Evaluation of mechanical and microstructural properties for laser powder-bed fusion 316L

UPTEC Q 18021

Examensarbete 30 hp Juni 2018

Evaluation of mechanical and microstructural properties for laser powder-bed fusion 316L

Philip Eriksson

Teknisk- naturvetenskaplig fakultet UTH-enheten

Besöksadress:

Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0 Postadress:

Box 536 751 21 Uppsala

Telefon:

018 – 471 30 03 Telefax:

018 – 471 30 00 Hemsida:

http://www.teknat.uu.se/student

Abstract

Evaluation of mechanical and microstructural properties for laser powder-bed fusion 316L

Philip Eriksson

This thesis work was done to get a fundamental knowledge of the mechanical and microstructural properties of 316L stainless steel fabricated with the additive manufacturing technique, laser powder-bed fusion (L-PBF). The aims of the thesis were to study the

mechanical and microstructural properties in two different building orientations for samples built in two different machines, and to summarize mechanical data from previous research on additive manufactured

316L.

Additive manufacturing (AM) or 3D-printing, is a manufacturing technique that in recent years has been adopted by the industry due to the complexity of parts that can be built and the wide range of materials that can be used. This have made it important to understand the behavior and properties of the material, since the material differs from conventionally produced material. This also adds to 316L, which is an austenitic stainless steel used in corrosive environments.

To study the effect of the building orientation, samples of 316L were built in different orientations on the build plate. The density and amount of pores were also measured. Tensile testing and Charpy-V testing were made at room temperature. Vickers hardness was also measured. Microstructure and fracture surfaces were examined using light optical microscope (LOM) and scanning electron microscope (SEM).

The microstructure of the 316L made with L-PBF was found to have meltpools with coarser grains inside them, sometime spanning over several meltpools. Inside these coarser grains was a finer

cellular/columnar sub-grain structure. The tensile properties were found to be anisotropic with higher strength values in the

orientation perpendicular to the building direction. Also high dense samples had higher tensile properties than low dense samples. The impact toughness was found to be influenced negatively by high porosity. Hardness was similar in different orientations, but lower for less dense samples. Defects due to lack of fusing of particles were found on both the microstructure sample surfaces and fracture surfaces. The values from this study compare well with previous reported research findings.

Tryckt av: Uppsala

ISSN: 1401-5773, UPTEC Q18 021 Examinator: Åsa Kassman Rudolphi Ämnesgranskare: Urban Wiklund Handledare: Mikael Schuisky

Utvärdering av mekaniska och mikrostrukturella egenskaper hos lasersmält-pulverbädds 316L

Philip Eriksson

Komplexa komponenter där flera delar sätts ihop är vanligt vid tillverkning av komponenter med traditionella tillverkningstekniker såsom gjutning eller bearbetning. Men på senare år har additiv tillverkning, eller 3D-printing som det även kallas, blivit en teknik där detta är ett minne blott. Istället för att tillverka en komponent i flera delar så har additiv tillverkning gjort det möjligt att tillverka den i ett stycke. Additiv tillverkning har även öppnat upp möjligheten att tillverka komplexa komponenter som tidigare inte har varit möjliga, med till exempel komplexa kylkanaler, och användningen av nya material. Man kan idag tillverka alltifrån plast till mänsklig vävnad. Då additiv tillverkning gjort det möjligt att tillverka nya typer av

utformningar på komponenter och material så krävs det även förståelse för hur materialen beter sig. För metaller är detta extra viktigt eftersom man oftast avser att tillverka fullt

fungerande delar som ska användas i t.ex. motorer eller som implantat. Det gör att man måste få en förståelse för vilken mikrostruktur materialet får och hur den påverkar de mekaniska egenskaperna.

Additiv tillverkning är till skillnad från traditionella tillverkningstekniker en

tillverkningsteknik där man adderar material lager-för-lager tills man får sin komponent. Man utgår från en datormodell av sin komponent som man delar upp tunna lager. Olika additiva tillverkningstekniker adderar materialet lager-för-lager på olika sätt. En av teknikerna heter smältning av pulverbädd och illustreras i figuren nedan. Ett tunt lager med metallpulver sprids ut på en platta, en laser- eller elektronstråle smälter sedan varje lager från den datormodell som man delat upp i lager. Sedan sänks plattan ner och processen upprepas tills att man har sin färdiga komponent. Pulvret som blir över tas bort från komponenten och återanvänds för tillverkning av nästa komponent.

Figur 1 - Smältning av pulverbädd med laser

När man tillverkar material med additiv tillverkning fås en mikrostruktur som är olika i de olika riktningarna. Detta gör att materialet tål olika mycket kraft eller beter sig annorlunda i olika riktningar. Denna rapport har undersökt hur mycket kraft som materialet klarar av när man drar i materialet, hur mycket energi som krävs för att slå sönder materialet, hur hårt materialet är och hur mikrostrukturen ser ut i olika riktningar för rostfritt stål 316L tillverkat med lasersmältning av pulverbädd i två olika maskiner. I rapporten finns även en

sammanställning av tidigare forskning på materialet som tillverkats med samma teknik och vilka mekaniska egenskaper som tidigare forskning har uppnått. De mekaniska egenskaper som undersökts är vanligt undersökta egenskaper som säger mycket om hur materialet kommer att bete sig när det används i färdiga komponenter.

