INDEX

1 Introduction……………………………………………………………………………………..2

2 Orientation………………………………………………………………………………………3

3 Concord CNC wire EDM working and SOP…………………………………………….3

4 Process of 3D printing a part……………………………………………………………….9

5 Digital Twin and Additive Manufacturing…………………………………………..…14

6 Robot Studio………………………………………………………………………………..…17

7 Additive manufacturing using WAAM (Robot arm) ………………………………..22

8 Advance-Hyrel 3D printer working and SOP…………………………………………34

9 Literature studies

9.1 Methodology for Experimental Investigation of Continuous Carbon Fibre…48

9.2 Structural Performance of Additively Manufactured Composite Lattice
Structures: Strain Rate, Cell Geometry, and Weight Ratio E ects……………57
9.3 Introduction to Nozzle Clogging in Fused Deposition Modelling (FDM)……..60

10 Markforged composite 3D printer working and SOP……………………………….70

11 Attempted designing of Scissor Lift…………………………………………………….74

12 Digital twin of a printer for material interaction study framework…………….82

13 Conclusion…………………………………………………………………………………….93

14 References……………………………………………………………………………………..94

1
Introduction
This internship was carried out over a period of 45 days at the Additive Manufacturing
and Digital Manufacturing Laboratory. The primary objective of the internship was to
gain practical exposure to modern manufacturing technologies, industrial equipment,
and current research areas in additive manufacturing.

During the internship, hands-on training and technical observations were conducted on
various manufacturing systems including CNC Wire EDM, polymer and composite 3D
printers, robotic Wire Arc Additive Manufacturing (WAAM), and ABB RobotStudio.
Standard operating procedures, machine setup, programming, and workflow of these
systems were studied to understand their industrial applications.

In addition to equipment operation, literature studies were carried out on continuous

carbon fibre reinforced composites, lattice structures, and nozzle clogging in fused
deposition modelling. Practical design work included the modelling and analysis of a
scissor lift mechanism, followed by the development of a conceptual digital twin
framework for studying layer-to-layer thermal interactions in the FDM process.

This report summarizes the knowledge acquired, practical activities performed,

challenges encountered, and the technical understanding developed throughout the
internship.

2
Orientation
Was informed about preparing daily report, daily ppt and about keeping a presentation
twice weekly (on Monday and Thursday). Was informed about the rules of the lab, about
whatever you publish online should first be taken permission from the lab, since
everything is lab property.

Concord CNC wire EDM

Concord CNC Wire EDM (Electrical Discharge Machining) is a high-precision industrial

machine tool used to cut and shape hard, conductive metals.

It acts like an incredibly precise electronic jigsaw. Instead of using a physical blade or
friction, it uses a thin, rapidly moving wire to create a continuous path of electrical
sparks. These sparks melt and erode microscopic amounts of metal to carve out
intricate shapes. And if molybdenum wire is being used instead of brass, it significantly
lowers cost.

Because it cuts metal via electricity rather than mechanical force, it is designed for
businesses that require extreme precision without distorting the material.

 Tool & Die Makers: For carving out complex injection Molds, extrusion dies, and
stamping tools.

 Aerospace & Defence: For high-strength, heat-resistant alloys (like Inconel or

Titanium) that dull traditional drill bits.

 Medical Equipment Manufacturers: For creating tiny, intricate components for

surgical instruments.

 Automotive Industry: For precision engine and transmission prototyping

components.

3
Concord machines are specifically engineered to balance precision with cost-
e ectiveness:

 Dramatic Cost Savings: Molybdenum wire can be wound back and forth and
reused multiple times, whereas brass wire EDM machines discard the wire after
a single pass.

 No Mechanical Stress: Because the wire never actually touches the metal (it
erodes it via a spark gap), thin or fragile parts can be machined without bending
or wrapping.

 Extreme Hardness Agnostic: It can cut through hardened steel, carbide, or

exotic alloys just as easily as soft aluminium, provided the material conducts
electricity.

 Tapers and Intricate Profiles: Its CNC control system allows the wire to tilt
(often up to ±6∘ or more), enabling complex 3D tapering and sharp internal
corners that a traditional milling bit cannot achieve.

Steps to setup
• Check if the machine is properly set first, devoid of any obstructions like clamps or
tools
• The workpiece is assembled properly, then the MCB is turned on
• Machine is then powered on

• Machine is then powered on by following the specific sequence: MCB  Voltage

Stabilizer  UPS  CPU.
• Control system boots up
• High voltage/pulse power supply.
• Dielectric fluid pump (The dielectric here, also acting as a coolant here, has to be
maintained daily by using it daily without a workpiece for 15-20 min) is to be checked.
The coolant is composed of 4L water (Demineralised water) and 1L gel JR3.
• The tension in the wire is checked (Manually).
• Wire: material is Molybdenum, 0.18 mm diameter
• Stretch the wire onto the spool and screw it into place

• When fixing the wire onto the roller, thread it from right to left, ensuring it moves
correctly between the left sensor and right sensor to prevent axis stops.
• A spark gap of 0.01 mm is maintained to cut the specimen without any contact
between the specimen and the wire

• Set the machine coordinates to parallel (X0, Y0), click 'Wire', turn on the pump, and
begin cutting from the Y-axis transitioning to the X-axis.
4
Part Program Setup and Path Configuration
This section covers importing the part geometry into NCCAD, defining the wire path
direction, and setting the machining start point prior to cutting.

Step 1: Insert the USB drive into the machine's control system.

Open the CAM software and navigate to the file import option.

 File Compatibility: Select the DXF format and browse to the USB directory to
locate your part file. Ensure the drawing file is saved in either AutoCAD 2004 or
AutoCAD 2007 DXF format to avoid import compatibility errors.

 Geometry Verification: Import the file and verify that the geometry appears
parallel and correctly on screen (X0, Y0). Check for any missing lines, open
contours, or scaling issues before proceeding.

 Wire O set Programming: Before processing the cut, you must fix and program
the wire diameter compensation (e.g., 0.18mm). When setting up the cutting
sizing, verify the path line colours: the programmed toolpath o set will turn Red
after processing the original blue geometry line.

 Send to Machine: Once the path is correctly sized, left-click on the red line to
assign the code and send the file data to the machine control unit.

Step 2: Create Path and Apply Wire O set (from X or Y Axis) Once the geometry is
loaded, generate the wire path by specifying the approach direction, either along the X
or Y axis depending on your part orientation and fixturing. Define whether the wire will
traverse clockwise or counterclockwise around the profile.

Before finalizing the path, apply the wire o set (kerf compensation) to ensure
dimensional accuracy of the finished part. Since the wire has a diameter of 0.18 mm
and the spark gap on each side is 0.01 mm, the total o set applied is:

O set = (0.18 / 2) + 0.01 = 0.10 mm

This o set is programmed in NCCAD so that the wire centreline travels 0.10 mm away
from the actual part profile, compensating for material removed by the spark erosion
process.

5
Step 3: Set Starting Point Define the starting point of the cut, which is typically at a pre-
drilled pilot hole on the workpiece. Set this as the program origin in NCCAD so the
machine knows where to begin the discharge. Confirm the coordinates match the
physical position of the wire on the machine before initiating the cut.

Wire Tension

Wire tension is maintained mechanically through a brake system on the supply spool. A
felt pad or friction brake presses against the spool to create resistance, keeping the wire
taut as it unwinds. To set tension manually, adjust the tension knob or nut on the supply
spool shaft, tighten to increase tension, loosen to reduce it. The wire should be firm but
not overtightened, as excessive tension can cause wire breakage during operation. On
the Concord machine, this is a passive mechanical brake system.

Wire Break Detection and Guide Assembly

The wire break detection assembly is located along the wire path between the supply
spool and the upper wire guide. It consists of the following components:

 Guide Rollers: direct the wire along the correct path from the supply spool to
the cutting zone.

 Stopper (Wire Break Sensor): a lever or contact pin through which the wire
passes. When the wire is intact, it holds the lever in position. If the wire breaks,
the lever drops and triggers an automatic machine stop.

 Proximity Sensor / Microswitch: detects the position of the stopper lever and
sends the stop signal to the controller.

6
  Upper and Lower Wire Guides: diamond or carbide guides that precisely
position the wire over the workpiece during cutting.

The stopper e ectively acts as a dead-man switch, as long as the wire is present and
running, the machine continues. If the wire breaks, the machine halts automatically.

Feed Rate and Wire Speed Settings

Feed rate and wire speed are configured in the machining parameters menu on the
NCCAD screen. Both are expressed in Hz, as they control the pulse frequency of the
respective drive motors.

 Wire Speed (Hz): controls the rate at which the wire travels from the supply
spool to the take-up spool. A higher Hz value means faster wire movement,
ensuring fresh wire is continuously present in the cutting zone, which reduces
the risk of wire breakage.

 Feed Rate (Hz): controls the rate of table advancement along the programmed
path. A higher Hz value results in faster cutting. This must be balanced carefully,
too high a feed rate causes wire breakage, while too low a rate reduces
machining e iciency.

Forward and Reverse Direction:

 Forward: the table advances in the programmed cutting direction under normal
machining conditions.

 Reverse: the table retracts slightly when a short circuit or wire jam is detected,
allowing the spark gap to recover. This can occur automatically or be triggered
manually.

Both forward and reverse feed rates are set independently in the settings menu, with the
reverse Hz typically kept lower than forward for controlled recovery.

Scheduled Maintenance Schedule

 Daily (Every Shift): Actuate the manual lubrication pump by pulling the lever
completely. Note: If the machine is sitting idle and not actively cutting, still pull
the lever 3 times a day to prevent the ways from seizing.

 Weekly: Apply grease to the guide wheel bearings. Check and clean the
upper/lower carbide power feed contacts.

7
Status Indicators & Error Handling

Status Meaning Required Action

Green Light Machine is actively Normal operation,

cutting monitor flushing action

Yellow Light Machine is in normal, Safe to set up workpiece,

Idle condition thread wire, or load files.

Red Light System Error or Check if E-stop is pressed

Emergency stop engaged or if a cutting error occurred

Handling "Job in Short" (Short Circuit Error): If a short-circuit error occurs, the spark
gap is bridged. The machine control system should automatically turn the pulse power
o and back the wire feed out reverse-radially (backward along the programmed path)
until the short clears. If it fails to clear automatically, stop the machine, check wire
tension, clean the carbide contacts, and ensure flushing nozzles are clearing sludge out
of the cut gap.

8
Process of 3D printing a part

Overview: Additive Manufacturing Workflow

The process of 3D printing a part involves three main stages: modelling, slicing, and
printing. Before beginning, it is important to understand the distinction between
designing and modelling, as well as the type of modelling approach used.

Designing and Modelling

Designing refers to the process of conceptualizing a part with an underlying engineering

purpose, accounting for function, fit, and intent. Modelling refers to the creation of the
digital 3D representation of that part using CAD software. In short, designing involves
the engineering thought behind the part, while modelling is the execution of that idea in
a virtual environment.

Parametric vs Non-Parametric Modelling

Modelling approaches can be classified as parametric or non-parametric. In parametric
modelling, the geometry of the part is defined and driven by parameters such as
dimensions, constraints, and relationships between features. In non-parametric
modelling, no such parameters are required, the geometry is defined freely without
dimension-based constraints.

1) Modelling (SolidWorks)

The part is modelled using SolidWorks. Once the model is complete, it is exported in
either .STL or .STEP format, which are standard file formats compatible with slicing
software.

A part made in SolidWorks

9
 2) Slicing (Ultimaker Cura)

The exported file is imported into Ultimaker Cura, a slicing software. The slicer converts
the 3D model into a series of horizontal layers and generates the corresponding
toolpath for the printer. The output is saved as a .gcode file, which contains all the
instructions the printer needs, including layer height, print speed, temperature, and
infill pattern.

Since the 3D printer used here is Anycubic Kobra 2, the settings for the slicing, most of
them, are set as default values. Build Plate Adhesion is given none since the specimen
to be printed is self-sustaining. In Materials option, the printing temperature and build
plate temperature are set to 200 °C (The filament material range is 195-210 °C) and 60
°C respectively.

Preview of the specimen in UltiMaker Cura

3) Printing
The generated .gcode file is transferred to the 3D printer. The printer reads the
file and fabricates the part layer by layer, depositing material along the
defined toolpath until the complete part is produced.

10
 Anycubic Kobra 2 3D printer

1) 3D Printing a part
When 3D printing a part, the following has to be kept in mind.

• Make sure the Magnetic PEI build plate is clean and doesn’t have any residual material
stuck on it. It can be cleaned using a metal scraper tool.
• Make sure the heated bed is not obstructed as it will stop printing midway.
• Make sure there is enough material (filament material) for printing.

Changing the filament

To change the filament, in the printer settings there will be an option in settings to
retract the filament. After that option has been selected, the filament will begin to
retract itself. After pulling out the old filament and inserting in the new one, in settings,
there will be an option called “extrude”. This will make the filament extrude out after the
temperature of the melting point of the filament is reached. After making sure the
filament is being extruded out properly, the extruding is paused/cancelled.

Starting the print

After selecting the field and starting the printing, ensure that the first layer of the print is
good, since this is where most of the time the print goes wrong

During the print

You can see clearly how each layer is when being printed (according to the thickness
given for each layer when printing, and the infill pattern given inside the object)

11
After the print

After the print is done, the object can be removed after the bed comes to its home
position. It can either be removed using a metal scraper or removing the PEI Build Plate
and twisting it, which will make the object pop out eventually.

Observations

Upon inspection of the 3D printed gear, the following observations were made regarding
its surface characteristics and print quality.

