainly controlled by the speed of the print­
head and the sand re-coater and the dimension of the x-y plane within
the job-box. This binder fluid’s viscosity is a function of temperature that
is hard to control within a continuously moving printhead; hundreds or
thousands of nozzles in the printhead are in µm scale. Therefore, a
frequent clogging problem is inevitable and needs attention.
The strength of the sand mould depends on the porosity, Fig. 7. When
perfect spheres of sand particles are compacted in a Face Centred Cubic
(FCC)/Hexagonal Closest Packed (HCP), Body-Centred Cubic (BCC),
Simple Cubic (SC)manner, the density of the printed part takes the
maximum achievable density of these structures, corresponding to
74.06 %, 68.02 %, 52.36 % of the bulk sand density respectively,
ignoring the binder density. The binder/activator is added to the sand
mix on a certain weight fraction, and the binder within the 3D printed
sand mould is distributed within the interstitial gaps. Therefore the extra
mass within a unit volume can be calculated, and the density or strength
related using Eq. (1) [76].
Ms ρb
σt = (1 − ε)σ s (1)
Mb ρs
T. Sivarupan et al.
Fig. 6. Processing oriented advantages and the research areas to be focused on improving the capabilities of 3DP.
Interparticle friction has a significant effect on filling fine-scale
particles, and the sphericity of these sand particles is in the range of
70 %; therefore, the packing fraction can be expected to deviate from
these predictions. Despite this, the maximum density achieved experi­
mentally was nearly 52 % of the bulk sand density [6], corresponding to
the density of a simple cubic packing arrangement of these sand parti­
cles. This structure’s primary advantage is that the capillary action of the
binder fluid may shift the sand particles from their original interlocked
status after the recoating iteration. Therefore, the binder should be
dispersed either in a ((b)− 1) pendular or ((b)− 2) funicular manner, as
shown in Fig. 7(b). The binder distribution depends on the fluid pene­
tration method within the compacted sand; therefore, the binder dis­
tribution can be controlled by optimising the voltage waveform of the
printhead [76]. Well-connected pore regions are the key to achieving
better permeability of the printed mould.
3.6. Functional sand mould
Significant drawbacks of 3D sand printing are the binder and sand
material capabilities and characteristics in creating functional soft sand
mould tools without hard tools. This section documents the research
methodology to overcome these critical points, using several traditional
and advanced characterisation approaches. The use of selectively prin­
ted micro-droplets of different foundry binders into a fine layer of
permeable activator coated traditional and advanced ceramic materials,
could lead to functionalise the sand mould. For this reason, there is a
need to focus further research on the process physics involved in the 3D
printing technology. Moreover, to address the characterisation of the
powder to form a primitive building element and the final printed
product, more attention should be paid to the characterisation tech­
niques regarding mechanical performance and transport properties of
3D sand printed moulds [74].
The current sand moulds are manufactured using only one type of
sand particles. This provides uniform thermal properties in the mould
assembly. As the molten metal enters the sprue, the temperature drop is
significant, and hence a very high superheat may be required for the
molten alloy during pouring. This not only results in wasted energy but
also negatively influences the solidification characteristics of the alloy.
If the mould is designed with geometry-dependent thermal properties,
the molten alloy can fill the mould without severe temperature loss but
can cool rapidly once entering the mould cavity where the actual part is
created. This means that the sprue can be made with materials of the low
volumetric heat capacity and low heat conductivity but the mould cavity
with exceptionally high volumetric heat capacity and high thermal
conductivity to speed up the solidification process. Shan et al. [77]
showed that the sand re-coater could be modified to spread various sand
layers during each iteration of the sand recoating process. In this way,
the mould can be manufactured with geometry dependent thermal
properties. Silica sand has very low thermal conductivity and volumetric
heat capacity, see Fig. 8. Therefore, it may be suitable sand for the sprue
9
Additive Manufacturing 40 (2021) 101889
section of the mould assembly. In contrast, zirconia has slightly higher
volumetric heat capacity and is, therefore, a better candidate for the
mould cavity section because it can provide better heat absorption from
the molten alloy. In addition, silicon carbide sand (currently used for
casting) has high volumetric heat capacity and high thermal conduc­
tivity, compared in Fig. 9. However, the particle shape is angular [78]
due to the production method and therefore presents difficulties in
finding suitable printing process parameters. Silica sand is cheap and
abundant on the earth’s crust. It has poor heat conductivity and lower
volumetric heat capacity, Fig. 8, compared to the currently used sands. It
would be a better choice for the gating system in the sand mould as­
sembly where the molten alloy should be kept in liquid form until the
solidification in the cavity is complete.
3.7. Hazardous chemicals
Furan binders are reportedly carcinogenic [81]. The acid activator
used to initiate the polymerisation process to bond the sand particles
within the sand mould is also unsuitable for human exposure. Therefore,
extreme care is required to work in the print room, e.g., a powered filter
mask.
There is a real need for alternative binding material or technologies
to eliminate the use of hazardous chemicals and minimise carbon
emissions. Furan and other binders are not new to the industry, having
been used for many decades, but 3D sand mould printing technology is
proven to optimise the consumption of all consumables and metals [10].
3D sand mould printing has been reported as an energy-saving [1]
alternative production route to the traditional casting process.
3.8. Hard tooling cost and time
In traditional hard tooling, the main cost contributors are the ma­
terial, maintenance, machining and labour. Each minor adjustment to
the design or the need for retooling can increase the overall tooling cost
by many time. Furthermore, identifying the mould defects at the time of
metal pouring and reiterating the design and tooling stages adds cost
and increases the total time elapsed for producing a casting [82]. 3D
printing techniques can bring down both the cost and time spent for
manufacturing a tool part. Also, unlike the conventional hard tooling
methods, the challenges involved in manufacturing complex geometries
can be dealt with ease using AM methods.
3D printing technologies claim to minimise the hard tooling cost and
time significantly. However, the initial cost of a printer and the labour
required to optimise the process parameters or design for additive
manufacturing (DfAM) are still economically challenging. The
comparative cost of a permanent mould casting die against a 3D printer,
and the related cost per part is still not precise. However, when the
market pull becomes dominant, the cost challenges are expected to be
solved, and the industry may benefit from 3D printing technologies.
T. Sivarupan et al. Additive Manufacturing 40 (2021) 101889

Fig. 7. Strength of the 3D printed sand mould as a function of porosity when considering C as a constant. Where C = Ms ρb/(Mb ρs) σs, ρp and ρb are the densities of
sand and binder, Ms/Mb is the weight rate of binder and sand particles [76]. (b) The binder filling methods within the 3D printed sand.

Fig. 8. Volumetric heat capacity (J.K-1.m-3) as a function of thermal conductivity (W.m-1K-1). The chart produced by CES-Edu pack software and material property
database [80] containing a collection of material property data from published sources.