För att undersöka egenskaper i de olika riktningarna orienterades testproverna i två olika riktningar på byggplattan, längs med byggriktningen och vertikalt mot byggriktningen.

Byggriktningen är den riktning som materialet byggs upp lager-för-lager. För att undersöka densiteten och mikrostrukturen så kapades prover längs de två riktningarna. För att undersöka densiteten så användes Archimedes princip och ett ljusoptiskt mikroskop, och för

undersökning av mikrostrukturen och ytorna på de prover som dragits och slagits sönder användes ett ljusoptiskt mikroskop och ett svepelektronmikroskop.

Resultaten visar att densiteten skiljer sig mellan de två maskinerna. Mikrostrukturen är olika i de olika riktningarna, där materialet har en tydlig lagerstruktur på ytan längs med

byggriktningen och där mikrostrukturen har större korn som utgörs av väldigt små korn. Men defekter som osmälta pulverpartiklar och porer hittades även på dessa ytor. De mekaniska egenskaperna för drag av provet är olika i de olika riktningarna där riktningen vertikalt mot byggriktningen visar på högre draghållfasthet än riktningen längs med byggriktningen.

Proverna från maskinen som ger högre densitet hade högre hållfastighetsvärden än för de proverna med lägre densitet. Störst påverkan av densitet kunde ses där energin mättes för att slå sönder proverna, väldigt tydligt att det krävdes mycket mindre energi att slå sönder ett prov med låg densitet.

Resultaten jämför sig väl med tidigare forskning och visar ibland på egenskaper bättre än tidigare rapporterat. Som slutsats kunde också dras att metoden för hur mycket energi som krävs för att slå av prover lämpar sig väl att använda vid optimering för hur man tillverkar materialet i maskinen.

Examensarbete 30 hp på civilingenjörsprogrammet Teknisk fysik med materialvetenskap

Uppsala universitet, juni 2018

Acknowledgement

First I would like to thank everyone at Sandvik AM Center for being supportive and help- ful during my thesis, especially people in the laser group. My supervisor Mikael Schuisky for discussing results with me and for trusting me in my work. I would like to thank Richard Nordström, Håkan Nylen, Agneta Östberg and Thomas Eriksson at Stålforskningen and Sand- vik Materials Technology for preparation and testing of my samples. Simon Lövquist, Sinuhé Hernández and Böret Lindblom for helping me with the SEM analysis of my samples.

