Journal of Food Engineering 215 (2017) 13e22

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier.com/locate/jfoodeng

3D printing complex chocolate objects: Platform design, optimization

and evaluation
Matthew Lanaro*, David P. Forrestal, Stefan Scheurer, Damien J. Slinger, Sam Liao,
Sean K. Powell, Maria A. Woodruff
Queensland University of Technology (QUT), Science and Engineering Faculty, Brisbane, QLD 4000, Australia

a r t i c l e i n f o a b s t r a c t

Article history: 3D printing of foods is an emerging technology that makes it possible to produce unique and complex
Received 24 January 2017 food items. A number of different foods have already been 3D printed, however, better characterisation of
Received in revised form fabrication parameters is still necessary to improve quality and efficiency. In this paper, we present
1 May 2017
details of the construction of a melt extrusion 3D printer based on readily available open source com-
Accepted 22 June 2017
Available online 24 June 2017
ponents. Several key fabrication parameters were investigated and optimised to enable printing complex
3D objects made from chocolate. The ability of an extruded chocolate fibre to span large distances
without collapsing was investigated by adjusting variables such as movement speeds, extrusion rates and
Keywords:
Chocolate
cooling rates. We found that the chocolate spanning distance was unaffected by movement speeds
3D printing ranging from 300 to 700 mm/min, and that the optimal extrusion rate was 10e20% leaner. Furthermore,
Digital gastronomy the spanning distance improved by directing air across the printing part, lowering the air temperature by
Additive manufacturing approximately 3.5
C. The results from this research can be applied to 3D printing of complex objects
Open source using chocolate and other similar foods.
© 2017 Published by Elsevier Ltd.

1. Introduction printing, or methods based on 2D inkjet printing principles (Godoi

Additive manufacturing and 3D printing technologies are
driving the emerging field of food printing with the promise of
highly customisable foods for applications in high end markets, and
for controlled nutrition intake. This high level of customisation has
garnered the interest of the confectionary industry, to expand
market-share with increasingly novel and commercially attractive
products (3D Systems, 2015). Furthermore, applications exist for
consumers who are required to eat highly controlled diets such as
athletes, the sick and elderly, or individuals who have trouble
chewing or swallowing food (Lipton et al., 2015; Kjeld, 2014;
Nathan, 2010). This application has been recognized by the
Netherlands Organisation for Applied Scientific Research (TNO) as a
mid-term vision, for the emerging food printing industry (Jelmer,
2014).
Similar to other 3D printing techniques, food printing builds up
a 3D object by layering a series of 2D patterns. The main 3D printing
techniques include powder bed techniques, extrusion based 3D
* Corresponding author. QUT Science and Engineering Faculty, GPO Box 2434,
Brisbane, QLD 4001, Australia.
E-mail address: m.lanaro@qut.edu.au (M. Lanaro).
et al., 2016; Sun et al., 2015a, 2015b, 2015c).
Powder bed techniques deposit a thin layer of powdered food
and then form the desired 2D pattern by selectively sintering or
bonding of the powder. Multiple 2D layers are then stacked to form
the 3D object. Sintering techniques typically use either lasers
(Nathan, 2010) or hot air guns (Evil Mad Scientist Laboratories,
2016) to selectively melt the powder. Liquid binding uses small
quantities of adhesive or a chemical curing fluid. By adding colours
to the binding liquid, complex full colour 3D objects can be formed
(3D Systems, 2016; Southerland et al., 2011). An example of this is
the ChefJet (3D Systems, 2016), a high end commercial liquid
binding printer capable of printing confectionary in high resolution
with multiple colours.
Extrusion techniques are based on fused deposition modelling
(FDM) or fused filament fabrication methods (FFF). These ap-
proaches use a translatable nozzle (or fixed nozzle/translatable
build plate) which extrudes hot or cold food onto a build plate in 2D
patterned layers, which are then stacked to build up the 3D object.
An advantage of extrusion methods over other approaches is the
wider range of available food materials which can be extruded
simultaneously to create an entire meal. Chocolate (Hao et al.,
2010), burrito (Weiss, 2012), pasta (Alec, 2015), pure ed pork

http://dx.doi.org/10.1016/j.jfoodeng.2017.06.029
0260-8774/© 2017 Published by Elsevier Ltd.
14 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22