Surface Finish and Layer Visibility

The side walls of the gear exhibited visible layer lines, which are a direct result of the
perimeter tracing performed by the nozzle at each layer during the FDM process. These
lines could be felt as a stepped texture when running a finger along the side wall.

Top Surface

The top surface displayed a clean, linear fill pattern with uniform deposition. Since the
top layer is printed as a fully dense solid layer without any infill pattern, the surface
appeared neat and consistent.

12
Bottom Surface

The bottom surface appeared noticeably flatter and more compressed compared to the
top. This is due to the intentional reduction in nozzle height during the first layer
printing, which squishes the filament onto the print bed to ensure proper adhesion. As a
result, the extrusion lines on the bottom face are wider and flatter, giving it a distinctly
di erent appearance compared to the top surface.

Internal Structure

The interior of the gear was printed with an infill pattern, which reduces material
consumption while maintaining structural integrity. This internal pattern is not visible
externally but directly influences the weight and strength of the part.

13
Digital twin and Additive Manufacturing

What is a Digital Twin?

A digital twin is a virtual replica of a real physical object or process that is continuously
updated in real time using data collected from the physical counterpart. Unlike a simple
simulation, a digital twin maintains a live connection with the real object. data flows
from the physical system into the virtual model, allowing it to mirror the actual state of
the object at any given moment. This enables real-time monitoring, prediction of future
behaviour, and testing of hypothetical scenarios without interfering with the physical
system.

What is Additive Manufacturing?

Additive manufacturing (AM), commonly known as 3D printing, is a manufacturing

process in which a part is built by depositing material layer by layer based on a digital
model. This is in contrast to traditional subtractive manufacturing processes such as
CNC machining or Wire EDM, where material is removed from a solid block to achieve
the desired shape.

The primary additive manufacturing technologies include:

FDM (Fused Deposition Modelling): melts and deposits thermoplastic filament layer by
layer. Widely used for prototyping and general-purpose printing

SLA (Stereolithography): uses ultraviolet light to cure liquid resin layer by layer

SLS (Selective Laser Sintering): uses a laser to fuse plastic powder particles layer by
layer

DMLS/SLM (Direct Metal Laser Sintering/Selective Laser Melting): uses a high-powered

laser to selectively melt metal powder layer by layer, producing fully dense metal parts

EBM (Electron Beam Melting): similar to SLM but uses an electron beam as the energy
source, performed in a vacuum environment

Among these, metal additive manufacturing processes such as DMLS and SLM are of
particular significance in high-stakes industries such as aerospace, defence, and
medical implants, where part integrity is critical.

14
Challenges in Metal Additive Manufacturing

Metal AM involves extreme thermal conditions during processing. As a result, defects

such as porosity (internal voids), micro-cracks, and delamination between layers can
form invisibly within the part during printing. These defects are di icult to detect after
the part is completed without destructive testing, making real-time monitoring during
the build process essential.

Types of Digital Twins

Digital twins can be classified into three levels based on their scope:

Component Twin: the most basic level, representing a single part or component.
Monitors the behaviour of that individual part, such as stress distribution or
temperature

Asset Twin: represents a complete machine or system composed of multiple

components. Monitors how all components interact and perform together

Process Twin: the most comprehensive level, representing an entire manufacturing

process or workflow. Monitors all stages of the operation simultaneously, from material
input to final output

Each successive level increases in complexity and scope, from a single part to an entire
process.

Optical Tomography and its Role in Metal AM

Optical tomography is a non-destructive imaging technique that captures cross-

sectional images of an object using light sensors and cameras, without physically
cutting into it. In the context of metal additive manufacturing, an optical tomography
system is integrated into the printing setup to photograph each layer of the metal part as
it is being built.

As every layer is deposited and melted by the laser, cameras capture the surface of that
layer and analyse it for irregularities such as uneven melting, porosity, or inconsistent
material distribution. These layer-by-layer images are stacked to produce a complete
three-dimensional picture of the internal condition of the part as it is being
manufactured.

This layer data is fed directly into the digital twin of the printing process, continuously
updating the virtual model with real information about the internal state of the part. The
digital twin can then visualise internal defects, predict whether a defect will propagate
further, and alert the operator to intervene before the build is completed — avoiding
material waste and potential part failure.

15
In summary, optical tomography serves as the sensing mechanism that gives the digital
twin its real-time awareness of what is happening inside a metal additive manufacturing
build.

16
RobotStudio
RobotStudio is a PC-based o line programming (OLP) and simulation software
specifically developed for ABB industrial robots.

I) The setup

• Click on project, then click on create (make sure the “include a Robot and Virtual
controller” is not checked)
• Then in the home tab, click on the ABB Library and select any one from the “articulated
robots” section
• after that
Import Library > Arc Welding Equipment >AW GUN PSF 25
• in the modelling tab, drag and drop the gun onto the robot to attach it onto the robot
• From the “ABB Library” again, select a positioner (IRBP A)
• position the positioner at a suitable distance from the robot arm by right-clicking the
positioner > Position > Set Position. Give suitable coordinates

II) Setting up virtual controller

• Click “Virtual Controller” > From Layout (only in premium version)

• then provide the necessary options and click next and then click finish
• after a while the virtual controller will be created

17
III) Creating Paths and targets

• In the home tab, Paths&Targets tab will be present on the left side of the window

• then right-click on the “Paths & Procedures” folder > Create Path
• Then using “Jog TCP” on the gun, move the gun around.
• Click on “Teach Instruction” to place a point in space and keep going till enough points
are present
• to complete the loop, copy “MoveI..” of the first point and then past it, the new copied
path should be placed at the end to close the loop.

(a) Mechanism joint jog VS Mechanism linear jog

• in MJJ, each joint can be moved separately

• in MLJ, it controls TCP (tool centre point) linearly, so multiple joints may

move together.

(b) Work-envelop of the robot

Layout > robot > show work-envelop (check 3D volume)

18
Work Object

• It’s the coordinate system used to describe a workpiece.

• All objects will be related to the object frame that’s related to the user frame that’s
related to the world coordinate system

(c) Creating a Work object

other > create work object > frame by points… > 3-point

• click “Surface Selection” & “Snap Edge”

• Select the 3 points, name the object, click create. The object will be created in “Work
Objects & Targets”

(d) Creating Target

Target > Create Target

• Make sure “Surface Selection” & “Snap End” is selected

• After Selecting all the points (4 points), click create. The 4 points will be created

Right-click (on one of the points) > View Tool at Target > My Tool

• This shows the tool orientation at that point. If the tool orientation is wrong and has to
be changed

Right-click point > Modify Target > Rotate (Rotate along the desired axis, here Y axis
(180°))

• For the other points,

Right-click 1st point > Copy Position > Orientation relative to world

19
Select all the other 3 points > Apply Position > Relative World

(e) Creating Path

Select all points > Right-click > add to a new path

• can also modify instructions like speed, zone, etc

Right-click > Modify instructions > speed/zone/etc

(f) Autoconfiguration Feature

• Allows automatic definition of the optimal robot configuration for the whole path

(g) Adding a Home Position

• Get the robot to Home Position

Teach Instruction > Create Path > Add to Path > Path_10 > First

For Circular path of the robot arm TCP

The following is the RAPID code for the circular path of the TCP
PROC circular_weld()

! Move to safe home position

MoveL h1,v300,fine,F_FRNS_CMT;

! Move to weld start point

MoveL p1,v100,fine,F_FRNS_CMT;

20
 ! Start welding

ArcStart p1,v10,seam1,weld1,fine,F_FRNS_CMT;

! First quadrant

ArcC p2,p3,v10,seam1,weld1,z0,F_FRNS_CMT;

! Second quadrant

ArcC p4,p5,v10,seam1,weld1,z0,F_FRNS_CMT;

! Third quadrant

ArcC p6,p7,v10,seam1,weld1,z0,F_FRNS_CMT;

! Fourth quadrant

ArcC p8,p1,v10,seam1,weld1,z0,F_FRNS_CMT;

! End welding

ArcEnd p1,v10,seam1,weld1,fine,F_FRNS_CMT;

! Return home

MoveL h1,v300,fine,F_FRNS_CMT;

ENDPROC

21
Additive manufacturing using WAAM (Robot arm)
Objective

To observe and understand the workflow of robotic additive manufacturing using an ABB
IRB 1200 industrial robot integrated with a Fronius CMT welding system and ARC
welding process.

Introduction

This session focused on robotic additive manufacturing using Wire Arc Additive
Manufacturing (WAAM) principles. In this process, metal is deposited layer by layer
using an ARC welding process controlled by an industrial robotic arm.

The complete setup consisted of:

• ABB IRB 1200 Robot Arm

• Fronius Welding Machine

• CMT (Cold Metal Transfer) System

• FlexPendant and IRC5 Controller

• ARC Welding Process

The robotic system provided accurate motion control, while the CMT welding system-
controlled wire feeding and metal deposition during the manufacturing process.

Figure 1: ABB IRB 1200 industrial robot used for additive manufacturing and ARC
welding operations.

22
Equipment Observed

ABB IRB 1200 Robot

The ABB IRB 1200 is an industrial robotic manipulator used for precision welding and
additive manufacturing applications. The robot observed during the session had:

• Payload Capacity: 5 kg

• IRC5 Controller Integration

• FlexPendant Programming Interface

The robot was programmed to move along predefined paths for metal deposition.

Figure 2: Fronius welding power source integrated with the robotic additive
manufacturing setup.

Fronius CMT Welding System

The welding setup used a Fronius CMT (Cold Metal Transfer) system integrated with the
robot arm.

Features observed:

• Controlled metal deposition

• Reduced heat input

• Improved welding stability

23
• ARC welding-based additive manufacturing

The welding torch was mounted at the robot end e ector for automated material
deposition.

Figure 3: Fronius WF 60i Robacta Drive CMT welding torch mounted on the ABB
robot arm.

Process Setup and Programming

The robot movement and welding sequence were programmed through modules and
routines using the FlexPendant and IRC5 controller.

Initial Parameters

• Welding/wire length to base: approximately 1.5 cm

• Minimum height maintained between welding head and base plate: 15 mm

24
Programming Procedure

1. Set the Home Position (h1)

2. Select and modify the starting point P1

3. Create or edit the required Module and Routine

4. Move the robot from Home Position to Starting Point

5. Verify minimum clearance height of 15 mm

6. Move horizontally towards endpoint P2

7. Perform welding deposition

8. Move the welding head upward after deposition

9. Save and set the routine

Program Execution Procedure

The following execution procedure was observed using the ABB FlexPendant interface:

1. Open the Production Screen

2. Select Welding Action

3. Navigate through:

o Debug

o PP (Part Program) to Routine

o OK

4. Hold the Dead Man Switch to enable motor operation

5. Press the Play Button to start execution

Programming operations were carried out in the “Program” section under the Dharwad
Module.

25
 Figure 4: ABB FlexPendant interface showing program execution and routine debugging.

Digital Output Configuration

The welding process was controlled using digital outputs in the controller.

Digital Output Used

• doFrlArcOn

Function

• Value 0 → Welding OFF

• Value 1 → Welding ON

Navigation Path

Home → Menu → I/O → View → Digital Outputs → doFrlArcOn → Set to 1

Weld Activation Procedure

To verify welding activation status:

Home Page → Menu → Production Screen → Next → ManoPRI

• Weld Active → ON

• Weld Blocked → OFF

This enabled the ARC welding process during additive manufacturing.

26
Flowchart of Additive Manufacturing Procedure
START

Set Home Position (h1)

Set Starting Point (P1)

Check Minimum Height (15 mm)

Move Horizontally to P2

Perform ARC Welding

Move Up

Save Routine

Open Production Screen

Debug → PP to Routine → OK

Enable Motor (Dead Man Switch)

Press Play Button

Set doFrlArcOn = 1

Welding Active

END

27
Final Additive Manufactured Layer

The additive manufacturing process resulted in layer-by-layer metal deposition using

robotic ARC welding. The process demonstrated practical implementation of Wire Arc
Additive Manufacturing (WAAM) using robotic automation.

Figure 5: Final deposited layer produced using robotic ARC additive manufacturing.

Conclusion

Today’s internship session provided practical exposure to robotic additive

manufacturing using the ABB IRB 1200 industrial robot and Fronius CMT welding
system. The session helped in understanding robotic programming, welding activation,
routine execution, digital output control, and automated layer deposition using ARC
welding processes.

The observation also provided insight into the practical implementation of Wire Arc
Additive Manufacturing (WAAM) and industrial robotic automation in modern
manufacturing applications.

28
Observed Additive Manufacturing Using ABB IRB 1200 Robot Arm

On this day, a live demonstration of the wire arc additive manufacturing process was
observed in the lab. The session involved a change in material from the previously used
aluminium to steel (stainless steel).

The following are the steps to setup the apparatus

Phase 1: Disconnecting the Robotic Torch Body

1. Torch Body Separation: The operators unlatched and removed the main robotic
torch body from the Fronius WF 60i Robacta Drive crash box/mounting.

2. Coolant Circuit Isolation: Removing this body broke the liquid circuit, causing
residual coolant (water-glycol mixture) to leak slightly until the integrated self-
sealing check-valves or manual plugs were engaged.

3. Material Extraction: The remaining aluminium wire from the previous operating
cycle was snipped and carefully extracted from the feed lines to clear the path.

Phase 2: Removing the Main Liners (Feeder to Torch)

4. Torch-End Liner Removal: The operators loosened a clamping nut at the torch
connection interface. They pulled out the wire guide liner (or steel/Teflon inner
liner) that runs directly into the welding torch assembly.