3.9. Towards mass production
The job-box size limits the freedom of design in 3D sand mould
printing. To overcome the limitations of the job-box size it is possible to
create large structure mould using assembly. The production cost can be
managed by minimising sand waste. Shan et al. [49] demonstrated an
integrated forming method for two-way coating and printing, Fig. 10.
There are two re-coaters on either side of the print head, and each of
them is unidirectionally active. The print head and the re-coater and an
infrared heating source perform their job simultaneously without the
time lag currently experienced by other 3D sand printers, due to the
re-coater and print head moving in perpendicular directions in the x-y
plane. The authors also proposed a possible method of using a
multi-material printing process by using high thermal conductivity sand
near the cavity of the mould and high permeable particles for the bulk of
the sand mould. This is a significant milestone in timesaving during
printing. EnvisionTEC, a robotics manufacturer, introduced their pro­
prietary technique to manufacture sand moulds and cores using their
Robotic Additive Manufacturing systems, Viridis3D, at a 1/3 of the
competitors’ machine costs [83].
4. Sustainability in sand casting operations
Sustainability is a specific term that has been embraced for decades
to represent society’s need to function through its own means and to
utilise energy and goods without impacting the health and well-being of

10
T. Sivarupan et al. Additive Manufacturing 40 (2021) 101889

Fig. 9. Volumetric heat capacity (J.K-1.m-3) × Thermal conductivity (W/m/◦ C) as a function of price in Australian dollars per volume (AUD/m3). The chart pro­
duced by CES-Edu pack software and material property database [80] containing a collection of material property data from published sources.

Fig. 10. Schematic of the integrated forming method for two-way coating and printing [49].

future generations [84]. The cost of sustainability must be integrated
into the manufacturing cost from conception [85]. These are incurred
across manufacturing time, quality and flexibility when pursuing sus­
tainable manufacturing practices. Such manufacturing operations are
characterised by the triple-bottom-line of sustainability - People (So­
cial), Planet (Environment) and Profit (Economy). Production processes
and systems have a significant impact on all three pillars of sustain­
ability [86].
Metal casting is considered an energy-intensive operation because of
the high amount of energy expended per product unit [87,88]. Typi­
cally, patterns and cores are produced with acceptable dimensional
tolerances to be placed inside a sand mould creating a mould cavity. This
contributes to 7 % to 20 % of the foundry’s overall energy consumption
[89]. A few studies have examined the sustainability of metal casting. Sa
et al. [90] discussed the relationship between management and energy
efficiency in analysing a Swedish foundry.
Similarly, another study was conducted by Carabili et al. [91] to
determine energy efficiency techniques within Colombian foundries. In
2016, around 36 % of the global CO2 emission was from the industrial
sector [92]. This indicates that metal casting can be among the critical
manufacturing technologies that will significantly affect combating
climate change.
Several energy-efficient techniques for manufacturing metal castings
have been suggested, that have also contributed to a decrease in non-
renewable energy consumption [93,94]. However, the primary
obstacle in reducing the overall energy consumption is improving

11
T. Sivarupan et al.
casting yields and reducing the use of metals, mainly due to the
non-optimised mould design in traditional metal-casting operations
[95].
From an emissions perspective, concerns on deteriorating air quality
by traditional foundries have been raised in the literature [96,97].
Release of toxins such as Volatile Organic Compounds (VOCs) and
hazardous chemicals and gases such as CO, SO2, NOx, H2S and HCN are
causing detrimental severe environmental effects. It is reported by Kmita
et al. [98] Reducing the contact period between the sand mould and the
hot metal can potentially bring down the toxic emissions. Furthermore,
dumping of the sand-casting waste in landfills is another environmental
concern, especially the furan dust [99]. The severity of the impact cor­
responds to one ton of sand being sent to landfill for every 2 tons of
casting [62,100,101]. Integration of additive manufacturing within
traditional metal casting can significantly improve the associated envi­
ronmental impact of the process [10]. Usually, a sand mould is a thick
heavy solid block with a cavity inside. Waste sand is one of the primary
pollutants in foundries. 3D printing technologies have been adopted to
make monolithic moulds and cores, but this is slow, materially ineffi­
cient and costly. Hollow moulds will reduce material usage and process
time to make mass production a possibility. Research works show that
more than 60 % of the sand, resin, time, and energy can be saved, and
the cost reduced by 75 % of current 3D mould techniques. This will
lower CO2 emissions and waste. Integrating cooling channels and noz­
zles will enable better microstructural control and present potential for
low-grade heat capture, e.g. Volvo’s G3 foundry, Fig. 11. This means
that the foundries have already started concentrating on energy opti­
misation in the processes. However, the 3D printed sand moulds and
cores can add another dimension to this improvement instead of
standing as a competing technology to the traditional casting tech­
niques. Alternative binders, e.g., ECOLOTEC [102], a water-based resole
process that uses CO2 as a hardening agent, are already in use. Never­
theless, these types of binder have just been introduced by one of the
major sand mould printer manufacturers, Voxeljet, as well as many
other manufacturers are in seeking a suitable binder; once these envi­
ronmentally friendlier binders have proven to show equivalent/better
properties of the furan mould a sustainable and environmentally
friendlier foundry processes should have a more significant positive
impact on the global carbon emissions.
Energy consumption for rapid mould manufacturing depends on the
type of 3D printer used. A range of commercial 3D printers is available
with average power consumption ranging between 5000 W and
Fig. 11. Schematic of the Volvo’s mould assembly for low grade heat capture as well as improved microstructure of the casting due to improved cooling rate [58].
12
Additive Manufacturing 40 (2021) 101889
10300 W. The choice of an appropriate 3D printer is dependent on the
size of the part to be produced, and the build speed required. The
accounted specific energy consumption is approximately 1.08 MJ/kg to
manufacture a sand mould and core of density 1738 kg/m3, using VX500
Voxeljet printer (with a direct power consumption of 10300 W). If the
carbon intensity is assumed as 325 g of CO2 per KWh of energy, the CO2
emissions can be estimated from the following equation.
( ) 325
CO2 kgCO2 = .E (2)
3600
Real-time data of carbon intensity for grid electricity in the UK can be
imported from the Ref. 103. Therefore, to select the best-suited process
from the sustainability perspective, the indicators need to be defined
with all three sustainability pillars in mind [104]. Some of the indicators
are listed below:
1. Social indicators: number of workplace accidents/deaths per annum,
number of people joining or leaving the foundry, long and short-term
illnesses, number of training seminars for the employees.
2. Economic indicators: cost of machines, cost of technology, energy
consumption, resource utilisation, net profit, overall manufacturing
costs, and waste-generation.
3. Environmental indicators: CO2 emissions, energy source, solid waste,
acidification, and heat transfer.
Pagone et al. [104] have formulated and discussed the
decision-making metrics for the mould manufacturing process selection,
considering the economic and environmental dimension of sustainabil­
ity. The process metrics are divided into four categories: environmental
sustainability, quality, cost and time. The metrics are shown in Table 4.