(Kjeld, 2014), cheese (Scott, 2016) and pizza (NASA, 2013) have all Table 1
been successfully printed using extrusion based techniques. How- Ingredients: Sugar, Vegetable Fat [Emulsifiers (492, Soy Leci-
thin)], Cocoa, Milk Solids, Emulsifier (Soy Lecithin), Flavours, Salt.
ever, two main limitations exist; achieving a smooth surface finish Contains Milk and Soy. Food Standards Australia and New Zea-
(Godoi et al., 2016) and printing objects with large overhangs land (FSANZ), consumer labelling ingredient lists and percentage
(higher overhang angles result in the food collapsing). labelling guidelines for Nestle dark chocolate cooking melts.
Inkjet printing is a technique where droplets of edible binding Ingredients Quantity per 100 g
materials are dispensed onto powdered food in liquid form using
Protein 2
small pressure pulses generated by piezoelectric or thermal ele-
Total Fat 6
ments (Foodjet, 2012). However, since inkjet printing is predomi- Saturated Fat 14
nantly used for 2D graphical decorating, surface filling, or cavity Sugars 9
depositing (Foodjet, 2012), it may be more appropriate to classify Sodium 0.8
Energy (Cal) 4
this as a 2D printing technique rather than a true free form method
of creating 3D edible objects (Pallottino et al., 2016).
Chocolate is the most commonly used material used in food
printing, predominantly because of its capacity for melt extrusion received. To insure uniformity of thermal history across all tests, the
and popularity in the high end food market. Despite this, there is a chocolate was heated to 47
C in the syringe and were left for
lack of published literature investigating the practicalities and de- 20 min under occasional stirring. Thereafter temperature was set to
tails of 3D printing chocolate, with only three relevant papers the working temperature, which ranged from 31 to 36
C.
accessible at the time of writing (Malone and Lipson, 2007; Sereno
et al., 2012) and (Hao et al., 2010). Hao et al. (2010), investigated the
2.2. Mechanical platform
effect of several fundamental parameters on 3D printing capabil-
ities including nozzle to plate distance, the relative volume of
Computer numerically controlled (CNC) machine movement in
chocolate extruded per unit time and movement speed. They found
X, Y and Z directions was achieved with the ORD Bot Hadron me-
that previous theoretical models for optimising the nozzle-to-plate
chanical platform shown in Fig. 1. This system is a commercially
distance were inconsistent with experimental findings. As the
available open source design derived from the RepRap project
result of this research, they went on to create the first commercially
(Jones et al., 2011) and describes the configuration of motors, belts,
available chocolate printerdthe Choc Edgedwhich can print a
linear guidance, extruder mounting plate and build plate mounting
range of 2D and 3D shapes. Other groups have also constructed
platform. A borosilicate glass build plate (200 mm  210 mm) was
chocolate printers such as the 3Drag (Open Electronics, Online),
fixed to the mounting plate with custom designed spring loaded,
QiaoKe (BEcoda, Beijing, China) and CocoJet (3D Systems, Rock Hill,
USA).
Chocolate, in spite of being the most widely used material in
many early 3D food printers, remains a challenging material to
work with. This is largely attributed to its varied composition
involving a mixture of cocoa, milk butter and of a number of
different triglycerides (Beckett, 2008). These triglycerides can as-
sume six different forms (I to VI) of varying thermodynamic sta-
bility, where the tightly packed and very stable form V or b2 is the
most desired in the food industry and is known as tempered
chocolate. This form exhibits superior physical properties such as
hardness and melting point, as well as more appealing visual aes-
thetics including colour and gloss. This form also increases the shelf
life, taste and structural integrity of printed or cast chocolate
(Beckett, 2008). The shear thinning behaviour of chocolate is
considered a favourable property for melt extrusion 3D printing
due to the tendency to enhance flow through the extrusion process
under pressure, and hold its shape when deposited on the print
surface (Afoakwa et al., 2007; Turner et al., 2014).
In this paper, we present the design of a low-cost chocolate
printing machine based on inexpensive, readily available and open
sourced components. To optimise the printing process, we evalu-
ated the effect of varying several printing parameters including
movement speed, extrusion rate and cooling rate on the printing
process.

