Enhancing the Interlayer Bond in Printed

Concrete Structures

Presented by:
Xavier A. Hines-Coombs

Submitted to:
​Professor Cagdas Onal and the Worcester Polytechnic Institute Robotic Engineering
Program in partial fulfillment of the requirements for the Degree of Bachelor of Science

​2019-2020

This report represents the work of a WPI undergraduate student submitted to the faculty as evidence of a
degree requirement. WPI routinely publishes these reports on its web site without editorial or peer review.
For more information about the projects program at WPI, see

http://www.wpi.edu/Academics/Project
Abstract
3D printing with concrete is a new, innovative technology with promising benefits for the
construction industry. Additive manufacturing of concrete has the potential to save costs on
labor, formwork, and material and is a more sustainable means of producing the built
environment for the future. Although potential exists, there are many obstacles to overcome
before concrete 3D printing can be applied on a construction scale.This project first aimed to
determine an optimal mixture design for printable concrete by developing a concrete 3D-printer
and experimenting with two different viscosity modifying admixtures. After complications arose,
the project then aimed to investigate improvements to the interfacial bond between layers of
concrete alongside developing a custom nozzle design that could perform inline mixing.
Different paste extrusion techniques were investigated and an innovative nozzle design was
created inorder to perform inline-mixing. Further, testing of the design is needed inorder to
achieve results based on inline mixing.

Acknowledgements ​*refer MQP.ReportFinal.CE for additional acknowledgements*

I would like to thank Professor Glenn Gaudettte for generously donating the gantry style
printer from the Biomedical Engineering department.

Table of contents

Abstract 1

Acknowledgements *refer MQP.ReportFinal.CE for additional acknowledgements* 1

Table of contents 1

Introduction *refer MQP.Report.Final.CE* 2

Background *refer MQP.Report.Final.CE for additional background* 2

2.1. Robotics 2
2.1.1. Concrete 3D printing in practice 2
2.1.2. Cartesian systems vs. robotic arms 3
2.1.3. Industry leaders 3
2.1.4. LOSTPED requirements 8
2.1.5. Nozzle geometry for concrete printing 9
2.2. Paste Extrusion 10
2.2.1 Process 10
2.2.2.1 Ram Extruder 12
2.2.2.2. Shutter Valve 12
2.2.2.3. Auger Valve 12
Methodology 13
3.1. Gantry System *due to complications with the mortar pump, gantry was never used* 13
3.1.1 Modifications 13
3.1.2 Motor Analysis 15
3.1.2.1 X and Y Axis Motors (Horizontal) 15
3.1.2.2 Z Axis Motor (Vertical) 16
3.2. Nozzle 17
3.3. Programming *due to complications with the mortar pump, gantry was never used* 17
3.3.1 CNC Shield 17
3.3.2 Firmware for Arduino 18
3.3.3 G-Code Sender 18
3.4. Optimization - Inline Mixing *unable to complete due to COVID-19* 19
3.4.1 Materials 19
3.4.2 Assembly 20
3.5 Budget 21
3.5.1 Gantry 21
3.5.2 Custom Nozzle 22

