Entry level 3D printers

Entry level 3D printers
Entry level 3D printers
Usage and Benchmarking
Bachelor’s thesis
Kone- ja tuotantotekniikka (Mechanical Engineering)
Riihimäki, 2014
Petri Heino
ABSTRACT
HAMK Riihimäki
Kone- ja tuotantotekniikka (Mechanical Engineering)
Author
Petri Heino
Subject of Bachelor’s thesis
Entry level 3D Printers
Year 2014
ABSTRACT
This thesis was made as a primer for getting more involved and specialized in 3D printing. The subject was proposed and the opportunity was
provided by Aalto University's Digital Design Laboratory (ADD) located
in the Otaniemi campus in Espoo city, in Finland.
Ten different printers implementing four different 3D printing methods are
introduced, with the machine prices ranging from 2000€ to 35000€.
The writing is based on two months of intensive use and testing. The aim
was to benchmark the printers and learn how to use them. A lot of new
knowledge was acquired and the differences between a 2000€ printer and
a 12000€ one became apparent.
A well established base of knowledge was established for considerations
of small-series manufacturing with the printers and their correct use and
maintenance.
Keywords
3D printers, FDM, FFF, SLA, SL, Stereolithography, Polyjet, inkjet.
3DP, powder printing, lamination
Pages
61 p.
TIIVISTELMÄ
HAMK Riihimäki
Kone- ja tuotantotekniikka (Mechanical Engineering)
Tekijä
Petri Heino
Työn nimi
Entry level 3D Printers
Vuosi 2014
TIIVISTELMÄ
Tämä päättötyö on tehty alustaksi erikoistumiselle 3D tulostimiin ja -menetelmiin. Aalto-yliopiston Digital Design Laboratory (ADD) ehdotti työn
aihetta sekä tarjosi tilaisuuden ja mahdollisuuden tehdä työtä. Digital
Design Laboratory sijaitsee Otaniemen kampuksella Espoossa.
10 erilaista tulostinta jotka käyttävät 4 erilaista 3D tulostus -menetelmää
esitellään. Laitteiden hinnat ovat välillä 2000 – 35000€.
Kirjoitus perustuu kahden kuukauden intensiiviseen käyttökokemukseen ja
testaamiseen.Tavoitteena oli benchmarkata tulostimet ja oppia käyttämään
niitä. Työssä kertyi paljon uutta tietoa ja erot 2000€ ja 35000€ laitteen välillä tulivat varsin selviksi.
Työssä kehittyi hyvä tietopohja ja valmiuksia piensarjatuotantoa varten tulostimia käyttäen. Lisäksi tulostimien oikea käyttö ja huoltaminen nousivat selkeästi esiin työtä tehdessä.
Avainsanat 3D tulostus, Stereolithografia, laminointi, pulveritulostus, pikavalmistus
Sivut
61 s
CONTENTS
1
INTRODUCTION......................................................................................................1
1.1 3D Printing nowadays.........................................................................................2
1.2 Terminology and technology...............................................................................3
1.2.1 Extrusion-based............................................................................................4
1.2.2 Photopolymerization-based..........................................................................5
1.2.3 Inkjetting-based............................................................................................6
1.2.4 Sheet lamination processes..........................................................................7
2
MATERIALS..............................................................................................................8
2.1 Wire filament materials.......................................................................................8
2.1.1 Build platforms for filament materials.......................................................14
2.2 Powder materials...............................................................................................15
2.3 Printable waxes.................................................................................................15
2.4 Liquid materials.................................................................................................15
3
THE PRINTERS......................................................................................................16
3.1 Extrusion-based printers....................................................................................17
3.1.1 Ultimaker 2.................................................................................................17
3.1.2 3D Touch....................................................................................................21
3.1.3 Gigabot.......................................................................................................22
3.1.4 UPrint SE Plus............................................................................................24
3.2 Stereolithography / Form 1+.............................................................................27
3.3 Inkjet printers....................................................................................................31
3.3.1 Objet 30 Scholar.........................................................................................31
3.3.2 Thermojet...................................................................................................35
3.3.3 ZPrinter 250 ..............................................................................................37
3.4 Sheet Lamination / Mcor Matrix 300................................................................39
4
BENCHMARKING THE PRINTERS.....................................................................42
4.1 Test prints..........................................................................................................43
4.1.1 Ultimaker 2.................................................................................................45
4.1.2 UPrint SE Plus............................................................................................51
4.1.3 Form 1 Plus................................................................................................53
4.1.4 Objet 30 Scholar.........................................................................................56
5
CONCLUSIONS......................................................................................................58
6
SOURCES................................................................................................................59
Appendix 1 Printer specifications
Appendix 2 Test round-up
Entry level 3D printers
1
INTRODUCTION
3D printing is a manufacturing technique that has also been
called additive manufacturing and rapid prototyping. It is a
method of manufacturing parts by adding material to build up a
work piece incrementally, conversely to removing material from
a blank as in machining. Making parts this way has several obvious advantages, such as saving material and reducing the need of
patterns for molds.
Aalto University's Digital Design Laboratory (ADD) offered a
chance for learning how to use their machines and to get to know
their capabilities. ADD laboratory had 10 different 3D printers
available at the time of writing (2014).
Besides believing that 3D printers will get much more popular in
the future as a manufacturing method, motivations for starting
this work were to gain knowledge how to coordinate a 3D printing shop/factory, would the printers be good for commercial
small scale manufacturing, and what are the strengths, weaknesses and best applications for the printers available. Also
learning how the different machines were constructed would
provide a good chance to learn machine design perspectives.
Before starting this thesis I had been interested in all kinds of additive manufacturing methods, but had only limited experience
of using them, limited to RepRap building projects and using
and improving a Minifactory 2 at my previous workplace. Therefore getting my hands on machines worth several thousand or
tens of thousands of Euros was a unique and highly interesting
opportunity for me.
1
Entry level 3D printers
1.1
3D Printing nowadays
Extrusion-based desktop 3D-printer are getting widely popular,
due to increasing visibility of the printers, but still industrial usage is dominant. These industrial machines are made by large
and professional companies, from which the biggest two are
Stratasys and 3D Systems. For the cheapest printers there are
many big and small companies making them around the world.
The onset of popularization of the extrusion-based 3D printers
was mostly set by expiration of crucial patents and the very successful open-source RepRap project. Also, freely distributed
designs and building instructions play a huge role with services
such as Thingiverse (www.thingiverse.com)
The open-source hardware and software on the internet is commonly distributed under either the GNU General Public License
(GPL) or Creative Commons (CC)-licences. The CC-licences
may have limitations to commercial use set by the publisher, the
GNU GPL licence is a good example of a licence set to try to ensure the work to remain in public domain and not hidden under
commercial use only. These licences have representatives in
many countries and have even been successfully used in court.
(GNU GPL, 2007, GPL Violations, n.d. See also Creative Commons Case Study, n.d.).
Huge advancements have been made lately in the methods where
a laser beam or an electron beam is used for joining powder metal together appropriately so that a work piece is formed, such as
Selective Laser Sintering (SLS) or Direct Metal Laser Melting
(DMLM) and are becoming competitors for traditional production methods with complicated designs especially in the
aerospace industry. Recently in the 2014, General Electronics
Aviation division established a 50 Million Dollar additive manufacturing facility in USA. (GE Press Release, 2014.)
There is a wide range of different additive manufacturing technologies besides the ones mentioned in this thesis, but researching all of them is out of the scope of this thesis and the technologies will be reviewed only briefly. Interesting methods would be,
for example, direct write methods including plasma ionization or
deposition of powder metal through a nozzle that melts it at the
same time with a laser beam. The reader is suggested to research
them online, or to acquire for example, the book Additive Manufacturing Technologies, 2010, Gibson, I. Rosen, D. Stucker, B.
E-ISBN: 978-1-4419-1120-9.
2
Entry level 3D printers
1.2
Terminology and technology
Several different abbreviations are commonly used in 3D printer
technology, some of which are reviewed here along with descriptions of the technology. (Gibson et al. 2010. See also Materialise,
n.d.)
As a general description, initially the part being manufactured is
modelled in a CAD program, a 3D-model is made, and then this
file is processed so that information for creating the part is made
for the printer.
3D models are universally, somewhat standardized, transferred
to the 3D printer software in a .stl format. In this format, the
model is described in vertexes, triangles, that close the surface of
the model. This format is supported by all the programs involved
in this thesis and should be so in any programs involved in 3D
printers.
This .stl file is “sliced” into multiple layers, describing the
toolpaths for the printers. The slicing generates standard g-code
or a similar language, that is then sent to the printer's electronics,
which control the different axis's and material deposition methods. The g-code is a standardized machine control language consisting of lines of commands, position coordinates, and settings
for the machine to read and operate accordingly.
3
Entry level 3D printers
1.2.1
Extrusion-based
FDM = Fused Deposition Modeling, a trademark by the Stratasys company. PJP, Plastic Jet Printing, a trademark by the 3D
Systems company. May also be called FFF, Fused Filament Fabrication or just plainly extrusion-based systems.
Figure 1
2010, 145).
Schematic describing extrusion-based systems (Gibson et al.