5. Robacta Drive-End Liner Removal: Moving to the opposite side of the robotic
wrist interface, a similar clamping nut was unscrewed. The long main wire guide
liner, which bridges the gap between the bulk wire feeder and the robotic drive,
was pulled free.

6. Infeed Spacer / O-Ring Inspection: During this step, small collars or retaining
sleeves were exposed. These are liner collets or infeed spacers used to centre
the liner perfectly within the drive-roll housing so the aluminium wire doesn't
bend or bird-nest.

29
Phase 3: Bulk Feeder Disassembly

7. Wire Feeder Interface Disassembly: Over at the primary wire feeding unit
(controlled via the ABB FlexPendant interface), the retaining nut on the wire-feed
block was unscrewed.

8. Infeed Guide / Collet Removal: The corresponding infeed nozzle / collar

assembly was removed from the drive-roll mechanism to ensure no old
aluminium shavings were trapped inside the feeding gears.

Phase 4: Torch Front-End (Consumables) Disassembly & Material Assessment

9. Spatter Guard & Gas Nozzle Removal: Using a specialized spatter-

cleaning/multitool, the operator unthreaded the outer gas nozzle from the front
of the torch body.

10. Contact Tip Verification: Underneath the gas nozzle, they exposed the contact
tip (the gold/yellow brass or copper-alloy component at the very end).

11. Torch Neck-Liner Inspection: The shorter brass neck-liner inside the torch
neck was inspected. Brass or graphite neck-liners are often retained at the very
tip of the torch because they handle high heat better than the long main liners
running back to the feeder.

Phase 5: Installing the Steel Wire Guide Liners (Reverse Assembly)

12. Main Steel Spiral Liner Installation: The new steel spiral liner was fed back
through the primary conduit connecting the bulk wire feeder to the Fronius WF

30
 60i Robacta Drive. This metal-wound liner provides the rigidity needed to guide
hard steel wire without buckling.

13. Securing the Liners and Collets: The operator reinserted the specialized liner
collets (retaining collars) at the entry points to centre the liner. The clamping
nuts were then tightened down on both the Robacta Drive end and the primary
wire feeder block to lock the new feed path securely in place.

14. Reassembling the Torch Front-End: The shorter internal torch neck-liner was
reseated. The contact tip and outer gas nozzle (nozzle cover) were threaded
back onto the torch body using the specialized tool and tightened to operational
torque.

15. Docking the Torch Body: The main robotic torch body was snapped back into
the Robacta Drive crash box, and the coolant lines were re-engaged to restore
the water-cooling loop.

Phase 6: Spool Loading and Automated Wire Feeding

16. Spool Mounting: The old aluminium wire spool was removed, and a new spool
of steel welding wire was mounted onto the de-reeler / wire spool carrier (the
standalone machine next to the robotic cell).

17. Manual Threading to Drive Rolls: The operator manually fed the leading edge of
the steel wire into the primary wire feeder's inlet guide, threading it directly
between the drive rolls (which grip and push the wire).

18. Automated Wire Inching (Jogging): Using either the ABB FlexPendant or the
local interface on the Fronius wire feeder, the operator initiated the Wire Inch (or
"Wire Jog") function. This spins the motorized drive rolls at a controlled speed to
automatically push the rigid steel wire through the entire length of the newly
installed metal liners.

19. Final Tip Alignment: The wire was fed continuously until it successfully emerged
from the copper contact tip at the very front of the welding gun. The operator
then snipped the excess wire to the correct stick-out length, completing the
material changeover from aluminium to steel.

31
Conclusion

It provided practical insights into the maintenance and execution workflows of Wire Arc
Additive Manufacturing (WAAM).

Observing the system changeover from aluminium to steel highlighted how di erent
material properties require specific internal hardware—specifically transitioning from
flexible Teflon liners to rugged Steel Spiral wire guide liners to handle abrasive wire.
Additionally, observing the multi-layer welding routine to build a vertical wall
successfully demonstrated how robotic path programming and Fronius CMT thermal
control come together to create 3D metal structures layer-by-layer.

This observation provided a solid, practical understanding of industrial robotic

automation and material handling in modern manufacturing.

32
Introduction to the Hyrel 3D Printing System
The Hyrel 3D printer is a modular, multi-tool manufacturing platform designed for
advanced research, prototyping, and material science. Unlike standard 3D printers that
only melt plastic, the Hyrel system uses a rugged, open-architecture design capable of
processing thousands of di erent materials. This versatility makes it an industry-
standard tool for universities, medical labs, and aerospace R&D facilities worldwide.

System Capabilities and Materials

The platform uses specialized, CANbus-controlled print heads that users can hot-swap
in seconds without tools. This allows the machine to switch seamlessly between
diverse fabrication methods:

 Standard Filaments: ABS, PLA, Nylon, and Polycarbonate.

 Liquid Pastes & Gels: Silicone, clay, ceramics, and conductive inks.

 Biological Materials: Hydrogels and biocompatible matrices for tissue

engineering.

 Subtractive Tools: Laser cutters, engravers, and light-duty spindle routers.

Multi-Nozzle Architecture

Hyrel systems feature a multi-head mounting array that accommodates up to five

independent tool heads simultaneously. This multi-nozzle configuration delivers several
key manufacturing advantages:

 Multi-Material Co-Extrusion: Combines rigid structural polymers with flexible

seals or conductive traces in a single print job.

 Soluble Support Integration: Dedicates one nozzle to a water-soluble matrix (like

PVA) to build complex, overhanging geometries.

 Clone & Mirror Printing: Runs multiple identical print heads simultaneously to
duplicate production throughput.

 E iciency Optimization: Uses large nozzles for rapid internal infill and fine-
diameter nozzles for high-resolution exterior walls.

33
Advance-Hyrel 3D printer working and SOP

1. Purpose

This SOP describes the procedure for setting up, calibrating, and operating the
Advance-Hyrel 3D printer for fused deposition modelling (FDM) additive manufacturing.
The procedure includes printer setup, filament loading, G-code preparation using
PrusaSlicer, Z-axis calibration, and print execution.

2. Scope

This procedure is applicable to students, researchers, and laboratory personnel

operating the Advance-Hyrel 3D printer for fabrication of polymer-based components.

3. Equipment and Software Required

Equipment

 Advance-Hyrel 3D Printer

 Extruder assembly

 PLA filament spool

 Glass build plate

 Glue stick

34
  A4 paper sheet (for Z-axis calibration)

Software

 Repetrel

 PrusaSlicer

STEP 1&2 – Machine Setup and Filament Loading

4 Connecting the Printer to Repetrel

1. Switch ON the Advance-Hyrel 3D printer.

2. Connect the extruder to the printer.

3. Ensure all 16 connector pins are properly aligned and connected.

4. Tighten the locking screw located behind the extruder.

5. Switch ON the tablet.

6. Launch the Repetrel application.

4.1 Establishing Communication

1. Verify that the Repetrel interface displays a grey status screen, indicating that the
printer is not yet connected.

35
 2. From the right-side communication panel, select COM4.

3. Then select COM3.

4. Click OK when the confirmation pop-up appears.

5. Verify that the interface colour changes, indicating successful connection

between the tablet and printer.

4.2 Understanding the Interface

 Column 1: Print Bed Controls

 Column 2: Slot 1 Controls

 Column 3: Slot 2 Controls

 Column 4: Slot 3 Controls

 Additional columns correspond to additional extruder slots.

4.3 Heating the Extruder

1. Set the required extrusion temperature according to the filament material.

2. Wait until the extruder reaches the specified temperature before loading
filament.

4.4 Preparing the Extruder

1. In Column 4, click the Play/Next button.

2. Click the OFF button to switch the extruder ON.

3. On the control panel, press +X 10 to raise the print bed.

4.5 Loading the Filament

1. Mount the filament spool on the holder.

2. Ensure that the spool rotates towards the operator during feeding.

3. Navigate to:

Settings → Print → Load Settings

4. Click Apply.
36
 5. Feed the filament into the extruder until proper extrusion is observed.

4.6 Homing the Printer

1. Press the Home button.

2. Wait for the print bed and print head to move to their home positions.

4.7 Manual Bed Movement

 Use 0.1 mm increments for fine adjustments.

 Use 1 mm or 10 mm increments for larger movements.

5. Detailed Filament Insertion Procedure

1. Ensure the extruder has reached the required operating temperature.

2. Rotate the extruder feed knob in the opposite direction to release tension on the
filament path.

3. Cut the filament end at approximately 45° to facilitate smooth insertion.

4. Insert the filament into the filament entry port.

5. Feed the filament through the complete filament path until it reaches the nozzle.

6. Ensure the filament passes smoothly through all guide sections without bending.

Filament Entry → Feed Mechanism/Drive Gear → PTFE Guide Tube (or Extruder Body) →
Hot End/Nozzle

7. Continue feeding until molten material begins to extrude from the nozzle.

8. Verify continuous and uniform extrusion before proceeding.

STEP 3 – G-Code Preparation Using PrusaSlicer

6. Software Installation

1. Download and install the PrusaSlicer software on the workstation.

2. Launch the software.

37
6.1 Importing the Model

1. Open PrusaSlicer.

2. Import the STL file.

3. Verify the model orientation on the build plate.

6.2 Configure Printer Profile

1. Select the Printer tab.

2. Open the printer profile drop-down menu.

3. Select either:

o ESR CA 0.6NZ

o HYDRA 16-AS

4. Change the interface mode from Beginner to Expert (top-right corner).

6.3 Machine Parameters

Build Plate Dimensions

Configure the bed dimensions as:

Parameter Value

X Dimension 200

Y Dimension 200

Maximum Print Rate

Set:

Maximum Print Rate = 200 mm/s

(This defines the upper limit of print speed)

38
Save Preset

Save the configuration using an appropriate preset name (e.g., ESRIITDh).

6.4 Filament Settings

1. Open the Filament Settings tab.

2. Select PLA from the material list.

Configure:

Parameter Value

Nozzle Temperature 210°C

Bed Temperature 60°C

Cooling

Enable:

Keep Fan Always On

This improves PLA print quality and reduces deformation.

6.5 Print Settings

Layer Height

Navigate to:

Print Settings → Layers and Perimeters

Select a layer height between 40% and 60% of the nozzle diameter.

Example:

If nozzle diameter is 0.6mm then the recommended Layer Height is 0.24 to 0.36 mm

Solid Layers

Configure:

Top Solid Layers = 3

Bottom Solid Layers = 3

39
These layers create a fully closed surface at the top and bottom of the print.

Infill

Adjust the infill percentage according to the required strength.

Higher infill → smaller internal mesh cells → greater strength and material usage.

Skirt and Brim

Skirt

A line printed around the object without touching it.

Purpose:

 Primes the nozzle

 Checks material flow before printing

Brim

Additional material attached to the base of the part.

Purpose:

 Improves bed adhesion

 Reduces warping

 Useful for small contact areas

Support Material

Enable support structures when:

 Overhangs are present

 Bridging distances are excessive

 Features are suspended above empty space

Print Speed

Recommended range:

20–50 mm/s

Lower speeds generally improve print quality.

40
Extrusion Width

Set extrusion width to approximately 10% more of nozzle diameter.

Example:

If nozzle diameter is 0.6mm then Extrusion Width will be 0.66mm

>This improves bonding between adjacent extrusion lines.

6.6 Generate G-Code

1. Verify all printer, filament, and print settings.

2. Click Slice Now.

3. Review the generated toolpath.

4. Check estimated print time and material consumption.

5. Export the generated G-code file.

7. Trial Print Preparation

7.1 Creating a Test Geometry

1. Open PrusaSlicer.

2. Right-click on the build plate.

3. Select Add Shape → Cylinder.

4. Position the cylinder at the desired location on the build plate.

7.2 Modifying Cylinder Dimensions

1. Select the cylinder.

2. Locate the dimension controls in the lower-right corner of the interface.

3. Use the Lock icon to control dimensional scaling.

41
Locked Mode

 All dimensions scale proportionally.

 Changing one dimension automatically updates the others.

 The cylinder maintains its circular cross-section.

Unlocked Mode

 Dimensions can be modified independently.

 The cylinder can be stretched in one direction.

 The circular cross-section may become elliptical.

7.3 Spiral Vase Configuration

1. Navigate to:

Print Settings → Layers and Perimeters → Spiral Vase

2. Enable Spiral Vase Mode.

This mode:

 Prints a continuous outer wall.

 Eliminates layer seams.

 Produces a thin-walled hollow cylinder.

 Is useful for trial prints and extrusion testing.

7.4 Exporting G-Code

1. Slice the model.

2. Review the generated toolpath.

3. Click Export G-Code.

4. Save the generated G-code file.

7.5 Editing the G-Code File

1. Open the exported G-code file using Notepad or another text editor.

2. Perform the required machine-specific modifications before printing.

42
Modification 1

Replace the default extruder/tool selection command with:

T3

Modification 2

Change: M109 T15 to M109 T14

Modification 3

Next line of height, add

G54 X30 Y0;

7.6 Importing the G-Code into Repetrel

1. Open the Repetrel software.

2. Navigate to:

Control → Project Composer → Object → Add G-Code File

3. Browse to the exported G-code file.

4. Select and import the file into the project.

5. Verify that the file has been loaded successfully before proceeding.

8. Z-Axis Calibration

Note: Z-axis calibration must be performed before every print job to ensure proper first-
layer adhesion and print quality.

8.1 Preparation

1. Obtain a standard A4 sheet of paper.

2. Ensure the printer is connected and ready for operation.

3. Position the nozzle approximately at the centre of the print bed.

8.2 Bed Height Adjustment

1. Using the tablet controls, manually raise the print bed toward the nozzle.

2. Place the A4 sheet between the nozzle and the print bed.

43
 3. Adjust the bed height until the paper can move with slight resistance.