Results *refer MQP.Report.Final.CE* 22

Conclusion 22

References 23

1. Introduction ​*refer MQP.Report.Final.CE*

2. Background ​*refer MQP.Report.Final.CE for additional background*

2.1. Robotics

2.1.1. Concrete 3D printing in practice

Concrete 3D printing is an emerging innovative practice that is revolutionizing the
construction process due to its sustainability and scalability in this area of manufacturing.
Concrete 3D printing is increasing productivity due to a shift in the industry towards a fully
digitalized construction process, combining Computer Aided Design (CAD) and BIM. This
decreases the construction time, labor, and assembly errors, while improving the overall quality
in the manufacturing process(Buswell, Leal de Silva, W. R., Jones, & Dirrenberger, 2018). The
scalability of concrete printing is also on the rise. Many companies emerging in the industry are
now able to print entire buildings as opposed to individual structural members, and have already
put this technology to use on construction scale projects.
2.1.2. Cartesian systems vs. robotic arms
3D printers can be categorized into two main types of devices: cartesian systems, also
known as gantry systems, and robotic arms. Gantry system printers are the most common style
because of their simplicity. These printers only move in linear motion along their fixed axes
which simplifies programming and increases precision. On the other hand, robotic arms are free
to move in six degrees of freedom, which allow them to perform translation on the X, Y, and Z
axes as well as rotate about the three different axes using yaw, pitch, and roll. These six
degrees of freedom enables them to execute complicated commands, but also means
programming is much more intensive (Vaugh, 2016). Gantry style robots are confined to their
given print area. This makes scalability with gantry systems somewhat challenging because the
printed object has to fit inside of the robot’s internal volume. In contrast, robotic arms can be
mobile, allowing them to move around a construction site and print much larger objects. Robotic
arms may be more suited for on-site construction, while gantry systems will be useful for
fabrication of structural components. Due to the stability of gantry style robots, the overall
strength, accuracy, and stiffness is greater than that of a robotic arm of similar size (Keating &
Oxman, 2013; Vosniakos & Matsas, 2010). Stiffness of the device is important because it
directly impacts what type of load the mechanism is able to carry. When compared to a gantry of
similar size, the lack of stiffness in robotic arms limit the materials with which it can accurately
print(Keating & Oxman, 2013; Vosniakos & Matsas, 2010). Future growth is needed to increase
the scalability of gantry systems and to enable robotic arms to work alongside laborers and
other robotic arms to complete a print(Keating & Oxman, 2013) .

2.1.3. Industry leaders

There are a variety of concrete printing companies that are part of this emerging market
including Contour Crafting (El Segundo, California), Total Kustom (Excelsior, Minnesota),
COBOD (Copenhagen, Denmark), and XTreeE (Rungis, France). All of these companies are
unique; each targeting different areas of construction. Contour Crafting, the first pioneer of
concrete extrusion technology, is a company that is working towards printing homes, multi-story
buildings, and large multi-unit structures for low-income populations and emergency shelters
using a gantry style concrete printer(Contour Crafting Corporation, 2017). The company is even
looking at integrating insulation, wiring, and other components into the printing process. Their
nozzle design includes a top and side trowel, so that as material is extruded, the traversal of the
trowels creates a smooth finished surface (Khoshnevis, 2004).
 Figures 1 & 2: One of Contour Crafting gantry style machines print a wall (Contour Crafting
Corporation, 2017)

Total Kustom is a company that manufactures various types of 3D printers including

multiple types of concrete printers. Total Kustom’s StoryBot 6.2 is a concrete printer that is
specifically designed for printing homes, while their Architect’s Printer is used to print 3D model
homes; both of these robots use a gantry system (Total Kustom, n.d.). Total Kustom partnered
with the Philippines’ Lewis Grand Hotel to print a hotel suite that measured 10.5 meters by 12.5
meters with a height of four meters. The print, including installation of plumbing and electricity,
took about a week to complete and consisted of two bedrooms, a living room, and a jacuzzi
room.

Figure 3: The intricate layout of the 3d printed hotel (Total Kustom, )

 Figure 4: The plumbing that is being added (Total Kustom, )

Figure 5: The printed hotel after it’s been painted (Total Kustom, )

Another leader in the industry, COBOD created a gantry style concrete printer known as
the Building on Demand (BOD). The company further innovated this design to offer a modular
setup with improved stability, as well as other key features in its successor, the BOD2. A
modular system is capable of being subdivided into independent components that can be added
or subtracted from one another. The modular design of the BOD2 allow the printer to be
adjusted to the projects size requirements overcoming the scalability issues with gantry printers.
This modular design also makes it possible to easily transport the gantry system to different job
sites(COBOD International, 2019). COBOD has printed a 50 square meter building in Nordhavn,
Copenhagen.
 Figure 6: Above on the left, shows COBOD machine printing the walls.
Figure 7: Above on the right is an up close picture of COBOD nozzle extruding cement (COBOD
International, 2019).

Figure 8: Above on the left, shows printed walls of the building.

Figure 9: Above on the right, shows the finished product (COBOD International, 2019).