Contents

1 Introduction 1

1.1 Background . . . . 1

1.2 Aim and objectives . . . . 1

1.3 Limitations . . . . 2

2 Theory 2 2.1 Additive Manufacturing . . . . 2

2.1.1 Laser Powder-Bed Fusion . . . . 3

2.2 Stainless Steels . . . . 4

2.2.1 316 stainless steels and 316L (EN 1.4404, SS2348) . . . . 5

2.3 Mechanical Testing . . . . 6

2.3.1 Tensile Testing. . . . 6

2.3.2 Impact Toughness . . . . 6

2.3.3 Hardness. . . . 6

2.4 Literature Study . . . . 7

2.4.1 Additive Manufactured 316L . . . . 7

3 Experimental procedure 10 3.1 Materials . . . 10

3.1.1 Concept Laser Mlab Cusing . . . 10

3.1.2 Renishaw AM 250 . . . 12

3.2 Density Measurement Techniques . . . 13

3.2.1 Archimedes’ Method. . . 13

3.2.2 Cross-sectional density. . . 14

3.3 Mechanical Testing . . . 15

3.3.1 Tensile Test . . . 15

3.3.2 Impact Test (Charpy V-notch impact test) . . . 15

3.3.3 Hardness Measurement . . . 16

3.4 Examination of Microstructure and Fracture Surfaces . . . 16

3.4.1 Sample preparation. . . 16

3.4.2 Light Optical Microscope . . . 16

3.4.3 Scanning Electron Microscope . . . 16

4 Results 16 4.1 Density and parts . . . 16

4.2 Microstructure . . . 19

4.3 Mecanical Testing . . . 22

4.3.1 Tensile Test . . . 22

4.3.2 Impact Test . . . 26

4.3.3 Hardness. . . 30

5 Discussion 31 5.1 Parts and density . . . 31

5.2 Microstructure . . . 32

5.3 Mechanical Properties . . . 32

5.3.1 Tensile properties . . . 32

5.3.2 Impact properties . . . 33

5.3.3 Hardness test . . . 33

6 Conclusions 34

7 Further Development 34

List of Figures

1 L-PBF system . . . . 3

2 Scan strategies, L-PBF . . . . 4

3 Tensile curve. . . . 6

4 Vickers indent . . . . 7

5 SEM image of powder, Concept Laser . . . 11

6 Concept Laser build plates. . . 12

7 Renishaw build plate . . . 13

8 Weigh used for density measurement . . . 14

9 Density and hardness sample preperation, Concept Laser . . . 15

10 Density and hardness sample preperation, Renishaw . . . 15

11 Micrographs of cross-sections, Concept Laser . . . 17

12 Residual stresses in samples, Concept Laser . . . 18

13 Micrograph of cross-sections, Renishaw . . . 19

14 Microstructure in SEM, Concept Laser . . . 20

15 Microstructure in LOM, Concept Laser. . . 21

16 Microstructure, Concept Laser - Defects . . . 21

17 Micrographs of cross-sections, Renishaw - Defects . . . 22

18 Tensile properties for Concept Laser samples . . . 23

19 Tensile curves for Concept Laser samples . . . 23

20 Fractographs of tensile samples, Concept Laser - Overview . . . 24

21 Fractograph of tensile samples, Concept Laser - Dimples . . . 25

22 Fractograph of tensile samples, Concept laser - Pores and unmelted particles . . . 25

23 Tensile properties for Renishaw samples . . . 26

24 Tensile curves for Renishaw samples . . . 26

25 Impact toughness for Concept Laser samples . . . 27

26 Impact fractographs, Concept Laser - Overview . . . 28

27 Impact surface, Concept Laser - fracture . . . 29

28 Defect impact surface, Concept Laser - Unmolten particles . . . 29

29 Defect impact surface, Concept Laser - Oxide inclusion . . . 30

30 Impact toughness for Renishaw samples. . . 30

31 Comparison tensile values with previous research . . . 33

32 Impact toughness comparison from this study . . . 33

List of Tables

2 Mechanical properties, conventional material . . . . 5

3 Tensile properties, previous research . . . . 8

4 Hardness, previous research . . . . 9

5 Impact toughness, previous research . . . . 9

6 Mechanical properties from machine manufactures . . . 10

7 Chemecical composition of powder, Concept Laser. . . 10

8 Particle size distribution, Concept Laser . . . 11

9 Energy density of process parameters, Concept Laser . . . 11

10 Build plate information, Concept Laser . . . 12

11 Chemical composition of powder, Renishaw . . . 13

12 Energy density, Renishaw . . . 13

13 Density for Concept Laser samples . . . 17

14 Density of Renishaw samples . . . 18

15 Tensile properties for Concept Laser samples . . . 22

16 Tensile properties for Renishaw samples . . . 26

17 Impact toughness for Concept Laser samples . . . 27

18 Impact toughness for Renishaw samples. . . 30

19 Vickers hardness for Concept Laser samples . . . 31

20 Vickers hardness for Renishaw samples . . . 31

Abbreviations

AM Additive Manufacturing BD Building direction BJ Binder Jetting

CAD Computer-Aided Design L-PBF Laser Powder-Bed Fusion PBF Powder-Bed Fusion

RP Rapid Prototyping

Nomenclature

ǫf Fracture Elongation (mm/mm) σUT Ultimate Tensile Strength (MPa) σy0.2 Offset Yield Strength(0.2%) (MPa)

1 Introduction

This master thesis project was performed at Sandvik Additive Manufacturing Center in Sandviken, Sweden, during 20 weeks. The aim was to investigate the material properties of additive manufactured 316L and summarize previous research.

1.1 Background

Additive manufacturing (AM), also known as 3D-printing or rapid prototyping (RP), is a manufac- turing technique where the component is built layer-by-layer with the help of computer-aided design (CAD) softwares that slices a 3D-model of a component in to 20-200µm thin 2D cross-sections. This means that certain components can consist of several thousand layers. Metal AM has found a broad use in medical, aerospace, jewelery and automotive industry. [1–3]. When the technique was invented only polymer was possible to use, but today AM techniques are used and developed for many different materials, from metals to living cells. The shift towards these types of materials has increased the need for process and material knowledge to be able to produce AM materials that can reach the standards required for certain industries.

AM has opened up the potential to produce complex parts from CAD-designs that before were im- possible to produce in one piece, for example the Seco Tools coolant clamp and LKAB Wassara sliding case for drill hammers [4]. The Seco Tools coolant clamp consists of complex water coolant channels that are not possible to drill or cast with conventional methods. The LKAB Wassara sliding case, when manufactured conventionally it consists of several pieces that are put together, But with AM everything is made in one piece. Usually AM uses already existing materials used conventionally, but AM has also opened up the possibility for researchers and companies to produce new types of materials suited only for AM techniques, such as high-entropy alloys and amorphous metals [5].

Materials broadly used are plastics, steel alloys, super-alloys, titanium, high-temperature metals and ceramics, cemented carbides etc. Most of these materials are also used in conventional manufacturing techniques. But due to the new way of producing materials with AM, new properties of the materi- als are achieved [6]. Different properties and new use cases for AM materials made it important to understand the fundamental properties of the materials and techniques used.

1.2 Aim and objectives

To keep its competitive advantage and become a market leader in AM, Sandvik needs to continue its fundamental research on material behavior of AM materials. This thesis aimed to investigate the me- chanical and microstructural properties of stainless steel 316L manufactured with laser powder-bed fusion (L-PBF) and summarize previous research made on the material and technique.

The following questions were to be answered by this study:

• With existing machine , what are the...

– mechanical properties in different directions of the built material?

– microstructural properties in different directions of the built material?

• What properties can be expected of the soon to be installed machine dedicated for 316L?

• How do the mechanical properties and microstructure compare with previous research in the field of L-PBF and other techniques?

In order to reach the aim and answer theses questions, I will...

• produce 316L test samples in two orientations in...

– a Concept Laser Mlab Cusing, installed at Sandvik.

– a Renishaw AM 250 at Renishaw ltd, similar to the machine that will be installed during the second half of 2018.

• perform density and porosity measurements for later comparison and connection with the me- chanical properties.

• perform tensile, impact and hardness test for collection of mechanical data.

• analyze fracture surfaces and polished samples under microscope.

• conduct a literature study for papers on 316L fabricated in L-PBF.

