"DDVSBDZ BOBMZTJT PG BEEJUJWF NBOVGBDUVSJOH PG % QSJOUFS VTJOH PQUJDBM EJHJUJ[BUJPO
%JQMPNPWÈ QSÈDF
4UVEJKOÓ QSPHSBN / o .FDIBOJDBM &OHJOFFSJOH
4UVEJKOÓ PCPS 5 o .BDIJOFT BOE &RVJQNFOU %FTJHO
"VUPS QSÈDF 1IBOFFOESB .BOUBEB
7FEPVDÓ QSÈDF *OH 3BEPNÓS .FOEDzJDLâ 1I%
-JCFSFD
"DDVSBDZ BOBMZTJT PG BEEJUJWF NBOVGBDUVSJOH PG % QSJOUFS VTJOH PQUJDBM EJHJUJ[BUJPO
.BTUFS UIFTJT
4UVEZ QSPHSBNNF / o .FDIBOJDBM &OHJOFFSJOH
4UVEZ CSBODI 5 o .BDIJOFT BOE &RVJQNFOU %FTJHO
"VUIPS 1IBOFFOESB .BOUBEB
4VQFSWJTPS *OH 3BEPNÓS .FOEDzJDLâ 1I%
-JCFSFD
1SPIMÈÝFOÓ
#ZM KTFN TF[OÈNFO T UÓN äF OB NPV EJQMPNPWPV QSÈDJ TF QMOǔ W[UB
IVKF [ÈLPO Ǐ 4C P QSÈWV BVUPSTLÏN [FKNÏOB f o ÝLPMOÓ EÓMP
#FSV OB WǔEPNÓ äF 5FDIOJDLÈ VOJWFS[JUB W -JCFSDJ 56- OF[BTBIVKF EP NâDI BVUPSTLâDI QSÈW VäJUÓN NÏ EJQMPNPWÏ QSÈDF QSP WOJUDzOÓ QPUDzFCV 56-
6äJKJMJ EJQMPNPWPV QSÈDJ OFCP QPTLZUOVMJ MJDFODJ L KFKÓNV WZVäJUÓ KTFN TJ WǔEPN QPWJOOPTUJ JOGPSNPWBU P UÏUP TLVUFǏOPTUJ 56- W UPN
UP QDzÓQBEǔ NÈ 56- QSÈWP PEF NOF QPäBEPWBU ÞISBEV OÈLMBEǾ LUFSÏ WZOBMPäJMB OB WZUWPDzFOÓ EÓMB Bä EP KFKJDI TLVUFǏOÏ WâÝF
%JQMPNPWPV QSÈDJ KTFN WZQSBDPWBM TBNPTUBUOǔ T QPVäJUÓN VWFEFOÏ MJUFSBUVSZ B OB [ÈLMBEǔ LPO[VMUBDÓ T WFEPVDÓN NÏ EJQMPNPWÏ QSÈDF B LPO[VMUBOUFN
4PVǏBTOǔ ǏFTUOǔ QSPIMBÝVKJ äF UJÝUǔOÈ WFS[F QSÈDF TF TIPEVKF T FMFL
USPOJDLPV WFS[Ó WMPäFOPV EP *4 45"(
%BUVN
1PEQJT
Liberec 2017 5 Acknowledgement
I would like to thank the following people who were instrumental in making this project a reality. I am grateful for the precious time they had dedicated to motivate and guide me towards successful completion of this project.
ING. RADOMIR MENDRICKY, Ph.D. – Department of manufacturing systems and automation, for his expert guidance, enormous patience, constant encouragement and inspiration to complete this challenging project.
ING. JIRI SAFKA, Ph.D. - Department of manufacturing systems and automation, for his expert guidance, for help with printing.
ING. PETR ZELENY, Ph.D. – Head of the Department, Department of manufacturing systems and automation, for his encouragement, support and guidance.
Last but not least, I would like to thank my family and loved ones for their continuous support, trust and help in the process Study.
Liberec 2017 6 Abstract
There is a certain advancing need to study the process parameters of trending 3D- printing technologies to cater specific needs. The present work focuses on production of parts by different 3D-printing technologies (FDM, Polyjet Matrix, SLA and SLS). This article acquaints its readers with data and results of research dealing with influence of production process parameter settings on the magnitude of internal material tension of the printed part, or its influence on dimension and shape precision of products manufactured by these technologies. The prototypes which we printed they may change their properties by change in time so we are inspecting all the printed samples and analyse the dimensional and shape accuracy of each and every sample in three time frames (0 days, 14 days and 84 days). The produced samples were measured and their shape precision was analysed by ATOS 400 contactless optical scanner based on 3D optical digitisation method, allowing comparison of the actual part to a nominal CAD model by using the GOM inspect software and therefore perform shape and dimension precisions of the produced prototypes in a complex and objective manner and the effect of "aging" was researched too. It was observed that all the above mentioned technologies excel in specific cases. When transparent or translucent parts are required, Polyjet Matrix is clearly the way to go. The same applies if you need to achieve a smooth surface or display tiny details on your print. On the other hand, if your design would use a lot of support material to cope with its intricacy, you need a seriously strong part, or you just want to save some money, then you may want to consider a SLS or SLA plastic print.
Key words: 3D printing, Optical 3D Scanner, FDM (Fused Deposition Modelling), Polyjet Matrix, SLA (Stereo lithography), SLS (Selective Laser Sintering)
Liberec 2017 7 TÉMA
Analýza přesnosti aditivní výroby 3D tiskáren pomocí optické digitalizace
ANOTACE
Existuje požadavek studovat parametry procesu 3D-tisku tak, aby vyhovovaly specifickým potřebám. Tato diplomová práce se zaměřuje na výrobu dílů různými technologiemi 3D tisku (FDM, Polyjet Matrix, SLA a SLS) a seznamuje s daty a výsledky výzkumu zabývajícího se vlivem nastavení parametrů výrobního procesu na přesnost rozměrů a tvarů výrobků vyráběných těmito technologiemi. Prototypy vyrobené metodou 3D tisku mohou změnit své vlastnosti též v čase. Proto jsou všechny tištěné vzorky kontrolovány a analyzovány na rozměrovou a tvarovou přesnost ve třech časových intervalech (0 dní, 14 dní a 84 dnů po tisku). Vytvořené vzorky byly měřeny a přesnost jejich tvaru byla analyzována bezdotykovým optickým skenerem ATOS II 400 založeným na metodě 3D optické digitalizace, umožňující srovnání reálného dílu s nominálním CAD modelem v softwaru GOM Inspect. Bylo zjištěno, že všechny výše uvedené technologie vynikají ve zvláštních případech. Když jsou vyžadovány průsvitné nebo průhledné části, je zřejmé, že vyniká technologie Polyjet Matrix. Totéž platí, pokud potřebujete dosáhnout hladkého povrchu nebo vyrobit drobné detaily. Na druhou stranu, pokud stavba modelu vyžaduje velké množství podpůrného materiálu nebo chceme ušetřit nějaké peníze, lze zvážit tisk technologií SLS nebo SLA.
KLÍČOVÁ SLOVA
3D tisk, Optický 3D skener, FDM (Fused Deposition Modelling), Polyjet Matrix, SLA (Stereo lithography), SLS (Selective Laser Sintering).