2. Material and methods

2.1. Chocolate

Dark cooking chocolate was used throughout the experiments

(Melts, Nestle, Australia). The compulsory consumer labelling (Food
Standards Australia and New Zealand, FSANZ) information is pro-
vided below in Table 1. Furthermore, the chocolate was character-
ized with both rheological and thermal analyses. No changes were
made to the composition of the chocolate which was used as
Fig. 1. Chocolate 3D printing machine (glass build plate is 200 mm  210 mm). The
arrows indicate the translations of the stage in the x,y and z directions. Also shown is
the design of the cooling mechanism which involves the use of a ducted fan and
thermoelectric device. The chocolate is melted in the melting chamber and extruded
via the extruding mechanism. All electronic components were contained in the black
acrylic box behind the printer.
 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22
height adjustable brackets. These brackets were designed in-house
and 3D printed in PLA thermoplastic to enable build plate levelling.
2.3. Electrical and software
A RUMBA integrated controller board (Reprapdiscount, Hong
Kong) was used to control all electrical components of the 3D
printer. Power was supplied to the board from a 200e240 V AC
(input voltage) to 24 V DC (output voltage) 360 W switching mode
power supply.
Repetier-Firmware open source software (V0.92, Willich, Ger-
many) was used to configure, compile and upload firmware to the
RUBMA control board. Repetier-Host free software (V1.6.1, Willich,
Germany) was used to operate the printer from an external PC
connected to the RUMBA board via USB.
2.4. Extruder and cooling system
A custom syringe extrusion system was designed and fitted to
the ORD Bot Hadron mechanical platform for the controlled de-
livery of chocolate. It consisted of a NEMA 17 (0.9”) stepper motor
(WanTai Motor Co, China) fixed by way of flexible coupling to a
trapezoidal threaded screw (10 mm nominal diameter, 2 mm
pitch). A brass nut housed within a PLA 3D printed carriage was
engaged with this screw. The carriage was restricted to only move
in the vertical direction using four LM8UU linear bearings running
on two stainless steel shafts (8 mm diameter). A 30 mm diameter
syringe was used to handle and extrude the chocolate, with two
additional 3D printed PLA components designed to lock the plunger
into the carriage. The body of the syringe was loaded into a
stainless-steel tubular jacket, located within the main PLA 3D
printed housing. This housing also held the bearing shafts and
motor, and was secured onto the mounting plate of the ORD Bot
mechanical platform. Because of this arrangement, the motor
output shaft rotated the trapezoidal screw which drives the nut and
carriage vertically, causing the syringe plunger to move vertically.
Movement of the syringe plunger downwards resulted in the
extrusion of chocolate from a nozzle fixed to the outlet of the sy-
ringe. Fig. 1 shows the complete 3D printer with included elec-
tronics enclosure, extruder and cooling systems.
To heat the syringe, two 24 V 40 W ceramic cartridge heating
elements were placed within two 6 mm diameter holes machined
into the custom stainless steel jacket. It should be noted that other
options for heating could be considered, including a band heater,
water bath or aluminium heater. A 100 k NTC thermistor temper-
ature sensor (EPCOS, Germany) was included in the base of the
stainless-steel jacket to allow temperature measurement and
Table 2