A positive or negative impact on the overall mould manufacturing
process (conventional or AM printed moulds) is defined for each quan­
tity. The authors concluded that AM tends to be the better overall
alternative when the categories are grouped and deemed equally
necessary to the decision-maker, demonstrating a substantial advantage
over the traditional mould manufacturing process. However, this holds
primarily for the small batch sizes. The cost of 3D printing begins to
dominate over conventional mould making process for larger produc­
tion volumes. Table 5 shows an estimation of the costs based on the lead
time and quantity.
The sustainability of additive manufacturing processes is also high­
lighted and discussed in the following studies [62,106,107].
T. Sivarupan et al. Additive Manufacturing 40 (2021) 101889

Table 4 solidification of molten alloys, but the sand mould may not compensate
Sustainability metrics for the process selection [105]. for this. Hence, a possibility for hot tearing, or the rapidly reducing
Quantity Impact Category cooling rate, especially at hotspots, becomes apparent due to the
mould’s temperature rise, which further results in coarser microstruc­
Total sand used in mould making (wm) Negative Environmental
sustainability ture and poorer mechanical properties consequently lower success rate.
Total sand used in core making (wc) Negative Environmental Some of these issues can be dealt with using metal chills within the sand
sustainability moulds. This is a strenuous activity and avoiding any inclusions within
Casting weight (wcast) Negative Environmental the mould is essential to speed up the availability of such a large-scale
sustainability
CO2 emissions (CO2) Negative Environmental
complex sand mould assembly; that is, to provide a sand mould
sustainability directly from the printer without any chill inclusion. Thermal properties
Total energy consumption in mould Negative Environmental can be altered using various sand types. Design of a sand mould as­
making (E) sustainability sembly, using various sand types, and identifying the right sand type for
Tensile strength of mould (σt) Positive Quality
each section of a mould assembly is expected to provide a viable solution
Surface roughness of casting (Ra) Negative Quality
Porosity of casting Negative Quality using 3D sand mould printers.
Compressive strength-casting (σc) Positive Quality However, there remain specific challenges in establishing smart
Hardness-casting (HV) Positive Quality foundries as highlighted by Ref. 113. Smart foundries are highly reliant
Cost of one mould Negative Cost on connectivity. If the foundry experiences a loss in the internet
Mould making time Negative Time
connection in a worst-case scenario, the entire production process will
halt. This will further lead to data loss. Challenges related to data
phishing and security should be taken care of well in advance.
Table 5 Furthermore, the supply chain will be severely impacted by the exten­
Overall cost of mould manufacturing. sive automation of the foundries. Thus, rather than changing the entire
3D printed moulds Conventional moulds production model, it is better to add small modules at once.
Quantity tlead = 5 days tlead = 21 days tlead = (4–6) weeks Hybrid manufacturing, Fig. 12, processes are inevitably vital due to
1 € 898 € 410 € 3600 the possible correlation of each technology’s benefit. For example, 3D
5 € 3080 € 1428 € 3684 sand mould printing contributes to design freedom, while traditional
10 € 5490 € 2525 € 3789
casting contributes to reliable mechanical properties and sustainability
50 € 22,275 € 10,300 € 4628
and cost of production. Strategically placing the manufacturing plants
Table adapted from [105]. can significantly minimise the cost of a metal part in the hybrid-AM
supply chain [114] due to the possibility of foundries maintaining a
5. Integrating traditional and 3D printing processes along with digital-twin inventory for a final product instead of a finished product
artificial intelligence [115]. Therefore, the focus must be on how well AM technologies can be
integrated into traditional manufacturing processes.
The fourth industrial revolution (Industry 4.0), has impacted the The authors have published articles related to 3D sand mould
entire manufacturing sector [108]. As a result, there has been much printing [6–8,10,13–18] and the energy-saving on-demand casting
emphasis on smart manufacturing products and services, and foundries process discussing the potential of the CRIMSON (Constrained Rapid
are no exception. The nine pillars of Industry 4.0 are (1) Additive Induction Melted Single Shot Up Casting Method) technique [88,94,113,
Manufacturing, (2) Industrial Internet of Things (IIoT), (3) Cyberse­ 116–121]. 3D sand mould printing [10] and the CRIMSON [88] process
curity, (4) The cloud, (5) Horizontal and Vertical Integration, (6) have been reported as energy-efficient (more than 50 %) alternative
Augmented reality, (7) Simulation, (8) Autonomous robots and, (9) Big production routes in place of the traditional casting process. The benefits
data. The transformation of traditional foundries to smart foundries has of the CRIMSON process are (1) the use of clean metal charges with low
already started with the integration of the 3M’s (Man, Machine and impurities thus reducing casting defects, (2) on-demand production and
Material) into all stages of the value chain. Hybrid Additive consumption of molten metal (current practice is to melt large quantities
Manufacturing processes have already been developed, which will fill and hold at a high temperature until filling multiple moulds), (3) rapid
the gap between the strength and weakness of AM and the traditional induction melting in a constrained crucible for high production rates
processes. Hence, multi-material and multi-functional capabilities can and low exposure to the environment (reducing the detrimental effects
be brought into the production of metallic parts [109]. on the castings), and (4) use of counter gravity filling to control molten
Sahoo et al. [110] highlighted the application of artificial intelli­ metal flow and to eliminate turbulence on the surface of the melt to
gence (AI) in determining the sand mould quality; thereby, the best reduce entrapped casting defects.
casting parameters can be identified. Integrating the process simulation The generatively designed CAD (computer-aided design) model can
and the casting process with the 3D printing of a functional sand mould be sent directly to a printer to print the sections of a mould. These are
(time- and temperature-dependent properties) may transform the tech­ then assembled with sensors (using AI to learn continuously), and the
nology into 4D printing. 3D refers to the length, width and height of an molten alloy is injected using the CRIMSON process that can be remotely
object while 4D would refer the time or temperature as the 4th dimen­ monitored and controlled. Integrating cooling channels and nozzles can
sion in addition to the 3D printing; 4D printing is the process through further enable better microstructural control and have the potential for
which a 3D printed object transforms itself into another structure over low-grade heat capture. Production of a functional sand mould with the
the influence of external energy input as temperature or time. While the required properties, such as permeability and cooling rate, can elevate
direct metal 3D printing research has progressed into the 4D stage [111], this technology to challenge any other production technique for large
the 3D sand mould printing technology is facing challenges in utilising scale, cleaner, and economic production without the laborious addition
the benefit of the technology due to the lack of options to produce a of chillers within the mould. The hybridisation of traditional casting
functional sand mould with time- or temperature-dependent properties; with 3D sand mould printing needs further development to absorb po­
functional mould could assist the solidification of molten alloy poured tential benefits. Controlled pressure casting techniques can be utilised to
into the cavity and produce a casting with optimal properties. Binder tailor the functional 3D printed sand mould to promote a cleaner pro­
degradation during casting reduces thermal properties such as heat duction with minimal or no post-processing to take this technology to
conductivity and volumetric heat capacity; mainly, the binder content is the 4D stage. The functional mould can contract or adjust properties
high [112]. Significant thermal contraction is inherent during the according to the solidifying molten alloy’s requirement to minimise or

13
T. Sivarupan et al.
Fig. 12. Schematic of the possible mould manufacturing routes for sand casting [11].
prevent defect formation; the sand mould should be able to shrink with
the casting and provide a constant/increasing cooling rate with the
reducing temperature.