Liberec 2017 8 Contents
1 Introduction ... 14
2 Rapid Prototyping ... 15
2.1 History and Development of AM ... 15
2.2 Production Processes ... 16
2.2.1 Subtractive process ... 16
2.2.2 Additive Process ... 16
2.2.3 Formative Process ... 16
2.3 Process chain ... 16
2.4 Classification of Addictive Manufacturing systems ... 17
2.4.1 Liquid based ... 17
2.4.2 Solid based ... 18
2.4.3 Powder Based ... 18
2.5 Data conversion and transmission ... 18
2.5.1 STL Format ... 19
2.5.2 RepRap ... 19
2.5.3 Pre-processing ... 20
2.5.4 Slicing and G-code ... 20
2.5.5 Post-Processing ... 20
3 USED 3D PRINTING TECHNOLOGIES ... 21
3.1 Fused Deposition Modeling (FDM) ... 21
3.2 Stereolithography (SLA) ... 22
3.3 Polyjet Matrix ... 23
3.4 Selective Laser Sintering (SLS) ... 24
4 MANUFACTURING THE TEST SAMPLES ... 25
5 MEASUREMENT METHODS AND EQUIPMENT USED ... 28
5.1 Types of 3Ddigitization ... 28
5.1.1 Contact type digitizing... 28
5.1.2 Non-contact type digitizing ... 28
Liberec 2017 9
5.2 History of scanning ... 28
5.3 Introduction to Atos II system ... 29
5.4 General explanation for Atos scanner ... 30
5.5 Spray painting ... 31
5.6 Adjustment and calibration of the device ... 31
5.7 Preparing parts for measurements ... 32
5.8 Rules for placing the reference points On flat or slightly curved surfaces. ... 33
5.9 Optical 3D digitizing ... 33
6 ANALYSIS OF MANUFACTURING ACCURACY ... 34
6.1 Analysis of Deviations Day 1 ... 34
6.2 Analysis of Aging Effect (14 Days After) ... 41
6.3 Analysis of Aging Effect (84 Days After) ... 48
6.4 Compression of Colour Maps Day 1 , Day 14 And Day 84 ... 55
7 Conclusion ... 60
8 REFERENCES ... 61
9 Appendix ... 64
Liberec 2017 10 List of Figures
Figure 1: Three types of fundamental fabrication process [4]. ... 16
Figure 2: Process chain for the addictive manufacturing systems [7]. ... 17
Figure 3: The principle of FDM method [15] ... 22
Figure 4 : The principle of SLA method [18] ... 23
Figure 5 : The principle of PolyJet Matrix method [21] ... 23
Figure 6 : The principle of SLS method [24] ... 24
Figure 7: Model production using FDM, PolyJet Matrix, SLA and SLS ... 25
Figure 8: The designed cad model ... 26
Figure 9: 2D drawing of the CAD model ... 26
Figure 10: Optical scaner Atos II 400 [28] ... 29
Figure 11: Spray painting on the models ... 31
Figure 12: Auxiliary laser pointer ... 32
Figure 13: Preparation for measurement ... 32
Figure 14: Inspection of diameters and the dimension of sphere and spacing between them ... 34
Figure 15: Graphical view of the diameters deviation ... 35
Figure 16: graphical view of cylindricity deviation ... 36
Figure 17: Graphical view of the dimension LX and LY deviation ... 37
Figure 18: Graphical view of the sphere deviation ... 38
Figure 19: Grapgical view of sphere spacing ... 39
Figure 20: Graphical view of Flatness ... 40
Figure 21: Deviation of the cylinder after 14 days... 42
Figure 22: Deviation of cylindricity after 14 days ... 43
Figure 23: Deviation of Dimensions LX and LY after 14 days ... 44
Figure 24: Deviation of the sphere after 14 days ... 45
Figure 25: Deviaion of spacing betweeb spheres after 14 days ... 46
Liberec 2017 11
Figure 26: Deviation fo Flatness after 14 days ... 47
Figure 27: Deviation of the cylinder after 84 days... 48
Figure 28: Deviation of cylindricity after 84 days ... 49
Figure 29: Dimensional deviation after 84 days ... 50
Figure 30: Deviation of the sphere after 84 days ... 51
Figure 31: Divination of spacing between spheres after 84 days ... 52
Figure 32: Deviation of Flatness after 84 days ... 53
Figure 33: Compression of models in different time sets ... 54
Figure 34: Colour maps of normal deviations Day 1 ... 55
Figure 35: Colour maps of normal deviations of real part after 14 days versus 1st day ... 57
Figure 36: Colour maps of normal deviations of real part after 84 days versus 1st day ... 58
Liberec 2017 12 List of Tables
Table 1: Properties of the model and the 3D printers used ... 27
Table 2: Deviation of the nominal cylinder diameters ... 34
Table 3: Deviation of the nominal cylindricity diameters ... 35
Table 4: Deviations of the nominal dimensions XL and YL ... 36
Table 5: Deviation of the nominal sphere dimension ... 37
Table 6: Deveation of spacing between the spheres ... 38
Table 7: Devication of Flatess... 39
Table 8: Deviation of cylinder diameters afetr 14 days ... 41
Table 9: Deviation of the cylindricity after 14 days ... 42
Table 10: Deviations of the Dimension LX and LY after 14 days... 43
Table 11: Deviation of sphere after 14 days ... 44
Table 12: Deviation of the spacing between the spheres after 14 days ... 45
Table 13: Deviation of Flatness after 14 days ... 46
Table 14: Deviation of cylinnder diameters after 84days ... 48
Table 15: Deviation of cylindricity after 84 days ... 49
Table 16: Dimentional deviations of LX and LY after 84 days ... 50
Table 17: Deviation of sphere after 84 days ... 51
Table 18: Deviation of spacing between spheres after 84 days ... 52
Table 19: Deviation of Flatness after 84 days ... 53
Liberec 2017 13 List of Abbreviation
3D - Three-dimensional
ABS - Acrylonitrile butadiene styrene
ASCII - American Standard Code for Information Interchange CAD - Computer Aided Design
FDM - Fused Deposition Modelling FFF - Fused Filament Fabrication LCD - Liquid Crystal Display
NTC - Negative Temperature Coefficient PLA - Polylactic acid (polylactic acid) PTC - Positive Temperature Coefficient PTFE - Polytetrafluoroethylene (Teflon) PVA - Polyvinyl alcohol
RAMBo - RepRap Arduino-compatible Mother Board RAMPS- RepRap Arduino Mega Polo Shield
RepRap- Replicating Rapid Prototype SD- Secure Digital
STL- Standard Triangulation Language USB- Universal Serial Bus
ATOS – Advanced Sensor Topometric
STEP - Standard for the Exchange of the product data NC – Numerical control
CNC – Computer Numerical Control STL – Standard Triangulation Language
Liberec 2017 14 1 INTRODUCTION
Additive technologies are recently on a huge rise in the past, the 3D printers were used mainly in various fields of industry with a great pressure on reduction of production time.
However, their scope of application is now much wider nowadays, one may encounter these printers not only in the field of medicine, arts, construction or gastronomy, but in the field of model making or in households. Among the best known and most widely used are for example Selective Laser Sintering (SLS) technology that utilises a high-power laser beam to melt and sinter fine grains of the print material to form a required shape, or a method similar in principle –Stereolitography(SLA) that draws the individual layers of an object by means of ultra-violet laser beam on a surface of a polymer liquid. Other widely spread technologies are Fused Deposition Modelling (FDM), Multi Jet Modelling (MJM), or PolyJet Matrix.
The aim of this study was to analyse and verify the accuracy of production 3D printers. For this purpose it was necessary to test printed samples and receives their real images. To obtain real model images we used contactless scanner Atos II and GOM Inspect software. On each of the named printers were created two models with identical geometry and different set parameters that affect the future quality of the sample. As a result, we were able to analyze how great influence on the accuracy of these settings has on the final printed model. The first part describes the general terms of dealing with the method of rapid prototyping manufacturing processes, rapid prototyping history, basic technology, which correlate with used printer software and formats. The second part focuses on non-contact scanning;
digitizing explains the concept, types of scanning, scanning history, types of scanner, describes in detail the non- contact scanner Autos II. And software utilized by the scan stage and the final inspection. The third part briefly explains concepts dealing with tolerances.
The practical part is described the design and creation of the model, actual printing on all the printers and importance phase of pre-processing and post processing affecting the print quality of the final printed model. It also deals with preparing the operations of scanning and scanning process itself. Then concerning the evaluation of the data obtained a detailed analysis of each individual issue. In this section find colour maps, determination of flatness and deviations model real versus nominal micron size. The last part contains the conclusions assessed through measurements. The analysis are placed colourful maps, Chart controlled aspect and content of the accompanying CD.
Liberec 2017 15 2 RAPID PROTOTYPING
Rapid prototyping formally known as Addictive manufacturing (AM), has become a general term to describe many prototype processes, generally AM known as “ a process of joining materials to make objects from 3D model data , usually layer up on layer, as opposed to subtractive manufacturing methodologies” also known as 3D printing in public literature [1].
Nowadays this is all about speed, efficiency, and productivity, so this term “Rapid prototyping” is usually applied to the process and generates Prototype "fast". An explanation of the term additive process will be described in more detail in the next chapter on manufacturing processes. However, we should not forget even closely related terms with this definition and these are Rapid Tooling or Rapid Manufacturing. Rapid Tooling, a fast-paced tool, is the production of tools, moulds and moulds directly or indirectly from Rapid Prototyping technology. Rapid Manufacturing, or fast manufacturing, deals with the production of finally-used parts directly or indirectly from Rapid Prototyping technology [2]
2.1 History and Development of AM
AM id developed in 1980s, when a man named Charles “Chuck” Hull invented the first form of 3D printing called Stereolithography (SLA) and for the first tine enabled users to generate a physical object from digital data. It was the advancement of the laser technology along with the Hull’s innovation regarding the materials and the process he used that first made this conceptual method a reality.The innovation of AM technology was a watershed event because of the tremendous time saving, especially for complicated and difficult to produce models. Chuck Hull formed a very first company “3D systems” in 1986 to manufacture the 3D printers. Chuck was also the first person to find a way to allow CAD file to communicate with the RP system in order to build computer modelled parts. Since then, other new AM technologies have been commercialised including Fused Deposition Modelling (FDM) and Selective Laser Sintering (SLS) [3].
New applications for the AM and new capabilities are emerging all the time, including the applications to bioengineering which held to development of machines such as the Bioplotter from Envision TEC GmbH. It produces scaffold using biocompatible materials. New machines with new functionalities such as multi-material, multi-colour printing are being released.
Liberec 2017 16 2.2 Production Processes
Manufacturing processes, both manual and automated, can be defined as subtractive, additive or formative.
Figure 1: Three types of fundamental fabrication process [4].
2.2.1 Subtractive process
In subtractive process, one start with a single block of solid material larger than the final size of the desired object. Portions of the material are removed until describe shape is reached [5].
2.2.2 Additive Process
In contrast, an addictive process is the exact reverse in that the end product is much larger than the initial material. Materials are manipulated so that they are successively combined to form the desired object [5].
2.2.3 Formative Process
Lastly, the formative process is one where mechanical forces or restricting forms are applied on a material so as to form in into the desired shape [5].
2.3 Process chain
All the AM systems generally have five steps in similar sort of process chain, such as
• 3D modelling
• Data conversion and transmission
• Checking and preparing
Liberec 2017 17
• Building step
• Post processing [6]
Figure 2: Process chain for the addictive manufacturing systems [7].
2.4 Classification of Addictive Manufacturing systems
While there are many ways in which one can classify the numerous AM systems in the market, one of the better ways to classify AM systems broadly by the initial form of its material. All the AM systems can be easily categorised into
• Liquid based
• Solid based
• Powder based 2.4.1 Liquid based
The initial form of the liquid based AM systems building material is the liquid state.
Through a process commonly known as cutting, the liquid is converted in to solid state. The following AM systems are fall in to this category [8].
• 3D systems’ stereolithography Apparatus (SLA)
• Stratasys’ PolyJet
• 3D systems’ Multijet Printing (MJP)
• Envision TEC’s Perfactory
Liberec 2017 18
• CMET’s Solid Object Ultraviolet-Laser Printing
• Envision TEC’s Bioplotter
• RegenHU’s 3D Bioprinting
• Rapid Freeze Prototyping 2.4.2 Solid based
Solid based AM systems are meant to encompass all forms of material in the solid state. In this context, the solid form can include the shape in the form of wires, rolls, laminates and pellets. The following AM systems are in to this category [8].
• Stratasys’ Fused Deposition Modelling (FDM)
• Solidscape’s Benchtop System
• MCOR Technologies’ Selective Deposition Lamination (SDL)
• Cubic Technologies’ Laminated Object Manufacturing (LOM)
• Ultrasonic Consolidation 2.4.3 Powder Based
Powder is by-and-large in the solid state. However, it is initially created as a category the solid-based AM systems to mean powder in grain-like form. The following AM systems are in to this category [8].
• 3D systems’ Selective Laser Sintering (SLS)
• 3D systems’ ColorJet printing
• EOS’s EOSINT Systems
• Optomec’s Laser Engineered Net Shaping (LENS)
• Arcam’s Electron Beam Melting (EBM)
• Concept Laser GmbH’s LaserCUSING
• SLM Solutions GmbH’s SLM
• 3D Systems’ Phenix PXMT
• 3D-Micromac AG’ MicroSTRUCT
• The Exone Company’s ProMetal
• Voxeljet AG’s VX System
2.5 Data conversion and transmission
The representation methods uses to describe CAD geometry vary from one system to another. A standard interface is needed to convey geometric descriptions from various CAD packages to Addictive Manufacturing (AM) systems. For the last three decades, STL
Liberec 2017 19
(StereoLithography) file format is the default standard, and it has been used in many AM systems to exchange information between design programs and AM systems [9].
2.5.1 STL Format
It is the representation method used to describe the CAD geometry varies from one system to another. A standard interface is needed to convey geometric descriptions from various CAD packages to Additive Manufacturing (AM) system. For the last three decades, the STL file format is used to exchange information between design programs and AM systems[10].
The STL file consists of the unordered list of triangular facets representing the outside skin of an object. There are two STL file formats. One is the ASCII format and the other is binary format, but it is human readable. In a STL file, triangular facets are described by a set of X, Y and Z coordinates for each of the three vertices and a unit normal vector with X, Y and Z to indicate the side of the facet. Moreover, many commercial CAD models are not robust enough to generate the facet model and frequently have problems as a result [10].
There are several advantages of the STL file:
• It provides a simple method of representing 3D CAD data.
• It is already a de facto standard and has been used by most CAD systems and AM systems.
• Finally, it can provide small and accurate files for data transfer for certain shapes.
There are several disadvantages for STL as well:
• The STL file is many times larger than the original CAD data file for the given accuracy parameter. The STL file carries much redundant information such as duplicate vertices and edges.
• The geometry flaws exist in STL files because many commercial tessellation algorithm used by CAD vendors, which are not sufficiently robust.
• The STL file carries limited information to represent colour, texture material, substructure and other properties of manufactured end object.
• The subsequent slicing of large STL files can take many hours.
2.5.2 RepRap
Replicating Rapid Prototype (RepRap) is a device that is capable of self-replication and fast Prototyping. The project was founded in 2004 Dr.Adrian Bowyer at the University of Bath in the UK. Gradually, RepRap became an international project on the principle of open hardware and software, which means the assembly instructions RepRap products are freely open to anyone in the world[5]. Under this, license it can also available for everyone to
Liberec 2017 20
improve the project and re-run it freely. A large part of the components needed to build a new 3D RepRap printer is possible Print to another 3D printer. The fact that the printer is capable of partial self-replication and instructions for building such a device are freely available, resulting in substantial drop in prices for commercial FDM devices [11].
2.5.3 Pre-processing
The first pre-production step is to get the model in STL (Standard Triangulation Language).
Most FDM devices require this format to be further processed. The majority, of CAD programs are supported by STL model storage. There are also several websites currently offering ready-to download models, in many cases free of charge.
2.5.4 Slicing and G-code
The basic information about the print, such as the diameter of the filament, the melting temperature of the material, or the temperature of the print environment, must be passed to the slicer. After loading the model in the slicer, the 3D model is split into individual layers and translated into G code, which contains all the information to control the entire process.
In this file, there are commands (G-codes) where the extruder is located, how much material should be pushed at the specified temperature and in how much speed it’s moving. The G- code also contains the print commands that have been configured in the slicer settings (for example, the required print media temperature before the print is started). Popular software open slicers are Cura, Slic3R, Simplify3D and Skeinforge.
2.5.5 Post-Processing
Ideally, it is possible to remove a finished printout from the pad after overheating and gentle cooling. However, it is often necessary to remove support, either mechanically or by dissolving support material. The next step may be the surface finish of the model. For example, ABS used to immerse the model in an acetone vapour bath, creating a shiny surface of the model. When printing larger models that are larger than the print area of the printer, it is possible to divide them into several smaller parts before printing, which then sticks together after printing.
Liberec 2017 21 3 USED3DPRINTINGTECHNOLOGIES
3D printing is an additive manufacturing process which creates a physical object from a digital design. There are different 3D printing technologies and materials in which you can print. All the technologies are based on the same principle which states that a digital model is turned into a solid three-dimensional physical object by adding material layer by layer. There are different types of technologies and some of them what we used stated below.
• Fused Deposition Modelling (FDM)
• Stereolithography (SLA )
• Polyjet Matrix
• Selective Laser Sintering (SLS)
3.1 Fused Deposition Modeling (FDM)
Fused Layer Modelling (FLM) or Fused Deposition Modelling (FDM) shown in figure 1, is one of the most widespread additive technologies. This method was developed by S. Scott Crump, who also patented it in 1989 and later founded a company – Stratasys. Most commonly, the principle of the FDM lies in melting a thermoplastic material in a form of a fibre inside an extrusion head that extrudes the melt onto a build platform [12]. Due to 2-axis movement it forms a layer of material in the product’s horizontal cross-section plane. The FDM printers usually use two print heads (figure 3). One head builds the supporting structures and the other for layering the model material [13]. Layer thickness usually ranges from 127 to 330 micrometres. After finish one layer, the build platform is vertically lowered by the layer thickness, followed by applying another layer [14], while this process repeats until the whole product is printed. The supporting structure is created while protruding parts required, after the model printed we can remove the supporting material by the hand or by using some chemical liquid. The most common materials for FDM are ABS and PLA thermoplastics. Also, polyamide, polyethylene, or other thermoplastic materials can be used for the manufacturing process.
Liberec 2017 22 Figure 3: The principle of FDM method [15]
3.2 Stereolithography (SLA)
Stereolithography (SLA) (figure 4) using widely now a days in the field of tissue engineering [16]. In this case, production of the object is based on the photo polymerization of the liquid resin in the solid form. Stereolithography resin (SLR) is deposited layer by layer on the prerequisite model and simultaneously solidified or polymerised by the UV laser or different light sources. Typical SLRs used for SLA attached with a 355 nm wavelength laser and it can produce wide verity of shapes, it is often expansive mainly due to the 355 nm wavelength laser and the cationic photo initiator. Nowadays, desktop level stereolithography apparatus like Formlabs SLA, digital light projection (DLP) and continuous liquid interface production were developed. This apparatus usually use 405 nm (blue ray) wave band laser devise or DLP projector, SLR as the printing materials. When using this laser source, the beam of light focused onto the bottom surface of a tank filled with SLR. The light beam draws the layer of the object on the surface of the SLR forming a cured layer due to the photonic polymerization [17]. This cured layer is attached to the base or the previous layer and can be peeled off from the silicon attached on the surface of the resin tank. After that the base is raised to creation height and subsequently more liquid resins are refilling the gap between the cured part and the silicon.
Liberec 2017 23 Figure 4 : The principle of SLA method [18]
By repeating the above steps the desired part can be printed layer by layer. The illustration of the 405 nm SLA 3D printer is as shown in the figure. These desktop 3D printers can fabricate the models faster the then the traditional 355 nm 3D printers and it reduce the cost both 3D printer and the SLR.
3.3 Polyjet Matrix
The Objet company (today Stratasys company),, which patented the PolyJet Matrix method (figure 5), comes from Israel and is the first RP technology, which allows the simultaneous dosing of two types of resin during one process of model manufacturing [19].
Figure 5 : The principle of PolyJet Matrix method [21]
Liberec 2017 24
The print head extruded photopolymer is cured using a UV lamp. Thanks to the simultaneous dosing and mixing of the two or more components of the mixture, it is possible to build physical models with different mechanical and physical properties in one production process [20]. It is possible to select the resin about properties that are as close as possible to the properties of the finely applied material. Objet PolyJet Matrix eliminates the need to build separate parts of the model from a variety of materials.
3.4 Selective Laser Sintering (SLS)
The Selective Laser Sintering (figure 6) is a commonly used 3D printing technology which uses the material polymer powder (mostly PA12) [22] to produce the parts. The powder is distributed evenly on the base and heated by the radiant heaters just below the melting temperature of the polymer powder. Then via scanner mirrors a laser will exposes the desired geometry in to powder bed surface then the powder heated to extent and the individual powder particles melt and interconnect with the layer and with layer bellow. The non- exposed material will remain on the bed and it will act as the support material [23].
Figure 6 : The principle of SLS method [24]
Once the complete geometry exposed and partially melted then the building platform will lowered one layer and the next layer will applied. This process will continue till the object get over. During the cooling state we will keep the object on the bed to prevent from the thermal distortion
Liberec 2017 25 4 MANUFACTURINGTHETESTSAMPLES
Although there is no standard for testing the dimension and geometrical accuracy of parts manufactured by means of additive technologies, an own model based on research and prior experience [25] was designed. The base of the model is 100 × 100 mm, the sides of the base are fitted with M6 threads allowing mounting to measurement equipment, here in the figure 7 we can see the mode is printing by different 3D printers. The model contains shapes for inspection of basic dimensions, i.e. lengths, distances, angles and diameters of spherical and cylindrical surfaces. Geometry of the model was designed so that it contains as many problematic shapes (elements) as possible. The key parts were two planes, five cylinders in different orientations, three identical spheres and distance between them. In addition, it is possible to inspect some deviations of shape and position, such as flatness, parallelism, concentricity of cylindrical surfaces, perpendicularity, etc. It is also possible to evaluate small details. For that purpose, the model is fitted with tiered rectangular through-grooves and circular holes. Distance and size of each of the geometrical object was selected with regard to the 3D scanning performed in the future. It is appropriate to place he objects so they do not unnecessarily overlap each other.
Figure 7: Model production using FDM, PolyJet Matrix, SLA and SLS
Liberec 2017 26 Figure 8: The designed cad model
For the model printed by PolyJet Matrix method, VeroGraymaterial (matte mode – model covered by support material) and one more model with VevoGray material without covered support material was used (manufacturer Stratasys, tensile strength 50 - 64 MPa, modulus of elasticity 2000 - 3000 MPa, flexural strength 75 - 110 MPa, flexural modulus 2200 - 3200 MPa, Shore hardness 83 - 86 Scale D - more information [stratasys.com 2016], while the layer thickness was set to both, 16 microns (referred to as HQ –High Quality) and 30 microns (referred to as HS –High Speed).
Figure 9: 2D drawing of the CAD model
Liberec 2017 27 Table 1: Properties of the model and the 3D printers used
Printer Material Layer
thickness mm Model
1 FDM Dimension ABS 0.25 mm Full solid
2 FDM Dimension ABS 0.25 mm Sparse light
3 FDM Fortus ABS 0.25 mm Full solid
4 FDM Fortus ABS 0.25 mm Sparse light
5 PolyJet Object 500 VevoGray 0.016 mm Matt
6 PolyJet Object 500 VevoGray 0.016 mm Glossy
7 SLS EOS P3SP PA 2200 0.1 mm Vertically printed
8 SLS EOS P3SP PA 2200 0.1 mm Horizontally printed
9 SLA Formlabs 2 ABS 0.05 mm Full model
TUL
10 SLA
Ultra ABS 0.05 mm Full model
Out side
In case of FDM technology for both Dimension and Fortus, the most common material was used –ABS-P400 manufacturer Stratasys, tensile strength 22 MPa, modulus of elasticity 1627 MPa, flexural strength 41 MPa, flexural modulus 1834 MPa – more information [Dimension 2011]), while the construction height was constantly 250 microns. Furthermore, a required type of sparse support was selected. In case of FDM printing, solid or sparse high construction materials with high material density was chosen. The structure of Solid offers full internal structure, while on the opposite, the Sparse High enables forming a lightweight internal structure, leading to decrease of construction material consumption and shortening the time necessary to print the model.
In total, we printed 10 types of models. Two models created using the FDM method with difference internal structure, two models created by the PolyJet Matrix with difference in support, two models created by the SLA method with different printer machines and two models created by SLS method with different orientation during printing processes. The following table 1 shows the comparison of individual technologies regarding the printing time and consumption of model and supporting material.
Liberec 2017 28 5 MEASUREMENTMETHODSANDEQUIPMENTUSED
5.1 Types of 3Ddigitization
• Contact
• Non-contact 5.1.1 Contact type digitizing
• We have studied about two types of contact machines; one is coordinate measuring machine and measuring arm.
• It is very accurate measurements and scanning, but quite slows- hundred points per minute, problem with compute of coordinates when the ball end of touch probe is used. [26]
5.1.2 Non-contact type digitizing
• Optical scanners- these machines uses laser for measuring. We have studied about two kind of optical scanners; those are Atos and Rev-scan [27]
• It is very quick and quite accurate scanning and measurements, but problem lies in measuring the surface.
• Glossy surfaces or transparent components usually cannot be scanned directly.
• Problems with some complicated shapes (deep holes and ribs etc.)
5.2 History of scanning
The touch probe was developed in the 1980s. However, this digitization was very slow.
Therefore, the development of optical surface sensing technology has occurred. An advantage was the possibility of scanning Soft objects because it was not a touch method, so there was no risk of mechanical Components. The first use of the 3D scanner was capturing people. Cyberware Laboratories is based in Los Angeles she introduced her in 1980. Later on the years, the head scanner was usable in the film industry for animation. Five years Later the device was modified and it was possible to scan the entire body. In 1994, progress was made in 3D scanning and the REPLIK scanner produced by the company 3D Scanners are developing scanning of detailed subjects. Another point of development was Digibotics's 4- axis laser scanner. The principle of his scan was based on laser point dyeing. But it lacked 6 degrees of freedom, so it was not possible to cover the entire surface of the subject. This device has not yet captured colour.
Liberec 2017 29 5.3 Introduction to Atos II system
Atos system is an optical measurement system whose measurement process is based on the principles of optical triangulation, photometry and fringe projection. It is used in various industries such as construction, manufacturing, quality control, design, etc. The Atos system can ensure fast and easy digitisation of the measured objects with the relatively high resolution and precision. The most important part of the system is the optical 3d-scanner (figure 8) itself which is consisting of a projector. Each configured sensors defines the size of the 3D area in which the measured object will be scanned- so called measurement volume.
Parameters of Atos II 400 optical scanner:
Measured volume 250 X 200 X 200 mm
Weight 5,200 g
Time of 1 scan 1 second
Number of points in one can Up to 1,400,00
Point density 0.18 mm
Measurement accuracy Approx. 30 μm
Figure 10: Optical scaner Atos II 400 [28]
Atos provides dimensional measurement data and analysis of industrial components, i.e.
sheet metal parts, tools, moulds, turbine blades, castings etc. Instead of measuring individual point or by laser, Atos captures all the geometry and surface components into dense cloud and polygon. Atos is the broadest use of the system in the areas of CAD, CAM and FEM, [29] where it is necessary for the measurement of real objects and their comparison with the
Liberec 2017 30
virtual model. The entire device is designed so that the operator puts the minimal requirements users. Handling sensitive device around the head of the subject is very easy.
Since the object is located on the adjustable tripod. Also, there is no need to scan the object after regular sections (e.g.: 200), but it is enough to create the irregular images and makes auxiliary software brands will assess its position. The scanner is supplemented by computer- controlled rotary table. It finds the application in repeating of the same parts.
5.4 General explanation for Atos scanner
Atos core system mainly consists of 3 main elements, the left camera, the right camera and the projector in the middle. The definition of the object left and right results from the sensor view. The left camera and the right camera are calibrated relatively each other. This means that the angle in- between the cameras is known, therefore you may have called them as stereo cameras. The angle in-between the cameras results in a distance from the sensor to the point at which the camera rays meet. This distance is called measuring distance. At the measuring distance, the stereo cameras span a three-dimensional area. This area is called measuring volume, in which 3D points can be computed. The size of the measuring volume depends on the sensor type. The sensor projects fringes onto the measuring objects. This measuring process is captured by the 2 cameras and will result in the 3D coordinates of the camera image pixels. The resolution of the stereo cameras depends on the sensor type, but the results up to 5 million points per scan. You will get the 3D data information from those areas which were visible for the sensors. In this data are captured only for the areas which are visible for both the cameras simultaneously and are within the measurement volume. In order to capture the object entirely, more scans from different positions of the objects are necessary. However, if you take another measurement, there is no information about the correct orientation of these measurements. Reference points are placed automatically to combine the different scans together in one common coordinates system. As soon as the measurements are taken, the Atos system determines the position of the reference points in the 3D space and the distance between the reference points.
Liberec 2017 31 5.5 Spray painting
While scanning the model By using the Atos 3D optical scanner there will be some glare or shine due the reflection of brightly exposed light and formation of the shadowed areas to avoid this reflection the models were spray coated with a 3D laser scanning anti-glare spray.
Figure 11: Spray painting on the models
5.6 Adjustment and calibration of the device
Calibration of the device is necessary to adjust before each measurement; some steps are performed prior to each measurement, others after changing the optics. All the settings and control of the scanning process is performed directly by means of GOM Atos professional software [30]. According to our software we have to fit the cameras and projectors with a suitable lens and selecting the optics. Adjusting the recommended measuring distance from the calibration etalon adjusting auxiliary laser, ‘pointer’ we can see in figures. Further sensor settings only when needed. Projected focus adjustment- focusing on both cameras, setting aperture of both cameras. All the functions are accessed from the toolbar; we can use the guide showing upon choosing the desired function. Calibration of the device by means of calibration etalon can be performed by means of a guide, lies in scanning approximately 20 images of the etalon from the given positions.
Liberec 2017 32 Figure 12: Auxiliary laser pointer
5.7 Preparing parts for measurements
• Modification of surface by means of anti-reflection coating, because the part which I choose is transparent and shiny.
• Placing reference points.
• After placing the reference point we must clean.
• Mounting the scanned part to a measurement table. [31]
Figure 13: Preparation for measurement
Liberec 2017 33 5.8 Rules for placing the reference points On flat or slightly curved surfaces.