Molten thermoplastic material is liquefied and then forced
through a nozzle, building up layers one after another. Usually
thermoplastics don't have a certain melting point and therefore
the usable temperature range can be high even for the same material. Higher temperatures may be necessary if printed at higher
speed, but it is important to remember that plastics can thermally
decompose if left at elevated temperature for long times.
The material may be deposited either in filament form with a
pinch wheel extrusion system or with a feed screw in granule
form. The parts are usually built hollow with an infill pattern.
4
Entry level 3D printers
1.2.2
Photopolymerization-based
In photopolymerization processes a thermoset plastic resin is
used (thermoset means that the plastic is cross-linked and “rigid”
after setting down and can not be thermally reformed as the
polymer chains arranged so). The resin is delivered in liquid
form and deposited differently in each method.
All of the machines implementing this method subject to this
thesis use ultraviolet light (UV, wavelengths ranging from approximately 300 to 400nm) to cure the resin. Resin also exist
that are curable with light in the visible spectrum used in different printers.
In Stereolithography (abbreviated SL, or SLA) the resin is in a
tank and the part is built inside the tank, touching the liquid resin.
Figure 2
Schematic of Stereolithography (Gibson et al. 2010, 65).
The SL system may work opposite to the schematic in Figure 2,
requiring less resin to be available, but having disadvantages
such as forming adhesive forces between the tank and the build
platform. In the Form 1 -printer, as an example, the build platform is lowered into the resin. The bottom of the tank is silicon
coated plastic, and a laser beam, moved through mirrors inside
the machine, is directed through the bottom of the tank.
A scanning mask could be used instead of a scanning beam, as in
some visible light -curing SL printers where a common household DLP-projector could be used (the LCD-screen blocking
light the dark places).
5
Entry level 3D printers
1.2.3
Inkjetting-based
The MJM (Multi-jet Modeling, an inkjet-printing method trademark by 3D Systems) and the PolyJet (inkjet-3D-printing method trademark by Stratasys) brand names are both photopolymerizing- as well as inkjetting-based methods.
The polymer resin is jetted (pushed by piezo actuators) through
small nozzles on to the build area. There may be multiple sets of
nozzles in rows of many, as over 100 in the Objet printers, for
the ability to deposit different materials. The resin may be heated
up to reduce viscosity (make it flow more easily).
A very thin layer of material is deposited with the inkjets and
then a UV light is moved over the resin to cure it. The next layer
is printed directly on top of the last one.
3DP is a method that implement powder and binder material to
build parts. Note that the term may also relate generally to any
3D printing method. (Gibson et al. 2010, 195-200. See also
Wikipedia 3DP, n.d.)
The Zcorp machine uses a method of inkjetting a binder material
(viscous fluid) in to a powder. The binder adheres adjacent
powder particles together. A new layer of powder is deposited
evenly with a spreader, that may be knife-like or a roller. Unused
powder may be recycled and used to print again.
Figure 3
Schematic of 3DP process. Note that with the photopolymer
inkjet-printers, the part is built straight to the build platform and support
material is also jetted and there is no powder used anywhere (Gibson et al.
2010, 196.)
6
Entry level 3D printers
1.2.4
Sheet lamination processes
There are several different kinds of sheet lamination processes
with different brand names associated with them.
For example, the Mcor Matrix-printer uses a process named Selective Deposition Lamination (SDL) , where a titanium knife is
used to cut a single layer of ordinary A4-paper that has been
glued on top of a previously bonded and cut sheet of paper.
The laminating processes are separated in two ways. The first
way would be to first bond the laminated sheet and then form it.
Conversely, the sheet can be first formed to the wanted shape
and then bonded to the previous layer.
Bonding means that a sheet is glued, thermally adhered, or, for
example, ultrasonically bonded (called Ultrasonic Consolidation)
to the previous one.
Forming means that the sheet is cut to the desired shape with a
knife, a laser beam, by milling it or by other methods.
The bonded sheets may be almost any material, ranging from paper to ceramics to sheet metals. (Gibson et al. 2010, 207-210.)
Figure 4
Schematic of one of the first commercialized lamination
processes, the Layer Object Manufacturing (LOM) -process, (Gibson et al.
2010, 208.)
7
Entry level 3D printers
2
MATERIALS
The used materials depend on the printing process as seen in the
previous chapter.
Notable about the materials is that the price can vary a lot. The
used photopolymerizing resins have been getting considerably
cheaper as the user-base is getting wider and with more printer
manufacturers entering the market.
This can also be seen on the wire filaments. The industrial-grade
materials can be very close in chemistry to the ones used in the
cheaper printers. However, the price difference can still be ten
times. It is also good to remember that neither of the two stand
up to comparison if compared to the price of thermoplastics used
in the injection molding industry, where the material cost can be
much lower, in the range of only a few Euros.
2.1
Wire filament materials
The industrial grade spools come in sealed packages and include
moisture absorbent material. The materials used in the uPrint
machines are called ABS Plus with different brand names that
are extruded at 300ºC according to the machine interface. The
special spool holder type that they use, plus the included circuit
board in the spools makes the use of custom filaments impossible or at least very difficult. The uPrint has fixed settings
for extrusion temperatures.
For lower price range printers there is a huge variety of different
materials available. Currently there is no standard for the spool
sizes and this causes problems with interchangeability between
the machines. The spool format is manufacturer dependent.
8
Entry level 3D printers
The specific chemistry and properties of each plastic type can
differ greatly depending on the grade and the additives (such as
plasticizers, colorants, fillers etc.). Very different properties have
been experienced depending on the filament color for example.
Figure 5
Filament spool fitted on the back of Ultimaker 2.
9
Entry level 3D printers
ABS (copolymer of Acrylonitrile, Butadiene, and Styrene).
Extruded usually at temperatures ranging from 220ºC to 260ºC.
The uPrint machines use a grade of ABS that is extruded at
300ºC. ABS is subject to even violent heat shrinkages and warping during printing and therefore it is very demanding on the
build platform. Printed parts can tolerate higher temperatures
than PLA, as the glass transition temperatures for ABS are higher.
For example MatWeb has over 2400 different grades of ABS listed on their website. There is no single universal form of ABS
that you can get, but it's properties are somewhat manufacturer
and grade dependent. (MatWeb ABS, n.d.)
PLA (Polyactic Acid, Biopolymer). Extruded at 170ºC to
220ºC, however some grades may need a higher temperature of
245ºC to flow properly. Notable properties of PLA are as follows:
•
easier to print than ABS because of less warping due to
lower extrusion temperature and it is less demanding on the build
platform. Biodegradable material. Harder surface than ABS. Extruder usually need less force on the idler pulley because of the
hardness of the plastic to avoid digging in to the filament. (MatWeb PLA, n.d.)
•
may degrade quite fast depending on the grade if the
spool is opened from packaging and left unattended. The material becomes very hard and fragile and impossible to print.
•
may be used as support material for ABS. The supports
can be removed by immersing the part in hot water for 48 hours
or by using an ultrasonic washing tank with a solution of caustic
soda for 3 hours. (3D Touch Ultrasonic support removal tank,
n.d.)
10
Entry level 3D printers
Wood filament. These filaments are mixtures of some ground
wood material and PLA or some other plastic. Extruder blockages may be experienced with smaller nozzle sizes. Spools of
wood filament were printed with extruder nozzle diameter
0,6mm without blockages or misfeeds. Material properties are
similar to PLA, but the objects are much softer and weaker.
Wood filament requires adjustments to the retraction settings on
the printers to avoid getting leftovers between layer changes – it
flows quite differently from PLA or ABS.
With the wood filaments, changing extrusion temperature has a
minor effect on the color of the resulting prints.
Figure 6
3D-printed owl in wood filament, Thingiverse thing no. 18218.
11
Entry level 3D printers
TPE (Thermoplastic Elastomer), Ninjaflex, Soft PLA.
Extruded at 180ºC to 230ºC. Flexible, rubber-like materials of
different varieties. Extruder has to be more rigid in construction
for these filaments because the materials are very flexible. The
extruder idler pressure also has to be adjusted correctly to be
holding it under just the right pressure to avoid squashing the filament.
Because of lower melting point and flexibility, printing settings
have to be adjusted correctly. This has to be almost certainly
done experimentally. When bridging (printing between perimeters, on top of nothing), a build area fan may prove necessary to
reach acceptable details. (MatterHackers, n.d.)
Experiments to print a red soft PLA material were made. A
150mm diameter ball, intended to be used as a corned guard, was
printed with 5% and 10% honeycomb infill. The resultant balls
were not as flexible as expected because of the honeycomb
structure making it rigid on the edges. However, if printing only
the perimeters (the outlines) of objects, much more flexible objects could be made.
Figure 7
Ball printed in flexible PLA.
12
Entry level 3D printers
PVA (Polyvinyl Alcohol). Water soluble. Extruded at 160ºC to
170ºC. Will absorb moisture once opened and may require drying before usage.
Described to be used as a support material and supposedly will
work best with PLA because it's extruded at lower temperatures
– ABS is extruded at 260ºC and may cause degradation if printed on PVA. (Bilbycnc PVA Fact Sheet, n.d.)