Design parameters for the chocolate extruder mechanism shown in Fig. 2.
Parameter Symbol
Lead Screw Pitch p 2
Number of Thread Starts n 1
Plunger Linear Travel Per Rotation l p.n
Motor Steps s 400
Driver Micro Steps sm
Syringe Internal Diameter Øs
Nozzle Diameter Øn
Lead Screw Travel Per Step
Lead Screw Travel per Micro Step t l/(s.sm)
Plunger Cross Section Area Ap
Extruded Volume Per mm Plunger Travel Ve
Extruded Volume per Micro Step
Nozzle Cross Section Area An
Extruded Material Length per mm Plunger Travel
Length of Extruded Material per Micro Step
15
proportional-integral-derivative (PID) controlled extrusion tem-
perature using the inbuilt capability of the Repetier-Firmware and
Repetier-Host software. A number of calculations were performed
to inform the design and specification process for key elements of
the extruder design; these calculations are shown in Table 2. The
elements of the extruder system are shown in Fig. 2.
Quenching of the extruded chocolate was achieved using a
custom designed cooling system. The system consisted of a ther-
moelectric heat transfer Peltier module (Wellen Tech, TEC1-12706)
assembled between two aluminium heat sinks. Each heat sink was
located within its own enclosed chamber formed inside a custom-
made 3D printed housing. The chamber holding the heat sink
connected to the hot side of the Peltier module included an axial
fan spec (ARX, CeraDyna, FD1260-A1151G2AL) at one end and an
opening to atmosphere at the other end; this provided cooling to
prevent the hot side from overheating. On the cold side of the
Peltier module a DC ceramic impeller fan capable of outputting
17 m3/h at 12 V (ARX, CeraDyna, FW1275-A1041C2AL) was used to
direct air through the heat sink and into an insulated duct on the
opposite side of the chamber. This duct made from multi-purpose
flexible hose, (Kinetic) with pipe insulation (Elasto Kflex) was
connected to the extruder assembly housing where the air was
directed to blow out of a ring-shaped outlet surrounding the sy-
ringe nozzle. In this way, fan forced cooling could be applied
directly to the extruded chocolate during printing. Fig. 2 shows a
schematic of the cooling device.
3. Results and discussion
3.1. Material analysis and characterization
To determine the melting point of chocolate, differential scan-
ning calorimetry (DSC) on the DSC Q100DSC (TA instruments) was
conducted. A sample of chocolate (8.1 g) was loaded into the
sample holder and evaluated across a temperature range of
0e60
C at 1.5
C/s in a nitrogen atmosphere (Fig. 3). The melting
point of chocolate occurred between ~29 and 31
C while the so-
lidification point occurred at 23.5
C. The chocolate was thermally
stable across the tested range.
Rheological data was gathered with an M302 Rheometer (Anton
Paar). Two tests were carried out according to both the Interna-
tional Office of Cocoa, Chocolate and Sugar Confectionery (IOCCC)
and the International Chocolate Awards (ICA) guidelines (Figs. 4
and 5). We modified our methodology from both guidelines by
using a flat plate instead of the required cylinder, which was not
available. Yield point (16 and 14 Pa s near 0 s1) and viscosity (2.35
and 2.37 Pa s at 50 s1) for both standards are very similar, as
Equation Value Unit
Mm
Starts
2 Mm
steps/rev
16 m_steps/step
30 Mm
1.37 Mm
l/s 0.005 Mm
0.00031 Mm
p.(Øs/2)2 707 mm2
1.Ap 707 mm3/mm
Ve.t 0.221 mm3
p.(Øn/2)2 1.47 mm2
Ve/An 479.5 mm/mm
(Ve.t)/An 0.15 mm/m_step
16 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22