1.3 Limitations

The project was done during 20 weeks in the beginning of 2018.

2 Theory

In this section the theory behind additive manufacturing with a focus on L-PBF technique, stainless steels, stainless steel 316L, mechanical test methods and analysis methods, is presented. In the end of this section previous research for L-PBF and 316L is presented.

2.1 Additive Manufacturing

Modern additive techniques have their start in 1980’s when Hideo Kodama at Nagoya Municipal Indus- trial Research Institute proposed a systems using photosensitive polymer resin, a mask and UV-light to form 3D objects in a layer-by-layer technique [7]. Around the same time, Charles Hull invented stereolitography, a technique that uses photosensitive polymer resin and a computer controlled laser to form 3D-objects with the same layer-by-layer approach. For many years AM, or RP as it was called at that time, was focused on polymers and the stereolitography technique, but in 1997 AeroMat produced the first additive manufactured metal using high-power laser to fuse titanium alloy powder. It was not until mid and late 2000’s that metal AM machines started coming and in recent years the industry has adopted the technology. In 2017,∼1700 metal AM machines were sold world-wide [8, 9].

Additive manufacturing is divided into 7 main categories [10]:

• Vat photopolymerization

• Material extrusion

• Material jetting

• Binder jetting

• Direct energy deposition

• Sheet lamination

• Powder-bed fusion

Some of these techniques only allow for certain materials to be used, while some are more diverse. The material properties of AM materials are strongly dependent on the AM technique used, process param- eters, material composition, etc. Common for all techniques is that the material is built layer-by-layer from a CAD-model. For metal three of the techniques are commonly used in industry, Binder jetting (BJ); Powder-bed fusion (PBF); and Direct energy deposition [3]. BJ and PBF uses a powder-bed of metal particles that are either bound together selectively using a binder agent, or melted together with the use of a high energy laser or electron beam. When using BJ, a binder agent is printed on each new powder-bed layer to form a green body that then needs to be sintered in an furnace to gain its full dense material properties [3, 11, 12].

The PBF techniques all work in a similar manner. The component is built from down and up in a protective atmosphere, an inert gas (Argon or Nitrogen) or in vacuum.

- A thin powder layer is spread on the building platform to form the powder-bed. This thin layer is only a few powder particles thick (between 20-200 µm).

- A computer controlled energy source (laser or electron beam) then selectively melts the particles to fuse them together according to the 2D cross-section given from the CAD-design.

- The building plate is then lowered a layer thickness to allow for a new powder layer to be spread on top of the previously built layer.

- This is repeated layer-by-layer until the component is fully done.

A typical laser powder-bed fusion (L-PBF) system can be seen in Figure1.

Figure 1: A laser powder-bed fusion system and its components¹

2.1.1 Laser Powder-Bed Fusion

L-PBF is the most popular techniques used for metals and with the highest growth of use [1]. L-PBF is a technique with many different abbreviations depending on machine manufacturer; Direct Metal Laser Sintering (DMLS), Selective Laser Sintering (SLS), Selective Laser Melting (SLM), LaserCusing [13, 14].

L-PBF uses a laser, usually a Nd:YAG or Yb-fiber laser [15], to selectively melt areas of the powder-bed to fuse powder particles and the underlaying structure together [16]. The power of the commercial lasers used in L-PBF systems range between 50 - 1000W. The mechanical properties of the material de- pends on many process parameters such as, laser settings (laser power, beam size, beam focus, scan speed, etc.), powder size distribution, thermal conductivity of powder, etc. [17–20]. A commonly used parameter is the volumetric energy density or energy density, that is calculated from the main laser pro- cess parameters. The unit for energy density is J/mm³and the main laser process parameters used for calculation is P (laser power), v (scan speed) or s (spot time) and l (point distance), h (hatching distance, distance between two scan lines) and t (layer thickness).

Scanning pattern used will also influence the density [21]. There are many different patterns, see Figure 2; unidirectional, zig-zag, cross-hatching (zig-zag and every layer is rotated 90^◦), island scan strategy (zig-zag or unidirectional in a checkboard formation on each layer). Using a checkboard pattern where the squares are small and the board pattern is alternated in angle reduces the residual stress within the component [3].

1SLS system schematic, url : https://commons.wikimedia.org/wiki/File:Selective_laser_melting_

system_schematic.jpg, accessed: 2018-06-01

Figure 2: Scanning strategies used in L-PBF. (a) unidirectional scan pattern and (b) zig-zag pattern. Z-axis is building direction

2.2 Stainless Steels

Stainless steel is a steel often used for its resistance to corrosion and heat. It is the gathering name of steels alloyed with a minimum of 10.5 % chromium (Cr), which forms a protective oxide layer on the surface of the steel which protect it from the corrosive environment. Alloying with other elements such as Ni, Mo, Mn, N, P, etc, is done to achieve different properties and stabilize certain phases [22]. The variety of different stainless steel alloys is based on the combination and composition of the phases ferrite (α-Fe), austenite (γ-Fe) and martensite (ǫ-Fe or α^′-Fe) [23].