Nylon (Polyamide). Most Nylons are extruded at 240ºC to
250ºC. Nylon will most certainly require removing moisture before printing to avoid air or water bubbles. Flexible and durable
material.
Printing on blue tape or a wooden surface is recommended by
the manufacturer. (Taulman 3D, n.d. See also Reprap Wiki Polyamide, n.d.)
HIPS (High Intensity Polystyrene). The HIPS plastic is soluble
in Limonene (an organic solvent), making it a possible support
material for ABS. (RepRap Wiki HIPS, n.d.)
13
Entry level 3D printers
2.1.1
Build platforms for filament materials
There are many methods to reach good adhesion to the surfaces
but there is still room for improvement and the best ones are usually experimentally determined. Not all plastics can be printed
on the same surface. Common problems are: first layer not adhering at all to the platform, parts getting detached from the platform during printing, parts warping under heat.
Printing PLA on a clean surface of glass or aluminium might be
possible after proper cleaning with dissolvant leaving no
residues such as Isopropyl Alcohol. (RepRap Forums, n.d.)
Table 1
Build platform types and properties, author's experiences.
Platform Temperatu Successes
material res
Failures
Other
comments
3D
LamiTouch nated
plywood
Room
temperature
PLA and
TPE with
coating of
glue stick
Double sided
tape, ABS
Weak
mechanical
construction
leading to
vibrations
UltiHeated
maker glass
2,
plate
Leapfrog
Creatr
Glass plate
heated
with a
PCB board
up to
100ºC
PLA and
Wood
filament
with plastic
coating
varnish
spray (bed
75ºC),
ABS with
ABS Juice*
(bed 100ºC)
ABS warping
with coating
spray (bed
90ºC), thicker
coating and
higher bed
temperature
help
The spray
coating may
be reused
many times.
Printing ABS
needs glass
plate
preparations.
Milled
Room
5mm
temperatuthick
re
aluminiu
m plate
Blue
painters tape
(3M) and
Kapton tape
successful
with PLA
and TPE.
Originally too
good adhesion
worked out by
setting z-axis
offset +0,1mm
No warping
or detaching
of parts
experienced
even with
very large
rafts
Gigabot
UPrint DispoSE
sable
Plus
ABS
plate
Air heated
to around
75ºC,
visible on
machine
interface
A raft of
Reusing the
Plates cost
support
plates may lead around 5€ a
material is
to parts getting piece.
printed on
detached from
ABS plate, the plate
then the
actual part in
ABS plastic
on top of
that
*ABS Juice, a term coined on the RepRap forums and wiki for a mixture of
ABS dissolved in Acetone and applied on glass plate.
14
Entry level 3D printers
2.2
Powder materials
Many different powder materials can be printed, either by fusing
or melting them together, or by using a binder material. The
ZPrinter subject in this thesis uses a binder material to form objects in a special powder.
The powder is, as suggested by it's Materials Safety Data Sheet,
mostly water with color additives whereas the powder is a mixture of plaster and polymers (supposedly PVA or other water
solvent polymer or glue).
There have been records of people using alternative materials on
their powder-binder-printers. For example, salt could be used.
(RepRap Wiki Powder Printer Recipes, n.d.)
2.3
Printable waxes
Waxes could be printed, such as those used in the Thermojet
printer. The Thermojet uses waxes described as thermoplastics,
made out of Hydrocarbons, Amides and Esters. The machine is
discontinued but there are newer commercial casting-wax printers on the market. (3D Systems, 1998.)
2.4
Liquid materials
The stereolithography (Form 1) and the inkjet printers (Objet)
use UV curable resins.
The resulting parts are of thermosetting plastic and very different
in properties to the thermoplastics used in the extrusion-based
printers.
Little information about the chemistry of the resin can be found,
mainly from the materials safety datasheets. Manufacturers give,
however, technical properties for the plastics such as tensile- and
impact strengths and glass transition temperatures.
Traditionally, end users were limited to outsourcing the resins
from the printer manufacturer. Nowadays there may be alternative options available for far cheaper prices, mainly because of
the open source printers becoming more popular. (Makerjuice
Labs, n.d.)
The Objet printer uses a different support material deposited
from a separate printing head during printing. The supports are
removed by using pressurized water after printing.
15
Entry level 3D printers
3
THE PRINTERS
The machines that were available for usage for this thesis are
listed in tables two and three. More in depth analysis of the machines and user experience gained from this thesis can be found
in the later chapters.
Table 2
3D Printers
Manuf-acturer
Machine Model
Cost
(estimate)
Extrusion
Leapfrog
Creatr, Creatr XL
1400€,
2012
4000€ (XL)
Extrusion
Ultimaker
Ultimaker 2
2000€
2012
Extrusion
Bits From
Bytes
3D Touch
2000€
2010
Extrusion
Re:3D
Gigabot
2000€
2012
Extrusion
Stratasys
UPrint Plus
15.000€
2010
SL
FormLabs
Form 1
3300€
2013
Inkjet/
MJM
3D Systems
Thermojet
50.000€ (in
the 1990s)
Some
time
in
1990s
Inkjet/
PolyJet
Stratasys
Objet 30 Scholar
35.000€
2012
Inkjet
/3DP
Zcorp
(currently
owned by 3D
Systems)
Zprint 250 (currently
Projet 260 by 3D
Systems)
10.000€
2010
SDL
Mcor
Technologies
Matrix 300
11.000€
2010
Table 3
Year
made
Materials processing equipment
Machine manufacturer
Machine Model
Cost
(estimate)
Year made
Stratasys
WaveWash
2000€
2012
(estimate)
Stratasys
WaterJet
2500€
2012
(estimate)
16
Entry level 3D printers
3.1
Extrusion-based printers
The most common printers nowadays are the cheap desktop FFF
or FDM (called differently depending on the trademark and p)
printers with prices ranging from 500 to 2000€. There are many
companies manufacturing these printers and several different
models were available for the thesis. Luckily also a more experienced one, the uPrint SE was also available and this provides a
good comparison point for the cheaper ones.
3.1.1
Ultimaker 2
Figure 8
The Ultimaker 2 3D printer.
The Ultimaker is a commercial spinoff from the RepRap project
that has been getting popular throughout the world. It's based on
the company's improved open source software and hardware,
and all the files and plans at least for the older machines can be
found online.
17
Entry level 3D printers
For this thesis, three Ultimaker 2s were available for use. They
proved to be constant in accuracy and reliable if serviced properly. Common problems in usage were:
•
Nozzles getting blocked when doing filament changes.
The ABS or other residues block the extruder when printing PLA because of lower extrusion temperature.
•
Improper extruder idler tension, resulting in slippages
when extruding. Problems may appear at the end of the
spool.
•
Belts and bolts getting loose, requiring periodical attention and tightening. For belts, adding tensioners might be
necessary.
•
Filament jams, the filament wire getting stuck in the
spool holder or tangled in itself.
•
Filaments degraded at the end of the spools if old material (notable with PLA)
•
Difficult nozzle change procedures. Disassembling of the
complete extruder assembly is required to change the
nozzle.
Besides these obvious problems faced with this kind of machine,
the Ultimaker has proved to be able to produce good quality
parts if serviced right and the used material was good as well.
Some filaments are much easier to print than others. The Ultimakers would print without problems for tens of hours in a row.
18
Entry level 3D printers
The Ultimaker has it's positive sides and it's negative sides.
Printing simple geometries rather slowly would produce very
satisfactory results reliably.
Correct preparations would be necessary – if printing in ABS,
the build plate would have to be prepared with ABS juice to
avoid warping. Even with the ABS juice, some very large parts
could still warp too much and become bent.
Figure 9
Milling machine parts printed in ABS with the Ultimaker.
The biggest piece was bent about 3 mm from the middle after removal.
Heated bed at 100°C, with ABS-juice
19
Entry level 3D printers
The printer has only one nozzle, but the Cura software is able to
produce supports for the printed parts built with the same material as the actual part.
Printing with supports works, but with too fine details removal
becomes very work intensive and not worthwhile if done manually with sidecutters, a knife and a file.
Usually every problem encountered with the prints came up with
some change to the machine. If there was a change of spool, the
new spool could get tangled. If the material type was changed
problems with the nozzle were encountered. Therefore the Ultimaker might be best suitable for single-purpose use and not as a
3D printing testbench.
Printing wood filament was tested on an Ultimaker. The prints
constantly got blocked with the original 0,4mm nozzle. As a
solution to this the nozzle had to be changed to a bigger one. A
decision was made to drill the nozzle to a bigger size. This was
done on a Bridgeport mill with a 0,6mm HSS drill. After one
month of use, this same printer would suffer from material jam
with the spool getting tangled and while servicing, the heater
wire got broken because of too much bending. A new one would
cost around 35€.
Figure 10
The Ultimaker nozzle assembly showing the hot end and cold
end with the cooler fan. Not showing the heater and temperature resistor in
the brass block below.
20
Entry level 3D printers
3.1.2
3D Touch
Figure 11
The 3D Touch printer
The 3D Touch is a discontinued but still a rather popular printer
in Finnish schools. It has it's own software tool path which is
quite easy to use, but it doesn’t have many features for customization.