• We cannot place the points too close to the edges – issue of filling the opening.
• Reference should be appropriately distributed throughout the whole length, width and height of the measurement volume.
• Use as many reference points for the given measurement volume, so that sensor can reliably identify at least three reference points from the previous measurements.
• Do not place reference points in straight line.
• If we want to scan the part from both the sides, we need to place at least three reference points around the part to connect the partial measurement series.
• When scanning the flat surfaces, we do not place the points directly on the opposite spot (risk of point substitution = transformation errors).
5.9 Optical 3D digitizing
The Atos system is depending on the triangulation principle. The sensor unit projects deal with different patterns on the object that to be measured and observe them with two cameras, based on the optical transformation equations, the computer automatically measure the 3D coordinates for each camera pixel with high precision. Depends on the camera resolution, a point cloud of up to 4 million surfaces points can results for each individual measurement [32].
To digitize an object, several individual measurements from various views are required.
Transformation of global coordinate system can be done automatically by means of reference points. So that we can observe the digitization processes continuously on the screen. Each individual completes the building up of the 3D model of the object to be scanned. Finally, at the end of the digitizing process, a high - resolution polygon mesh of surface completely describe the object.
Liberec 2017 34 6 ANALYSISOFMANUFACTURINGACCURACY
Firstly, an analysis of dimensional accuracy was performed. The analysis consisted of inspecting the diameters of spherical and cylindrical surfaces, length dimensions or spacing of the individual elements. Basic geometrical elements (cylinders, spheres, planes, etc.) were calculated by interlacing the fitting elements with Gauss BestFit for 3σ (Figure 14) [33]. In addition to external and internal diameters, horizontal and vertical cylinders were also evaluated.
6.1 Analysis of Deviations Day 1
Figure 14: Inspection of diameters and the dimension of sphere and spacing between them Table 2: Deviation of the nominal cylinder diameters
Printer Cylinder 1 Ø [mm]
Cylinder 2 Ø [mm]
Cylinder 3 Ø [mm]
Cylinder 4 Ø [mm]
FDM Dimension 0.16 0.13 0.09 -0.21
FDM Dimension
Spare 0.12 0.09 0.04 -0.18
FDM Fortus 0.16 0.16 -0.01 -0.1
FDM FortusSpare 0 0.06 -0.07 -0.08
Poly jet Matte -0.21 -0.27 -0.21 0.22
Poly jet Glossy -0.02 -0.04 -0.03 0.01
SLS Vertical 0.04 -0.11 -0.02 0.05
SLS Horizontal -0.15 -0.32 -0.21 -0.05
SLA 0.1 0.03 0.03 -0.04
SLA Ultra 0.26 0.23 0.02 -0.19
Liberec 2017 35 Figure 15: Graphical view of the diameters deviation
Table 3: Deviation of the nominal cylindricity diameters
Printer
Cylinder 1
⌭[mm]
Cylinder 2
⌭[mm]
Cylinder 3
⌭[mm]
Cylinder 4
⌭[mm]
FDM Dimension 0.2 0.14 0.21 0.13
FDM Dimension
Spare 0.19 0.15 0.13 0.22
FDM Fortus 0.16 0.14 0.11 0.1
FDM FortusSpare 0.23 0.15 0.11 0.11
Poly jet Matte 0.2 0.21 0.11 0.16
Poly jet Glossy 0.08 0.12 0.08 0.07
SLS Vertical 0.25 0.3 0.22 0.18
SLS Horizontal 0.12 0.27 0.13 0.11
SLA 0.11 0.08 0.09 0.06
SLA Ultra 0.32 0.26 0.06 0.11
mm
Liberec 2017 36 Figure 16: graphical view of cylindricity deviation
Table 4: Deviations of the nominal dimensions XL and YL
Printer Dimension X1 LX [mm]
Dimension X2 LX [mm]
Dimension Y1 LY [mm]
Dimension Y2 LY [mm]
FDM Dimension -0.06 -0.05 0.01 -0.01
FDM Dimension
Spare -0.1 -0.08 -0.03 -0.04
FDM Fortus 0.09 -0.05 0.01 -0.1
FDM FortusSpare 0.1 -0.03 0.03 -0.04
Poly jet Matte -0.01 0.07 -0.02 0.08
Poly jet Glossy 0.06 0.06 -0.22 -0.1
SLS Vertical 0.16 0.15 -0.12 -0.06
SLS Horizontal -0.26 -0.09 -0.23 -0.08
SLA 0.1 -0.07 -0.11 0.07
SLA Ultra -0.21 0.06 -0.15 -0.03
mm
Liberec 2017 37 Figure 17: Graphical view of the dimension LX and LY deviation
Table 5: Deviation of the nominal sphere dimension
Printer sphere 1 [mm] sphere 2 [mm] sphere 3 [mm]
FDM Dimension -0.17 -0.17 -0.2
FDM Dimension Spare -0.19 -0.19 -0.17
FDM Fortus -0.11 -0.12 -0.13
FDM FortusSpare -0.09 -0.11 -0.12
Poly jet Matte 0.2 0.19 0.21
Poly jet Glossy 0.05 0.05 0
SLS Vertical 0.09 0.04 0.03
SLS Horizontal -0.08 -0.01 0.05
SLA -0.01 0 0.01
SLA Ultra -0.16 -0.13 -0.12
mm
Liberec 2017 38 Figure 18: Graphical view of the sphere deviation
Table 6: Deveation of spacing between the spheres
Printer spacing X [mm] spacing Y [mm]
FDM Dimension 0.07 0.05
FDM Dimension Spare 0.05 0.03
FDM Fortus 0.07 0.02
FDM FortusSpare 0.05 0
Poly jet Matte 0.01 -0.05
Poly jet Glossy 0.06 -0.06
SLS Vertical -0.011 -0.07
SLS Horizontal -0.16 -0.16
SLA -0.06 -0.07
SLA Ultra 0.22 0.02
mm
Liberec 2017 39 Figure 19: Grapgical view of sphere spacing
Table 7: Devication of Flatess Printer Flatness◊[mm]
FDM Dimention 0.15
FDM DimentionSpare 0.15
FDM Fortus 0.17
FDM FortusSpare 0.19
Poly jet Matte 0.1
Poly jet Glossy 0.11
SLS Vertical 0.23
SLS Horizontal 0.27
SLA 0.11
SLA Ultra 0.13
mm
Liberec 2017 40 Figure 20: Graphical view of Flatness
At first glance, the diameter deviation of the cylindrical elements printed by FDM Dimension shown in table 2 & figure 15 is 0.16 mm to -0.21 mm, and the cylindricity is ranges between 0.13 to 0.21 showed in the table 3 figure 16, which is little above the tolerance level and the other inspected geometries were diameters of spherical elements is(- 0.17mm to –0.2), their spacing (0.07-0.05), and absolute dimensions of the sample (table 5&6 and figure 18&19), and flatness is 0.15 as showed in the table 7 and figure 20, and inspection of absolute width showed that the samples printed by FDM Dimension method (see LX and LY dimensions in Table 4 and figure 17 ) the deviation ranged from -0.06 to 0.01 mm. Here we can observe that spacing between the spheres X and Y, and the dimension LX and LY are within the tolerance limit. In the observation the models from FDM Dimension sparse are also having the similar results.
In the FDM Fortus cylinder diameter and cylindericitytolerancesare between -0.01 mm to 0.16 mm and 0.1 mm to 0.16 mm little better than FDM dimension but it is also with in the tolerance limit, also in the sphere (-0.11 to -0.13) and flatness (0.17). The deviation of spacing between the spheres (0.02 to 0.07) and the dimensions of the LX and LY (-0.05 to 0.09) are within the tolerance limit. Compared to models from FDM Dimension, Fortus gave better results. Even models from FDM Fortus spares gave similar results like Fortus.