The sand mould printing process’s speed can also be improved by
possible application of machining the patterns for a traditionally made
block of sand in addition to the 3D printed fine structures for a large-
scale mould assembly. In other words, it is not necessary to print all
sections of the sand mould assembly using the 3D printing process. The
fine details that are difficult to achieve by the machining process can be
manufactured using the printer. The rest of the simple large shapes that
can easily be machined can utilise the traditional machining process.
Finally, the sections produced by 3D printing and machining can then be
assembled to make the large-scale mould. The hybridisation of tradi­
tional processes with new technologies assists in the rapid production of
parts and thus help minimise the carbon emission [11].
The vast range of process parameters in the 3D printing processes
poses challenges these technologies’ ultimate benefits. Implementation
of AI to continuously learn and decide the optimal process parameters is
expected to significantly improve the manufacturing industry, which is
facing challenges with centralised manufacturing. Researchers have
reported the application of machine learning to identify in-situ melt pool
signatures indicative of flaw formation in a laser powder bed fusion
additive manufacturing process [122]. Centralised data sharing, rather
than centralised manufacturing, is expected to provide a better
manufacturing industry solution. A decision-making tool to select the
traditional manufacturing process, an integrated manufacturing process
or a complete 3D printing process is the better route to manufacture
from a CAD model. A complete digital twin (CAD model, process pa­
rameters, and the optimal production method with time, energy, and
financial information) of a part can be prepared.
Controlling melt turbulence or speed is a determining factor of
casting quality, especially porosity, and this can be measured using
sensors embedded within a sand mould to remotely collect real-time
data and to compare against the computational simulation results,
therefore helping to improve the gating system design [123] and to
monitor the core shift during casting [124]. Son et al. [125] identified
the reliability of the flexural strength of the fabricated specimens
14
Additive Manufacturing 40 (2021) 101889
evaluated by a Weibull plot1 when the weight fraction of 0.2 % activator
was used with the binder content of 2 %. However, similar activator
content (weight fraction of the sand 0.18 %) was also used in previous
studies [6–8]. Machine learning can help cluster this type of experi­
mental data and help decide the optimum parameters for future needs.
Manufacturing guidelines and life cycle analysis databases [2] should be
made by incorporating the process parameters that influence the mould
design. This would make the dynamic process parameter adjustment
possible. For example, a mould section that requires just volume crea­
tion may be printed with higher layer thickness as the dimensional
tolerance is not an issue when printing that particular region of the
mould.
In contrast, a surface feature with complex geometry may be printed
with lower layer thickness to improve the tolerance for that region. In
each case, the sand particle size and supply rate, and x-/y- resolution of
the binder should be able to be altered by applying the machine learning
algorithms. This requires improvements in the design of the printing
machines. The GrantaMI module [126] of the materials selection soft­
ware package [80] has initiated this but needs to be strengthened with
data related to 3D sand mould printing. A CAD model could be used to
access materials data within the HyperMesh simulation software.
6. Conclusions
This review has identified some opportunities for the development of
3D sand printing. The printing speed of large-scale parts could be
improved by design improvements to the re-coater and printhead.
Environmental friendlier binders should be identified. Focusing on the
mould design for functional properties can help improve the casting
yield. A high-speed binder jet technology may be challenging, but this
can be achieved by having an array of printheads operating together and
having re-coaters on both sides of the printhead, where the printing
process is near-continuous for mass production. Generative design
1
Though the number of specimens used for each case in this study was less
than 10, a limiting factor of confidence.
T. Sivarupan et al.
methodologies can be utilised in designing sand mould assemblies, and
consequently, energy and mass consumption can be optimised. Appli­
cation of machine learning should be utilised on a global user scale so
that the printing process parameters can be optimised or shared between
users; this can be achieved using materials selection databases. Decision
making and process parameter identification algorithms should be
developed. All the traditional and modern manufacturing technologies
should be integrated based on the part complexity, volume, and other
customer requirements to find the optimal manufacturing route and
benefit from the integration of these technologies.
7. Research agenda
Finding environmentally friendlier consumables such as binders and
hardeners are essential to optimise the benefits of the 3D sand mould
printing technology. Possible ways of recycling the sand, at least the
unbound sand, should be identified such as a washout method and
drying after each casting, possibly using heat energy from the solidifying
melt within a continuous production cycle. Production of functional
sand moulds with the required properties, such as permeability and
cooling rate, can help this technology to challenge any other production
techniques, even for large scale economic production, without the need
for adding chills within the mould as these may be labor intensive to
produce. The hybridisation of traditional casting with 3D sand mould
printing for large-scale production needs further research; controlled
pressure casting techniques can be utilised to tailor the functional 3D
printed sand mould in order to promote cleaner production with mini­
mal or no post-processing and these could take this technology to the 4D
(four-dimensional) stage. The functional mould can contract or adjust
properties according to the solidifying molten alloy’s requirements to
minimise or prevent defect formation, therefore increasing the success
rate up to 100 % with the help of mould filling and solidification
simulation. In this case, the resin used to print the sand mould should be
identified and should be able to shrink with the decreasing temperature.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
The support of the Centre for Advanced Material Processing and
Manufacturing (AMPAM), The University of Queensland, Australia, is
much acknowledged. This paper includes research that was supported
by DMTC Limited (Australia). The authors have prepared this paper in
accordance with the intellectual property rights granted to partners
from the original DMTC project. Research referenced from Cranfield
University authors has been funded by UK EPSRC projects EP/G060096/
1/2, EP/M013863/1 and EP/P012272/1.
References
[1] X.H. Wang, J.Y.H. Fuh, Y.S. Wong, Y.X. Tang, Laser sintering of silica sand –
mechanism and application to sand casting mould, Int. J. Adv. Manuf. Technol.
21 (12) (2003) 1015–1020.
[2] T.A. Le Néel, P. Mognol, J.-Y. Hascoët, A review on additive manufacturing of
sand molds by binder jetting and selective laser sintering, Rapid Prototyp. J. 24
(8) (2018) 1325–1336.
[3] D. Bourell, J.P. Kruth, M. Leu, G. Levy, D. Rosen, A.M. Beese, A. Clare, Materials
for additive manufacturing, CIRP Ann. 66 (2) (2017) 659–681.
[4] A. Mostafaei, A.M. Elliott, J.E. Barnes, C.L. Cramer, P. Nandwana, M. Chmielus,
WITHDRAWN: Binder jet 3D printing – process parameters, materials, properties,
and challenges, Prog. Mater. Sci. (2020), 100684.