The prints are very slow and the reasons are apparent. The printer is set to periodically ooze and clean the nozzle by purging it to
a special container by wiping it against a plastic plate. The
corner guard ball mentioned in chapter 2.1 took over 50 hours to
print with the 3D Touch, compared to the Ultimaker 2, which
took 13 hours on it.
The build area is a sheet of laminated plywood. It can be challenging to find the best methods for proper adhesion to this surface.
Printing ABS is practically impossible because of parts getting
detached and warping.
21
Entry level 3D printers
3.1.3
Gigabot
Figure 12
The Gigabot printer. Notice the missing extruder. The printer
had suffered from extruder meltdown and molten plastic is visible on the build
platform.
The Gigabot was delivered as assembled from USA. It had been
standing for long times without usage. It uses RepetierHost and
common open-source components similar to the Leapfrog. The
printer is poorly documented and no manual could be found. The
settings for PC communication were iterated manually.
Special about this printer is its promised large build volume of
600 x 600 x 600mm (compared to the usual 200 x 200 x 200mm
of the other printers). However, printing big objects will take
long times. An estimate of 180 hours was given by the software
to print an almost full volume object with 0,3mm layers and
35mm/s speed with a 0,4mm nozzle.
22
Entry level 3D printers
The Gigabot had many problems when taken into use. Firstly of
all, prints were suddenly stopping after couple of layers.
Secondly, the y-axle would skip 50 mm in a positive direction
after 10 or so layers. These problems would appear randomly.
The y-axle skipping resulted finally in a situation of apparent
fire-threat. The thermistor from the extruder nozzle got separated
from the heater block, possibly by getting stuck in the floating
filament extruded in to air. This caused the heater to turn on
maximum for almost 10 minutes, which melted the extruder barrel made of some temperature resistant thermoplastic.
Luckily, the firmware aborted the print automatically after the
said 10 minutes. Otherwise the situation could have resulted in
the plastic catching fire. Now the only loss was the melted extrusion cold end that was plastic.
The reasons for the skipping were investigated. Firstly, a too low
belt tension was suspected, but apparently, the stepper motor
drivers' currents were not adjusted by the manufacturer. This
may have resulted in the motors skipping steps (losing positional
accuracy). The motors would skip steps if loaded by hand when
the printer was on.
While investigating the printer for faults, it was clear that the
stepper motor driver boards' current trim-pots had not been adjusted properly. They were all in the same orientation in the
middle of adjustment range.
23
Entry level 3D printers
3.1.4
UPrint SE Plus
Figure 13
uPrint SE Plus 3D Printer
The uPrint SE Plus is an entry level industrial FDM printer. It
has an air circulation system to heat up the enclosed build platform, disposable build plates and a purging and nozzle wipe system to keep the nozzles clean (the nozzle is wiped against a metal brush periodically).
24
Entry level 3D printers
Compared to Ultimaker, for example, the uPrint has more electronics and sensors in use which detect, for example, material
misfeeds and failures. This compared with the rigidity and it's
custom-made parts make it a quite different kind of a printer than
the Ultimaker.
The printed supports make this printer a lot more versatile than
the other FDM machines in this thesis. It is possible to print almost any shapes with enough supports, but this may be seen as a
waste of the support materials, and printing both is time consuming.
Figure 14
A complicated art-piece being printed with lots of supports to
facilitate the printing of a shell structure. Total duration of this print was 40
hours, pictured in halfway
The uPrint has got much better reliability and constancy than the
other extrusion-based printers. Overall the printer requires less
adjustments and tinkering with than the cheaper ones. The machine is constructed of high quality machined or fabricated metal
parts. The supplied materials are expensive but of guaranteed
high quality and packaged well.
25
Entry level 3D printers
There are occasional errors especially in the material feed system. There can be material breakages in the extruder. In the
worst case, the printer may print without either one of the materials because of breakage. This all requires attention from the
user.
The uPrint software, CatalysEX is rather limited in its customizability. Open source software such as Cura have many more options for manipulating the parts and changing print parameters
such as supports and the part's wall thicknesses. This might seem
negative at first, but it does eliminate the time-consuming experimentation quite effectively. This might be seen as positive or
negative thing depending on the use.
The software automatically generates support structures to be
printed. The parts can be washed with the WaveWash system
supplied by Stratasys to remove the supports. The washing system works by circulating tap water and an added mixture of Sodium Percarbonate and a mixture of Tetrasodium N, n-bis
(carboxylatomethyl)-l-glutamate and Citric Acid added before
starting the wash cycle. Mechanical removal is possible and in
simple cases more effective.
The filament spools are supplied by Stratasys. They are delivered in a sealed bags and there is moisture absorbent material
inside the package. The spools include a small circuit board described as an EEPROM chip. This is, in effect, makes it very difficult if not impossible to use second hand materials on the printer.
The spool prices can be as high as 190€ per spool if bought separately. The price can be slightly lower if the spools are bought
in large amounts. The build plates cost around 5€ per piece.
These are remarkably higher prices than with the materials used
in the Ultimaker machines and the like, but of the same type of
plastics.
All-in-all, the uPrint is a reliable machine, but it's options are
limited and it is very expensive to use when compared to the
cheaper ones made by smaller companies. The combination of a
well-maintained heated build platform and high quality plastics
makes the print-results very consistent and good in mechanical
properties.
26
Entry level 3D printers
3.2
Stereolithography / Form 1+
Figure 15
The Form 1+ stereolithography 3D printer
The Form 1 works by lowering a metal plate build platform in
the resin and using a 405nm wavelength UV laser that is directed
by mirrors to polymerize a layer next to the previous one. The
polymer tank comes pre-coated with a silicone gel to prevent the
parts sticking to the tank. Tanks are disposable and are recommended to be used for 1 or 2 liters of printing. The Form 1+ version is a faster, later version than the Form 1 printer from Formlabs. (Formlabs Form1+, n.d.)
There were two Form 1's for use for this thesis. The other Form
1 in use had a particular problem with printing from the start.
The parts being printed would adhere too strongly to the resin
tank and fail because of that. This problem would suffice despite
changing the tank, the build platform, resin and fine tuning the zaxis.
27
Entry level 3D printers
The other, a few months newer, would work much better producing constantly good prints with the cuboid. The difference
between the two machines raised many questions, as if this
would be a machine property, why would the same part print
successfully on one but not on the other.
Figure 16
An art piece printed with the Form 1. The printer works best
with small surface area pieces, and doesn’t have problems producing fine
details if they are properly supported. The pictured piece was broken during
support removal, requiring gluing. Also sanding was needed for surface finish
The Form 1 uses a notably user friendly PreForm software. It
generates supports and estimates the successfulness of the prints
automatically.
The automatic support generation is a critical feature in this
printer. Every trial made for this thesis to print without supports
would fail. Also the part orientation plays a big role and reorienting may solve printing problems (but introduce precision problems, see later tests with the cuboids in this thesis).
28
Entry level 3D printers
Figure 17
Form 1+ at 0,1mm layers on left, FDM printer at 0,1mm layers
on right (Formlabs Layer height, n.d.)
The Form 1's minimum feature size is defined as 0,3mm. It is
capable of much finer details than the FDM machines. However,
overhangs and intricate features present challenges and may result in failures.
The printer needs to be maintained well and cleaned periodically.
The resin tank especially has to be kept clean.
Figure 18
n.d.)
27x magnification of the comparison (Formlabs Layer height,
29
Entry level 3D printers
Notable about the build processes is that the scanning laser
works very fast in the x and y-axes and print time depends
mostly on the height of the objects (printing many similar sized
ones does not increase time notably, whereas in the extrusionbased printers the time increases linearly with the volume).
Unlike the extrusion-based machines, the pieces are printed solid
as a default. To have hollow pieces, drain holes should be designed in to the piece.
After printing, the parts are washed in Isopropyl Alcohol to remove uncured resin. The supports are mechanically removed by
the user. Formlabs has provided very good documentation for all
procedures on their website, including advanced tips for parts
design, applications etc.
Overall, the Form 1 is a good machine for fine details and small
pieces. However, it has it's problems and correct use is a necessity. Keeping everything clean is time consuming, but with suitable pieces, the Form 1 is a very notable printer. The competing
stereolithography 3D printers several times more expensive making the Form 1 a considerably choice at around 3500€ if it's features are needed. It is, however, not a clear winner over the extrusion-based printers as imagined by the writer before starting
to use it.
30
Entry level 3D printers
3.3
Inkjet printers
The inkjet 3D printers available are characterized as expensive
and large in size. There are no cheaper printers available currently and the two available for the thesis were much more expensive ones than the other printers.
The inkjet 3D printers work quite similar to the 2D inkjet printers as in material deposition, but the materials are of course different and there is an added z-axis and the inkheads also move in
y-axis instead of moving the paper.
3.3.1
Objet 30 Scholar
Figure 19
Objet 30 Scholar 3D printer pictured.
The Objet printers work by depositing a very thin layer of resin
on top of a previous layer and curing it with UV light. Layers are
deposited next to each other to build up parts.
There are two different nozzles, each one with over 100 small
holes in a row for jetting the resin. The other nozzle is for the
build material and the other for printing support materials.