Acceptance Criteria

The paper should:

 Move freely when pulled.

 Experience slight friction from the nozzle.

 Not be tightly clamped.

 Not move without any resistance.

This indicates an appropriate nozzle-to-bed gap for printing.

8.3 Large Z-Axis Movement

1. Enable the Enable Big Z Move option.

2. Use this mode when larger bed movements are required.

Note:

 When enabled, the bed can move in increments of approximately 10 mm.

 When disabled, movement is limited to approximately 1 mm increments.

8.4 Applying Z Calibration

1. Once the correct gap has been established, select Enable Z Calibrate.

2. Confirm that the calibration settings are applied.

3. Verify that the nozzle maintains the correct clearance from the build plate.

8.5 Bed Surface Preparation

1. Clean the glass build plate if necessary.

2. Apply a thin and uniform layer of glue stick to the printing surface.

3. Ensure complete coverage of the intended print area.

Purpose:

 Improves first-layer adhesion.

44
  Reduces print detachment during operation.

 Minimizes warping of PLA components.

9. Starting the Print

1. Verify that:

o The G-code file has been loaded.

o Filament has been loaded correctly.

o Z-axis calibration has been completed.

o Glue has been applied to the build plate.

2. Start the print job.

3. Observe the first layer carefully.

4. Ensure that the filament adheres uniformly to the build plate.

Temperature Override Function

Normally, the printer waits until the build plate reaches the specified temperature
before beginning the print.

To bypass this waiting sequence:

1. Press Pause.

2. Press Start.

This overrides the waiting condition and allows the print to begin before the bed reaches
its target temperature.

(Use with caution. Printing before the bed reaches the required temperature may
reduce first-layer adhesion and increase the likelihood of print failure.)

45
For Test Printing

1. Experimental Print: Open-Top Conical Fulcrum

Using PrusaSlicer, a conical model was sliced and prepared using the Spiral Vase
(Single-Wall) feature. The physical and slicing configuration parameters included:

 Geometric Geometry: The component was printed as a conical fulcrum (a

truncated cone with a flat top and bottom).

 Orientation: The model was oriented with its smaller radius face seated directly
on the build plate and its larger radius face forming the open top.

 Spiral Vase Configuration: This option forces the extruder to move in a

continuous, uninterrupted helical path along the Z-axis, eliminating layer change
retraction lines.

 Base Layer Adhesion: To ensure a solid foundation without a ecting the

continuous wall path above, the printer laid down 3 solid bottom layers before
initiating the single-wall spiral path, leaving the top side entirely open.

2. Theoretical Analysis: Overhang Limits & Support Structure Calculation

Following the test prints, an analysis was conducted regarding how overhanging
structures fail without adequate foundation material. Using the engineering principles
from the lecture data, the threshold angle before a print requires sacrificial support was
calculated using facet normal vectors and layer dimensions.

46
Mathematical Formulation for Support Generation:

For a given STL file where the build direction is vertical along the positive Z-axis (+Z), the
necessity of a support structure is determined by calculating the angle α between the
build direction vector (k) and the outward unit normal vector of the triangle facet (n):

𝛼 = cos 𝑛

Where 𝑛 represents the Z-component coordinate of the outward normal of the i-th
facet

Critical Angle Condition (θ):

If a printer has a maximum allowable lateral layer overhang (d) and a specified layer
thickness (t), the critical self-supporting boundary angle (θ) is modelled as:
𝑡
tan(𝜋 − 𝜃) =
𝑑
𝑡
𝜃 = 𝜋 − tan
𝑑
No Support Zone: If 𝛼 < 𝜃, the facet normal angle is safe, meaning the layer su iciently
steps out onto the previous layer's perimeter without collapsing.

Support Required Zone: If 𝛼 ≥ 𝜃, the angle becomes too steep for the extruded plastic
to stay suspended in mid-air, resulting in print failure unless explicit sacrificial support
material is generated during pre-processing.

47
Methodology for Experimental Investigation of Continuous Carbon Fibre

Reinforced Onyx Composites1. Introduction

Continuous fibre reinforced polymer composites have gained significant attention in

engineering applications due to their high specific strength, high sti ness, corrosion
resistance, and lightweight characteristics. Among the various additive manufacturing
techniques available, Continuous Fiber Fabrication (CFF) enables the production of
composite components by embedding continuous reinforcement fibres within a
thermoplastic matrix during the printing process. This allows the mechanical properties
of printed components to be tailored by controlling parameters such as fibre
orientation, fibre volume fraction, reinforcement pattern, and laminate stacking
sequence.

In the present work, Onyx was selected as the matrix material and continuous carbon
fibre (CF) was used as the reinforcement material. Onyx is a nylon-based engineering
thermoplastic containing chopped carbon fibres, providing improved strength,
sti ness, and dimensional stability compared to conventional nylon. The addition of
continuous carbon fibre further enhances the load carrying capability of the composite
by transferring tensile loads through the continuous reinforcement.

The primary objective of this work is to investigate the influence of fibre orientation,
fibre volume fraction (FVF), and reinforcement pattern on the tensile and flexural
behaviour of Onyx–continuous carbon fibre composites manufactured using the
Markforged Mark Two composite 3D printer.

2. Materials and Equipment

The composite specimens were manufactured using the following materials and
equipment.

Matrix Material

The matrix material used in this study was Onyx, a micro carbon fibre reinforced nylon
developed for additive manufacturing. Onyx provides improved sti ness, surface finish,
dimensional stability, and wear resistance while acting as the binding matrix for the
continuous fibre reinforcement.

Reinforcement Material

Continuous carbon fibre was selected as the reinforcement material because of its high
tensile strength, high elastic modulus, low density, and excellent fatigue resistance.
During printing, the continuous carbon fibre is embedded within the molten Onyx
matrix, allowing the fibres to carry a significant portion of the applied tensile load.

48
Printing System

All specimens were fabricated using the Markforged Mark Two desktop composite 3D
printer. The printer operates using the Continuous Fiber Fabrication (CFF) process,
where two independent print systems are used: one nozzle deposits the Onyx
thermoplastic matrix, while the second print head places continuous reinforcement
fibres within selected layers.

The printing process was prepared using Markforged Eiger software, which was used to
generate the toolpath, define fibre orientation, select reinforcement pattern, and
specify the fibre volume fraction.

3. Specimen Design

The tensile specimens were designed according to ASTM D3039, while the flexural
specimens were prepared with reference to ASTM D7264.

The tensile specimen dimensions were maintained at 250 mm × 25 mm × 3 mm. The

grip section measured 50 mm at each end, leaving a gauge length of 150 mm for tensile
loading. These dimensions provide su icient gripping area while ensuring that failure
occurs within the gauge section rather than at the grips.

The laminate thickness was fixed at approximately 3 mm throughout the investigation.

Maintaining a constant specimen thickness ensures that any variation in mechanical
behaviour can be attributed to changes in fibre orientation, fibre volume fraction, or
reinforcement pattern rather than changes in specimen geometry.

A layer height of 0.125 mm was adopted, resulting in a total of 24 printed layers. The
layer height was not selected manually but was automatically fixed by the Markforged
Mark Two printer when continuous carbon fibre reinforcement was enabled. The
printer's Continuous Fiber Fabrication process requires continuous fibre to be
embedded within an Onyx layer of matching thickness, thereby locking the layer height
to 0.125 mm. Consequently, a laminate thickness of 3 mm corresponds to a total of 24
layers.

Continuous carbon fibre reinforcement cannot be deposited on the outermost layers of

the laminate. The Markforged printing process requires the first and last layers to
consist solely of Onyx, forming protective cover layers around the reinforcement. These
cover layers encapsulate the carbon fibres, improve interlayer bonding, protect the
reinforcement from environmental exposure, and enhance the overall print quality and
dimensional accuracy of the specimen.

49
4. Design of Experiments (DOE)

The experimental programme was designed to investigate the influence of three

independent variables on the mechanical performance of the composite specimens:

• Fibre orientation

• Fibre volume fraction

• Fibre reinforcement pattern

A full factorial experimental design was adopted during the first phase of the
investigation. Three parameters were varied simultaneously, producing eight unique
experimental cases.

Fibre Orientation

Two fibre orientations were selected for the first phase of the study:

• 0°

• 90°

These orientations represent the two extreme loading conditions of a unidirectional

composite laminate. At 0°, the fibres are aligned parallel to the loading direction,
enabling the continuous carbon fibres to carry the majority of the applied load and
therefore producing the highest expected tensile sti ness and strength. At 90°, the
fibres are oriented perpendicular to the applied load, causing the Onyx matrix to carry
most of the tensile load while allowing the fibre–matrix interface and interlaminar
behaviour to be evaluated.

Investigating these two extreme cases establishes a baseline for understanding the
mechanical behaviour of the composite before studying intermediate fibre orientations
in future work.

Fibre Volume Fraction (FVF)

Two reinforcement levels were selected:

• Low Fibre Volume Fraction

• High Fibre Volume Fraction

The low fibre volume fraction consisted of four continuous carbon fibre reinforcement
layers positioned at Layers 5, 9, 16, and 20.

The high fibre volume fraction consisted of eight reinforcement layers positioned at
Layers 3, 6, 9, 12, 13, 16, 19, and 22.

50
The total laminate thickness remained constant for both cases. Varying only the number
of reinforcement layers allows the influence of fibre volume fraction on the mechanical
properties to be investigated while eliminating the e ect of specimen thickness.

The reinforcement layers were intentionally distributed throughout the laminate

thickness rather than grouped together. This arrangement promotes a more uniform
distribution of sti ness and load transfer across the laminate while reducing stress
concentrations and minimizing the possibility of premature interlaminar delamination.

Reinforcement Pattern

Two reinforcement patterns available within the Markforged Eiger software were
selected:

• Isotropic

• Concentric

The isotropic reinforcement pattern distributes continuous fibres uniformly across the
interior of the specimen, providing relatively balanced mechanical properties. The
concentric reinforcement pattern follows the outer profile of the geometry, reinforcing
the specimen boundaries and improving load transfer around the perimeter. Comparing
these two reinforcement strategies allows the influence of fibre routing on the
mechanical behaviour of the composite to be evaluated.

The complete Design of Experiments consisted of eight experimental cases generated

using a full factorial combination of fibre orientation, fibre volume fraction, and
reinforcement pattern.

5. Printing Parameters

The composite specimens were fabricated using the Markforged Mark Two desktop
composite 3D printer based on the Continuous Fiber Fabrication (CFF) process. The
printing process was prepared using the Markforged Eiger slicing software, which
generates the toolpaths for both the Onyx matrix material and the continuous carbon
fibre reinforcement.

Unlike conventional FDM printers, the Markforged system employs two independent
print heads. The first nozzle deposits the molten Onyx thermoplastic, while the second
nozzle places the continuous carbon fibre reinforcement within selected layers of the
laminate. During deposition, the reinforcement fibres are completely encapsulated by
the Onyx matrix, ensuring adequate bonding between the matrix and reinforcement.

When continuous fibre reinforcement is enabled within the Eiger software, the layer
height is automatically fixed at 0.125 mm. This restriction is imposed by the Continuous

51
Fiber Fabrication process to ensure proper fibre placement and complete
encapsulation of the reinforcement material. Consequently, the layer height cannot be
modified by the user when printing continuous fibre reinforced composites.

The Markforged system also restricts the placement of continuous fibre reinforcement
in the outermost layers of the laminate. The first and last printed layers are deposited
entirely using Onyx, forming protective cover layers that improve surface finish, protect
the reinforcement fibres from environmental exposure, and enhance interlayer bonding.
Therefore, continuous carbon fibre reinforcement is positioned only within the interior
layers of the laminate.

The primary printing parameters considered during this investigation are summarized in
Table 5.1.

Table 5.1 Printing Parameters

Parameter Value

Printer Markforged Mark Two

Printing Technology Continuous Fiber Fabrication (CFF)

Matrix Material Onyx

Reinforcement Material Continuous Carbon Fibre

Layer Height 0.125 mm

Total Layers 24

Laminate Thickness 3 mm

Reinforcement Pattern Isotropic / Concentric

Fibre Orientation 0° / 90°

Fibre Volume Fraction Low (4 CF layers), High (8 CF layers)

6. Tensile Testing Methodology

The tensile behaviour of the printed composite specimens will be evaluated in

accordance with ASTM D3039, the standard test method for determining the tensile
properties of polymer matrix composite materials reinforced with continuous fibres.

52
Prior to testing, the grip section and gauge section will be identified on each specimen.
The grip section provides su icient area for secure clamping by the Universal Testing
Machine (UTM), while the gauge section represents the region in which deformation and
failure are expected to occur. Failure occurring within the gauge section ensures that
the measured mechanical properties accurately represent the material behaviour rather
than the gripping arrangement.

The tensile specimens will be mounted centrally within the grips of the Universal Testing
Machine to minimize bending during loading. The load will be applied along the
longitudinal axis of the specimen at a constant crosshead speed of 2 mm/min. This
loading rate is recommended by ASTM D3039 to provide controlled tensile loading while
minimizing dynamic e ects that may influence the measured mechanical response.

Testing will be carried out under standard laboratory environmental conditions of

approximately 23 ± 3 °C and 50 ± 10% relative humidity. These conditions are
particularly important for nylon-based materials such as Onyx because moisture
absorption can significantly influence sti ness, strength, and failure behaviour.

During testing, load and displacement data will be continuously recorded by the
Universal Testing Machine. The acquired data will subsequently be converted into
engineering stress and engineering strain for construction of the stress-strain curve.

The following tensile properties will be determined from the experimental results:

• Ultimate Tensile Strength (UTS)

• Young's Modulus

• Ultimate Tensile Strain

• Poisson's Ratio

• Maximum Tensile Load

• Stress-Strain Curve

These parameters provide a comprehensive understanding of the tensile behaviour of

the continuous fibre reinforced composite and enable comparison between di erent
reinforcement configurations.

53
7. Flexural Testing Methodology

The flexural behaviour of the printed composite specimens will be evaluated according
to ASTM D7264, the standard test method specifically developed for determining the
flexural properties of polymer matrix composite materials.

Unlike ASTM D790, which is primarily intended for plastic materials, ASTM D7264 is
specifically designed for fibre reinforced polymer composites. Therefore, ASTM D7264
provides a more suitable method for evaluating the flexural behaviour of continuous
carbon fibre reinforced Onyx composites.

The flexural specimens will be tested using a three-point bending configuration on a

Universal Testing Machine. The support span will be selected according to the
recommended span-to-thickness ratio of 32:1 specified by the standard. Since the
specimen thickness is maintained at approximately 3 mm, the corresponding support
span will be approximately 96 mm.

During testing, the specimen will be placed symmetrically on the supporting rollers
while the loading nose applies force at the centre of the span. The applied load will
generate compressive stresses on the upper surface and tensile stresses on the lower
surface of the specimen, enabling evaluation of the flexural performance of the
composite laminate.

The load and mid-span displacement will be continuously recorded throughout the test.
These measurements will be used to determine:

• Flexural Strength

• Flexural Modulus

• Maximum Load

• Load-Deflection Curve

Comparison of these properties among the di erent specimen configurations will

provide insight into the influence of fibre orientation, fibre volume fraction, and
reinforcement pattern on the flexural response of the composite.

54
8. Experimental Matrix

A full factorial experimental design was adopted for Phase I of the investigation. Three
independent variables were considered: fibre orientation, fibre volume fraction, and
reinforcement pattern. The combination of these variables resulted in eight
experimental cases.

Table 8.1 Experimental Matrix

Case Fibre Orientation Fibre Volume Fraction Reinforcement Pattern

1 0° Low Isotropic

2 0° Low Concentric

3 0° High Isotropic

4 0° High Concentric

5 90° Low Isotropic

6 90° Low Concentric

7 90° High Isotropic

8 90° High Concentric

The selected design enables the individual and combined influence of fibre orientation,
fibre volume fraction, and reinforcement pattern on the mechanical properties of the
composite to be systematically investigated while maintaining all other processing
parameters constant.

9. Expected Experimental Outcomes

The primary objective of the experimental investigation is to quantify the influence of

fibre orientation, fibre volume fraction, and reinforcement pattern on the tensile and
flexural behaviour of continuous carbon fibre reinforced Onyx composites.

Based on the tensile and flexural testing, the mechanical properties obtained from each
specimen configuration will be compared to identify the reinforcement strategy that
provides the highest structural performance.

55
The tensile test will provide information regarding the load carrying capability, tensile
sti ness, deformation behaviour, and failure characteristics of the composite. Similarly,
the flexural test will evaluate the bending resistance, flexural sti ness, and structural
integrity of the laminate under transverse loading.

The experimental results obtained from all eight specimen configurations will be
analysed to determine the e ect of each design parameter and to identify the optimum
reinforcement configuration for achieving improved mechanical performance in
continuous carbon fibre reinforced Onyx composites manufactured using the
Markforged Mark Two additive manufacturing system.

56
Structural Performance of Additively Manufactured
Composite Lattice Structures: Strain Rate, Cell Geometry,
and Weight Ratio E ects

Methodology and Design of Experiments

The study investigated the structural performance of additively manufactured
composite lattice structures by examining the e ects of unit cell geometry,
reinforcement weight ratio, and strain rate on their mechanical behavior. Two lattice
configurations, namely cylindrical and hexagonal honeycomb structures, were designed
using computer-aided design (CAD) software and fabricated using a masked
stereolithography (MSLA) vat photopolymerization 3D printing process. The lattice
structures consisted of 8 × 9 unit cells enclosed within an outer wall to ensure uniform
load distribution during compression. Each specimen had overall dimensions of 46 mm
× 46 mm × 25 mm, with a unit cell diameter/width of 5 mm and a minimum strut
thickness of 0.5 mm.

The printing material consisted of a flexible ultraviolet-curable resin modified with

hollow SiO₂ glass microballoons. Four material compositions were prepared: neat resin
(0 wt%), 5 wt%, 10 wt%, and 20 wt% glass microballoon reinforcement. The reinforcing
particles were manually mixed with the resin for 10 minutes and subsequently
sonicated for 20 minutes to remove entrapped air before printing. Fabrication was
carried out using an Anycubic Photon S LCD resin printer with a layer thickness of 50
μm. After printing, all specimens underwent a standardized post-processing procedure
consisting of compressed air cleaning, ultrasonic cleaning in a water–isopropyl alcohol–
detergent solution, ultraviolet post-curing for 80 minutes, and conditioning in a dark
ambient environment for two weeks prior to testing.

The dimensional accuracy and microstructural characteristics of the printed specimens

were evaluated using optical microscopy and scanning electron microscopy (SEM).
These analyses were performed to examine the printed cell geometry, dimensional
deviations, distribution of reinforcing particles, and possible defects such as porosity
and particle agglomeration.

Mechanical characterization was performed under both quasi-static compression and

dynamic impact loading conditions. Quasi-static compression tests were conducted
using an Instron 5843 universal testing machine at a crosshead speed of 50 mm/min,
corresponding to an approximate strain rate of 0.02 s⁻¹. Dynamic impact tests were
carried out using an instrumented drop-weight impact system by dropping a 950 g mass
from a height of 1.03 m, producing an impact energy of 9.6 J and an approximate strain

57
rate of 97 s⁻¹. Five specimens from each material composition were tested under each
loading condition, resulting in ten specimens for every reinforcement level.

Digital Image Correlation (DIC) was employed throughout both quasi-static and
dynamic experiments to obtain full-field strain measurements. The technique enabled
the measurement of longitudinal, transverse, and shear strain distributions within the
lattice structures, providing detailed information regarding local deformation
mechanisms and their relationship to the overall structural response.

The experimental design consisted of three independent variables: unit cell geometry
(cylindrical and hexagonal), glass microballoon reinforcement content (0 wt%, 5 wt%,
10 wt%, and 20 wt%), and loading condition (quasi-static and dynamic impact). The
primary response variables included dimensional accuracy, stress–strain behavior,
plateau stress, densification strain, tangent modulus, specific energy absorption (SEA),
energy absorption e iciency, force–time response, and local strain distributions
obtained through Digital Image Correlation. This methodology enabled the researchers
to systematically evaluate the combined influence of material composition, lattice
geometry, and strain rate on the structural performance of additively manufactured
composite honeycomb structures.

Summary of Findings
The study successfully demonstrated the feasibility of manufacturing reinforced
composite lattice structures using vat photopolymerization with flexible photocurable
resin reinforced by hollow glass microballoons. The printed structures exhibited high
dimensional accuracy at the macroscopic level, although minor dimensional deviations
were observed in the strut thickness due to printing-related factors such as ultraviolet
light scattering, resin shrinkage, and flexible resin deformation during fabrication. These
manufacturing imperfections had only a limited influence on the overall structural
performance.

The experimental investigation showed that the geometry of the lattice unit cell
significantly influenced the mechanical response of the structures. Cylindrical lattice
structures exhibited higher compressive strength, greater plateau stress, and higher
structural sti ness than hexagonal lattices. However, the cylindrical structures also
reached densification at lower strains, indicating earlier collapse of the cellular
structure during compression. In contrast, hexagonal lattices sustained larger
deformations before densification, providing improved deformation capability and
higher energy absorption e iciency over a wider strain range.

The addition of glass microballoons increased the sti ness and load-bearing capability
of the lattice structures under quasi-static loading. However, increasing the

58
reinforcement content generally reduced the densification strain, indicating that the
reinforced structures became less deformable before complete collapse. The 10 wt%
reinforced specimens displayed inconsistent mechanical behavior due to particle
agglomeration, entrapped air bubbles, and poor bonding between the resin matrix and
the reinforcing particles, highlighting the importance of proper material preparation
during composite fabrication.

Under dynamic impact loading, the structural response was strongly dependent on
both cell geometry and reinforcement content. Cylindrical lattices exhibited broader
and lower force peaks as the reinforcement percentage increased, suggesting improved
impact mitigation characteristics through progressive energy absorption. Hexagonal
lattices demonstrated longer impact durations and more stable deformation patterns,
making them more suitable for applications requiring controlled energy absorption
during impact loading.

Digital Image Correlation revealed that the deformation mechanisms di ered

considerably between the two lattice geometries. Cylindrical cells primarily
experienced compression-dominated deformation with localized strain concentrations
at the cell junctions, whereas hexagonal cells deformed mainly through bending of their
inclined struts, resulting in a more distributed strain field. During impact loading, strain
localization was significantly reduced because the higher strain rate produced a more
uniform deformation throughout the lattice structure.

Overall, the study concluded that structural performance is governed by the combined
e ects of unit cell geometry, reinforcement weight ratio, and strain rate. By
appropriately selecting these parameters, the mechanical response, sti ness, energy
absorption characteristics, and impact performance of additively manufactured
composite lattice structures can be tailored for specific engineering applications. The
research also demonstrated the potential of vat photopolymerization for producing
lightweight, reinforced lattice structures suitable for impact mitigation and other
structural applications.

59
Introduction to Nozzle Clogging in Fused Deposition Modelling (FDM)
Fused Deposition Modelling (FDM) is one of the most widely adopted additive
manufacturing technologies due to its simplicity, a ordability, and ability to fabricate
complex three-dimensional components directly from digital models. The process
involves feeding a continuous thermoplastic filament into a heated extruder, where the
material is melted and forced through a nozzle before being deposited layer by layer to
produce the desired component. The quality, dimensional accuracy, and mechanical
properties of the printed part are highly dependent on maintaining a stable and
continuous flow of molten material throughout the printing process.

One of the most common operational issues encountered during FDM printing is nozzle
clogging. Nozzle clogging occurs when the flow of molten polymer through the nozzle
becomes partially or completely restricted, resulting in reduced or interrupted material
extrusion. Depending on the severity of the obstruction, clogging may lead to under-
extrusion, inconsistent layer deposition, poor surface finish, dimensional inaccuracies,
weak interlayer bonding, or complete print failure. Since the nozzle is the final stage of
the extrusion system, even minor disturbances in material flow can significantly a ect
the overall printing performance.

The occurrence of nozzle clogging is influenced by numerous interacting factors rather

than a single cause. Parameters such as nozzle temperature, print speed, volumetric
flow rate, filament quality, moisture content, nozzle diameter, nozzle material, cooling
conditions, retraction settings, and prolonged heating during print pauses all contribute
to the likelihood of clog formation. In addition, thermal degradation of the polymer,
contamination within the filament, heat creep, and excessive extrusion pressure can
each initiate di erent clogging mechanisms. As a result, nozzle clogging should be
considered a multi-physics phenomenon involving heat transfer, polymer rheology,
material degradation, and fluid flow.

Polylactic Acid (PLA) is one of the most commonly used thermoplastics in FDM because
of its low printing temperature, minimal warping, and ease of processing. However,
despite these advantages, PLA remains susceptible to several forms of nozzle clogging
when printing conditions deviate from the recommended operating range. A brass
nozzle with a diameter of 0.4 mm is widely regarded as the standard configuration for
desktop FDM printers due to its excellent thermal conductivity and balance between
print quality and extrusion e iciency. Consequently, understanding clogging behaviour
under these commonly used conditions is of practical importance for both industrial
and academic applications.

The objective of this study is to investigate the mechanisms responsible for nozzle
clogging in FDM printing using PLA filament extruded through a 0.4 mm brass nozzle.
The study examines the influence of various process parameters on material flow,

60
identifies the physical mechanisms leading to partial and complete clogging, analyses
the e ects of clogging on extrusion characteristics and printed part quality, and
discusses methods for diagnosing, preventing, and recovering from nozzle blockage. By
understanding the relationship between printing parameters and clogging behaviour,
more reliable printing processes can be achieved while reducing material waste,
machine downtime, and failed prints.

Furthermore, the findings presented in this report provide a technical foundation for
future development of predictive monitoring systems and digital twin frameworks
capable of identifying potential clogging conditions before complete nozzle blockage
occurs.

2. Fundamentals of PLA Flow Through a 0.4 mm Brass Nozzle

In Fused Deposition Modelling (FDM), the printing process is based on the controlled
melting and extrusion of a thermoplastic filament through a heated nozzle. For this
study, the extrusion system consists of a standard 0.4 mm brass nozzle and Polylactic
Acid (PLA) filament. Understanding the normal flow behaviour of PLA through the nozzle
is essential before analysing the mechanisms that lead to nozzle clogging.

The extrusion process begins when the filament is driven into the hot end by the
extruder mechanism. The extruder motor applies a controlled force to push the solid
PLA filament through the heat break and into the heater block. Inside the heater block,
thermal energy supplied by the heating cartridge is transferred to the filament primarily
through conduction from the brass nozzle and surrounding metal components. As the
filament absorbs heat, it gradually transitions from a rigid solid to a softened state
before reaching a fully molten condition near the nozzle tip.

The temperature of the nozzle must remain within the recommended processing range
for PLA to ensure proper melting and continuous flow. PLA is typically printed between
190 °C and 220 °C, although the exact temperature depends on the filament
manufacturer, printing speed, and nozzle design. Within this temperature range, the
viscosity of molten PLA decreases significantly, allowing it to flow through the narrow
nozzle opening under pressure generated by the extruder.

The flow of molten PLA inside the nozzle is governed by the combined e ects of
temperature, pressure, and viscosity. Viscosity is a measure of a fluid's resistance to
flow and is one of the most important material properties influencing extrusion. As
nozzle temperature increases, the viscosity of PLA decreases, enabling smoother flow
and reducing the force required for extrusion. Conversely, lower temperatures increase

61
viscosity, requiring greater extrusion force and increasing the likelihood of flow
restriction.

The pressure required to extrude molten PLA is generated by the extruder motor pushing
the filament into the heated region. Since the nozzle outlet has a diameter of only 0.4
mm, a pressure di erence is created between the molten material inside the nozzle and
the surrounding atmosphere. This pressure forces the molten polymer through the
nozzle opening, producing a continuous filament that is deposited onto the build
platform or previously printed layer.

The volumetric flow rate is another important parameter governing extrusion behaviour.
It represents the volume of material passing through the nozzle per unit time and is
given by:

Q = A*v

where:

 (Q) = Volumetric flow rate (mm³/s)

 (A) = Cross-sectional area of the deposited extrusion (mm²)

 (v) = Print speed (mm/s)

The volumetric flow rate determines the amount of molten material that must be
supplied by the hot end during printing. If the required flow rate exceeds the melting
capacity of the heater block, portions of the filament may remain partially unmelted
before reaching the nozzle. This increases extrusion pressure and may eventually
contribute to partial or complete clogging.

The nozzle diameter also plays a significant role in determining extrusion behaviour. A
0.4 mm nozzle is widely used because it provides a balance between print resolution
and material flow. Smaller nozzles improve printing accuracy but require higher
extrusion pressure and are more susceptible to clogging. Larger nozzles allow higher
material flow with lower pressure but reduce printing resolution.

Brass is the most commonly used nozzle material for printing PLA due to its excellent
thermal conductivity. High thermal conductivity enables rapid and uniform heat transfer
from the heater block to the filament, promoting consistent melting throughout the
extrusion process. Brass is also relatively inexpensive and easy to machine, making it
suitable for general-purpose FDM printing. However, brass nozzles wear more rapidly
than hardened steel nozzles when printing abrasive composite filaments containing
carbon fibre, glass fibre, or metal particles. Since standard PLA is non-abrasive, brass
remains the preferred nozzle material for this application.

During normal extrusion, the molten PLA exits the nozzle as a continuous cylindrical
strand that is slightly compressed against the build surface or previous layer. The
62
deposited filament generally has a width slightly greater than the nozzle diameter due to
compression during deposition. For a 0.4 mm nozzle, typical extrusion widths range
from approximately 0.40 mm to 0.48 mm depending on slicer settings, while common
layer heights range from 0.12 mm to 0.28 mm. Maintaining consistent extrusion
geometry is essential for achieving accurate dimensions, strong interlayer bonding, and
high-quality surface finish.

The extrusion process therefore depends on maintaining a stable balance between

thermal energy input, extrusion pressure, material viscosity, and volumetric flow rate.
Any disturbance to one or more of these factors may alter the flow characteristics of
molten PLA, leading to unstable extrusion, partial blockage, or complete nozzle
clogging. Understanding these fundamental principles provides the basis for analysing
the various clogging mechanisms discussed in the following sections.

3. Classification of Nozzle Clogging Mechanisms

Nozzle clogging in Fused Deposition Modelling (FDM) is not caused by a single factor
but can occur through several di erent mechanisms. Each mechanism has a distinct
cause, mode of development, and e ect on the extrusion process. Identifying the type
of clogging is important because the appropriate corrective action depends on the
underlying cause. The major clogging mechanisms encountered while printing PLA
using a 0.4 mm brass nozzle are described below.

3.1 Thermal Clogging

Thermal clogging occurs when the nozzle temperature is lower than the recommended
printing temperature for PLA. Under these conditions, the filament does not melt
completely before reaching the nozzle orifice. The partially melted material has a higher
viscosity and requires greater extrusion pressure to flow through the nozzle. As pressure
increases, the flow becomes unstable, leading to under-extrusion and eventually
complete blockage if the condition persists.

3.2 Heat Creep Clogging

Heat creep occurs when heat from the heater block travels upward through the heat
break into the cold end of the hot end assembly. Instead of remaining solid until it
reaches the melting zone, the PLA softens prematurely inside the heat break. The
softened filament expands and sticks to the inner walls, increasing friction and
preventing smooth movement through the hot end. Heat creep is commonly caused by
inadequate cooling, a malfunctioning hot-end fan, or prolonged exposure to high nozzle
temperatures.

63
3.3 Contamination Clogging

Contamination clogging results from foreign particles entering the extrusion system.
Dust, dirt, burnt polymer residues, fragments of previously printed materials, or
impurities present in low-quality filament may accumulate inside the nozzle. Since the
nozzle opening is only 0.4 mm in diameter, even very small particles can partially or
completely obstruct the flow of molten PLA.

3.4 Polymer Degradation Clogging

PLA is a biodegradable thermoplastic that undergoes thermal degradation when

exposed to elevated temperatures for extended periods. During long print pauses or
excessive nozzle temperatures, the molten polymer may begin to decompose, forming
carbonised deposits inside the nozzle. These deposits gradually reduce the e ective
nozzle diameter and restrict material flow, eventually leading to clogging.

3.5 Flow-Induced Clogging

Flow-induced clogging occurs when the required volumetric flow rate exceeds the
melting capacity of the hot end. This commonly happens when printing at very high
speeds, large layer heights, or high extrusion widths. Because the filament passes
through the heater block too quickly, it may not absorb su icient heat to melt
completely. The resulting increase in extrusion pressure can cause inconsistent flow,
partial blockage, or complete nozzle clogging.

3.6 Mechanical Clogging

Mechanical clogging is associated with physical damage or wear of the nozzle and
extrusion system. Internal scratches, burrs, deformation of the nozzle opening, or
manufacturing defects may disturb the smooth flow of molten PLA. Although brass
nozzles provide excellent thermal conductivity, they gradually wear when used with
abrasive composite filaments, increasing the possibility of flow irregularities. For
standard PLA printing, however, mechanical clogging due to nozzle wear is relatively
uncommon.

3.7 Moisture-Related Clogging

Although moisture absorbed by PLA does not directly block the nozzle, it significantly
a ects extrusion behaviour. When wet filament is heated, absorbed water rapidly
vaporises into steam, producing bubbles within the molten polymer. These bubbles
create fluctuations in extrusion pressure, leading to inconsistent material flow, surface
defects, and an increased likelihood of partial blockage when combined with other
clogging mechanisms.

64
Summary of Nozzle Clogging Mechanisms

Clogging Mechanism Primary Cause Typical Symptoms

Under-extrusion, extruder
Thermal clogging Low nozzle temperature
clicking

Inadequate cooling of the hot Filament jam above the heat

Heat creep clogging
end break

Contamination Dust, debris, or foreign Sudden blockage, inconsistent

clogging particles extrusion

Polymer degradation Prolonged heating or Brown or black deposits,

clogging excessive temperature reduced flow

High extrusion pressure,

Flow-induced clogging Excessive volumetric flow rate
inconsistent extrusion

Restricted flow, irregular

Mechanical clogging Damaged or worn nozzle
extrusion

Moisture-related Bubbles, popping sounds,

Wet PLA filament
clogging unstable extrusion

Each of these clogging mechanisms develops under di erent operating conditions and
may produce similar symptoms during printing. Therefore, understanding the
characteristics of each mechanism is essential for accurately diagnosing nozzle
problems and selecting the appropriate preventive or corrective measures. The
following section discusses the various process parameters that influence the
occurrence of these clogging mechanisms.

4. Variables A ecting Nozzle Clogging

Nozzle clogging in FDM printing is influenced by several process parameters that a ect
the melting, flow, and extrusion of PLA through the nozzle. These variables alter the
temperature distribution, melt viscosity, extrusion pressure, and material behaviour
inside the hot end. Understanding the influence of each parameter is essential for
selecting appropriate printing conditions and minimizing the occurrence of nozzle
clogging.

4.1 Nozzle Temperature

65
Nozzle temperature is one of the most significant parameters a ecting extrusion
performance. PLA is typically printed within a temperature range of 190–220 °C. At
temperatures below the recommended range, the filament does not melt completely,
resulting in high melt viscosity and increased extrusion resistance. This requires greater
force from the extruder and may eventually lead to partial or complete clogging.

Conversely, excessively high temperatures reduce the viscosity of PLA, allowing it to

flow more easily. However, prolonged exposure to temperatures above the
recommended range may cause thermal degradation of the polymer, producing
carbonised deposits inside the nozzle that gradually restrict material flow.

4.2 Volumetric Flow Rate

The volumetric flow rate represents the volume of material extruded per unit time and
depends on print speed, layer height, and extrusion width. When the required flow rate
exceeds the melting capacity of the hot end, the filament cannot absorb su icient heat
before reaching the nozzle. This results in partially melted material, increased extrusion
pressure, and a greater risk of clogging.

4.3 Print Speed

Print speed directly a ects the amount of material that must be melted within a given
time. Higher print speeds require greater volumetric flow rates and reduce the time
available for heat transfer to the filament. Excessively high speeds may therefore lead to
incomplete melting and unstable extrusion, particularly when combined with low
nozzle temperatures.

4.4 Layer Height and Extrusion Width

Increasing the layer height or extrusion width increases the volume of material
deposited during printing. Consequently, the hot end must melt more filament within
the same time interval. If the extrusion demand exceeds the heating capacity of the
nozzle, extrusion pressure increases and the likelihood of clogging becomes greater.

4.5 Nozzle Diameter

The nozzle diameter determines the size of the flow passage available for molten PLA.
Smaller nozzles provide higher printing resolution but require greater extrusion pressure
because of the reduced flow area. They are therefore more susceptible to clogging by
contaminants, partially melted polymer, or degraded material. A 0.4 mm nozzle is
widely used because it provides a practical balance between print quality and extrusion
reliability.

4.6 Nozzle Material

The nozzle material influences heat transfer to the filament. Brass nozzles possess high
thermal conductivity, allowing e icient and uniform heating of PLA. This promotes
66
stable melting and consistent extrusion. Nozzles made from materials with lower
thermal conductivity may require slightly higher printing temperatures to achieve similar
extrusion performance.

4.7 Filament Quality

The quality of the filament has a significant influence on nozzle reliability. Filaments
with inconsistent diameter, embedded contaminants, manufacturing defects, or poor
winding may produce irregular extrusion and increase the probability of clog formation.
High-quality filaments with consistent dimensions provide more stable material flow.

4.8 Filament Moisture

PLA is hygroscopic and gradually absorbs moisture from the surrounding environment
during storage. When heated inside the nozzle, absorbed moisture rapidly vaporises
into steam, producing bubbles within the molten polymer. These bubbles disturb the
extrusion process, causing inconsistent flow, poor surface finish, and increasing the
possibility of partial blockage.

4.9 Cooling Performance

Proper cooling of the hot end is necessary to maintain a clear separation between the
melting zone and the cold end. Insu icient cooling allows heat to travel upward through
the heat break, causing premature softening of the filament. This phenomenon, known
as heat creep, can result in filament expansion and jamming before the material
reaches the nozzle.

4.10 Retraction Settings

Retraction is used to reduce stringing by pulling molten filament slightly away from the
nozzle during non-printing movements. Excessive retraction distances or high retraction
speeds may pull softened PLA into the cooler regions of the hot end, where it solidifies
and increases friction. Repeated occurrences may eventually cause extrusion blockage.

4.11 Print Pauses

During a print pause, the nozzle often remains at the printing temperature while molten
PLA remains inside the hot end. Prolonged exposure to elevated temperatures without
extrusion may lead to thermal degradation of the material, forming burnt residues that
reduce the e ective nozzle diameter. The longer the pause duration, the greater the
possibility of degradation and clog formation.

4.12 Environmental Conditions

Ambient conditions such as dust, humidity, and surrounding temperature also

influence nozzle performance. Dust particles may adhere to the filament before
entering the extruder, while high humidity increases moisture absorption by PLA.

67
Although these factors have a smaller influence than temperature or flow rate, they
contribute to long-term nozzle reliability.

Summary of Variables A ecting Nozzle Clogging

E ect on
Variable Reason
Clogging

Nozzle Temperature High A ects melting and polymer degradation

Determines material demand and extrusion

Volumetric Flow Rate High
pressure

Print Speed High Influences melting time and flow rate

Layer Height Medium Increases extrusion volume

Extrusion Width Medium Increases material flow requirement

Nozzle Diameter High Smaller diameters increase flow resistance

Nozzle Material Medium Influences heat transfer characteristics

Contaminants and diameter variation a ect

Filament Quality High
flow

Produces steam bubbles and unstable

Filament Moisture Medium
extrusion

Cooling Performance High Poor cooling causes heat creep

Excessive retraction may cause filament

Retraction Settings Medium
jams

Promotes PLA degradation during idle

Print Pause Duration Medium
periods

Environmental Dust and humidity indirectly increase

Low–Medium
Conditions clogging risk

The variables discussed above do not act independently. In most practical printing
situations, nozzle clogging results from the combined influence of multiple parameters.
For example, high print speed together with low nozzle temperature can significantly
increase extrusion pressure, while prolonged print pauses at elevated temperatures can
accelerate polymer degradation. Therefore, successful prevention of nozzle clogging

68
 requires maintaining an appropriate balance between all printing parameters rather
than optimizing a single variable.

Nozzle
<2 mm³/s 2–5 mm³/s 5–10 mm³/s 10–15 mm³/s >15 mm³/s
Temperature

Very poor Cold Immediate

<170 °C High clog risk Jam likely
melting extrusion failure

Acceptable Under-extrusion
170–185 °C Moderate High pressure Jam likely
for slow prints possible

Recommended Monitor May under-

190–210 °C Stable Stable
operating region extrusion extrude

Recommended
Slight Near hot-end
210–225 °C Stable Stable for high-speed
stringing limit
printing

Stable if Good flow but

Heat creep
225–240 °C Oozing Oozing continuously monitor
risk
printing degradation

Polymer
Reduced viscosity High
degradation Degradation Temporary
>240 °C but degradation degradation
during idle risk operation
increases and clog risk
periods

69
Markforged composite 3D printer working and SOP

Purpose: Standard workflow for preparing, slicing, reinforcing, and sending a print job
to the Markforged Mark Two desktop printer via Eiger.

1. File Upload and Initial Configuration

a) Export STL: CAD Environment.

Export your 3D model from your CAD software as a .STL file. Ensure units are correct
(typically millimetres).

b) Import to Eiger: Eiger Cloud Platform.

Log into your organization's account at [Link]. Click Import STL in the top right,
choose your file, and assign it to your project folder.

c) Configure General Settings: Right Panel.

In the Part View on the right panel, select your exact setup:

 Printer: Mark Two

 Base Material: Onyx, Nylon, etc.

 Reinforcement Material: Carbon Fiber, Fiberglass, HSHT Fiberglass, or Kevlar

(or 'None' if purely plastic).

2. Part Orientation and Slicing Basics

Before you touch individual layers, you need to set up the macro layout of the print.

70
  Orienting the Part: Use the orientation tools to lay your part flat on the optimal
face. You can click on the face of the part on which you need it to rest on

Rule of Thumb: Remember that continuous fibre routes horizontally along the XY plane.
Position the part so that the structural loads will pull along the length of the fibres, not
across them.

 Settings Panel (General tab):

o Layer Height: Usually 0.10 mm or 0.125 mm (depends on the selected

fibre type).

o Infill: Choose your pattern (Triangular is standard and strongest for Onyx)
and fill percentage (typically 37% for general use).

o Raised on Raft: Toggle this ON if your part has a delicate or complex

bottom surface, though printing directly on the bed is preferred for
smooth surface finishes.

3. Internal View: Fiber Reinforcement & Layer Editing

Once the general settings look good, click Save and then click Internal View (bottom
right) to open the layer-by-layer slice viewer. This is where the true power of the Mark
Two is utilized.

Reinforcement Types

You can click and drag on the 2D layer timeline bar at the bottom to select specific
layers or groups of layers ("groups") to apply reinforcement:

 Concentric Fill: The printer routes continuous fibre rings along the outer walls of
the part. This is best for outer edge impacts, side-load bending, and
reinforcing holes/walls. You can specify how many rings to place.

 Isotropic Fill: The printer mimics a solid sheet of metal by laying down parallel
sheets of fibre, alternating angles by 45° or 90° on subsequent layers. This is best
for resisting heavy twisting (torsion) or overall crushing forces.

 Fiber Angles: For Isotropic fill, you can edit the orientation angle (e.g., changing
from a standard 0°/90° weave to specialized angles to fight specific directional
stresses).

71
Adding/Removing Layers

 To strip fibre completely from a section, highlight those layers on the timeline,
open the reinforcement settings, and toggle Use Fiber to OFF.

 Always keep an eye on the Fiber Layers indicator at the bottom to make sure
you have enough "sandwich" strength (e.g., putting fibre at the very bottom and
top layers creates an incredibly strong I-beam e ect).

4. Final Review & Sending to Print

a) Verify Slicing Estimate: Bottom Status Bar.

Check the total print time, plastic volume, and fibre volume usage estimated at the
bottom of the screen. Ensure it fits your material availability.

b) Click 'Print': Bottom Right Corner.

Click the blue Print button. This takes you to the Build Device screen.

c) Select Printer and Position: Build Bed Layout.

Select your online Mark Two printer from the dropdown menu. Drag your part to the
desired location on the virtual build plate (it's best to rotate locations slightly over time
to prevent wearing down one specific spot on your print bed).

d) Physical Bed Prep: At the Machine.

Go to the physical Mark Two printer:

 Scrape o any old glue from the print bed.

 Apply a fresh, thin, even layer of Markforged glue stick to the printing zone.

 Ensure the bed is locked securely into place with its magnetic mounts.

e) Confirm Materials & Run: Eiger & Printer UI.

Confirm that the physical spools loaded into the dry box (and the fibre spool inside the
printer chassis) match what you selected in Eiger. Hit Print on your browser or confirm it
on the printer's touchscreen.

72
5. Post-Print Cleanup

1. Use the provided scraper tool carefully under the edges of the part to lift it o the
bed.

2. Wash the glue o the bottom of the part and the print bed using warm water (if
necessary)

73
Design of Scissor lift

Objectives
 Develop a scissor lift mechanism using SolidWorks.

 Create an assembly capable of opening and closing using a central screw

mechanism.

 Progress from external-link configuration to internal-link configuration.

Work carried out

 Developed the initial CAD model of a scissor lift with externally mounted links.

 Created assembly mates and motion to demonstrate lifting through a central

lead screw mechanism.

 Modified the design to an internally mounted link configuration.

 Produced multiple design iterations to improve mechanism arrangement and

movement.

 Performed preliminary design calculations for link dimensions and loading.

 Began structural design considering load capacity, link sizing, and geometric
parameters.

 Investigated the influence of design variables such as link dimensions, platform

height, and lifting load.

Challenges encountered
As the design progressed, it became evident that modelling alone was insu icient to
develop a practical lifting system. Determining the load capacity required the selection
of suitable design parameters including material, link dimensions, pin sizes, safety
factors, and joint configurations. Changes to one parameter frequently required
recalculation of several others, highlighting the iterative nature of mechanical design.
Although complete optimisation of the design was not achieved during the internship
period, the work provided valuable insight into systematic engineering design
methodology and the importance of establishing design requirements before detailed
CAD modelling.

74
Design of the Scissor lift

 Then after that a suitable mechanism was to be decided to operate the

scissor lift. So, at first the rake and pinion were used but then later found out
that it doesn’t have self-locking mechanism.

 Modelled a di erent variant for the lift, where the links fold inside the lift

 Then a central screw was used as the mechanism for the lift

 Had to face a lot of trial and error for the above models, like the dimensions
to be used, especially for the lead screw assembly since it was becoming
really di icult to make a thread inside the sliding cylinder. At the end an

75
 alternate solution was found to compensate for the intended way of
implementing the mechanism.

 After the models were done, calculations for the forces acting on the links,
pins, etc had to be done. So, a new model for carrying a 25Kg load to 1 meter
high had to be made.

 The calculations for the force needed to be withstood by the base is

calculated. The shear stress, crushing stress and the bending stress are
calculated for the pin. The calculations are done below

Calculations
(Assumptions/Requirements (PDS) = 25kg load to be raised to 1 meter height, SS 304
material used, 1 stage scissor lift with central lead screw mechanism)

Load mass (m) = 25Kg

Height = 1m

Material, SS 304 (E = 193 GPa, Yield strength ≈ 215 MPa)

• Maximum angle (θmax) = 55° [since it becomes unstable at 50° to 60°] (reference
taken)

Since its 1 stage, arm length, H = L*Sin (θ)

L=H/Sin (θ) => 1000mm/Sin (55°) ≈ 1220 mm

• Minimum angle (θmin) = 15°

Closed horizontal length, (Xclosed):

Xclosed = L*Cos (15°)

= 1220 * 0.9659 ≈ 1178 mm

 Length of base at least 1178 mm, so l = 1300 mm (length of base)

• Plate width = 400 mm (assumption)

load, W = 25 * g = 250 N

76
At minimum angle, FLink = W/2Sin (θ)
= 250 / 2Sin (15) ≈ 237 N
So to give an FOS of 2, we can give 237 * 2 ≈ 473 N (for one pair of links)

• Euler Buckling
PG = π2EI/(KL)2 (Where, K = 1)
then, link is 50mm wide, 6 mm thick (through assumptions and trial and error)

Wd = 490N (W*FOS)
this means that 473N is the least required force to be handled by the links (Individually)
to withstand the force. So

50mm x 6mm (width x thickness) can be taken since it will be 693N which is more than
490N.

Pin

Force on pin = FLink = 473N

Pin in shear, τ = F/A

A = 2(πd2/4) [double shear]

τallow = 0.3σy = 215 * 0.3 ≈ 65MPa

d = (2F/ π τallow)1/2 ≈ 2.152355 mm

dcalc ≈ 2.15 mm

Bearing (crushing) stress

σb = F/dt = 473/6*5 = 15.8 MPa (low compared to SS 304 yield strength)

Since 2.15 mm isn’t practical, 6mm diameter can be used too,

Pin length

The scissor lift consists of two scissor assemblies connected by common shafts. These

77
shafts pass through both assemblies and support the pivot as well as the sliding motion
of the mechanism.

Shaft Diameter

The shaft diameter was selected based on shear, bearing, and bending stress
calculations.

Selected shaft diameter:

d = 6 mm

Shaft Length

The shaft length was determined from the assembly dimensions.

Thickness of one link = 5 mm

Number of links = 4

Total link thickness = 4 × 5 = 20 mm

Spacing between scissor assemblies = 360 mm

Washer allowance = 3 mm

Locking allowance = 5 mm

Therefore,

Shaft Length = 20 + 360 + 3 + 5 = 388 mm

Nearest standard length selected: L = 390 mm

Screw

FScrew = W/Tan (θ), at θ = 15 (Very large force)

Fslider = W * Cot(θ) (Force balance on an ideal scissor lift, F * Tan (θ) = W)

FSlider ≈ 1830N

since σy ≈ 215 MPa, we can take σallow as 107.5 MPa (FOS = 2)

σ = F/A
A = 1830/107.5 = 17mm2

78
  d = ((4*17)/ π)1/2 ≈ 4.7 mm ≈ 5 mm

 This screw becomes tiny for the scissor lift, so an 8 mm screw can be used to
make it better, since that is the nearest standard

 So, for the 25 Kg lift, Tr8 x 2 or M8 lead screw can be used

[The minimum screw diameter is 5mm obtained from calculations. To account

for the thread reduction, wear and manufacturing tolerances and to provide
additional rigidity, 8 mm standard diameter screw was selected]

 If we want better sti ness and reduction possibility to buckling, while

maintaining compact direction, Tr 10 x 2 can be used.

Screw Length

LScrew ≈ Slider travel + end tolerance

Bottom joint spacing

At minimum height = L Cos (15°)
= 1220 (0.960) = 1178 mm
At maximum = 1220 * Cos (55°)

= 700 mm
therefore slider travel = 1178 – 700 ≈ 478 mm 500 mm
to adding allowance of 50 on both sides, 500 + 100 ≈ 600 mm

Nut length

Common rule for engagement length

Ln = 1.5d to 2d
screw, d = 10mm
Ln = 15 mm to 20 mm,
We can take 20 mm,

Therefore, the nut block dimension,

30 mm x 30 mm x 25 mm (Width x Height x Length)

79
Height of Lift

Hactual = H + tbase + ttop + o set

Base/top plate = 5 mm

Roller/pin o set = 20 mm total

Hminimum = L * Sin (15°) = 1220 * Sin (15°) = 316 mm

HMaximum = L * Sin (55°) = 1220 * Sin (55°) = 999 mm

Hactual min = 316 + 5 + 5 + 20 = 346 mm

Hactual max = 999 + 5 + 5 + 20 = 1029 mm

Plate Thickness

Maximum bending moment,

MMax = WL/4 = (490*1300)/4

MMax = 159250 Nmm

 Section modulus of rectangular plate,

Z = bt2/6 = 400t2/6

Therefore, Bending stress,

σ = M/Z (σallow = 107.5 MPa) (FOS=2)

107.5 = 159250/(400t2/6)

t = 4.71 mm ≈ 5 mm

80
Learning Outcomes
 Learned advanced SolidWorks part and assembly modelling.
 Gained practical experience with assembly constraints and motion.
 Understood the di erence between CAD modelling and engineering design.
 Learned that design begins with defining requirements before creating detailed
models.
 Experienced the iterative process of engineering design, where calculations,
modelling, and revisions are performed repeatedly until the design satisfies the
required criteria.
 Recognised the importance of considering material selection, joint design,
loading conditions, and manufacturability during the early stages of product
development.

81
Digital twin of a printer for material interaction study framework

About Pronterface
Pronterface is a lightweight desktop application used for direct, wired control of a 3D
printer. It is part of the Printrun software suite.

Why it is used

• Hardware testing: Moving individual axes to check motor health.

• Calibration: Adjusting steps-per-millimetre (E-steps) and tuning PID temperature

values.

• Troubleshooting: Sending raw G-code text commands and viewing error responses.

• O line printing: Running a 3D print directly from a computer without using an SD

card.

How it is used

• Tethered connection: Connect a computer to the 3D printer using a physical USB

cable.

• Control panel: Click on-screen buttons to heat nozzles, extrude plastic, or home
axes.

• G-code console: Type commands into a command line to modify internal printer
settings.

About OctoPrint
OctoPrint is a powerful web-based server application that turns a standard 3D printer
into a smart, network-connected device.

82
Why it is used

• Wireless management: Starting, pausing, and monitoring 3D prints over a home Wi-
Fi network.

• Remote viewing: Monitoring the printer using a connected webcam to watch for
failures.

• Automation: Creating automated time-lapse videos of the printing process.

• Extensibility: Adding third-party plugins for features like AI spaghetti detection or

custom user interfaces.

How it is used

• Dedicated server: Runs continuously on a dedicated device (like a Raspberry Pi)

plugged into the printer.

• Web dashboard: Access the control panel by typing the printer's local IP address into
any phone or computer browser.

• File dropping: Upload sliced G-code files directly into the web browser to start a print
instantly.

The above information about OctoPrint and Pronterface has been found out and plans
regarding which to use first was being decided.

The following plan was thought to be executed to check the various levels in which data
can be acquired from the Printer

1) Using Pronterface, we can find the actual temperature of nozzle and bed. It
was considered the first step because it doesn’t have to do with interfering
with the firmware of the Printer

2) OctoPrint was thought to be used the 2nd time since it gave a little more data
about the actual temperature of nozzle and bed, print speed can be
controlled and found out, layer time can also be calculated.
Extrusion rate, even though it can be calculated mathematically after each
layer, can be found out using cameras and sensors for better accuracy.
for all this Raspberry Pi is recommended for use. The laptop can also be
used but may lead to reboot issues. The firmware isn’t a ected at all in these
steps

83
 3) Klipper Firmware can be installed which can help in getting the data even
more accurately and can easily help in maintaining the parameters and
decreasing the print time a lot. This may take time to setup and may lead to a
lot of errors if not done properly.

After Pronterface installation

Initial e orts were spent installing Pronterface to establish direct communication with
the 3D printer. However, after researching OctoPrint setups online, I determined that
OctoPrint o ers a more streamlined, centralized method for data collection. And since
a Raspberry Pi is available, I pivoted the workflow to utilize OctoPrint directly, rendering
Pronterface redundant and simplifying the overall system architecture.

Setup of Octoprint
Hardware setup

• Raspberry Pi (model 3 and above) (alternatively a laptop with Windows or Linux)

• USB Power supply

• If strong 5V printer, hotwire into a USB cable. (Make sure to shut down the Raspberry Pi
cleanly in OctoPrint before powering o printer)

• USB cable

• SD card (and card reader)

• Laptop

• Webcam (Optional) (Either USB webcam or Raspberry Pi camera modul

Set Up a Local Virtual Printer Environment

• Opened OctoPrint on the laptop using [Link] Because it runs locally

(localhost), it stays entirely on the computer's memory and bypasses any campus Wi-Fi
security issues.

• Connected OctoPrint to an imaginary serial port named VIRTUAL to emulate the

Anycubic Kobra 2 Neo without needing the physical machine plugged in.

• Sent a test command (M105) and verified that the virtual machine responds with live
temperature data.

84
Digital twin of a printer for material interaction study framework

Objective
To make a digital twin framework of an FDM process (concentrating only on the nozzle
and the bed part) to study the material interaction between the layers of the additive
manufacturing process.

Work carried out

1) Setting up of Octoprint for Data acquisition.

This was done previously and tested with a virtual printer (refer previous section).
The Octoprint software allows the connection of a 3D printer with any device
(computers, Raspberry pi and even mobile phones) and using the plugins that
are available in the Octoprint website. The printer can be directly connected to a
laptop and using plugins like “DisplayLayerProgress” and other plugins, the
display shows that much amount of information during a particular print
process. Here, since there is a lack of printer, a virtual printer setup is made

Octoprint interface

After the virtual printer is setup, the python code for extracting/compiling data is
made. The software provides an API to get information of the print process. The
API used here is REST API. Then using command prompt, the code is run during
the printing of the virtual printer. The progress of the print is displayed in the
command prompt window and the layers can be seen being completed.

85
 2) The Python Code

The python code is given below for extracting the data from the Octoprint
software.

Python [Link]

After the python code is set up, make sure the API key and the Octoprint URL are
typed into the code properly. This python code will make an excel sheet where
the following data will be store on each row of the sheet.

1) Timestamp (from the device time)

2) Elapsed time (in seconds, elapsed from the time the code is executed)
3) Job state (if the printer is printing, paused or idle)
4) Layer Completion (How much of the print is complete in percentage)
5) Current layer
6) Total layers
7) Current Z height (height of the nozzle above the bed in mm)
8) Nozzle temperature
9) Bed Temperature
10) Target nozzle temperature
11) Target Bed Temperature
12) Thermal gradient (Nozzle Temperature – Bed Temperature)(how much
thermal stress the part is under. High value = large temperature di erence
between top and bottom of print)
13) Layer elapsed
14) Previous layer duration
15) Fusion Factor

The fusion factor estimates the quality of bonding between two successive FDM
layers as a percentage (0–100%), based on whether the layer interface retains
su icient thermal energy for polymer chain interdi usion (reptation) to occur
before cooling below the glass transition temperature. It is calculated as:

𝑇 − 𝑇 𝑇 − 𝑇
𝐹𝑢𝑠𝑖𝑜𝑛 (%) = ∗ ∗ 100
𝑇 − 𝑇 𝑇 − 𝑇

86
 where 𝑇 is the estimated interface temperature at deposition, 𝑇 is the
interface temperature when the following layer arrives (after exponential cooling
over one layer duration), and 𝑇 is the glass transition temperature of the
material. A fusion factor above 80% indicates strong bonding conditions; values
below 60% indicate delamination risk.

16) Interface Temperature

It's the temperature of the surface of the previous layer; at the exact moment the
new layer is being deposited on top of it. The actual contact point between old
and new material. This is the temperature that decides whether bonding
happens or not.

Tᵢ = Tbed + (Tnozzle − Tbed) × e^ (−α × z) (α: Decay constant (0.05mm-1))

17) Cooling Rate

How fast the part is losing heat per second

Rc = (Tnozzle – Tbed) / layer duration

18) Z Thermal Gradient

How much the temperature changes for every 1mm going up in the print.

∇T = (Tnozzle − Tbed) / zheight

After this, another python code is made to visualize the layers from this raw data
which is still in a work in progress.

Python Code for

[Link]

87
 Octoprint interface during virtual 3D printing

Cmd Window during printing

88
 Excel sheet after data acquisition

The visualisation code after data acquisition (Work-in-progress)

Digital Twin Framework

The data acquisition system developed using OctoPrint and Python serves as the
foundation for the proposed digital twin framework. The objective of the
framework is to create a virtual representation of a simplified Fused Deposition
Modelling (FDM) process to study the thermal interaction occurring between
successive deposited layers.
Rather than modelling the complete 3D printer, the framework focuses only on
the critical region where material interaction takes place, namely the heated
build plate, the extrusion nozzle, and the first three deposited layers. This
simplified approach allows the thermal behaviour at the layer interfaces to be
analysed while reducing the complexity of the overall system.
The proposed framework consists of the following stages:

89
 1. Physical FDM Process
A simplified physical model consisting of the heated bed, nozzle and three
deposited layers is considered. During printing, molten thermoplastic material is
extruded through the nozzle and deposited layer-by-layer onto the heated bed.
Each newly deposited layer transfers heat to the previously deposited layer,
resulting in reheating, cooling and inter-layer bonding.
2. Data Acquisition
Real-time process information is collected from OctoPrint through its REST API.
The acquired parameters include nozzle temperature, bed temperature, current
layer number, Z-height, layer completion, print state, layer duration and
timestamps. The collected data are stored automatically in an Excel file using
Python for further processing.
3. Thermal Parameter Estimation
Using the acquired process data, thermal parameters governing inter-layer
bonding are calculated. These include interface temperature, cooling rate,
thermal gradient and fusion factor. These parameters provide an estimate of the
thermal condition at the interface between two successive layers during
deposition.
4. Material Interaction Analysis
During deposition, the hot extruded filament comes into contact with the
previously deposited layer. Heat is transferred across the interface, allowing
polymer chains to di use between adjacent layers before cooling occurs. The
quality of this interaction determines the strength of the bond between
successive layers. The calculated thermal parameters are used to estimate
these material interactions throughout the printing process.
5. Digital Twin Output
The processed thermal data provide a virtual representation of the deposition
process. The framework is intended to predict the thermal state of each
deposited layer, estimate inter-layer bonding conditions and identify regions that
may experience weak bonding or delamination due to insu icient thermal
energy.

Physical Model
To support the proposed framework, a simplified physical model of the FDM
deposition process is proposed. The model consists of the following
components:
 Heated build plate
 Standard FDM extrusion nozzle
 Three deposited material layers
The model represents only the region immediately surrounding the deposition
process, as this is where thermal interaction between layers occurs.

90
 Representative dimensions based on a standard FDM printer are adopted for the
conceptual model.

Component Representative Dimension

Nozzle Diameter 0.4 mm

Layer Height 0.2 mm

Extrusion Width 0.4 mm

Layer Length 40 mm

Heated Bed Representative rectangular platform

Material Interaction During Deposition

The principal objective of the proposed framework is to study the thermal
interaction occurring between successive deposited layers during the FDM
process.
When molten material exits the nozzle, it is deposited onto the surface of the
previous layer. Since the newly deposited material is at a much higher
temperature than the underlying layer, heat is transferred across the interface.
This reheats the surface of the previous layer, allowing polymer chains from both
layers to di use across the interface. As the material cools, the bond between
the layers is established.
The quality of this bond depends on several process parameters, including
nozzle temperature, bed temperature, layer height, print speed and cooling time.
If the interface temperature remains su iciently high for an adequate duration,
stronger inter-layer bonding is achieved. Conversely, rapid cooling or insu icient
interface temperature may result in weak bonding and possible delamination.
The proposed digital twin framework estimates these thermal conditions using
real-time printer data, enabling the material interaction during deposition to be
analysed without direct measurement.

Challenges and Limitations

The current work represents the initial stage of the proposed digital twin
framework and is subject to the following limitations.

Virtual Printer Environment

The framework was developed and tested using OctoPrint's virtual printer due to
the unavailability of a physical FDM printer. Although suitable for developing the

91
 data acquisition system, the virtual printer does not accurately represent the
thermal behaviour of an actual printing process.

Unrealistic Layer Timing

The virtual printer completed several layers within a few seconds, resulting in
unrealistic layer durations. Since thermal parameters such as interface
temperature, cooling rate and fusion factor depend on layer time, the calculated
values are intended only to demonstrate the methodology rather than represent
actual printing conditions.

Lack of Physical Validation

The absence of a physical printer prevented experimental validation of the
calculated thermal parameters.
Absence of Additional Sensors
The current framework relies solely on process data obtained through the
OctoPrint API. Thermal cameras, thermocouples or machine vision systems
were not available to directly measure interface temperatures or monitor layer
formation during deposition.

7. Future Work
Future work will focus on completing the proposed digital twin framework by:
 Developing the complete CAD representation of the nozzle, heated bed and
deposited layers.
 Completing the Python-based visualization of layer-wise thermal behaviour.
 Integrating the physical model with the thermal calculations developed in this
work.
 Validating the framework using an actual FDM printer.
 Incorporating thermal sensing and machine vision for real-time monitoring of the
deposition process.
 Comparing experimentally measured temperatures with predicted values to
improve the accuracy of the thermal model.
 Extending the framework to predict print quality and inter-layer bond strength
under di erent printing conditions.

92
Conclusion
The internship provided valuable practical exposure to advanced manufacturing
technologies and their industrial applications. It strengthened the understanding of
additive manufacturing processes, CNC machining, robotic automation, and composite
manufacturing through both practical demonstrations and technical studies.

Working with di erent 3D printing systems, RobotStudio, Wire EDM, and WAAM
provided insight into the complete workflow involved in manufacturing, from design and
process planning to machine operation and quality considerations. The literature
studies further enhanced the understanding of current research topics related to
composite materials, lattice structures, and process defects in additive manufacturing.

The design exercise involving the scissor lift improved practical CAD modelling and
engineering problem-solving skills, while the development of a digital twin framework
introduced the concepts of process monitoring, thermal behaviour, and predictive
manufacturing. Overall, the internship bridged theoretical knowledge gained during
academic studies with real industrial and research practices, providing a strong
foundation for future work in mechanical engineering and additive manufacturing.

93
References

1) Kliment. (n.g.). Printrun: Pure Python 3D printing host software (Pronterface).

[Link]

2) Markforged, Inc. (2026). Markforged design guide for 3D printing with composites
(Version 1.7). [Link]

3) OctoPrint. (n.g.). [Link] - The snappy web interface for your 3D printer.
[Link]

4) Ben Amor, S., Elloumi, N., Eltaief, A., Louhichi, B., Alrasheedi, N. H., & Seibi, A.
(2024). Digital Twin Implementation in Additive Manufacturing: A Comprehensive
Review. Processes, 12(6), 1062. [Link]

5) "A Study on the Static Stability of Scissor Lifts," Journal of Mechanical Engineering
Research, vol. 42, no. 3, pp. 115-122, 2015

6) Saxena, A. (2016). Deriving a Generalized, Actuator Position-Independent Expression

for the Force Output of a Scissor Lift. arXiv. [Link]

7) Spackman, H. M. (1990). Mathematical Analysis of Scissor Lifts (Technical Document

1550). Naval Ocean Systems Center.

8) ASTM D7264/D7264M-21. (2021). Standard Test Method for Flexural Properties of

Polymer Matrix Composite Materials. ASTM International.
[Link]

9) ASTM D790-25. (2025). Standard Test Methods for Flexural Properties of Unreinforced
and Reinforced Plastics and Electrical Insulating Materials. ASTM International.
[Link]

10) ASTM D3039/D3039M-00. (2002). Standard Test Method for Tensile Properties of
Polymer Matrix Composite Materials. ASTM International.
[Link]

94