Unlike the other companies previously stated, XTreeE uses a robotic arm to create
complex concrete structures that would be nearly impossible to create without additive
manufacturing(XTreeE, n.d.). For example, XTree 3D printed a concrete artificial coral reef that
is now being used to restore the lost ecological habitat in Calanques National Park (France).
The intricate design of the coral reef, reiterates the ability of robotic arms to create complex
structures using its six degrees of freedom.
 Figure 10: The robotic arm that XtreeE used to print an artificial coral reef (XTreeE, )

Figure 11: The artificial coral reef printed by XtreeE, submerged in the ocean near Calanques
National Park (France) (XTreeE, )

The example structures from these companies indicate that the gantry style system is
more suitable for printing simple structural components and small scale buildings, while robotic
arms, because of their freedom of movement, are more useful for creating intricate architectural
designs.
2.1.4. LOSTPED requirements
There are many varying factors when deciding what type of 3D printer to use because
concrete printers can differ in size, shape, and complexity depending on the purpose of the
device. When designing any robotic mechanism, it is important to know the required LOSTPED
parameters. LOSTPED refers to the load, orientation, speed, travel, precision, environment, and
duty cycle of the robot. Load is the amount of weight that the robot will be required to support
(Joseph Macura, Deng, & Allen Powell, 2018). Knowing the load is important for the overall
construction of the device. The load dictates what material the device needs to be made out of
as well as how much power it will need. For the purpose of concrete printing, a more reinforced
printer design is required because of the large unit weight of concrete in comparison to other
printing materials such as polymers. Orientation is the motion of the end-effector in 3D space,
which can be defined by the 3 axes, X,Y, and Z. In concrete 3D printing, the end-effector is the
extrusion nozzle, and knowing the required orientation will dictate whether the nozzle needs to
rotate or remain stationery (Joseph Macura et al., 2018). The conventional layering toolpath,
shown in Fig. 12a, is planar layering. Planar layering does not require the nozzle to change
orientation, whereas multi-plane layering requires the nozzle to rotate (Fig. 12b) (Bin Ishak,
Fisher, & Larochelle, n.d.).

Figure 12: Depicts the two different ways of printing a desired object. 12a shows the
process of printing by using planar layering, 12b shows the process of printing with multi-plane
layering (Bin Ishak et al., ).

Speed refers to the required acceleration and deceleration of the nozzle movement and
gives insight into what type of motor is needed (Joseph Macura et al., 2018). In the concrete
printing process the robot usually travels anywhere between 30-60 mm/s (1.18-2.36 in/s), which
is relatively slow compared to a regular 3D printer that can move up to three times that speed
(Buswell et al., 2018). Travel is how far the nozzle is required to move. This component dictates
the overall dimensions of the concrete printer. Precision is how accurately the robot is able to
execute given commands. This is directly affected by the load, orientation, speed and travel of
the robot (Joseph Macura et al., 2018). Environment is the surrounding area of the robot, which
includes the immediate ambient environment as well as any hazards in the space itself (Joseph
Macura et al., 2018). In regards to 3D printing of concrete, the environment gives insight into
what type of sensors will be needed to detect any potential hazards, such as laborers on a
construction site, as well as if there will need to be restraints programmed into the robots
motion. Duty cycle refers to the operating and non-operating time and how it affects the
mechanical components (Joseph Macura et al., 2018). For instance, the lifetime of a concrete
printer depends on how long it has been in use and the conditions in which it has operated. The
primary factors that will be considered in this project are the load due to the weight of concrete,
the speed at which the material is extruded, and the environment in which the concrete will set
and cure. Travel, orientation, precision, and duty cycle are not considered for the scope of this
project.

2.1.5. Nozzle geometry for concrete printing

A crucial mechanical component of a concrete printer is the nozzle. The design of the
nozzle directly affects the geometry and properties of the printed filament. The placement of the
nozzle, in regards to height, is also important because it affects the precision of the layering
when printing. The construction of the nozzle can be customized to have unique additional
features such as inline mixing. However, this requires that the nozzle have an internal mixing
process to evenly distribute the accelerator throughout the mix. For this reason, inline mixing is
not used in this project. Further, the nozzle will need to control the amount of accelerator added
prior to extrusion. The shape of the nozzle affects the flow rate and extrudability. Additionally,
because printable concrete is often quite viscous, pressure builds up in the nozzle during
extrusion, so it is important for the nozzle to be made of strong material in order to resist this
pressure build up.
The two shapes that are often used for nozzle designs, circular and rectangular, each
have their own benefits and drawbacks. The choice of nozzle geometry depends largely on the
distribution load area, as shown in Fig. 13b. The distribution load area is important when printing
because without it the structure will become deformed, as shown in Fig. 13a. Circular nozzles
often provide a faster and more free-form design, meaning that you can print faster and create
free-flowing shapes because the nozzle does not need to rotate to accommodate angles in a
design, but the distribution area is diminished because of the round surface (Shakor, Nejadi, &
Paul, 2019). Rectangular nozzles may take longer to print when compared to circular nozzles,
but their distribution area when layering is greater so it results in a more structurally sound
design and has a better surface finish (Shakor et al., 2019).
 Figure 13a: A planar layered print using a circle nozzle vs. a rectangle nozzle.
Figure 13b: Expresses the distribution load area of a circle nozzle vs. a rectangular
nozzle (Shakor et al., 2019)

The optimal nozzle size for projects vary, depending on the size of the fine aggregate
and fibers in the concrete mix. As a result, creating an optimal nozzle head is important for the
printing process. For this project, the optimal nozzle head is a rectangular because we are
testing printable concrete, and a rectangular nozzle allows for a more stable design because of
the greater distribution area.

2.2. Paste Extrusion

2.2.1 Process
Paste extrusion is normally carried out in three steps: paste preparation, forming, and
finishing. During preparation, powdered and liquid components are mixed into a paste and
deposited into a container. In the forming process, pressure is generated inorder to force the
paste through the nozzle. After the paste has been extruded layer by layer, the paste solidifies
during the finishing stage. The quality of a printed design depends on the dwell time, paste
property, and the extrusion mechanism (Pitayachaval & Baothong, 2018). Dwell time refers to
the amount of time that the gantry remains stationary during the start and stop of extrusion . An
excess of dwell time can lead to the accumulation of material at the start and endpoint of printed
designs. Paste has both solid and liquid properties and experiences non-Newtonian behavior
because liquid particles move much faster than solid particles during the extrusion process; As
a result, paste regions of low liquid become drier (Pitayachaval & Baothong, 2018; Rough,
Bridgwater, & Wilson, 2000). The extrusion mechanism affects the process because different
mechanisms use different methods of extrusion and as result affect the precision of the printed
piece.
 Concrete is extruded as a thick paste-like substance during the 3D-printing fabrication
process, so the same techniques that are used to print ceramic paste can be paralleled with
printed concrete. The fabrication of dense ceramic parts, by paste extrusion-based methods,
require a precise control of the dispensed material from start and stop.This precise control of
extrusion can be referred to as extrusion-on-demand (EOD) (Oakes, Kulkarni, Landers, & Leu,
2009). The EOD process for high solids loading pastes, in regards to ceramic paste extrusion,
is difficult to control due to the pastes non-Newtonian behavior, compressibility, and
inhomogeneity. Precise control of the flow-rate, in regards to ceramic paste, is important
because fluctuations in the extruded flow-rate alongside inaccurate extrusion start and stop will
lead to the generation of pores (Li, Armani, Leu, & Landers, 2017). The level of precise control
that is required to print ceramic is not necessary when printing concrete because concrete is
naturally permeable.
The paste extrusion process involves two phases, steady-state and transient.
Steady-state extrusion occurs when the filament is printed at a consistent rate. Transient
extrusion occurs when the flow-rate changes, which is most commonly during the start and stop
of extrusion(Li et al., 2017).
Paste extrusion is usually accomplished with a ram extruder, which is typically arranged
in the same manner as the end-effector of a traditional CNC style 3D-printer. The ram extruder
is used in conjunction with a syringe and a plunger. Two comparable techniques of the ram
extruder, that are used in the ceramic paste industry, are the shutter valve based extruder and
the auger extruder(Li et al., 2017; Pitayachaval & Baothong, 2018).