Fig. 2. (A) Printing head, highlighted slot where thermocouple is placed. (B) Exploded view of chocolate extruder mechanism showing syringe, housing for stainless steel heating
jacket, bearings, traveling nut, coupling, motor and screw-thread plunger depressor. (C) Air cooler design consisting of a Peltier module situated between two heat sinks. Air is
pushed over the heatsinks, which is then directed to the chocolate extruder through an insulated duct. The hot side of the Peltier module is cooled with a heat sink and fan, with the
exhaust air vented to atmosphere. (D) Cut view of printing head, insulated duct connects to housing where a channel directs air down onto printing filament, also seen is location of
heating cartridges.
 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22 17

Fig. 3. Differential scanning calorimetry curve for chocolate samples. Sample was evaluated across 0 to 60
C at 1.5
C/s. Solidification peak (A), 23.5
C; first melting peak (B), 29
C;
second melting peak (C) 31
C.

Fig. 4. Viscosity data obtained from International Office of Cocoa, Chocolate and Sugar Confectionery (IOCCC, 2000) and the International Chocolate Awards (ICA) testing guidelines.
*Modified to use flat instead of cylindrical testing piece.

measured by the IOCCC 2000 and ICA standards respectively
(Table 3).
Viscosity data was also gathered for the chocolate at tempera-
tures of 30, 32, 34, 36 and 38
C, using the modified IOCCC meth-
odology (Fig. 6). Yield point and viscosity data is consistent across
samples 32 to 40
C with a yield point at low shear rate of 16 Pa s
and operating shear rate viscosity of 2.35 Pa s. The DSC analysis
showed (Fig. 3) solidification should occur between 23.5 and 31
C
which was confirmed by the rheology testing, which showed
significantly higher viscosity for chocolate at 30
C. Fig. 6 shows
that chocolate is a shear thinning material with shear rates below
6 s1 at temperatures from 32 to 40
C. Therefore, to print in the
18 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22

Fig. 5. Shear stress data obtained from International Office of Cocoa, Chocolate and Sugar Confectionery (IOCCC, 2000) and the International Chocolate Awards (ICA) testing
guidelines. *Modified to use flat instead of cylindrical testing piece.

Table 3
Yield point at almost rest was measured using the windhab model according to IOCCC 2000 guidelines.

Measuring standards Yield point at almost rest Viscosity at operating shear rate (50 s-1)

IOCCC 2000 16 Pa s 2.35 Pa s

ICA 14 Pa s 2.37 Pa s

Fig. 6. Viscosity data for chocolate samples at temperatures of 30, 32, 34,36,38 and 40
C. Methodology was modified from the IOCCC 2000 standards to use a flat plate across
various temperatures.
 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22
newtonian region of chocolate is important to insure that the
temperature is keep above the solidification point of ~30
C and
above a shear rate of 10 s1. To calculate the shear rate of Newto-
nian liquid through a pipe, which is analogous to a needle, Equation
(1) can be used. Under normal operating conditions the shear rate is
50 s1 (inner diameter of 1.37 mm and velocity of 500 mm/min)
which is within the Newtonian region of chocolate.
Shear Rateðdotted gammaÞ ¼ 8*v=d;
where v is velocity through a pipe and d is diameter of the pipe for a
Newtonian fluid.
3.2. Evaluation of chocolate printer
To evaluate whether the 3D printer could extrude chocolate
accurately, line tests were conducted where the calculated amount
of extruded filament could be compared to the actual amount of
extruded filament. The amount of extrusion was made to equal the
inner diameter of the needle using Equation (2). To conduct the line
tests, the printer was directed to deposit chocolate in a back-and-
fourth motion where the diameter of the fibre could be
measured. To measure the diameter of the chocolate the image
analysis program ImageJ was used. As the chocolate was extruded,
the expected diameter could be compared to the actual diameter of
chocolate extruded (Fig. 6). The filament started to flatten out at a
tip to collector distance between 1.2 and 1.4 mm, which correlates
with the needle diameter of 1.37 mm. This result confirmed that by
Fig. 7. The line test was conducted (A) where the diameters of the filament were measured using ImageJ. (B) A simulation of the test is shown. (C) The effect of varying tip to
collector distance when compared to the fibre width. (D) To evaluate the effect of leaner or heavier extrusion, the calculated flow (Equation (2)) was multiplied by a scaler value and
compared to fibre width.
Table 4
Experimental conditions.
Chocolate temperature (
C) Movement speed (mm/min)
31e32 300e700
33e34 300e700
35e36 300e700
19
applying Equation (2), we were able to extrude an amount of
chocolate that equalled the diameter of the needle from which it
was extruded. Next, we tested whether changing the amount of
extrusion would have a concomitant change in the rate of chocolate
extruded. Equation (2) was again used to determine the amount of
chocolate to extrude, which was then multiplied by a scaler value
between 0.6 and 1.4, where an extrusion multiplier of 0.6 is anal-
ogous to a lean chocolate flow (40% less chocolate solids) and a
(1) multiplier of 1.4 is analogous to a heavy chocolate flow (40% more
chocolate solids). The results in Fig. 7 show an expected linear
relationship between the amount of extrusion and the fibre
diameter. Both tests indicate that our melt extrusion printing head
and software produced accurate and repeatable results.