In the PolyJet glossy mode cylinder diameter and the cylindricity tolerances are between - 0.04 mm to 0.01 mm and 0.7 mm to 0.12 mm, in the sphere 0.0 mm to 0.05mm, and flatness (0.11 mm). The deviation of spacing between the spheres (-0.06 mm to 0.06 mm) and the
mm
Liberec 2017 41
dimensions of the LX and LY are (-0.01 mm to 0.06 mm) which is well within the tolerance limit. Compared to models from FDM Dimension &Fortus, PolyJet glossy mode gave better results. But models from PolyJet Matt mode gave poor results.
In the SLS vertical cylinder diameter and the cylindricity tolerancesare between -0.11 mm to 0.05 mm and 0.03 mm to 0.25mm, in the sphere (0.03 mm to 0.09 mm), and flatness (0.23 mm). The deviation of spacing between the spheres (-0.11 mm to -0.07 mm) and the dimensions of the LX and LY are (-0.12 mm to 0.16 mm), and some measurements are extended to the tolerance limit. Compared to models from FDM Dimension &Fortus, SLS vertical gave the better results. From the above analysed data we can observe that models from PolyJet glossy mode gave the best results so far, but models from SLS horizontal gave very poor results.
In the SLA cylinder diameter and cylindricitytolerances are -0.04 mm to 0.1 mm and 0.06 mm to 0.11 mm, in the sphere (-0.01 mm to 0.01 mm), and flatness (0.11 mm). The deviation of spacing between the spheres (-0.06 mm to -0.07 mm) and the dimensions of the LX and LY are (-0.11 mm to 0.01 mm) all the dimensions are the tolerance limit. Compare to models from the remaining technologies SLA gave the best results up to now and PolyJet glossy mode is afterword’s. But models from SLA-ultra gave comparatively poor results.
6.2 Analysis of Aging Effect (14 Days After)
Table 8: Deviation of cylinder diameters afetr 14 days
Printer
Cylinder 1 Ø [mm]
Cylinder 2 Ø [mm]
Cylinder 3 Ø [mm]
Cylinder 4 Ø [mm]
FDM Dimension 0.01 0.03 -0.02 -0.13
FDM Dimension
Spare 0.13 0.08 0.03 -0.17
FDM Fortus 0.13 0.11 -0.03 -0.09
FDM FortusSpare -0.03 0 -0.07 -0.06
Poly jet Matte -0.21 -0.25 -0.19 0.23
Poly jet Glossy -0.03 -0.04 -0.04 0.03
SLS Vertical 0.02 -0.12 -0.05 0.06
SLS Horizontal -0.15 -0.33 -0.21 -0.05
SLA 0.06 0 0.01 -0.02
SLA Ultra 0.24 0.21 0.15 -0.19
Liberec 2017 42 Figure 21: Deviation of the cylinder after 14 days
Table 9: Deviation of the cylindricity after 14 days
Printer
Cylinder 1
⌭[mm]
Cylinder 2
⌭[mm]
Cylinder 3
⌭[mm]
Cylinder 4
⌭[mm]
FDM Dimension 0.15 0.12 0.12 0.17
FDM Dimension
Spare 0.2 0.13 0.12 0.2
FDM Fortus 0.15 0.14 0.11 0.08
FDM FortusSpare 0.23 0.15 0.11 0.1
Poly jet Matte 0.19 0.22 0.12 0.15
Poly jet Glossy 0.08 0.11 0.08 0.05
SLS Vertical 0.25 0.28 0.2 0.16
SLS Horizontal 0.14 0.27 0.09 0.09
SLA 0.11 0.07 0.07 0.05
SLA Ultra 0.32 0.25 0.13 0.1
mm
Liberec 2017 43 Figure 22: Deviation of cylindricity after 14 days
Table 10: Deviations of the Dimension LX and LY after 14 days
Printer
Dimension X1 LX [mm]
Dimension X2 LX [mm]
Dimension Y1 LY [mm]
Dimension Y2 LY [mm]
FDM Dimension -0.01 -0.01 0.05 0.01
FDM Dimension
Spare -0.09 -0.07 -0.01 -0.03
FDM Fortus 0.02 -0.09 -0.06 -0.13
FDM FortusSpare 0.04 -0.09 -0.03 -0.08
Poly jet Matte 0.05 0.12 0.04 0.12
Poly jet Glossy 0.14 0.1 -0.11 -0.08
SLS Vertical 0.2 0.19 -0.06 -0.02
SLS Horizontal -0.23 -0.06 -0.19 -0.06
SLA 0.11 -0.05 -0.06 0.09
SLA Ultra -0.27 0.04 -0.16 -0.04
mm
Liberec 2017 44 Figure 23: Deviation of Dimensions LX and LY after 14 days
Table 11: Deviation of sphere after 14 days
Printer sphere 1 [mm] sphere 2 [mm] sphere 3 [mm]
FDM Dimension -0.13 -0.12 -0.13
FDM Dimension Spare -0.17 -0.17 -0.16
FDM Fortus -0.11 -0.1 -0.12
FDM FortusSpare -0.08 -0.1 -0.11
Poly jet Matte 0.23 0.23 0.24
Poly jet Glossy 0.04 0.05 0.01
SLS Vertical 0.11 0.07 0.05
SLS Horizontal -0.04 0.02 0.06
SLA 0.01 0.02 0.04
SLA Ultra -0.15 -0.13 -0.12
mm
Liberec 2017 45 Figure 24: Deviation of the sphere after 14 days
Table 12: Deviation of the spacing between the spheres after 14 days
Printer spacing X [mm] spacing Y [mm]
FDM Dimension 0.05 0.03
FDM Dimension Spare 0.05 0.03
FDM Fortus 0.04 -0.01
FDM FortusSpare -0.01 -0.04
Poly jet Matte 0.03 -0.03
Poly jet Glossy 0.08 -0.05
SLS Vertical -0.1 -0.06
SLS Horizontal -0.15 -0.16
SLA -0.06 0.07
SLA Ultra 0.21 0.01
mm
Liberec 2017 46 Figure 25: Deviaion of spacing betweeb spheres after 14 days
Table 13: Deviation of Flatness after 14 days
Printer Flatness ◊ [mm]
FDM Dimension 0.14
FDM Dimension Spare 0.15
FDM Fortus 0.14
FDM FortusSpare 0.2
Poly jet Matte 0.1
Poly jet Glossy 0.11
SLS Vertical 0.19
SLS Horizontal 0.24
SLA 0.11
SLA Ultra 0.14
mm
Liberec 2017 47 Figure 26: Deviation fo Flatness after 14 days
These are the results of the models after 14 days maintaining in room temperature and comparing with the day 1 model. Here in table 8 and figure 21 we can observe that errors after 14 days are similar to Day 1 errors. So the difference (change in dimensions) is minimal for example the models from SLA and PolyJet glossy mode had change the dimensions very less (0.04 to 0.05 mm) compare to others.
Even in table 9 and table 11 and figure 22 and 24 we can see that SLA and PolyJet glossy mode gave the best results. In the figure 14 FDM Dimension and SLS vertical also gave the better results but these machines are not maintaining the same minimal deviations in all areas. SLA and PolyJet glossy mode are maintained in all the geometries.
Here in this observation we can see that even the models from the best technologies have some geometrical and dimensional changes and by changing the time these deviations of some of the models are within the limit. Among these, the sparse models show greater deviation in the aging effect.
mm
Liberec 2017 48 6.3 Analysis of Aging Effect (84 Days After)
Table 14: Deviation of cylinnder diameters after 84days
Printer Cylinder 1 Ø [mm]
Cylinder 2 Ø [mm]
Cylinder 3 Ø [mm]
Cylinder 4 Ø [mm]
FDM Dimension 0.01 0.02 -0.02 -0.14
FDM Dimension
Spare 0.02 0.03 -0.02 -0.13
FDM Fortus 0.12 0.09 -0.03 -0.09
FDM FortusSpare -0.04 -0.02 -0.08 -0.07
Poly jet Matte -0.21 -0.26 -0.19 0.22
Poly jet Glossy -0.03 -0.04 -0.03 0.03
SLS Vertical 0.03 -0.12 -0.06 0.05
SLS Horizontal -0.16 -0.34 -0.21 -0.05
SLA 0.07 0.01 0.02 -0.03
SLA Ultra 0.23 0.2 0.18 -0.21
Figure 27: Deviation of the cylinder after 84 days mm
Liberec 2017 49 Table 15: Deviation of cylindricity after 84 days
Printer
Cylinder 1
⌭[mm]
Cylinder 2
⌭[mm]
Cylinder 3
⌭[mm]
Cylinder 4
⌭[mm]
FDM Dimension 0.15 0.14 0.13 0.15
FDM Dimension
Spare 0.14 0.13 0.12 0.16
FDM Fortus 0.16 0.14 0.1 0.08
FDM FortusSpare 0.23 0.15 0.11 0.11
Poly jet Matte 0.19 0.22 0.13 0.14
Poly jet Glossy 0.07 0.12 0.08 0.04
SLS Vertical 0.25 0.29 0.22 0.16
SLS Horizontal 0.14 0.28 0.1 0.08
SLA 0.11 0.07 0.07 0.04
SLA Ultra 0.32 0.25 0.12 0.1
Figure 28: Deviation of cylindricity after 84 days mm
Liberec 2017 50 Table 16: Dimentional deviations of LX and LY after 84 days
Printer Dimension X1 LX [mm]
Dimension X2 LX [mm]
Dimension Y1 LY [mm]
Dimension Y2 LY [mm]
FDM Dimension -0.05 -0.05 0.01 -0.02
FDM Dimension
Spare -0.07 -0.06 -0.01 -0.02
FDM Fortus -0.03 -0.12 -0.12 -0.16
FDM FortusSpare -0.01 -0.12 -0.08 -0.1
Poly jet Matte 0.01 0.1 0.01 0.1
Poly jet Glossy 0.1 0.08 -0.13 -0.09
SLS Vertical 0.18 0.18 -0.08 -0.04
SLS Horizontal -0.27 -0.08 -0.22 -0.08
SLA 0.08 -0.06 -0.06 0.07
SLA Ultra -0.33 0.01 -0.21 -0.06
Figure 29: Dimensional deviation after 84 days mm
Liberec 2017 51 Table 17: Deviation of sphere after 84 days
Printer sphere 1 [mm] sphere 2 [mm] sphere 3 [mm] spacing X [mm]
FDM Dimension -0.13 -0.13 -0.14 0.03
FDM Dimension
Spare -0.14 -0.15 -0.14 0.03
FDM Fortus -0.12 -0.12 -0.13 0
FDM FortusSpare -0.09 -0.11 -0.12 -0.03
Poly jet Matte 0.21 0.22 0.23 0.02
Poly jet Glossy 0.06 0.06 0 0.07
SLS Vertical 0.11 0.06 0.06 -0.11
SLS Horizontal -0.05 0.02 0.06 -0.17
SLA -0.01 0 0.01 -0.07
SLA Ultra -0.17 -0.15 -0.14 0.2
Figure 30: Deviation of the sphere after 84 days mm
Liberec 2017 52 Table 18: Deviation of spacing between spheres after 84 days
Printer spacing X [mm] spacing Y [mm]
FDM Dimension 0.03 0.01
FDM Dimension Spare 0.03 0.02
FDM Fortus 0 -0.03
FDM FortusSpare -0.03 -0.05
Poly jet Matte 0.02 -0.04
Poly jet Glossy 0.07 -0.05
SLS Vertical -0.11 -0.07
SLS Horizontal -0.17 -0.17
SLA -0.07 0.06
SLA Ultra 0.2 0
Figure 31: Divination of spacing between spheres after 84 days mm
Liberec 2017 53 Table 19: Deviation of Flatness after 84 days
Printer Flatness ◊ [mm]
FDM Dimension 0.14
FDM Dimension Spare 0.14
FDM Fortus 0.16
FDM FortusSpare 0.18
Poly jet Matte 0.1
Poly jet Glossy 0.11
SLS Vertical 0.2
SLS Horizontal 0.29
SLA 0.12
SLA Ultra 0.15
Figure 32: Deviation of Flatness after 84 days mm
Liberec 2017 54 Day 1
Day 14
Day 84
Figure 33: Compression of models in different time sets
Here after 84 days we made the scanning and compared the results with the models scanned on day 1 and we got the same results like what we got on 14 days. Here we can see the figure 27, figure 28, figure 30 and figure 33 we got the deviations same like day 14. We can also observe that there are no geometrical changes, if you go through figure 33 where you can observe there are no deviations, and therefore geometrical and dimensional deviations after 14 days.
Liberec 2017 55 6.4 Compression of Colour Maps Day 1 , Day 14 And Day 84
Figure 34: Colour maps of normal deviations Day 1
Liberec 2017 56
The analysis shows, the highest accuracy was reached by PolyJet glossy mode with the HQ setting –16 μm layer thickness. The manufacturer states that construction accuracy should range between 0.02 to 0.085 mm depending on used material, geometry of individual parts, model orientation and settings of construction parameters. This requirement was met on most cases (figure 34). FDM Fortus gave better results and overall higher than in case of the samples printed by means of PolyJet Matrix glossy mode. The declared tolerance of construction accuracy provided by the manufacturer of this printer is 0.127 mm. This requirement was met on most cases (Figure 34). Unfortunately, the other models failed to keep within the tolerance limit for construction accuracy. The analysed results of FDM Dimension samples shows that, deviations from the CAD model did not meet the required conditions. The remaining models from SLA and SLS also failed to keep within the tolerance limit the manufacturer provided the accuracy of printers for both SLS and SLA is 0.10mm, (figure 34).
Liberec 2017 57 Figure 35: Colour maps of normal deviations of real part after 14 days versus 1st day
Liberec 2017 58 Figure 36: Colour maps of normal deviations of real part after 84 days versus 1st day
Liberec 2017 59
As the analysis shows now we are going to find the deviations after 14 days and after 84 days. By the analysis of figure 34,figure 35 and figure 36 there was some deviations from all the models, compare PolyJet glossy mode day 1, day 14 and day84 there is a miner deviation it’s not even 0.005 mm, it’s a very lees deviation and this is the best result we got. And in FDM Fortus there are some deviations on figure 34 and figure 35 where we can observe there was a deviation in day 1 and day 14 of around 0.04 mm the deviation is within the tolerance limit, but in the day 14 and day 84 the deviation is almost same. In the models SLS there are some deviations figure 34 and figure 35 where we can observe there was a deviation in day 1 and day 14 of around +0.21 mm. we can observe that the geometry of the model has been expanded and the deviation has extended the tolerance limit, but in the day 14 and day 84 the deviation is almost same, and we can say there is no deviation after day 14. In SLA also there are some deviations here in figure 34 and figure 35 we can observe there was a deviation in day 1 and day 14 are around +0.15 mm here we can observe that the model has been expand the geometry and the deviation is extended the tolerance limit but in the day 14 and day 84 the deviation is almost same. Compared to all the printers, models form FDM Fortus and SLA gave the deviations in the time difference here in figure 34, figure 35 we can observe that the models are compressed in the time it maybe caused because of the poor internal structure and the deviations after 14 days are almost same figure 35 and figure 36.