Figure 14. Schematic of paste extrusion-based material deposition(Li et al., 2017)

2.2.2. Methods

Figure 15. Schematic of three paste extrusion mechanisms: (a) Ram Extruder, (b) shutter
valve-based extruder, (c) Auger extruder (Li et al., 2017)

2.2.2.1 Ram Extruder

The ram extruder, as previously stated, consists of a ram-driven plunger and a syringe.
The syringe encases the material and the plunger’s movement regulates the paste flow-rate.
More specifically, the plunger’s force and velocity exerted on the material control the start and
stop of the extrusion(Pitayachaval & Baothong, 2018). The control parameters for extrusion start
and stop need to be adjusted inorder to accommodate for different paste solids and nozzle
diameters.

2.2.2.2. Shutter Valve

The shutter valve-based extrusion method consists of a plunger and syringe, similarly to
the ram extruder, but a shutter needle is added to the flowpath. The needle is located directly
above the material outlet and is either lifted up or pressed down by a controller, resulting in the
start and stop of the extrusion (Li et al., 2017; Pitayachaval & Baothong, 2018). The extrusion
flow-rate is controlled by velocity and force applied by the plunger. The addition of the shutter
needle allows for a more controlled start and stop of the flowrate and the calibrated control
parameters can remain constant with differing paste solids and nozzle diameters(Li et al., 2017).

2.2.2.3. Auger Valve

The auger based extrusion method consists of a syringe and is preloaded with pressure
by compressed air. The preloaded pressure is used to deliver the paste to the auger chamber,
rather than be used for extrusion. The extrusion and flow-rate are both controlled by the auger.
The extrusion, as well as the start and stop, are controlled by rotating the auger using
servo-motors and the flow-rate is regulated by controlling the auger's angular velocity(Jianping
& Guiling, 2004; Pitayachaval & Baothong, 2018). Similarly to the shutter-valve, the addition of
the auger valve leads to a more accurate start and stop of the flowrate and the calibrated control
parameters can remain constant with differing paste solids and nozzle diameters (Li et al.,
2017).

3. Methodology
3.1. Gantry System *due to complications with the mortar pump, gantry
was never used*
3.1.1 Modifications
As mentioned in the background, a cartesian gantry style robot was chosen based on
the simplicity of design and programming for the scope of this project. The gantry system, seen
in Fig. 16, was received from the biomedical engineering lab at Worcester Polytechnic Institute.
The system was originally used to mass produce sutures and fibers. It was then determined that
the gantry could be modified to suit the needs of this project. These modifications included
removing all the processing boards, wires, and sensors from the gantry.

Figure 16: Photo of the donated gantry system.

Next, the CTP12ELF10MAA00 stepper motors were integrated into an Arduino Mega
Board using a CNC Shield. The CNC shield was used alongside three A4988 Stepper Drive
chips because the shield allowed a single Arduino board to control three separate stepper
motors without complicated wiring (Fig. 17). All of the wiring was synced together by attaching
the CNC Shield to the arduino as shown in Fig. 18. The motors received power from the
external DC power supply which was connected to the CNC shield.
Figure 17: Photo of the A4988 chip (left) and CNC Shield (right).​

Figure 18: Photo of wiring system.

3.1.2 Motor Analysis
List of Variables:
m1 = T able M ass (kg) V max = M aximum Speed (m/s) n = N umber of reciprications (min)
m2 = W ork M ass (kg) t1 = Acceleration T ime (s) t2 = U nif orm M otion T ime (s)
t3 = Deceleration T ime (s) t4 = Stationary T ime (s) b = Backlash (mm)
ls = Stroke Length (mm) μ = F riction Coef f icient p = positioning accuracy (mm)
f = Guide Surf ace Resistance r = Reduction Gear g = gravitational pull (m/s2 )
P h = Lead (mm) η = N ormal Ef f iciency
s = M inimum f eed amount (mm/pulse)
pr = P ositioning Accuracy Repeatability
J m = Inertial moment of the motor
J s = Inertial M oment give be Screw Shaf t × length (kg × m2 )