31
Entry level 3D printers
The Scholar is a 3D printer marketed specially for educational
use but it is almost identical to the Objet 24 and Objet 30 Pro
versions. The difference between different Objet machines is
mostly in the availability of the materials. The resolutions are the
same but the build area varies a little and the layer thickness is
thinner in more expensive machines.
Figure 20
Objet 30 materials availability. (AIPWorks Objet, n.d.)
The machine for use had been standing for over a month without
usage during the summer holidays. This had caused the nozzles
to get blocked, most probably because of plastic polymerizing
inside the nozzle - rendering them useless. Methods to open the
nozzles were researched and burning them open in an oven at
400ºC seemed feasible.
This has not yet been tried, because there are electronics, suspectedly piezo elements, embedded in the nozzle. New nozzles cost
around 1000€ per piece, plus service fees. After the nozzle
change, recalibration of the axes has to be made because of
minor changes in the nozzle positions.
One of the reasons adding up to the cause of blockage is clearly
the price of materials. Flushing the nozzles spends much of the
material and doing this has been avoided at all costs.
32
Entry level 3D printers
Table 4
The Objet 30 Maintenance Schedule Program (Objet 30
Manual, n.d. 101).
As described in the manual, the printer should not be turned off
if in idle mode, but should be kept idling for up to one week. If
the printer is not used for over one week, a shutdown wizard
should be run on it – involving a flush/fill cycle. For shutdown,
the nozzles should be first emptied, and then flushed with a
cleaning fluid, sold separately. (Objet 30 Manual, n.d. 92-125.)
If changing materials, flushing cycles are also necessary to have
proper colors in the print if the color if changed. Changing to a
darker color, the economy cycle will spend 65 grams (± 10%)
with one cartridge. If changing to a clearer color, the recommended High-performance cycle will take 300 grams (± 10%) when
replacing one cartridge. The washing cycle dumps the material
in the waste collector. 300 grams of material costs around 100 to
150€. (Objet 30 Manual, n.d. 92-125.)
33
Entry level 3D printers
Besides the materials used in this specific printer, Stratasys offers different materials, like the flexible Tango-materials.
However, the usage of these is limited to the more expensive and
advanced Objet Connex -family of printers. The prices for these
may be over a hundred thousand Euros. The Objet 30 Scholar
belongs to the Objet desktop -family. (Stratasys System Matrix,
n.d.)
The material resin prices are high compared to any other printer
used. Prices per kilo range from 250€ to 350€. Support material
costs 110€ per kilo. The materials containers have fixed last
dates of usage and they have a RFID tag taped on their bottoms.
The Objet is the most expensive machine to use of all the printers in this thesis. Not many pieces were printed with it besides
ordered and paid work. The Objet can produce fine details and
with supports, it should be able to print almost any shape desired. Testing these capabilities more deeply was, however, left
outside this thesis. As seen in the Cuboid tests -chapter later on,
the Objet seems to keep up to its promises as an accurate printer.
Figure 21
The WaterJet washing station. Not pictured: pressure
water pump (standard hardware-store type) and foot pedal for controlling
the pressure.
34
Entry level 3D printers
3.3.2
Thermojet
Figure 22
The Thermojet 3D printer.
The Thermojet is a long discontinued inkjet printer made by 3D
Systems several tens of years ago. The thermoplastic waxes used
in the ThermoJet are particularly suited for creating patterns for
investment castings. The melting temperatures of the thermoplastic waxes are 55 C for TJ2000 and 95 C for TJ88.
The materials are fed in a solid-wax-form to the printer. Heaters
are used to lower the viscosity of the wax and the machine has a
10 minute heat-up period before printing.
Production of a similar material seems highly challenging. The
materials safety datasheet in Finnish reveals little about the characteristics of the material, whereas the English datasheets tell
even less. (3D Systems, 1998 & 2005.)
The machine was taken into use. It had been standing in idle
mode for unknown time. The test print cycle functioned. Connection to PC was made with a private LAN connection.
35
Entry level 3D printers
64 bit Windows 7 was used to communicate with the printer,
with VirtualBox emulating Windows NT. After making the proper IP settings, the printer responded to ping-commands. For file
sharing, a Windows private network sharing folder configurations needed to be made.
After this it was discovered that the ThermoJet program crashed
when trying to send a file to be printed. The most probable reason for this was some incompatibility error with the emulated
Windows NT operating system.
Because of time limitations, the software problems remained unsolved. The printer had been used several years ago for investment casting patterns successfully.
36
Entry level 3D printers
3.3.3
ZPrinter 250
Figure 23
The Zprinter Powder printer pictured. Notice the powder
recycling station next to it.
The ZPrinter 250 builds up parts by using a binder material to
adhere powder particles together to form up parts in sequential
layers. It uses cartridges from normal HP 2D-printers, HP11 and
HP57. The HP57 cartridge is used to enable printing in 64 different colors in the ZPrinter 250. This is not available on the cheaper ZPrinter 150.
37
Entry level 3D printers
The resulting parts are fragile and porous by nature and there are
different methods that are recommended for infiltrating the part
to reduce the porosity and increase the parts' strength. These include salt water spraying methods, Cyanoacrylate (super glue)
treatments or using a specific ZBinder compound.
Note that all links to the Zprinter 250 product page were redirected at the time of writing to the 3D Systems product page of their
machine ProJet 260C. Zcorp, the original company, was bought
by 3D Systems in 2012. (3D Systems press release, 2012)
Solheim Additive Manufacturing Laboratory in the Mechanical
Engineering Department on the University of Washington campus has been developing the binders and the powder materials on
their ZCorp ZPrinter and these results are available online on
their Open3DP-blog. They have been successfully printing objects in salt powder and other ground-up materials using custom
binder-solutions. (Open3DP Blog, 2011a, 2011b)
The ZPrinter 250 available for use had been unused for several
months because of prints stopping after couple of layers and the
machine giving error codes such as ”Slow axis pos err. Error
2303:0”.
After reviewing the ZPrinter user and service manuals, it became
apparent that the machine had not been serviced or cleaned according to instructions at this time. The service station especially
was dirty and the axes hadn’t been lubricated as instructed. This
was suspected as the source of the errors, but at a service visit
from 3D System's Finland's representative, Multiprint Oy, a
faulty idler pulley was discovered. A service component was
ordered and the delivery took around 2 weeks.
Because of the shortage in time, no test prints on the ZPrinter
made it to this thesis. However, the prints should be accurate and
gain some strength after infiltrating (means curing) with the
ZBond- or a Cyanoacrylate-adhesive.
38
Entry level 3D printers
3.4
Sheet Lamination / Mcor Matrix 300
Figure 24
The Mcor Matrix 300 3D printer pictured.
The Mcor Matrix works by gluing paper consequently on top of
each other and cutting the desired geometry on it after gluing.
The blade used is a titanium blade, scheduled for replacement
after cutting 7000 meters. The price of the blades is running
around 50 to 100 Euros. The adhesives cost about the same per
litre.
The Mcor must be left to idle to avoid blockages in glue delivery. The machine does a service routine of applying one layer of
paper every 24 hours. Manuals recommend that it would never
be shut down.
It soon became apparent that servicing the machine is important.
The blade may get stuck if the assembly is dirty. The glue delivery line might works badly if the deposition roller has any remains of adhered glue on it.
39
Entry level 3D printers
Figure 25
Test piece tried with Mcor.
Several trials were made to try to print the thin object in Figure
25 with 1mm features that were especially challenging to the
Mcor. Printing the object with the extrusion-based machines or
with the Form 1 printer presented practically no problems with
it, the Mcor was constantly failing.
Figure 26
Paper layers getting detached during printing.
40
Entry level 3D printers
After proper maintenance of the glue deposition head and the
blade, these similar problems were still present but lesser. Printouts have very different properties to the other printers. The layers are easily detachable and the parts are fragile but can hold up
to compressive forces.
Figure 27
A small gear of approximately 10mm in diameter, printed with
the Mcor, showing easily separable layers.
However, after the minor improvement in print quality after allaround cleaning and maintenance, every part tried on the printer
would fail, even the semi-complex. If paper objects are a must,
the printer should be investigated more, but with the obtained
results the printer seems very limited in it's capabilities and very
expensive to use overall. The machine is slowly consuming it's
glue and blade when not in use, and both are expensive, raising
questions about the future of this printer.
41
Entry level 3D printers
4
BENCHMARKING THE PRINTERS
A test piece was used to contest the printers. The piece consisted
of three cubes connected together in different planes. The cubes
measured 16 x 16 x 16 mm each. The faces of the cubes were
along the same planes. The diagonal of the squares connecting
the cubes was 2mm and the fillets were 1mm. The cuboid was
designed in Open Source CAD-program called FreeCad, version
0.14 and exported in .stl-file format.
Figure 28
The cuboid visualized in FreeCad.
It may be difficult to have an universal test piece for all the different printers. For example, the SL and inkjet printers, are able
to print very small features, under 1mm, accurately, whereas
FDM printers may struggle greatly. Different printers have their
strengths and weaknesses.
Zprint, Thermojet were both non-operational during the time of
tests. Mcor was producing too inconsistent results and gave the
impression of a very unreliable machine. None of these three
could suggestively produce strong enough parts for purposeful
handling of the pieces straight out of the printer. Therefore
Zprint, Mcor and Thermojet were left out of the tests. These
printers are most suitable for the specific parts designed for
printing with them.