.

Dp ¼ ðØn =2Þ2 Pd ðØs=2Þ2 P ; (2)
where Dp is the syringe plunger displacement, Øn is the nozzle
diameter, d is the displacement of the travel move and Øs is the
syringe internal diameter.
While we showed that our printer could deposit accurate and
repeatable amounts of chocolate, there was a delay when begin-
ning and ending chocolate flow. Due to the high yield point of
chocolate, there was a high degree of stress on the flexible coupling
joining the stepper motor (Fig. 2) to the threaded screw. This dy-
namic led to a lag between the beginning and end of the chocolate
flow. We were able to mitigate these effects by first extruding a
small amount of chocolate before any prints were started.
Extrusion multiplier Quenching
Fan speed Temperature (
C)
0.6e1.4 20e100% 19.9e23.7
0.6e1.4 20e100% 19.9e23.7
0.6e1.4 20e100% 19.9e23.7
20 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22

3.3. Optimization of printing parameters
Several printing parameters were investigated to optimize the
ability to print complex 3D objects including chocolate tempera-
ture, movement speed, extrusion rate and cooling rate as shown in
Table 4. The method we selected to evaluate the effectiveness of
printing parameters on the ability to form complex objects is called
bridging, which describes the maximum distance a strand of
unsupported filament can span without breaking. The printing
parameters which were left unchanged throughout all tests were
the tip to collector distance and nozzle diameter (1.37 mm). Unless
otherwise specified, the movement speed was 500 mm/min and
printing was done at room temperature with no forced air flow. The
test parameters are shown in Fig. 8 and physical results are seen in
Fig. 9.
The cooling rate or quenching of chocolate is of particular