3.1.2.1 X and Y Axis Motors (Horizontal)

Calculating the Maximum Axial Load:
acceleration : α = V tmax
1

F orward acceleration : F a1 = μ × (m1 + m2 )g + f + (m1 + m2 ) × α

F orward unif orm motion : F a2 = μ × (m1 + m2 )g + f
F orward deceleration : F a3 = μ × (m1 + m2 )g + f − (m1 + m2 ) × α
B ackward acceleration : F a4 = − μ × (m1 + m2 )g − f − (m1 + m2 ) × α
B ackward unif orm motion : F a5 = − μ × (m1 + m2 )g − f
B ackward deceleration : F a6 = − μ × (m1 + m2 )g − f + (m1 + m2 ) × α
F amax = F a1

Maximum Rotational Speed:

V max ×ls×m1
N max = Ph

Required Rotational Torque:

Fa × Ph
T orque w/ External Load : T 1 = 2π ×η × A
I nertial M oment : J = (m1 + m2 ) × ( P2πh )2 × A2 × 10−6 + J s × A2
Angular Acceleration : ω ′ = 2π60×t
× Nm
1

T orque required f or acceleration : T 2 = (J + J m )

- During acceleration : T k = T 1 + T 2
- During unif orm motion : T t = T 1
- During deceleration : T g = T 1 − T 2
- W hen Stationary : T s

√
T k 2 ×t1 +T t 2 ×t2 +T g 2 ×t3 +T s 2 ×t4
E f f ective/ Rated T orque : T rms = t1 +t2 +t3 +t4
 3.1.2.2 Z Axis Motor (Vertical)
Calculating the Maximum Axial Load:
acceleration : α = V tmax
1

U pward acceleration : F a1 = (m1 + m2 )g + f + (m1 + m2 ) × α

U pward unif orm motion : F a2 = (m1 + m2 )g + f
U pward deceleration : F a3 = (m1 + m2 )g + f − (m1 + m2 ) × α
Downward acceleration : F a4 = (m1 + m2 )g − f − (m1 + m2 ) × α
Downward unif orm motion : F a5 = (m1 + m2 )g − f
Downward deceleration : F a6 = (m1 + m2 )g − f + (m1 + m2 ) × α

F amax = F a1

Maximum Rotational Speed:

V max ×ls×m1
N max = Ph

Required Rotational Torque:

Fa × Ph
T orque w/ External Load : T 1 = 2π ×η × A
I nertial M oment : J = (m1 + m2 ) × ( P2πh )2 × A2 × 10−6 + J s × A2
Angular Acceleration : ω ′ = 2π60×t
× Nm
1

T orque required f or acceleration : T 2 = (J + J m )

- During acceleration : T k = T 1 + T 2
- During unif orm motion : T t = T 1
- During deceleration : T g = T 1 − T 2
- W hen Stationary : T s

√
T k 2 ×t1 +T t 2 ×t2 +T g 2 ×t3 +T s 2 ×t4
E f f ective/ Rated T orque : T rms = t1 +t2 +t3 +t4
3.2. Nozzle
While the gantry system was being integrated with the arduino and wires, a custom
nozzle head was designed. A rectangular nozzle head, as opposed to a circular nozzle head,
was ultimately chosen because it provided more stability and a larger distribution area. The
nozzle dimensions were four centimeters by two centimeters. Similar to the nozzle design used
by Contour Crafting, trowels, with a height equivalent to one printed layer, were added to each
side of the nozzle to provide a smooth finished surface as well as to improve cohesion between
printed layers. Another nozzle design was created with a grooved edge to promote mechanical
interlocking between layers, as well as with trowels Fig. 23.

Figure 19: Photo nozzle head with grooved extrusion head and trowels

3.3. Programming *due to complications with the mortar pump, gantry was
never used*
3.3.1 CNC Shield
After the gantry system was built with the necessary components, it was coded to run a
program that would allow the system to print CAD designs. The CNC shield’s A4988 chip
potentiometers were each individually adjusted to be about 0.4896v, as shown in Equation 11.

Eq. 11)
V ref = Rated M otor Current × 8 × Rsense
V ref = 1 × 8 × .068 = .544 → 10% saf ety f actor = .544 − (.544 × .1) = 0.4896v

Figure 20: Photo of adjusting the A4988 chips’ potentiometer using a Phillips head screwdriver
3.3.2 Firmware for Arduino
In order to use the CNC shield effectively, a control firmware, known as GRBL, was
downloaded onto the arduino. GRBL is an open-source software that enables the Arduino to
take G-code commands via serial and then turns the commands into motor signals. GRBL was
downloaded from Github, and uploaded onto the Arduino board.(GRBL-mega.)

3.3.3 G-Code Sender

In order to send the CNC Shield commands, a universal G-Code Sender was
downloaded into a computer. The G-Code Sender was downloaded from Github as previously
mentioned.(Universal-G-code-sender.) The file was opened and the port connection was set to
COM3, and the Baud rate was set to 115200 in accordance with the GRBL firmware Fig. 21.

Figure 21: Screenshot of G-Code Sender layout

After it was downloaded, “Jog Control” tab was selected and allowed the program to run
different step sizes and times to each individual Motor, Fig. 25. The step size number box
allowed the step size to be changed and the X±, Y±, and Z± , respectively, controlled how many
steps and in which direction the nozzle would move.

Figure 22: Screenshot of G-Code Sender “Jog Control” control panel

3.4. Optimization - Inline Mixing *unable to complete due to COVID-19*

3.4.1 Materials
Based on the background, the ram-extruder model was used as a base for the design in
conjunction with performing inline mixing due to its simple and easily adaptive design. The
modification to this design included the addition of a custom nozzle head. The custom nozzle
head would include motor, bevel gears, a shaft, an Arduino liquid pump, and tubing inorder to
provide inline mixing. Also two differential pressure sensors for arduino were added inorder to
ensure a constant flow-rate and provide feedback control to the system if the flow rate were to
vary due to the inline mixing of the accelerant.

Figure 23: Updated nozzle design, to allow inline mixing

Since the inclusion of inline mixing, an additional part needed to be printed and the
previous nozzle design needed to be adapted in order to support the addition of a motor and
tubing. In Fig. 23, the new nozzle design includes a housing unit, for the motor and bevel gears,
as well as a housing unit for the tube. The housing unit for the tube will be printed along with the
nozzle, while the housing unit will be printed separately. The housing unit, Fig. 25, is designed
to allow the bevel gears to be unaffected by the concrete mix. In order to add the housing unit
into the nozzle, the nozzle design is printed into parts, and then connected by a hinge, Fig. 26.

Figure 24: Split Nozzle design, allows for the addition of the housing unit
 Fig 25: Housing Unit part

3.4.2 Assembly
Assembly of the nozzle begins after the nozzle and housing unit are finished printing.
The first process involves connecting the motor, bevel gears, and shaft to the nozzle, Fig 26.
The motor is first connected to bevel gear A. The shaft is then fed through the housing unit and
connect to bevel gear B. The final step in this process involves connecting the two bevel gears
and screwing in the housing unit into the nozzle.

Figure 26: Side view of the housing unit, connected to the motor. Bevel gear A is connected to
bevel gear B. Bevel gear B is connected to the shaft.

The second process involves connecting all the sensors and tubing. The differential
pressure sensors are attached before and after the accelerant, as shown in Fig. 27. Next, the
two nozzle pieces are connected and sealed around the pvc pipe. Then the tubing is ran
through the tubing unit and sealed, Fig. 27. The tubing is connected to the Arduino liquid pump.
Finally, the sensors and the pump are connected to the Arduino Mega board.