The Ultimaker 2 stands out positively among the 2000€ price
range printers. The other printers were not actively used for
mechanical weaknesses or problems, or just because setting them
up would be time consuming and would require a great amount
of tinkering with to get them printing.
Objet, Ultimaker, uPrint and Form 1 were selected for final tests.
These were all in operation most of the time during the making
of this thesis and proved to be the best printers in house.
42
Entry level 3D printers
4.1
Test prints
During the preliminary tests on the printers, it became apparent
that the cuboid should be printed in a 45 degree tilted position
towards the build plate. This made the layer-wise properties of
the print much better in the thin connection points.
When the connecting ribs were printed parallel to the machine's
build area's plane, they became noticeably weaker against bending in this direction because of layer properties. Also, it is obvious that this connection point is the weakest link and very low
infill should not affect the overall material strength noticeably.
The same behaviour has been observed in tensile tests performed
on 3D printed pieces – strength of ABS test pieces printed in horizontal orientation could reach up to 80% of the tensile strength
of injection moulded pieces, whereas parts printed in vertical orientation might reach only 25% of the strength before plastic deformation.
Notable about the tensile tests is also that injection moulded
ABS might deform (before breaking) up to 32%, while results
for horizontal and vertical 3D prints are approximately 7,5% and
2,5% respectively. (Aula et al. Kon-41.4005 Loppuraportti,
2014, 21.)
Printing in the 45-degree orientation had its positive and negative sides. FDM printers could knock the pieces down easily if
the piece was not properly adhered to the build platform or
warped due to heat distortions.
After printing the cuboids, visual tests to notice distortions and
differences and measurements with a Vernier's caliber were
made.
43
Entry level 3D printers
Figure 29
Bad surface finish as seen on the planes where supports were
printed on preliminary tests on the Form 1. This is also the piece that is
described as the “preliminary test piece” in the thesis.
The surface quality of the parts is dependent on the orientation.
If there is a support structure printed, the surface plane that the
part is created on, after the supports, can have very rough finishes due to non-uniform material deposition plane. The deposited material doesn’t have a flat surface to be start with. This
makes the later finishing operations necessary.
With the Form 1, as the part is being peeled off the resin tank,
there is some adhesive forces that may pull the piece out from
the build platform, making the generated supports a necessity to
keep the 45 degree angle oriented cuboid from falling down.
44
Entry level 3D printers
4.1.1
Ultimaker 2
Both 0,4mm and 0,6mm nozzles were used in these tests. It is
hard to notice any difference between the two. With default orientation the connection point for the overhang would print
stronger and a little bigger with 0,4mm nozzle, while the 0,4mm
nozzle might have a little better surface finish than the bigger
one, but this is hard to tell by visual inspection.
None of the prints came out perfect from the printer but required
finishing. The best results could be got with a small layer size
and a 45-degree orientation with full 45-degree supports and
probably a lower speed also. The used filament may also play an
important role, as to not heat warp too much and to adhere well
to the build platform because the surface area is low under the
cubes.
Several methods were tried with the Ultimaker, developing as
more and more experiments were made.
First tests were done with PLA filaments. There was White PLA,
supplied by Ultimaker, and black PLA supplied by Weistek. The
machine using the white filament used a 0,4mm nozzle and the
black a redrilled 0,6mm nozzle. The machines were similar otherwise and purchased during the same time.
Later on Natural ABS from Ultimaker would be used with the
0,4mm nozzle and a black spool of ABS from Bits from Bytes
with the 0,6mm nozzle.
45
Entry level 3D printers
Table 5
Ultimaker 2 test settings and approximations
Printer Material Layer
model
size
Print
speed
Infill
Temperature
Other notes
Ultima PLA
ker 2
and
ABS
3mm
spools
40 and
50
mm/s
20%
Nozzle
210°C or
260°C,
Bed 75°C or
100°C
Several
methods
tried.
0,06 to
0,2 mm
Print time (generated by software)
Material cost
30min to 1h 30min per piece, depending
on the settings.
1h 30min with layer size 0,1mm, 1,2mm
solid layers on shells, 20% infill, print
speed 50mm/s and generated supports
30€/kg (overall), 1,15 meters per
cube, 9 grams (Cura estimate), so
approximately 0,03€ or 3 cents.
Figure 30
The cuboid oriented in 45 degrees angle in Cura.
46
Entry level 3D printers
Notable about the PLA tests is that each one of the prints had
some imperfections, less so with the slow but finer 0,06mm layer
prints. Sanding and finishing with a knife or side cutters was necessary. If printed in default orientation, the overhang's lower
surface had very poor surface finish with both grid- and line-type
supports.
Figure 31
in PLA.
Cuboids printed at default orientation with grid type supports
With the second series of tests, ABS filaments were used. The
preliminary tests all failed for not being adhered properly due to
warping on the build surface when it was treated with plastic
coating spray.
Therefore the ABS juice method was tried, which also failed
when done with the Ultimaker Natural ABS -filament. This
would get separated from the glass and form weak points with
air bubbles between the film and the glass.
47
Entry level 3D printers
Figure 32
ABS prints with Ultimaker Natural ABS getting detached
when knocked over by the extruder nozzle due to heat distortions.
Figure 33
Visual description of the thickness of ABS juice used.
Acetone is fully evaporated in picture. A very runny solution was used,
prepared with ABS Plus filament used in the uPrint SE Plus. Care must be
taken not to break the glass by too fast cooling or heat-up.
48
Entry level 3D printers
Table 6
PLA test prints results on Ultimaker 2. Build platform was sprayed
with plastic coating spray and heated at 75°C. Extruder temperature was 210°C
for each test. Each test was printed with both two printers one with white PLA
and one with black. For each machine the toolpaths with correct nozzle
diameters were generated using Cura 14.07.
Test Pieces Cube
no. at
orienonce
tation
Layer
size(s)*
Support
structures
Results
1.
1
45
0,1mm, Touching
degrees 0,25mm buildpla-te
(T.b.), 60 deg*2,
and brim
With many different
settings, the cube failed
many times and the idea
of printing single at once
was abandoned.
2.
3
Default 0,2mm
3.
4
Default 0,06mm T.b. 60 deg. and Connection point between
brim
overhung cube much
stronger than with 0,2mm
layers. The printed pieces
measured 15,9mm each.
Connections 1,1mm on
overhang and 1,4mm
elsewhere.
4.
4
45
0,06mm T.b. 45 deg. and Best quality PLA prints
degrees
brim
reached, measuring 16mm
by sides and 1,3 to 1,5mm
on connections.
T.b. 60 deg. and The parts printed out
brim
completely, but the
connecting point for the
overhung square was
unacceptably weak. Cubes
measured 16 to 16,1mm.
Connection points 1,1
(weak on both) to 1,9mm,
strongest with nozzle
0,6mm between parallel
ones.
*Note that every print used 1,2mm for shell and bottom/top thicknesses, fill
density 20% and 50mm/s for speed, except 40mm/s for the last tests with very
low layer size. “Lines”-setting was used for generating supports.
*2 The 60 degree support setting would generate support only for the flat surface between the two cubes on the sides whereas 45 degree generates in also
on the sides of the cubes. See figure 25. Brim means printing a flat surface of
one layer extending over the part to reach better surface adhesion.
49
Entry level 3D printers
Table 7
ABS test prints results on Ultimaker 2. Build platform was
treated with ABS+Acetone mixture and heated to 100°C. Extruder
temperature was 260°C for each test. Each test was printed with both
two printers one with white ABS and one with black.
Test Pieces Cube
at
orienonce tation
Layer Support
Results
size(s) structures*
1.
3
45 deg. 0,2mm T.b. 60
deg. and
brim
Printed without ABS juice
– the outer cubes ended up
bending and making a little
round if compared to a flat
surface.
2.
10
45 deg. 0,1mm T.b. 45
deg. and
brim.
White filament would
distort from heat too much
and fail easily, see figure
27. Black ABS would need
machine fine tuning, see
figure 29. Finished parts
need post-processing after
support removal, sanding
etc.
ABS Juice effective if
made with suitable ABS.
*See the table 5 for descriptions.
Figure 34
A tray of 10 cuboids in ABS freshly removed from printer. Notice the
amount of material oozing after layers suggesting either problems with the printer settings
or the used material..
50
Entry level 3D printers
4.1.2
UPrint SE Plus
Printing with the uPrint was really straightforward and trouble
free compared to the Ultimaker. The only problem faced was that
when the part was oriented as seen in figure 35, the software
would generate bad toolpaths and not print it correctly. Using
auto-orientate resulted in successful results as seen in figure 36.
The connection points have a strong feel to them. No further processing or experimentation was needed.
Figure 35
Cuboid printed with uPrint SE Plus.
Figure 36
Uprint test settings and approximations
Printer Materi Layer size Print
model al
speed
Infill
Temperature
Other notes
Uprint ABS
SE
Plus
Plus
0,254 mm Not
selectable
Sparse- low
density
300ºC
Support fill
setting set
to SMARToption
Print time
(generated by
software)
Material cost
1h 1min per one
cube.