[5] D. Günther, F. Mögele, Additive manufacturing of casting tools using powder-
binder-jetting technology, in: New Trends in 3D Printing, 2016.
15
Additive Manufacturing 40 (2021) 101889
[6] T. Sivarupan, M. El Mansori, N. Coniglio, M. Dargusch, Effect of process
parameters on flexure strength and gas permeability of 3D printed sand molds,
J. Manuf. Process. 54 (2020) 420–437.
[7] T. Sivarupan, M. El Mansori, K. Daly, M.N. Mavrogordato, F. Pierron,
Characterisation of 3D printed sand moulds using micro-focus X-ray computed
tomography, Rapid Prototyp. J. 25 (2) (2019) 404–416.
[8] N. Coniglio, T. Sivarupan, M. El Mansori, Investigation of process parameter
effect on anisotropic properties of 3D printed sand molds, Int. J. Adv. Manuf.
Technol. 94 (5–8) (2017) 2175–2185.
[9] I. Gibson, D. Rosen, B. Stucker, Binder Jetting, Additive Manufacturing
Technologies, Springer, New York, New York, NY, 2015, pp. 205–218.
[10] T. Sivarupan, M. Upadhyay, Y. Ali, M. El Mansori, M.S. Dargusch, Reduced
consumption of materials and hazardous chemicals for energy efficient
production of metal parts through 3D printing of sand molds, J. Clean. Prod. 224
(2019) 411–420.
[11] J. Zheng, A. Chen, W. Zheng, X. Zhou, B. Bai, J. Wu, W. Ling, H. Ma, W. Wang,
Effectiveness analysis of resources consumption, environmental impact and
production efficiency in traditional manufacturing using new technologies: case
from sand casting, Energy Convers. Manag. 209 (2020), 112671.
[12] S.R. Sama, T. Badamo, G. Manogharan, Case studies on integrating 3D sand-
printing technology into the production portfolio of a sand-casting foundry, Int. J.
Met. 14 (1) (2019) 12–24.
[13] H.R. Dana, M. El Mansori, Mechanical characterisation of anisotropic silica sand/
furan resin compound induced by binder jet 3D additive manufacturing
technology, Ceram. Int. 46 (11) (2020) 17867–17880.
[14] S. Mitra, M. El Mansori, M. Costin, A. Rodríguez de Castro, On quality assessment
of 3D printed sand mold using non-destructive X-ray µ-CT.
[15] S. Mitra, A. Rodríguez de Castro, M. El Mansori, The effect of ageing process on
three-point bending strength and permeability of 3D printed sand molds, Int. J.
Adv. Manuf. Technol. 97 (2018) 1241–1251.
[16] S. Mitra, A. Rodríguez de Castro, M. El Mansori, On the rapid manufacturing
process of functional 3D printed sand molds, J. Manuf. Process. 42 (2019)
202–212.
[17] S. Mitra, M. El Mansori, M. Costin, A.R.D. Castro, On quality assessment of 3D
printed sand mold using non-destructive X-ray µ-CT, in: International Symposium
on Digital Industrial Radiology and Computed Tomography – DIR2019, 2019.
[18] M. Upadhyay, T. Sivarupan, M. El Mansori, 3D printing for rapid sand casting—a
review, J. Manuf. Process. (2017) 211–220.
[19] C. Hu, W. Du, Research on the method for improving mechanical properties of
sand mold based on 3D printing process, IOP Conf. Ser. Mater. Sci. Eng. 394 (3)
(2018) 0–6.
[20] H. Shangguan, J. Kang, C. Deng, Y. Hu, T. Huang, 3D-printed shell-truss sand
mold for aluminum castings, J. Mater. Process. Technol. 250 (2017) 247–253.
[21] H. Shangguan, J. Kang, J. Yi, X. Zhang, X. Wang, H. Wang, T. Huang, The design
of 3D-printed lattice-reinforced thickness-varying shell molds for castings,
Materials 11 (4) (2018).
[22] H.-l Shangguan, J.-w Kang, J.-h Yi, C.-y Deng, Y.-y Hu, T. Huang, Controlled
cooling of an aluminum alloy casting based on 3D printed rib reinforced shell
mold, China Foundry 15 (3) (2018) 210–215.
[23] C. Deng, J. Kang, H. Shangguan, Y. Hu, T. Huang, Z. Liu, Effects of hollow
structures in sand mold manufactured using 3D printing technology, J. Mater.
Process. Technol. 255 (2018) 516–523.
[24] J. Kang, H. Shangguan, C. Deng, Y. Hu, J. Yi, X. Wang, X. Zhang, T. Huang,
Additive manufacturing-driven mold design for castings, Addit. Manuf. 22 (2018)
472–478.
[25] P. Lechner, G. Fuchs, C. Hartmann, F. Steinlehner, F. Ettemeyer, W. Volk,
Acoustical and optical determination of mechanical properties of inorganically-
bound foundry core materials, Materials 13 (11) (2020).
[26] G.T. Kridli, P.A. Friedman, J.M. Boileau, Manufacturing processes for light alloys,
in: P.K. Mallick (Ed.), Materials, Design and Manufacturing for Lightweight
Vehicles, Woodhead Publishing, 2010, pp. 235–274.
[27] I. Polmear, D. StJohn, J.-F. Nie, M. Qian, Casting of light alloys, in: I. Polmear,
D. StJohn, J.-F. Nie, M. Qian (Eds.), Light Alloys, Butterworth-Heinemann,
Boston, 2017, pp. 109–156.
[28] E.S. Almaghariz, B.P. Conner, L. Lenner, R. Gullapalli, G.P. Manogharan,
B. Lamoncha, M. Fang, Quantifying the role of part design complexity in using 3D
sand printing for molds and cores, Int. J. Met. 10 (3) (2016) 240–252.
[29] J.-H. Wang, A. Qu, T.R. Langlois, D.L. James, Toward wave-based sound synthesis
for computer animation, ACM Trans. Graph. 37 (4) (2018) 1–16.
[30] M. Aghaei Meibodi, R. Giesecke, B. Dillenburger, 3D printing sand molds for
casting bespoke metal connections-digital metal: additive manufacturing for cast
metal joints in architecture, intelligent & informed, in: Proceedings of the 24th
International Conference of the Association for Computer-Aided Architectural
Design Research in Asia (CAADRIA), Association for Computer-Aided
Architectural Design Research in Asia (CAADRIA), Hong Kong, 2019, pp.
133–142.
[31] D. Snelling, H. Blount, C. Forman, K. Ramsburg, A. Wentzel, C. Williams, A.
Druschitz, The effects of 3D printed molds on metal castings, in: Proceedings of
the Solid Freeform Fabrication Symposium, 2013, pp. 827–845.
[32] M. Chhabra, R. Singh, Obtaining desired surface roughness of castings produced
using ZCast direct metal casting process through Taguchi’s experimental
approach, Rapid Prototyp. J. 18 (6) (2012) 458–471.
[33] M. Chhabra, R. Singh, Mathematical modeling of surface roughness of castings
produced using ZCast direct metal casting, J. Inst. Eng. Ser. C 96 (2) (2014)
145–155.
T. Sivarupan et al.
[34] V.V. Drokina, V.D. Belov, S.N. Chekhonin, Obtaining casts of aluminum alloys by
foundry in loose molds fabricated on installations of three-dimensional printing,
Russ. J. Non-Ferr. Met. 52 (1) (2011) 24–28.
[35] P.M. Hackney, R. Wooldridge, Characterisation of direct 3D sand printing process
for the production of sand cast mould tools, Rapid Prototyp. J. 23 (1) (2017)
7–15.
[36] N. Hawaldar, J. Zhang, A comparative study of fabrication of sand casting mold
using additive manufacturing and conventional process, Int. J. Adv. Manuf.
Technol. 97 (1–4) (2018) 1037–1045.
[37] M. Vaezi, C.K. Chua, Effects of layer thickness and binder saturation level
parameters on 3D printing process, Int. J. Adv. Manuf. Technol. 53 (1–4) (2010)
275–284.
[38] D. Snelling, C. Williams, A. Druschitz, A comparison of binder burnout and
mechanical characteristics of printed and chemically bonded sand molds, in: SFF
Symposium, 2014, pp. 197–209.
[39] J. Tóth, J.T. Svidró, A. Diószegi, D. Stevenson, Heat absorption capacity and
binder degradation characteristics of 3D printed cores investigated by inverse
Fourier thermal analysis, Int. J. Met. 10 (3) (2016) 306–314.
[40] S.S. Gill, M. Kaplas, Efficacy of powder-based three-dimensional printing (3DP)
technologies for rapid casting of light alloys, Int. J. Adv. Manuf. Technol. 52 (1–4)
(2010) 53–64.
[41] R. Kumar, R. Singh, I.P.S. Ahuja, Modeling and analysis for hardness and
structure of nonferrous alloy castings produced using Zcast metal casting process
through response surface methodology, Adv. Mater. Res. 1137 (2016) 101–116.
[42] N. McKenna, S. Singamneni, O. Diegel, D. Singh, T. Neitzert, J. St George, A. Roy
Choudhury, P. Yarlagadda, Direct metal casting through 3D printing: a critical
analysis of the mould characteristics, in: Global Congress on Manufacturing and
Management, 2008.
[43] E. Bassoli, E. Atzeni, Direct metal rapid casting: mechanical optimization and
tolerance calculation, Rapid Prototyp. J. 15 (2009) 238–243.
[44] D. Snelling, R. Kay, A. Druschitz, C. Williams, Mitigating gas defects in castings
produced from 3D printed molds, in: 117th Metalcasting Congress, 2012.
[45] D. Zhao, W. Guo, B. Zhang, F. Gao, Research on key technique of line forming for
3D sand mould printing based on quantitative analysis of binder content, Rapid
Prototyp. J. 25 (1) (2019) 62–75.
[46] W.G. Donghua Zhao, Baibing Zhang, Feng Gao, Investigation on moist silica
sand’s circulation for sand mold printing with line-forming, Rapid Prototyp. J.
(2019).
[47] D. Zhao, W. Guo, B. Zhang, F. Gao, 3D sand mould printing: a review and a new
approach, Rapid Prototyp. J. 24 (2) (2018) 285–300.
[48] Z. Guo, Z. Shan, F. Liu, D. Du, M. Zhao, Experimental investigation on the
performance and mesostructure of multi-material composite 3D-printed sand
mold, Rapid Prototyp. J. 26 (2) (2019) 309–318.
[49] Z. Shan, Z. Guo, D. Du, F. Liu, W. Li, Digital high-efficiency print forming method
and device for multi-material casting molds, Front. Mech. Eng. 15 (2) (2020)
328–337.
[50] M.A.A. Khan, A.K. Sheikh, A comparative study of simulation software for
modelling metal casting processes, Int. J. Simul. Model. 17 (2) (2018) 197–209.
[51] M. Papanikolaou, P. Saxena, E. Pagone, K. Salonitis, M. Jolly, Optimisation of the
filling process in counter-gravity casting, IOP Conf. Ser. Mater. Sci. Eng. 861
(2020), 012031.
[52] S.R. Sama, T. Badamo, P. Lynch, G. Manogharan, Novel sprue designs in metal
casting via 3D sand-printing, Addit. Manuf. 25 (2019) 563–578.
[53] D.W. Kang, C.M. Lee, A study on the development of the laser-assisted milling
process and a related constitutive equation for silicon nitride, CIRP Ann. 63 (1)
(2014) 109–112.
[54] D. Snelling, Q. Li, N. Meisel, C.B. Williams, R.C. Batra, A.P. Druschitz,
Lightweight metal cellular structures fabricated via 3D printing of sand cast
molds, Adv. Eng. Mater. 17 (7) (2015) 923–932.
[55] S.R. Sama, J. Wang, G. Manogharan, Non-conventional mold design for metal
casting using 3D sand-printing, J. Manuf. Process. 34 (2018) 765–775.
[56] J.-w Kang, Q.-x Ma, The role and impact of 3D printing technologies in casting,
China Foundry 14 (3) (2017) 157–168.
[57] ExOne, 3D Printed Sacrificial Tooling —Washes Out with Water, 2020. 〈http
s://www.exone.com/en-US/parts-and-service/washout-tooling〉. (Accessed 25
May 2020).
[58] V.G.T.O.-P.T. Production-Skovde, Volvo- A foundry with a vision, 2014.
[59] A.V. Koltygin, V.E. Bazhenov, Development of a substitute for Z cast molding
sand used on installations of 3D printing for obtaining aluminum, magnesium,
and iron casting, Russ. J. Non-Ferr. Met. 53 (1) (2012) 38–41.
[60] R. Ramakrishnan, B. Griebel, W. Volk, D. Günther, J. Günther, 3D printing of
inorganic sand moulds for casting applications, Adv. Mater. Res. 1018 (2014)
441–449.
[61] K.J. Hodder, R.J. Chalaturnyk, Bridging additive manufacturing and sand casting:
utilizing foundry sand, Addit. Manuf. 28 (2019) 649–660.
[62] J. Mitterpach, E. Hroncova, J. Ladomersky, K. Balco, Environmental analysis of
waste foundry sand via life cycle assessment, Environ. Sci. Pollut. Res. Int. 24 (3)
(2017) 3153–3162.
[63] P. Lynch, C.R. Hasbrouck, J. Wilck, M. Kay, G. Manogharan, Challenges and
opportunities to integrate the oldest and newest manufacturing processes: metal
casting and additive manufacturing, Rapid Prototyp. J. 26 (6) (2020) 1145–1154.
[64] L. Xue, C. Hu, X. Li, W. Li, R. Zhang, Research on the influence of furan resin
addition on the performance and accuracy of 3D printing sand mold, IOP Conf.
Ser. Mater. Sci. Eng. 392 (6) (2018) 0–6.
[65] S.S. Gill, M. Kaplas, Comparative study of 3D printing technologies for rapid
casting of aluminium alloy, Mater. Manuf. Process. 24 (12) (2009) 1405–1411.
16
Additive Manufacturing 40 (2021) 101889
[66] S.S. Gill, M. Kaplas, Efficacy of powder-based three-dimensional printing (3DP)
technologies for rapid casting of light alloys, Int. J. Adv. Manuf. Technol. 52 (1–4)
(2011) 53–64.
[67] S. Singamneni, N. McKenna, O. Diegel, D. Singh, T. Neitzert, J.S. George, A.
R. Choudhury, P. Yarlagadda, Rapid casting: a critical analysis of mould and
casting characteristics, Aust. J. Mech. Eng. 7 (1) (2009) 33–43.
[68] B. Primkulov, J. Chalaturnyk, R. Chalaturnyk, G. Zambrano Narvaez, 3D printed
sandstone strength: curing of furfuryl alcohol resin-based sandstones, 3D Print.
Addit. Manuf. 4 (3) (2017) 149–156.
[69] M. Upadhyay, T. Sivarupan, M. El Mansori, 3D printing for rapid sand casting—a
review, J. Manuf. Process. 29 (2017) 211–220.
[70] J. Walker, E. Harris, C. Lynagh, A. Beck, R. Lonardo, B. Vuksanovich, J. Thiel,
K. Rogers, B. Conner, E. MacDonald, 3D printed smart molds for sand casting, Int.
J. Met. 12 (4) (2018) 785–796.
[71] D.-H. Kim, J. Lee, J. Bae, S. Park, J. Choi, J.H. Lee, E. Kim, Mechanical analysis of
ceramic/polymer composite with mesh-type lightweight design using binder-Jet
3D printing, Materials 11 (10) (2018) 1941.
[72] H. Shangguan, J. Kang, C. Deng, J. Yi, Y. Hu, T. Huang, 3D-printed rib-enforced
shell sand mold for aluminum castings, Int. J. Adv. Manuf. Technol. 96 (5–8)
(2018) 2175–2182.
[73] J. Zheng, A. Chen, W. Zheng, X. Zhou, B. Bai, J. Wu, W. Ling, H. Ma, W. Wang,
Effectiveness analysis of resources consumption, environmental impact and
production efficiency in traditional manufacturing using new technologies: case
from sand casting, Energy Convers. Manag. 209 (2020), 112671.
[74] S. Mitra, M. El Mansori, A. Rodríguez de Castro, M. Costin, Study of the evolution
of transport properties induced by additive processing sand mold using X-ray
computed tomography, J. Mater. Process. Technol. 277 (2020), 116495.
[75] H. Liu, Y. Gao, S. Ding, F. Peng, D. Zhu, Research on the effect of the waveform on
the droplet injection behavior of a piezoelectric printhead and the forming
accuracy of casting sand molds, Int. J. Adv. Manuf. Technol. 100 (1–4) (2018)
251–261.
[76] Y.Y.颜. Yang Weidong (杨伟东)**, Zhang Renji (张人佶)†, Liu Libing (刘丽冰),
Effective Factors on Forming Precision in Patternless Casting Manufacturing
Technique, Tsinghua Science and Technology 14 (Number S1, June 2009).
[77] Z.-d Shan, Z. Guo, D. Du, F. Liu, Coating process of multi-material composite sand
mold 3D printing, China Foundry 14 (6) (2018) 498–505.
[78] S. Chastain, Metal Casting: A Sand Casting Manual for the Small Foundry, Steve
Chastain, 2004.
[79] ExOne, Sand Casting 3D Printing Materials & Binders, 2020. 〈http://www.exone.
com/Resources/Materials〉.
[80] Granta, A selection of publications referencing CES Selector, 2020. 〈http://www.
grantadesign.com/userarea/papers/〉.
[81] N. Bakhiya, K.E. Appel, Toxicity and carcinogenicity of furan in human diet, Arch.
Toxicol. 84 (7) (2010) 563–578.
[82] HoosiernPatternInc., Calculating the Cost of 3D Sand Printing, 2018. 〈https://
hoosierpattern.com/news/cost-analysis-breakdown-of-3d-sand-printing〉.
(Accessed 04 June 2020).
[83] EnvisionTEC, The Robotic Additive Manufacturing systems from EnvisionTEC
and Viridis3D are the fastest, most flexible robotic 3D printing options in the
foundry and 3D printing industries today, 2020. 〈https://envisiontec.co
m/3d-printers/robotic-additive-manufacturing/〉. (Accessed 15 June 2020).
[84] e. G.H. Bruntland, Report of the World Commission on Environment and
Development, Our Common Future, General United Nations Assembly Resolution,
42/187, 1987.
[85] F. Badurdeen, I.S. Jawahir, Strategies for value creation through sustainable
manufacturing, Procedia Manuf. 8 (2017) 20–27.
[86] P. Saxena, P. Stavropoulos, J. Kechagias, K. Salonitis, Sustainability assessment
for manufacturing operations, Energies 13 (11) (2020) 2730.
[87] E. Pagone, K. Salonitis, M. Jolly, Energy and material efficiency metrics in
foundries, in: 5th Global Conference on Sustainable Manufacturing,
Manufacturing Engineering Society International Conference 2017, MESIC, 2018.
[88] E. Pagone, K. Salonitis, M. Jolly, Energy-Efficient Casting Processes, Innovations
in Manufacturing for Sustainability, Springer, Cham, 2019, pp. 77–98.
[89] U.S.D.o. Environment, Energy and Environmental Profile of the U. S. Aluminum
Industry, Office (July), 1997.
[90] A. Sa, S. Paramonova, P. Thollander, E. Cagno, Classification of industrial energy
management practices, Energy Procedia 75 (2015) 2581–2588.
[91] D.M. Carabalí, C.R. Forero, Y. Cadavid, Energy diagnosis and structuring an
energy saving proposal for the metal casting industry: an experience in Colombia,
Appl. Therm. Eng. 137 (2018) 767–773.
[92] I.E. Association, CO2 Emissions from Fuel Combustion 2018—Highlights;
International Energy Association, Paris, France, 2018. 〈https://webstore.iea.org/
co2-emissions-from-fuel-combustion-2018-highlights〉. (Accessed 30 April 2020).
[93] O. Estrada, I.D. López, A. Hernández, J.C. Ortíz, Energy gap method (EGM) to
increase energy efficiency in industrial processes: successful cases in polymer
processing, J. Clean. Prod. 176 (2018) 7–25.
[94] K. Salonitis, M.R. Jolly, B. Zeng, H. Mehrabi, Improvements in energy
consumption and environmental impact by novel single shot melting process for
casting, J. Clean. Prod. 137 (2016) 1532–1542.
[95] K. Salonitis, B. Zeng, H.A. Mehrabi, M. Jolly, The challenges for energy efficient
casting processes, Procedia CIRP 40 (2016) 24–29.
[96] N. Tiedje, R. Crepaz, T. Eggert, N. Bey, Emission of organic compounds from
mould and core binders used for casting iron, aluminium and bronze in sand
moulds, J. Environ. Sci. Health A Toxic Hazard. Subst. Environ. Eng. 45 (14)
(2010) 1866–1876.
T. Sivarupan et al.
[97] B.C.S. Incorporated, Energy and Environmental Profile of the U.S. Mining
Industry, (December), 2002, 11–11.
[98] A. Kmita, C. Fischer, K. Hodor, M. Holtzer, A. Roczniak, Thermal decomposition
of foundry resins: a determination of organic products by thermogravimetry–gas
chromatography–mass spectrometry (TG–GC–MS), Arab. J. Chem. 11 (3) (2018)
380–387.
[99] M. Holtzer, R. Dańko, J. Dańko, Thermal utilization of the post reclamation dust
from molding sand with furan resin in test unit, Arch. Foundry Eng. 18 (4) (2018)
15–18.
[100] S. Dalquist, T. Gutowski, Life cycle analysis of conventional manufacturing
techniques: sand casting, in: ASME 2004 International Mechanical Engineering
Congress and Exposition, 2004, pp. 631–641.
[101] R. Huang, M.E. Riddle, D. Graziano, S. Das, S. Nimbalkar, J. Cresko, E. Masanet,
Environmental and economic implications of distributed additive manufacturing:
the case of injection mold tooling, J. Ind. Ecol. 21 (S1) (2017) S130–S143.
[102] Foseco, ECOLOTEC waterbased resole CO2 process, 2020.
[103] G.s.h.p. association/ICAX, Carbon Emissions Calculator Carbon emissions from
electric heating reduce as the Grid decarbonises each year, the efficient way to
use electricity is by heat transfer, 2020. 〈https://www.icax.co.uk/Carbon
_Emissions_Calculator.html〉. (Accessed 03 June 2020).
[104] E. Pagone, P. Saxena, M. Papanikolaou, K. Salonitis, M. Jolly, Sustainability
Assessment of Rapid Sand Mould Making Using Multi-criteria Decision-Making
Mapping, Springer Singapore, Singapore, 2021, pp. 345–355.
[105] D. Chen, S. Heyer, S. Ibbotson, K. Salonitis, J.G. Steingrímsson, S. Thiede, Direct
digital manufacturing: definition, evolution, and sustainability implications,
J. Clean. Prod. 107 (2015) 615–625.
[106] T.D. Ngo, A. Kashani, G. Imbalzano, K.T.Q. Nguyen, D. Hui, Additive
manufacturing (3D printing): a review of materials, methods, applications and
challenges, Compos. Part B Eng. 143 (2018) 172–196.
[107] S. Singh, S. Ramakrishna, M.K. Gupta, Towards zero waste manufacturing: a
multidisciplinary review, J. Clean. Prod. (2017) 1230–1243.
[108] T. Stock, G. Seliger, Opportunities of sustainable manufacturing in industry 4.0,
Procedia CIRP 40 (2016) 536–541.
[109] A. Bandyopadhyay, B. Heer, Additive manufacturing of multi-material structures,
Mater. Sci. Eng. R Rep. 129 (2018) 1–16.
[110] P.K. Sahoo, S. Pattnaik, M.K. Sutar, A state-of-the-art review on manufacturing
and additive influences on sand-cast components, Arab. J. Sci. Eng. 44 (12)
(2019) 9805–9835.
[111] A. Mitchell, U. Lafont, M. Hołyńska, C. Semprimoschnig, Additive manufacturing
— a review of 4D printing and future applications, Addit. Manuf. 24 (2018)
606–626.
17
Additive Manufacturing 40 (2021) 101889
[112] D. Martinez, C. Bate, G. Manogharan, Towards functionally graded sand molds for
metal casting: engineering thermo-mechanical properties using 3D sand printing,
JOM 72 (3) (2020) 1340–1354.
[113] P. Saxena, M. Papanikolaou, E. Pagone, K. Salonitis, M.R. Jolly. Digital
Manufacturing for Foundries 4.0, Springer International Publishing, Cham, 2020,
pp. 1019–1025.
[114] D. Strong, M. Kay, B. Conner, T. Wakefield, G. Manogharan, Hybrid
manufacturing – integrating traditional manufacturers with additive
manufacturing (AM) supply chain, Addit. Manuf. 21 (2018) 159–173.
[115] B. Berman, 3-D printing: the new industrial revolution, Bus. Horiz. 55 (2) (2012)
155–162.
[116] M. Papanikolaou, E. Pagone, K. Georgarakis, K. Rogers, M. Jolly, K. Salonitis,
Design optimisation of the feeding system of a novel counter-gravity casting
process, Metals 8 (10) (2018) 817.
[117] K. Salonitis, M. Jolly, B. Zeng, Simulation based energy and resource efficient
casting process chain selection: a case study, Procedia Manuf. 8 (2017) 67–74.
[118] K. Salonitis, B. Zeng, H.A. Mehrabi, M. Jolly, The challenges for energy efficient
casting processes, Procedia CIRP (2016) 24–29.
[119] B. Zeng, M. Jolly, K. Salonitis, Manufacturing cost modeling of castings produced
with CRIMSON process, in: M. Tiryakioğlu, J. Campbell, G. Byczynski (Eds.),
Shape Casting: 5th International Symposium 2014, Springer International
Publishing, Cham, 2014, pp. 201–208.
[120] B. Zeng, M. Jolly, K. Salonitis, Investigating the energy consumption of casting
process by multiple life cycle method, in: Proceeding of 1st international
conference on sustainable design and manufacturing (SDM14) (April) (2014)
28–30.
[121] X. Dai, M. Jolly, X. Yang, J. Campbell, Modelling of liquid metal flow and oxide
film defects in filling of aluminium alloy castings, IOP Conf. Ser. Mater. Sci. Eng.
33 (2012).
[122] L. Scime, J. Beuth, Using machine learning to identify in-situ melt pool signatures
indicative of flaw formation in a laser powder bed fusion additive manufacturing
process, Addit. Manuf. 25 (2019) 151–165.
[123] MacDonald Sama, Manogharan Voigt, Measurement of metal velocity in sand
casting during mold filling, Metals 9 (10) (2019) 1079.
[124] J.M. Walker, A. Prokop, C. Lynagh, B. Vuksanovich, B. Conner, K. Rogers, J. Thiel,
E. MacDonald, Real-time process monitoring of core shifts during metal casting
with wireless sensing and 3D sand printing, Addit. Manuf. 27 (2019) 54–60.
[125] H.J. Son, S. Jang, H.J. Lee, J.J. Yang, Y.G. Jeong, C.-J. Bae, Mechanical properties
and reliability of sand casting 3D printing materials, Korean J. Mater. Res. 30 (1)
(2020) 38–43.
[126] Granta, GRANTA MI Materials Gateway, Materials information, when and where
you need it, 2020.