Figure 27: Complete Nozzle Design

3.5 Budget

3.5.1 Gantry

Item Price/Unit Quantity Total

Materials

Mega Arduino Board $38.50 1 $38.50

CNC Shield Board $17.55 1 $17.55

Bipolar Stepper Cables $8.99 1 $8.99

Controlled DC Power Supply $50.95 1 $50.95

A4988 Stepper Drive $7.45 3 $7.45

Wiring 13.08 1 $13.08

Total $136.52
3.5.2 Custom Nozzle

Item Price/Unit Quantity Total

Materials

Motor $50.00 1 $50.00

Bevel Gears $51.00 2 $102.00

Shaft $8.00 1 $8.00

Arduino Liquid Pump $25.00 1 $25.00

Tubing $6.00 1 $6.00

PVC Pipe $7.00 1 $7.00

Differential Pressure Sensor $60.00 2 $120.00

Total $318.00

4. Results ​*refer MQP.Report.Final.CE*

5. Conclusion
The first objective of the research discussed was to develop a 3D-Printer inorder to test
printable concrete. The 3D-Printer was created but could not be tested alongside concrete,
because the mortar pumps that were ordered could not pump the concrete. The next objective
of this project was to test the improvement of interlaying bonding by performing inline mixing. An
innovative nozzle was designed, but could not be built or tested due to the COVID-19 pandemic.
References
Bin Ishak, I., Fisher, J., & Larochelle, P. (n.d.). ​Robot arm platform for additive manufacturing

using multi-plane toolpaths​ American Society of Mechanical Engineers.

doi:10.1115/DETC2016-59438

Buswell, R. A., Leal de Silva, W. R., Jones, S. Z., & Dirrenberger, J. (2018). ​3D printing using

concrete extrusion: A roadmap for research​ doi://doi.org/10.1016/j.cemconres.2018.05.006

COBOD International. (2019). Modular 3D construction printers - 3D printed buildings. Retrieved

from https://cobod.com/

Contour Crafting Corporation. (2017). Building construction. Retrieved from

http://contourcrafting.com/building-construction/

Jianping, L., & Guiling, D. (2004). Technology development and basic theory study of fluid

dispensing-a review. Paper presented at the ​Proceedings of the Sixth IEEE CPMT

Conference on High Density Microsystem Design and Packaging and Component Failure

Analysis (HDP'04), ​198-205.

Joseph Macura, M., Deng, R., & Allen Powell, C. (2018). In The Procter & Gamble Company

(Ed.), ​Robotic unwind stand

Keating, S., & Oxman, N. (2013). ​Compound fabrication: A multi-functional robotic platform for

digital design and fabrication​ doi://doi.org/10.1016/j.rcim.2013.05.001

Khoshnevis, B. (2004). ​Automated construction by contour crafting—related robotics and

information technologies​ doi://doi.org/10.1016/j.autcon.2003.08.012

Li, W., Armani, A., Leu, M., & Landers, R. (2017). Extrusion-on-demand methods for high solids

loading ceramic paste in freeform extrusion fabrication.​ Virtual and Physical Prototyping, ​,

1-13. doi:10.1080/17452759.2017.1312735
Oakes, T., Kulkarni, P., Landers, R. G., & Leu, M. C. (2009). Development of

extrusion-on-demand for ceramic freeze-form extrusion fabrication. Paper presented at the

Proceedings of Solid Freeform Fabrication Symposium, Laboratory for Freeform Fabrication,

Austin, TX, ​206-218.

Pitayachaval, P., & Baothong, T. (2018). A review of process parameters that effect to extrusion

on demand. Paper presented at the ​MATEC Web of Conferences, , 213​ 01013.

Rough, S. L., Bridgwater, J., & Wilson, D. I. (2000). Effects of liquid phase migration on

extrusion of microcrystalline cellulose pastes.​ International Journal of Pharmaceutics,

204​(1-2), 117-126.

Shakor, P., Nejadi, S., & Paul, G. (2019). A study into the effect of different nozzles shapes and

fibre-reinforcement in 3D printed mortar.​ Materials (Basel, Switzerland), 12​(10), 1708.

doi:10.3390/ma12101708

Total Kustom. (n.d.). 3D concrete printers. Retrieved from

http://www.totalkustom.com/3d-concrete-printers.html

Vaugh, R. (2016). The difference between cartesian, six-axis, and SCARA robots. Retrieved

from

https://www.machinedesign.com/motion-control/difference-between-cartesian-six-axis-and-sca

ra-robots

Vosniakos, G., & Matsas, E. (2010). ​Improving feasibility of robotic milling through robot

placement optimisation​ doi://doi.org/10.1016/j.rcim.2010.04.001

XTreeE. (n.d.). The large-scale 3D. Retrieved from https://www.xtreee.eu/news/