200€/kg
(overall price, around 500cm3 per spool)
Materials usage:
Model Material: 6,9cm3, Support Material: 1,51cm3
adding up to approximately 4€ per piece.
51
Entry level 3D printers
Figure 37
Cuboid printed with uPrint before support removal. The
piece would easily break while removing if bent too much.
Table 8
UPrint SE Plus print results.
Machine /
Visual
Material and inspection
supplier
notes
Uprint SE
Plus / ABS
Plus by
Stratasys
Caliber
Finishing
measurements
min/max cube
Layers easily 16 / 16,1mm.
visible but
Connections
constant
1,5mm
finish
52
Surface has
smooth feel
and no
disturbing
imperfections
Finishing
operations
needed
Support
material easily
removed with a
knife
Entry level 3D printers
4.1.3
Form 1 Plus
The cube was printed with automatically generated supports.
After the pieces were printed out, the part was washed in Isopropyl Alcohol and the supports were removed with sidecutters.
Some sanding would be needed for a perfect finish. Imperfections were easily visible in the transparent resin.
Figure 38
The cuboid oriented 45 degrees with automatically generated
supports in PreForm software.
Table 9
Form 1+ test settings and approximations.
Printer Material
model
Layer Print
size
speed
Infill
Other notes
Form 1 Clear
Plus
resin
0,1m Depends
m,
on layer
0,05m size
m
and
0,025
mm
Solid /
no
option
Trials to print with
reduced supports led to
failures with part falling
off from the build
platform.
Print time (generated by software)
Material cost
1h 34min with 0,1mm layers (actually
1h or less on the printer, also multiple
pieces would print at just minor
increase if printed simultanously)
125€/Liter (overall)
Volume of one piece 18,5mL with
supports
Approximate cost 2,3€ per piece
53
Entry level 3D printers
Figure 39
Picture of the skewness of the cubes printed with
0,1mm layer thickness. Part was oriented 45 degrees, suggesting either
improperly tuned z-axis or some other fault.
Table 10
Form 1+ print results.
Machine / Layer
Material size
and
supplier
Visual
inspection
notes
Form 1+ / 0,1mm
Clear
resin by
FormLab
s
Squares badly 15,8 / 16,1mm.
skewed up*,
Connections 1,1mm.
connection
ribs very
flexible and
thin*, lots of
imperfections
and material
voids*.
Clear
resin
Caliber measurements Finishing
min/max cube
operations
needed
0,05mm Some
15,9 / 16,3mm.
improvement Connections 1,3mm.
from 0,1mm
layers.
Suspected air
bubbles or
cured resin in
tank making
more
imperfections.
Washing the
part in
Isopropyl
Alcohol to
remove
uncured resin.
Support
removal and
sanding.
Same as above.
Clear
0,025m Best out of the 16,1 / 16,3mm.
Same as above.
resin
m
three.
Connections 1,5mm.
*The properties get better as layer size decreases and part becomes
stronger/better quality/more rigid.
54
Entry level 3D printers
All of the prints from the Form 1 included some visible imperfections and the reason for those is unknown. The resin tank was
visibly clean and it had been on low usage.
Other prints and preliminary tests with the Form 1 suggest that
the machine is very accurate in the pure x-y axis motions. A preliminary test version was printed with default-orientation and it
measured from 15,9mm to 16,1mm by the cubes' faces.
Figure 40
Cuboid printed with Form 1 with 0,025mm layers. Notice
the imperfections inside the cubes and the bottoms where the support
structure was built.
55
Entry level 3D printers
4.1.4
Objet 30 Scholar
Printing with Objet was easy, but expensive. When starting to
print the cuboid, the build material resin bottles had expired by
two days and the machine gave error codes for this and required
material change before printing (wasting 1 bottle of unused resin
worth 300€). The bottles have a RFID tag taped on to them that
the printer reads the date from.
The print was rather tall and that’s why it took a long time on the
Objet. The Objet is fast if parts of the same height are built at
once.
Figure 41
Table 11
Printer
model
The finished cuboid after support removal printed on Objet.
Objet 30 test settings and approximations.
Material
Layer size
Infill
Other notes
Objet
VeroWhite 900 DPI
Full
30
(default, no (no
Scholar
options)
options)
Support material has no
options and is generated
automatically.
Print time (generated by software)
Material cost
4h 26min (1041 layers, multiple pieces
would print at just minor increase in
time if printed simultaneously)
300€/kg (overall model material),
150€/kg (overall support material),
Consumption 27g model, 20g
support,
Approximately 11€ per piece.
56
Entry level 3D printers
Figure 42
The same cuboid before support removal, covered in
support material. The support material is fragile and sticky wax like
substance. It is difficult to remove with a knife and best done on the
WaterJet machine supplied by Stratasys.
Table 12
Objet 30 Scholar print results.
Machine /
Material
and
supplier
Visual
inspection
notes
Caliber measurements
min/max cube
Objet 30
Scholar /
VeroWhite
by
Stratasys
Very
16,1 / 16,2mm. Connections Washing the
constant,
1,6mm.
support material
milk-like
off.
texture with
some surface
roughness
57
Finishing
operations needed
Entry level 3D printers
5
CONCLUSIONS
To start with, there were 10 different 3D printers to experiment
with – 6 of them extrusion-based, 1 SL printer, 3 inkjets, and a
sheet laminating one.
In reflection, I spent most of my time experimenting with the 5
sub-10000€ extrusion-based printers, spending countless hours
wanting to get something done on them that they would not produce with default settings.
Besides the setbacks and unsatisfactory results with the cuboid
tests, I would conclude that these kinds of printers can still be
really capable printers, with experiences relating to using and
building RepRap printers. It is only a question of what is demanded of them and how well you know your machine and materials. Possibly the Ultimaker could produce similar quality to
the uPrint with the cuboids, but it may require more experimenting and possibly modifications to the printer.
Also in the sub-10000€ range was the Form 1+. The test results
were really disappointing because the printer was perceived as a
high-precision printer.
What was learned is that the more expensive printers are really
worth consideration but also have their problems. Materials are
expensive and modifications and repairs made by users are limited. Time usage is a big factor here – with industrial machines
you spend less time on the machine but pay more – whereas on
the cheaper printers you spend more time but pay sufficiently
less. Finally by building your own printer and spending time to
learn how everything works you end up using less time to do
maintenance or changes. The selection of a suitable printer
seems like a real personal and case dependent task.
All-in-all, none of the printers that were reviewed gave completely satisfactory results. As for my own personal opinion and
view on continuing this work, constructing my own printer
seems like the best option cost and time-wise, either with a heavily modified RepRap, Ultimaker or similar extrusion-based printer. For anyone's particular solution, time, money and quality are
the biggest questions.
See attachment 2 for round-up of the tests.
58
Entry level 3D printers
6
SOURCES
Additive Manufacturing Technologies, 2010, Gibson, I. Rosen, D.
Stucker, B. E-ISBN: 978-1-4419-1120-9. E-book accessed through
school network, also available as a paper version.
GNU GPL, 2007, General Public License. Accessed 6.10.2014
http://www.gnu.org/copyleft/gpl.html
GPL Violations, Vendor FAQ. Accessed 6.10.2014.
http://gpl-violations.org/faq/vendor-faq.html
Creative Commons Case Study, n.d. Technical Case Study. Accessed
6.10.2014.
https://wiki.creativecommons.org/Special:SearchByProperty?
title=Special:SearchByProperty&property=Tag&value=technical+details
3D Systems, 1998, Material Safety Data Sheet in Finnish. Accessed
6.10.2014.
http://www.3dsystems.com/products/datafiles/thermojet/msds/MSDSTJ88_Finnish.pdf
3D Systems, 2005, Material Safety Data Sheet in English. Accessed
6.10.2014.
http://www.3dsystems.com/products/datafiles/thermojet/msds/22647S12-00_SDS_TJ88_EU.pdf
3D Systems, 2001, Thermojet Datasheet. Accessed 6.10.2014.
http://www.3dsystems.com/products/datafiles/thermojet/datasheets/TJ
_Pr_DesignComm.pdf
GE Press Release, 2014. GE Aviation Selects Auburn, AL for High
Volume Additive Manufacturing Facility. Accessed 6.10.2014.
http://www.geaviation.com/press/other/other_20140715.html
Materialise, Technologies & Materials Overview, n.d. Accessed
6.10.2014.
http://manufacturing.materialise.com/all-available-materials
Wikipedia 3DP, n.d. Accessed 6.10.2014.
http://en.wikipedia.org/wiki/Powder_bed_and_inkjet_head_3D_printing
59
Entry level 3D printers
Stratasys Objet 30 data sheet, n.d. Accessed 6.10.2014.
http://www.stratasys.com/~/media/Main/Secure/System_Spec_SheetsSS/PolyJet/objet30%20pro%20brochure_letter%20stratasys%20low
%20res.pdf
MatWeb ABS, n.d. ABS Property data page. Accessed 6.10.2014.
http://www.matweb.com/reference/abspolymer.aspx
MatWeb PLA, n.d. PLA Material overview. Accessed 6.10.2014.
http://www.matweb.com/search/DataSheet.aspx?