Stainless steels are usually divided into five categories in respect to their microstructure:

• Austenitic stainless steels

• Ferritic stainless steels

• Duplex stainless steels (combination of ferritic and austenitic)

• Martensitic stainless steels

• Percipitation-hardening stainless steels

Austenitic

Austenitic stainless steels has a faced-center cubic (FCC) crystal structure and a grain structure consist- ing of austenite. It has a Cr-level of 16 - 25 % and an addition of up to 35 % Ni to stabilize the austenitic structure. Since so much Ni is added it also makes the austenitic stainless steels more expensive than ferritic stainless steels. They are non-magnetic and have a good formability and weldability. The tem- perature use-span is wide, from temperature below 123 K up to red-hot temperatures. Austenitic stain- less steels has a wide use and it often used in aircraft applications, food and diary industry and pulp and paper manufacturing just to mention a few [23, 24].

Ferritic

Ferritic stainless steels has a body-center cubic (BCC) crystal structure and a grain structure consisting of ferrite. The Cr-level spans from 10.5 % to greater than 25 %. Their low cost made it grow in use and they are well suited for use as light-gauge sheets. They are ferromagnetic, contains a low amount of carbon and are for some alloys poor for welding in thicker-walled pieces due to the formation of brittle martensite. They cannot be hardened by heat treatment, but strengthen by annealing. Ferritic stainless steels has a wide use in automotive exhaust systems and kitchen applications [23, 25].

Duplex

Duplex stainless steel is the newest type of stainless steel alloy. It is a combination of austenitic and ferritic structure to achieve a high strength. They have a high corrosion resistance due to the amount of Cr being more than 20 %. They are less expensive than austenitic stainless steels due to the low amount of Ni used. They are often used in -100 – 300^◦Capplications where austenitic stainless steels have been used before and where high strength is needed [23].

Martensitic

Martensitic stainless steels have either a martensitic α^′- or ǫ-phase. Martensite forms at high cooling rates, when the transformation of austenite happens diffusionless. They have a low corrosion resistance due to the low amount of alloying elements used to keep the martensitic phase stable. This reduces the corrosion resistance, but they still fill a use since they have very good mechanical properties such as high strength and hardness. They are perfect for high-temperature applications where a strong and stable material is needed [23].

Precipitation-hardening

Precipitation-hardening (PH) stainless steels are made for applications where very high strength is needed. One type of PH alloys are made from low austenitic Cr/Ni stainless steel and then transformed into martensite by either mechanical or heat treatment. The martensitic material is then hardened by precipitations based on added elements. Typical elements that are used to form precipitations are cop- per, niobium and aluminum. PH alloys can also be formed from austenite to either form semi-austenitic PH stainless steels or austenitic PH stainless steels. Compared to normal martensitic alloy grades, PH grades have a higher toughness and corrosion resistance [23, 26]. They are often used in gears, turbine blades, valves and engine components, etc [27].

2.2.1 316 stainless steels and 316L (EN 1.4404, SS2348)

Stainless steels 316 is a family of austenetic chromium-nickel stainless steels with molybdenum added for good corrosion resistance. There are several different types of 316 alloys depending on the chemical composition. 316N has added N for increased strength, 316F has added S and P for improved machin- ability, and 316L with reduced carbon content for good weldibility [23].

Stainless steel 316L has a low carbon content and 2-3 % Mo. The molybdenum increases the resistance against corrosion. It has a common use in corrosive environments such as in boilers for pulp produc- tion, furnace parts, valve and pump parts, etc. Since austenitic steels are non-magnetic they are also used as an implant material. The low carbon content reduces the formation of carbides during welding which makes it suitable for L-PBF [28]. The full density of wrought 316L is 8 g/cm³. For chemical composition of wrought 316L see Table1and for mechanical properties see Table2.

Chemical composition and mechanical properties of wrought 316L

Table 1: Chemical composition of 316L according to ASTM A240.

Element Fe C Cr Ni Mn Mo P S Si N

wt% Balance 0.03 16.0−18.0 10.0−14.0 2.0 2.0−3.0 0.045 0.03 0.75 0.1

Table 2: Mechanical values for conventionally manufactured Stainless steel 316L.

σy0.2= Offset yield strength (0.2%), σ_UT= Ultimate tensile strength, ǫ_f = Fracture elongation. These are de- scribed in section 2.3.1 Tensile testing.

Reported by Condition¹ σy0.2[MPa] σUT[MPa] ǫ_f [%] Hardness Impact toughness [J]

ASTM A276 [29] AN 170 485 40 — —

Penn stainless products [30] AN 170 485 40 217 HB (Max)

95 HRB (Max)

Song et al. [31] ST

Rolled

245 220

585 565

61.2 64.5

75.8 HRB 73.2 HRB

1AN = Annealed, ST = Solution treated

2.3 Mechanical Testing

This section will explain all the mechanical testing methods used for evaluation of the mechanical prop- erties: tensile properties, impact toughness and hardness.

2.3.1 Tensile Testing

Tensile testing is used to get information about the materials elastic and plastic properties under load.

The result is used to compare materials elastic and plastic behavior, used in the process when picking the right material for engineering applications. Tensile testing will tell about the elasticity of the ma- terial, the maximum stress the material can withstand, the maximum stress before it reaches a certain plasticity and more. The method is used for testing many materials, such as metals, alloys and plastics.

For brittle materials like ceramics, tensile testing is not suitable because they do not yield like metals.

Tensile test are performed at low strain rates ( ˙ǫ) or load rates according to existing tensile testing stan- dards. The output from the tensile test is recorded on a computer which than derives the yield strength at 0.2 % offset (σ_Y0.2), ultimate tensile strength (σUT) and elongation at fracture (ǫ_f) in Figure3[32].

Figure 3: Tensile curve and the information extracted from the curve.

2.3.2 Impact Toughness

Impact test measures the absorbed energy of a sample during fracture at high strain rate, which is called impact toughness. There are two methods for impact testing, Charpy impact test och IZOD impact test.

The testing is made according to impact test standards that exist depending on material and method used [33].

2.3.3 Hardness

Vickers hardness test uses a small diamond in the shape of a sqaure-base pyramide to make an indenta- tion with a constant load. The area of the indentation is then measured to get the hardness. The method was invented by engineers at Vickers Ltd in 1922 [34]. The weights used to make the indentation varies from 25 gf to 100 kgf (1 kgf≈9.80665 N). The two diagonals of the indentation is measured, see Figure4,

and then used to calculate the hardness using Equation1. The unit used for reporting Vickers hardness is HV.

HV≈^1.8544∗F

d² [kg f/mm²] (1)

F is the load used in kgf and d is the average value of the two measured diagonals in mm.

Figure 4: How a Vickers indentation looks like. The two diagonals, d₁ and d₂are measured for calculating the Vickers hardness. d is in mm.

2.4 Literature Study

In this section previous research on AM stainless steel 316L will be presented with a focus on L-PBF technique.

2.4.1 Additive Manufactured 316L

A lot of research have been made on the subject of mechanical and microstructural properties of AM 316L fabricated with L-PBF technique [17, 35–46].

For L-PBF technique, higher tensile strength properties than for conventionally manufactured (CM) wrought 316L is widely reported and is described to be due to the fine cellular sub-grain microstruc- ture, that is formed due to the high solidification/cooling rates, and that dislocations get concentrated at these boundaries [17, 35–37, 44, 47]. The fine microstructure is also reported to be the cause for a higher hardness for L-PBF 316L than for CM 316L [42]. The cellular sub-grains have the same orienta- tion within the same coarse grain. But among the coarser grains the cellular sub-grains are oriented in different directions. This is due to the change in direction of the heat flux and the temperature gradients during melting [48].

The orientation of the test specimens highly influence the tensile properties due to the high anisotropy between the horizontal and vertical microstructure due to meltpool boundaries between each layer.

Specimens where the load is applied in the building direction will have a lower yield and tensile strength [17, 37, 39, 40]. Suryawanshi et al. [37] also showed that scanning strategy influence the me- chanical properties and that the checker-board scanning strategy gives higher yield and tensile strength.

Fracture surfaces after tensile tests has shown by many to consist of partially molten particles, on the other hand these do not seem to have any influence on the tensile properties [17, 37, 39–41, 44]. Zhong et al. [17] conclude that the L-PBF fabricated 316L is tolerant against defects when it comes to ten- sile properties, but that the number of defects could be lowered by using a smaller layer thickness(eg.

20µm).

Cherry et al. [49] and Tucho et al. [43] both showed that the density of the material depends on the laser energy density. The density of the material will have a optimum with changed energy density. A

energy density of∼100−150 J/mm³has been reported to be the ideal energy density when using a hatch spacing around 25-130 µm and a layer thickness of 20-50 µm [38, 41, 49]. Sun et al. [38] showed that the density increased in the combination of a high scan speed and a small hatch spacing when keeping the energy density fixed. The density of the material affects the hardness of the material and mechanical properties, where high porosity gives a lower hardness [38, 43, 49]

A summary of reported mechanical properties in literature for L-PBF can be seen in Table3 (tensile properties), Table4(hardness) and Table5(impact toughness). Also a summary of tensile properties given by additive manufacturing machine manufacturers and their material is given in Table6.

Table 3: Tensile properties of AM Stainless steel 316L from literature.

σ_y0.2= Offset Yield strength (0.2%), σUT= Ultimate tensile strength, ǫ_f = Fracture elongation.

Load applied, during tensile test, along the building direction is denotedkand load applied perpendicular to the building direction is denoted⊥.

Process Reported by Machine used σy0.2[MPa] σ_UT[MPa] ǫ_f[%]

L-PBF^a Zhong et al.[17] Renishaw AM250 487±3 (k, RT) 376±3 (k, 250^◦C)

594±4 (k, RT) 461±3 (k, 250^◦C)

49±4 (k, RT) 31±4 (k, 250^◦C) L-PBF^a Saeidi [50] EOSINT M 270 456±17 (k, as-built)

419±17 (k, AN at 1100^◦C)

703±8 (k, as-built) 674±30 (k, AN at 1100^◦C)

∼48 (k, as-built)

∼55 (k, AN at 1100^◦C) L-PBF Wang et al.[35] Concept Laser M2 590±5 (unknown direction) ∼700 (unknown direction) —

L-PBF Wang et al.[35] Open architecture Fraunhofer 450±10 (⊥andk) ∼640 (⊥andk) —

L-PBF^a Bartolomeu et al.[36] SLM Solutions 125HL ∼490±5 (unknown direction) ∼640±10 (unknown direction) ∼25±2 (unknown direction) L-PBF^a Suryawanshi et al.[37] Concept Laser Mlab Cusing R

511.6±14 (SM^e-⊥) 430.4±11 (SM-k) 536.4±4 (CB^f-⊥) 448.5±20 (CB-k)

621.7±12 (SM-⊥) 509.0±3 (SM-k) 668.4±5 (CB-⊥) 527.9±7 (CB-k)

20.4±3 (SM-⊥) 12.4±1 (SM-k) 24.7±2 (CB-⊥) 11.6±1 (CB-k) L-PBF^a Casati et al.[39] Renishaw AM250 554.0±4.6 (⊥)

- (k)

684.7±4.7 (⊥) 580.7±14.5 (k)

36.3±2.1 (⊥) 25.7±12.2 (k) L-PBF^b Martens et al.[40] ReaLizer SLM 250

534±5.7 (⊥x-direction) 528±3.9 (⊥y-direction) 444±26.5 (k)

653±3.4 (⊥x-direction) 659±3.2 (⊥y-direction) 567±18.6 (k)

—

L-PBF^b Liverani et al.[41] SISMA MYSINT100 ∼525 (45^◦)

∼490 (k)

∼650 (45^◦)

∼575 (k)

∼40 (45^◦)

∼70 (k) L-PBF^c Zhang et al.[44] EOS M 290 ∼625 (0.7P₀) (⊥) ∼750 (0.7P₀) (⊥)

L-PBF^b Buchanan et al.[45] EOS M 270 493±11 (⊥) 480±13 (45^◦) 410±26 (k)

636±15 (⊥) 647±11 (45^◦) 564±17 (k)

35.1±5.0 (⊥) 55.8±3.0 (45^◦) 39.2±15.8 (k)

L-PBF^d Brytan[46] Renishaw AM125 539±3 600±3 28.0±0.5

L-PBF^b Wang et al.[51] SYNDAYA Dimetal-100 — ∼590 (⊥) ∼21.1 (⊥)

EBM®^a Zhong et al.[52] Arcam A2 253±3 (RT)

152±3 (250^◦C)

509±5 (RT) 386±3 (250^◦C)

59±3 (RT) 46±3 (250^◦C) LENS^g Zietala et al.[53] Optomec LENS MR-7 ∼576 (⊥XY)

∼479 (k)

∼776 (⊥XY)

∼703 (k)

∼33 (⊥XY)

∼46 (k) aGeometry of specimen according to standard ASTM E8

bGeometry of specimen according to standard ISO 6892-1

cGeometry of specimen according to standard ASTM E466

dGeometry of specimen according to standard ISO 2740

eSM = Single melt laser pattern

fCB = Checkboard laser pattern

gLENS = Laser Engineered Net Shaping (a direct energy deposition technique)

Table 4: Hardness of AM Stainless steel 316L from literature

Process Reported by Machine used Hardness

L-PBF Zhong et al.[17] Renishaw AM250

239±5 HV_1/10(Bottom of side surface) 219±5 HV_1/10(Top of side surface) 228±4 HV_1/10(Cross-section) L-PBF Bartolomeu et al.[36] SLM Solutions 125HL ∼229 HV_3/15(Plane⊥to BD) L-PBF Sun et al.[38] SLM Solutions 250 HL ∼216 HV_1/15

L-PBF Liverani et al.[41] SISMA MYSINT100 ∼210−240 HV₁ L-PBF Tucho et al.[43] SLM Solutions 280 HL ∼213±3 HV_5/10

L-PBF Brytan[46] Renishaw AM125 215±10 HV₁

L-PBF Wang et al.[51] SYNDAYA Dimetal-100 ∼281.6 HV_0.1/20

EBM®^a Zhong et al.[52] Arcam A2 EBM®machine ∼165 HV_1/10 (inner part)

∼153 HV_1/10 (edge part) LENS^b Zietala et al.[53] Optomec LENS MR-7 289±16 HV_0.1/10(⊥to layers)

272±35HV_0.1/10(kto layers)

aEBM = Electron Beam Melting

bLENS = Laser Engineered Net Shaping (a direct energy deposition technique)

Table 5: Impact toughness of AM Stainless steel 316L from literature Process Reported by Machine used Energy difference [J]

L-PBF^a Zhong et al.[17] Renishaw AM250 103±4 (RT, machined,⊥) 144±15 (250^◦, machined,⊥) L-PBF^a Brytan[46] Renishaw AM125 253±10 (un-notched,⊥)

L-PBF^a Yasa et al.[54] Concept laser M3 59.2±3.9 (as-built material and notch,k^b) L-PBF^a Zhukov et al.[55] EOSint M270 ∼55−96 J/cm²(as-built, 20 µm layer,⊥)

∼12−25 J/cm²(as-built, 20 µm layer,k) L-PBF Kuznetsov et al.[56] EOSint 270 112−140 (as-built, machined notch,⊥)

aSpecimen geometry according to ASTM E23

bNotch in the building direction

Table 6: Mechanical properties for 316L acquired from machine manufactures material data sheets (MDS).

Process Company σy0.2[MPa] σUT[MPa] ǫf [%] Hardness

L-PBF EOS

EOS M 100 Flexline [57]

EOSINT M 280 and 290 [58]

EOS M 290 Flexline [59]

EOS M 400-4 [60]

∼535(⊥),∼490(k) 530±60(⊥), 470±90(k)

∼500(⊥andk)

∼550(⊥),∼490(k)

∼650(⊥),∼590(k) 640±50(⊥), 540±55(k)

∼590(⊥andk)

∼650(⊥),∼590(k)

∼35(⊥),∼45(k) 40±15(⊥), 50±20(k)

∼46,7(⊥andk)

∼40(⊥),∼45(k)

—

∼89 HRB

—

∼90 HRB

L-PBF Renishaw [61] 547±3 (⊥)

494±14 (k)

676±2 (⊥) 624±17 (k)

43±2 (⊥) 35±8 (k)

198±8 HV0.5(⊥) 190±10 HV0.5(k) L-PBF Concept Laser [14]

374±5(⊥) 385±6(45^◦) 330±8(k)

650±5(⊥) 640±7(45^◦) 529±8(k)

65±4(⊥) 63±5(45^◦) 63±5(k)

20 HRC

L-PBF SLM Solutions [62] 519±25 633±28 31±6 209±2 HV10

L-PBF Farsoon Technologies [] 550±50 ≥600 ≥35 —

L-PBF 3D-systems [63]

After Stress Relief:

530±20 (⊥) 440±20 (k) Full Anneal:

370±30 (⊥) 320±20 (k)

After Stress Relief:

660±20 (⊥) 570±30 (k) Full Anneal:

610±30 (⊥) 540±30 (k)

After Stress Relief:

39±5 (⊥) 49±5 (k) Full Anneal:

51±5 (⊥) 66±5 (k)

After Stress Relief:

90±6 HRB Full Anneal:

83±4 HRB

Binder jetting Digital Metal [64]

(Höganäs Group) ∼180 ∼520 ∼50 55 HRB

Previous research and data from machine manufactures show that the mechanical properties differs a lot within the same technique and between different machines.

3 Experimental procedure

Here all the methods used for producing, testing and analyzing samples are described.

3.1 Materials

In this study two different L-PBF machines, Concept Laser Mlab Cusing and Renishaw AM250, were used to manufacture tensile, impact and hardness/density samples. The Concept Laser samples were manufactured on-site at Sandvik Additive Manufacturing Center and the Renishaw samples were man- ufactured at Renishaw in England.

3.1.1 Concept Laser Mlab Cusing Powder

Concept Laser stainless steel 316L CL20ES 10-45 µm powder was used in the Concept Laser Mlab Cus- ing. Both virgin and used powder were used. The used powder was virgin powder that had been used for one build and than sieved (62 µm sieve) to remove clusters of particles melted together.

Particle size of the virgin CL20ES powder ranged from∼12−76 µm according to the test report given by Concept Laser for the specific powder batch. The chemical composition of CL20ES virgin powder can be seen in Table7, particle size distribution in Table8and SEM images of virgin and one time used powder in Figure5.

Table 7: Chemical composition of CL20ES given by the information sheet from Concept Laser

Element Fe Cr Ni Mo Mn Si C P S

wt.% balance 17.9 12.8 2.40 1.40 0.64-0.66 0.19 <0.045 0.006

Table 8: Particle size distribution of CL20ES from Concept Laser

Volume distribution Dv10 Dv50 Dv90

µm 18.946 29.484 46.172

(a) (b)

Figure 5: (a) Virgin stainless steel 316L CL20ES powder. (b) Used once and sieved stainless steel 316L CL20ES powder.

Process parameters and machining

A Concept Laser Mlab Cusing R is a commercial machine with a 100 W continues wave fiber laser and a building area of 9 cm x 9 cm. The parameters used when producing the specimens were developed on site at Sandvik AM Center. Calculated energy density from the parameters can be seen in Table 9. Argon was used as a protective gas during the process, giving an oxygen level of ∼0.2−0.3 %.

Checkboard pattern was used with Concept Laser own LaserCUSING® scanning strategy.

Table 9: Energy density of process parameters used in Concept Laser Mlab Cusing for 316L.

Parameter Value

Calculated energy density, E_d[J/mm³] 105.2

Rods were made and then machined in to tensile specimens according to ASTM E8M (small-size spec- imens proportional to standard, specimen 4) and bars to be machined in to impact specimens with a V-notch according to ASTM E23. 6 different build plates were made, see Table10and Figure6. The position for each sample on the building plate was not denoted because of the small building area in the Concept Laser Mlab Cusing. Building direction (BD) is defined as the direction as of which the layer is built up on each other.

Table 10: Stainless steel 316L samples built in Concept Laser Mlab Cusing Built name Sample type Powder Orientation (to BD) 150 Tensile and impact Sieved 1 kand⊥

153 Tensile and impact Virgin k

154 Tensile Virgin ⊥

155 Impact Virgin ⊥

156 Hardness and density Sieved 1 No specific orientation 157 Tensile and impact Sieved 1 ⊥

(a) (b) (c)

(d) (e) (f)

Figure 6: 6 different builds with vertical and horizontal tensile and impact test specimens, and density and hardness specimens made in Concept Laser Mlab Cusing. Powder rake from left to right in the pictures. (a) 150 (k- and⊥-BD tensile and impact test specimens), (b) 153 (k-BD tensile and impact test specimens), (c) 154 (⊥-BD tensile test specimens), (d) 155 (⊥-BD impact test specimens), (e) 157 (⊥-BD tensile and impact test specimens),(f) 156 (density and hardness test specimens). The marking on each sample were made to keep track of each sample during testing.

Two of thek-BD impact test specimens in build (b) were discarded due to lack of powder during the building process but did not influence the other samples.

3.1.2 Renishaw AM 250 Powder

Renishaw stainless steel 316L-0407 powder was used in the Renishaw AM 250. The chemical composi- tion was according to the SS 316L-0407 data sheet, see Table11.