Fig. 8. (A) A simulation of a simple bridging experiment where maximum bridging distance is 8 mm. (B) This test was scaled up using a custom 3D printed testing rig. (C) The effect
of varying extrusion multiplier from 0.6 to 1.4. (D) The effect of movement speed was varied over 300e700 mm/min (EeG) The effect of increasing the air flow from 20 to 100%
using both room temperature 23.7
C and chilled air 19.9
C. Note that all tests were conducted for chocolate at temperatures of 31e32
C, 33 to 34
C and 35 to 36
C.
 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22
Fig. 9. 3D Printed Chocolate. (A) No air cooling used, simple 2D structures were made (height is 9 mm). (B) More complex structures were attempted, such as a chocolate bunny
without cooling (height is 30 mm and 42 mm). (C) Using cooling, a complex three-dimensional bunny was made (height is 94 mm).
importance to avoid the melting and collapse of chocolate struc-
tures. The extrusion head features a channel (Fig. 2) which can
direct air flow down onto the workspace where a fan can vary the
airflow speed from 20 to 100%, thus lowering the air temperature
from 23.7
C to 19.9
C. A fan speed of 100% refers to a power level
of 100% which is rated to move 17 m3 of air per hour (ARX, Cera-
Dyna, FW1275-A1041C2AL).
Movement speed of the chocolate deposition appeared to have
no effect on the ability of the extruded chocolate fibres to span
distances within the range of speeds tested (between 300 and
700 mm/min). Note that the same amount of chocolate was
extruded for each test, only the speed was varied.
When testing the extrusion multiplier against bridging distance,
three observations were made. First, the presence of a local
maximum in bridging length at an extrusion multiplier of 0.8
indicated that the ideal extrusion multiplier for bridging was be-
tween 0.8 and 0.9. This may be attributed to a slightly leaner
extrusion, where the mass of the fibre was minimized while
maintaining sufficient structural integrity. Second, at the lowest
extrusion multiplier setting we observed that the chocolate did not
always extrude in a consistent strand, which resulted in very large
error bars. This is likely due to the surface cooling and tension of the
chocolate causing the material to lump and glob. Finally, the ability
for chocolate to bridge was unchanged at heavier extrusions, which
was not expected since a higher mass of chocolate would theo-
retically cause the fibre to collapse at smaller spanning distances.
We also investigated the effect on bridging capability due to the
application of forced ambient and actively cooled air onto the
nozzle during extrusion. The results shown in Fig. 8 suggest that the
air flow volume onto the extruder (controlled by varying the fan
speed) did not appreciably alter the ability of the chocolate to
bridge. For all three chocolate temperatures, forced with ambient
and actively cooled air, we found no discernible trend across fan
speeds of 20e100%. However, we did observe that an air temper-
ature decrease around the nozzle from 23.7
C to 19.9
C had a large
impact on bridging, with an increase in spanning distance of
around 1e2 mm consistent across all tests. We also observed that
by applying a cooled airflow directly towards the extrusion nozzle,
a greater variation in the maximum bridging distance was possible,
even greater than when varying only the movement speed. This is
likely due to the turbulent airflow adding variability to the system
21
and altering the solidification of the chocolate.
4. Conclusion
As an emerging field within additive manufacturing, food
printing will continue to gain momentum as it evolves from the
niche high-end novelty market to a practical method used for in-
dividuals with specific food requirements, such as people with
chewing difficulties or those requiring nutritionally precise meals.
In this paper, we presented a design of a melt extrusion 3D printer
of chocolate, with optimized printing parameters based on the
evaluation of printing performance. Using this design, we suc-
cessfully printed complex 3D objects such as a chocolate bunny
consisting of steep build angles and overhanging parts (Fig. 8). We
identified two important areas for optimising the design: (1)
designing the extruder assembly to be as rigid as possible, thereby
reducing flexion and enabling a more accurate deposition of
chocolate, and (2) improving the design of the active cooling sys-
tem to quench the chocolate at lower temperatures. Cooling the
chocolate during extrusion was investigated to find the optimal
cooling parameters that enabled the fabrication of self-supporting
layers, critical for producing complex 3D geometries. We found
that quenching the chocolate with cooler air significantly increased
its ability to form self-supporting layersda temperature difference
of 3.8
C resulted in an increased bridging span by up to 2 mm.
While movement speed was shown to have a negligible effect on
forming self-supporting layers, the extruded chocolate volume to
translation speed ratio resulted in superior performance at 0.8 to
0.9.
Future improvements to the system involve the introduction of
a screw thread extrusion process or a pressure driven pump to print
the material from a remotely situated holding tank, overcoming
capacity limitations of the system. Further testing of chocolate
could also be undertaken, such as assessing how much cooling is
necessary to resist chocolate melting and investigating the
maximum unsupported build angle.
Funding
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
22 M. Lanaro et al. / Journal of Food Engineering 215 (2017) 13e22

References the food industry. Trends Food Sci. Technol. 43 (1), 114e123.
Malone, E., Lipson, H., 2007. Fab@Home: the personal desktop fabricator kit. Rapid
Prototyp. J. 13 (4), 245e255.
3D Systems, 2015, January 15. 3D systems previews New chocolate 3D printer
Nathan, G., 2010. Looking to the future: creating novel foods using 3D printing.
CocoJetTM at 2015 international CES. Retrieved from http://www.3dsystems.
Retrieved October 20, 2016 from http://www.foodnavigator.com/Science/
com/de/node/7563. .
Looking-to-the-future-Creating-novel-foods-using-3D-printing. .
3D Systems, 2016. Bringing the power of 3D printing to culinary. Retrieved from 3D
National Aeronautics and Space Administraion, 2013. 3D printing: food in space.
Systems, Inc. website http://www.3dsystems.com/culinary. .
Retrieved November 22, 2016 from https://www.nasa.gov/directorates/
Afoakwa, E.O., Paterson, A., Fowler, M., 2007. Factors influencing rheological and
spacetech/home/feature_3d_food.html. .
textural qualities in chocolateea review. Trends Food Sci. Technol. 18 (6),
Pallottino, F., Hakola, L., Costa, C., Antonucci, F., Figorilli, S., Seisto, A., Menesatti, P.,
290e298.
2016. Printing on food or food printing: a review. Food Bioprocess Technol. 9
Alec, B., 2015, May 5. Pasta maker Barilla to show off its 3D pasta printer at the
(5), 725e733.
Milan EXPO 2015. Retrieved October 25, 2016 from http://www.3ders.org/
Scott, C., 2016. Would you eat 3D printed Cheese? Dutch researchers hope so.
articles/20150505-pasta-maker-barilla-to-show-off-its-3d-pasta-printer-at-
Retrieved October 10, 2016 from https://3dprint.com/130500/3d-printed-
the-milan-expo-2015.html. .
dairy/.
Beckett, S.T., 2008. The Science of Chocolate, “The Science of Chocolate”, second ed.
Sereno, L., Vallicrosa, G., Delgado, J., Ciurana, J., 2012. A New application for food
Cambridgeshire: The Royal Society of Chemistry, Cambridge.
customization with additive manufacturing technologies. In: AIP Conference
Evil Mad Scientist Laboratories, 2016. The CandyFab. Retrieved October 7, 2016 from
Proceedings. Cadiz, Spain.
http://candyfab.org/. .
Southerland, D., Walters, P., & Huson, D. (2011). Edible 3D Printing. In Digital
Foodjet, 2012. Foodjet printing systems. Retrieved from Foodjet, website http://
Fabrication 2011 Conference, NIP 27, 27th International Conference on Digital
foodjet.nl/. .
Printing Technologies. Minnesota, USA.
Godoi, F.C., Prakash, S., Bhandari, B.R., 2016. 3d printing technologies applied for
Sun, J., Peng, Z., Yan, L., Fuh, J., Hong, G.S., 2015a. 3D food printingdan innovative
food design: status and prospects. J. Food Eng. 179 (2016), 44e54.
way of mass customization in food fabrication. Int. J. Bioprinting 1 (1), 27e38.
Hao, L., Mellor, S., Seaman, O., Henderson, J., Sewell, N., Sloan, M., 2010. Material
Sun, J., Peng, Z., Zhou, W., Fuh, J.Y.H., Hong, G.S., Chiu, A., 2015b. A review on 3D
characterisation and process development for chocolate additive layer
printing for customized food fabrication. Procedia Manuf. 1 (1), 308e319.
manufacturing. Virtual Phys. Prototyp. 5 (2), 57e64.
Sun, J., Zhou, W., Huang, D., Fuh, J.Y.H., Hong, G.S., 2015c. An overview of 3D printing
Jelmer, L., 2014. EU works on food printer to create personalized meals for the
technologies for food fabrication. Food Bioprocess Technol. 8 (8), 1605e1615.
elderly. Retrieved December 1, 2016 from http://3dprinting.com/news/
Turner, B., Strong, R., Gold, S., 2014. A review of melt extrusion additive
eufoodprinter/. .
manufacturing processes: I. Process design and modelling. Rapid Prototyp. J. 20
Jones, R., Haufe, P., Sells, E., Iravani, P., Olliver, V., Palmer, C., Bowyer, A., 2011.
(3), 192e204.
RepRapethe replicating rapid prototyper. Robotica 29 (1), 177e191.
Weiss, C.C., 2012. BurritoBot: 3D printing meets fast food. Retrieved November 11,
Kjeld, V.B., 2014. 3D food. Retrieved October 10, 2016 from http://vo-ho.nl/wp-
2016 from http://newatlas.com/burritobot-burrito-printer/23026/. .
content/uploads/2014/06/TNO-3D-food-printing-Leeuwarden.pdf. .
Lipton, J.I., Cutler, M., Nigl, F., Cohen, D., Lipson, H., 2015. Additive manufacturing for