MatGUID=ab96a4c0655c4018a8785ac4031b9278
Ultimaker, n.d. Ultimaker 2 specifications. Accessed 8.10.2014.
https://www.ultimaker.com/products/ultimaker-2
Taulman 3D, n.d. 618 Material specifications. Accessed 8.10.2014.
http://taulman3d.com/618-features.html
RepRap Forums, n.d. Forum discussion. Accessed 8.10.2014.
http://forums.reprap.org/read.php?1,137841
Bilbycnc PVA Fact Sheet, n.d. Accessed 8.10.2014.
http://www.bilbycnc.com.au/AboutPVABilbyCNCFactSheet.pdf
Reprap wiki Polyamide, n.d. Accessed 8.10.2014.
http://reprap.org/wiki/Polyamide
RepRap Wiki HIPS, n.d. Accessed 1.11.2014.
http://reprap.org/wiki/HIPS
MatterHackers n.d. Webstore, Soft PLA description. Accessed
9.10.2014.
https://www.matterhackers.com/store/3d-printer-filament/300mm-redsoft-pla-threequarter-kg
Makerjuice Labs, n.d. Hobbyist resins. Accessed 9.10.2014.
http://www.makerjuice.com/
Formlabs Form 1+, n.d. Form 1 tech specs. Accessed 9.10.2014.
http://formlabs.com/en/products/form-1-plus/tech-specs/
Formlabs Layer height, n.d. Comparison of printers. Accessed
9.10.2014.
http://formlabs.com/en/support/help/advanced/layer-heights/
Leapfrog, n.d. Creatr product page. Accessed 9.10.2014.
http://www.lpfrg.com/creatr
Cubify, n.d. 3D Touch data sheets, PDF. Accessed 8.10.2014.
http://cubify.com/en/Products/LegacyDownloads
Stratasys System Matrix, n.d. Available as PDF. Accessed 9.10.2014.
http://www.stratasys.com/~/media/Main/Secure/Material%20Specs
60
Entry level 3D printers
%20MS/PolyJet-Material-Specs/System_Matrix.pdf
Mcor technologies, n.d. Matrix 300 plus product page. Accessed
9.10.2014.
http://mcortechnologies.com/3d-printers/matrix-300-plus/
Stratasys uPrint SE, n.d. uPrint SE Plus product page. Accessed
9.10.2014.
http://www.stratasys.com/3d-printers/idea-series/uprint-se-plus
Open3DP Blog, 2011a. Salty parts blog post. Accessed 9.10.2014.
http://open3dp.me.washington.edu/2011/03/salty-parts-3dp-in-salt/
AIPWorks Objet, n.d. Objet Materials (in Finnish). Accessed
22.10.2014. http://www.aipworks.fi/objet_materiaalit
Open3DP Blog, 2011b. Salt printing Research report.
Accessed
9.10.2014.
http://open3dp.me.washington.edu/wpcontent/uploads/2011/03/Salt-Printing.pdf
3D Systems press release, 2012. Accessed 9.10.2014.
http://www.3dsystems.com/press-releases/3d-systems-completes acquisition-z-corp-and-vidar
ZPrinter 250 User Manual, n.d. Available for online viewing as a
Scribd upload. Accessed 9.10.2014.
https://www.scribd.com/doc/233066646/ZPrinter-150-and-250-UserManual
RepRap Wiki Powder Printer Recipes, n.d. Accessed 10.10.2014
http://reprap.org/wiki/Powder_Printer_Recipes
3D Touch Ultrasonic support removal tank , n.d. User guide and safety
manual . Available as a PDF-file. Accessed 10.10.2014.
http://cubify.com/en/Products/DDDTouch
Objet 30 Manual, n.d. Available as a PDF-file. Accessed 10.10.2014.
http://www.ece.ubc.ca/~leos/pdf/tools/objet/manual.pdf
Aula J, Qiongge T, Koskinen H, Isomaa T, Kinnunen A. Kon-41.4005
Loppuraportti, 2014, 21. 3D-tulostetun muovin materiaaliominaisuuksien määrittäminen. Aalto University, unpublished final report from a
Mechanical Engineering course.
61
Title of thesis
Appendix 1
Printer Specifications, in order of presentation.
Ultimaker 2 (Ultimaker, n.d.)
Speed: 10 - 300 mm/s (user definable, default 50mm/s)
Layer thickness: 0,02mm to 0,25 mm (user definable, nozzle
size limits the maximum, default 0,1mm)
Build volume: 230 x 225 x 205 mm
Files transferred via an SD-card.
Maximum nozzle temperature: 260ºC
Maximum heated bed temperature reached: 100ºC
Build platform type: Heated glass plate
Used filament: any thermoplastic of diameter 2,85 – 3 mm (not
able to print 1,75mm filament because of molten plastic reaching and blocking the cold-end of the nozzle, tested)
• Settings are completely customizable in the Cura software by the
user, making it quite versatile machine for experienced users. Other softwares could also be used for slicing, such as Slic3r. Besides
this, the generated gcode can be customized manually. There are
many materials available. PLA and ABS are the most common materials.
Bits From Bytes 3D Touch (Cubify, n.d.)
Speed: maximum 15 mm/s
Layer thickness: 0,125mm
Build volume: 185 x 275 x 210 mm (triple extruder)
Files transferred via an USB memory stick.
Maximum nozzle temperature: 260ºC
Build platform type: Laminated plywood
Used filament: any thermoplastic of diameter 2,85 – 3 mm
Re:3D Gigabot
Speed: 10 - 300 mm/s (user definable)
Layer thickness: 0,02 – 0,40 mm (user definable)
Build volume: 600 x 600 x 600 mm
Files transferred via an microSD-card or USB cable.
Build platform type: Aluminium plate
Used filament: any thermoplastic of diameter 2,85 – 3 mm
Title of thesis
Leapfrog Creatr and Creatr XL (Leapfrog, n.d.)
Speed: 10 – 60 mm/s (user definable)
Layer thickness: 0,05 – 0,35 mm (user definable)
Build volume: 230 x 270 x 200 mm (z-height 500mm on XL)
Has to be constantly connected to computer via USB
Maximum nozzle temperature: 260ºC
Maximum heated bed temperature reached: 100ºC
Build platform type: Heated glass plate
Used filament: any thermoplastic of diameter 1,75 mm
Stratasys uPrint SE Plus (Stratasys uPrint SE, n.d.)
Speed: undefined, similar to default Ultimaker
Layer thickness: 0,254 mm or 0,330 mm
Build volume: 203 x 203 x 152 mm
Files transferred via Ethernet (Windows 7 only).
Build platform type: Heated build area by circulated hot air, disposable injection molded ABS plastic tray.
Used filament: Diameter 1,75mm ABSplus in 9 colors (only 1
color in the basic uPrint model) and support material, SR-30XL.
• Set timers for parts replacements and maintenance, like the
nozzles, which are changed every 2000 hours or the 500h maintenance routine to keep things in good condition.
• Interface displays the temperatures the time of printing as: Build
material extruder 300 C, Support material extruder 300 C and
build chamber 75C.
Formlabs Form 1+
Speed:, fast in x-y
Layer thickness: 0,025mm, 0,050mm or 0,100mm
Build volume: 125 x 125 x 165 mm
Files transferred via USB connection to from a PC.
• Preform -software has very good user interface, but lacks machine
settings, such as laser beam power and speed settings that you
might be interested in if you use custom resins.
Objet 30 Scholar. (Stratasys Objet 30 data sheet, n.d.)
Speed: fast in x-y
Build Resolution (from Objet30 Pro datasheet):
X-axis: 600 dpi, Y-axis: 600 dpi, Z-axis: 900 dpi
Size:
82.5
x
62
x
59
cm
Weight:
93
kg
Build
Size:
294
x
192
x
148.6
mm
Layer Thickness: Horizontal build layers down to 0.016mm
Connection by Ethernet (includes embedded computer)
• Seen on the printer software, the print area is kept at ambient temperature of 31°C and the print heads are kept at around 73°C to
76°C during printing.
Title of thesis
3D Systems ThermoJet (3D Systems, 2001.)
Build materials: ThermoJet 2000 and ThermoJet 88 Thermoplastic
Wax
build
material
Resolution
300
x
400
x
600
DPI
(xyz)
Maximum model size 250 x 190 x 200 mm
Machine dimensions W1.37 x D0.76 x H1.12 m
Machine
weight
375kg
Connection by Ethernet
Zcorp ZPrinter 250 (ZPrinter 250 User Manual, n.d. 2-3.)
Build
speed:
2-4
layers
per
minute
Build material: zp 150 Powder with zb 63 Binder
Build Size: 185 x 236 x 132 mm
Machine dimensions: 74 cm wide x 79 cm deep x 140 cm high
Machine weight: 166 kg
Resolution: 0,102mm, minimum feature size 0,41mm
Connection by Ethernet
Mcor Matrix 300 (Mcor technologies, n.d.)
Build material: standard A4 Paper
Build Size: A4 Paper: 256 x 169 x 150 mm
Resolution
theoretically
0,1mm
(paper
thickness)
Machine
dimensions
950
x
700
x
800mm
Machine
weight
~150kg
Connection by Ethernet
Title of thesis
Appendix 2
Test round-up
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement