Postadress: Besöksadress: Telefon:
Box 1026 Gjuterigatan 5 036-10 10 00 (vx) 551 11 Jönköping
Surface Finishing and Corrosion
Resistance of 3D Printed Maraging
Steel
PAPER WITHIN Product development and Materials Engineering
AUTHOR: Yinan Shao
TUTOR:Donya Ahmadkhaniha
Postadress: Besöksadress: Telefon:
Box 1026 Gjuterigatan 5 036-10 10 00 (vx) 551 11 Jönköping
Scope: 30 credits
Date: 2020/5/19
1
Abstract
3D printing, also known as additive manufacturing (AM), has got rapidly developed since 1987. Compared with conventional manufacturing methods, 3D printing provides some advantages such as increasing material utilization and less waste of material. Maraging steel provides good strength and toughness without losing ductility, which has been used for the 3D printing technique.
Selective laser melting (SLM) is one of the 3D printing methods, which is mostly used for metal and alloy powder. In this thesis, selective laser melting will be used for maraging steel. 3D printing maraging steel is a new material, the research about the properties of 3D printing maraging steel is still ongoing. Corrosion resistance is one of the most important properties of maraging steel due to the high cost of corrosion. So this thesis will focus on the corrosion behavior of 3D printing maraging steel. The purpose of this thesis was to find the best heat treatment condition for high corrosion resistance and to find the relationship between microstructure and corrosion behavior of maraging steel. In this thesis, several kinds of maraging steel samples with different heat treatment conditions were used. SLM, SLM
austenized&quenched, SLM aged, conventional austenized&quenched, and
conventional aged. Besides, two kinds of solutions were produced, NaOH (pH=11.5) and Na2SO4 (pH=6.5).
To observe the microstructure, an optical microscope was used. The grain size is different between SLM and conventional samples, and also different between the samples with different heat treatment conditions.
The potentiodynamic polarization method was used to measuring the corrosion behavior. SLM samples have much lower current density, and the passivation
potential and the corrosion rate are similar compared with conventional samples. But due to the lack of further experiments, the relationship between corrosion behavior could be affected by the combined effect of several factors.
Keywords
Surface finishing, corrosion resistance, selective laser melting, microstructure, maraging steel.
2
2 Theoretical background ... 6
2.1 3D printing ... 6
2.2 Selective laser melting ... 6
2.3 Maraging steel ... 8 2.3.1 Heat treatment ... 8 2.3.1.1 Austenitizing ... 8 2.3.1.2 Quenching ... 9 2.3.1.3 Aging ... 10 2.3.1.4 Tempering ... 11 2.3.2 Martensite transformation ... 11 2.3.3 Microstructure ... 11 2.4 Corrosion ... 12 2.4.1 Corrosion types ... 13
2.4.1.1 General or uniform corrosion ... 13
2.4.1.2 Pitting ... 13 2.4.1.3 Crevice corrosion ... 14 2.4.1.4 Intergranular corrosion ... 15 2.4.1.5 Galvanic corrosion ... 15 2.4.2 Passivation ... 16 2.4.3 Corrosion test ... 16
2.4.3.1 Potentiodynamic polarization test ... 16
2.4.4 Corrosion of 3D printed samples ... 18
2.4.5 Corrosion of maraging steel ... 18
3 Method and implementation ... 20
3.1 Preparation of samples ... 20
3.1.1 Chemical composition ... 20
3.1.2 Process and post-treatment ... 20
3.1.3 Schematic diagram of samples ... 21
3.2 Experiments ... 21 3.2.1 Porosity measurement ... 21 3.2.2 Mounting ... 22 3.2.3 Grinding ... 24 3.2.4 Polishing ... 24 3.2.5 Etching ... 25 3.2.6 Microstructure obseration ... 25 3.2.7 Polarization ... 26
3
4 Findings and analysis ... 28
4.1 Porosity ... 28
4.2 Metallography after etching ... 28
4.2.1 Metallography of SLM samples ... 28
4.2.2 Metallography of Conventional samples ... 31
4.3 Metallography before polarization ... 34
4.4 Polarization ... 35
4.4.1 Polarization in Na2SO4 solution (pH=6.5) ... 35
4.4.2 Polarization in NaOH solution (pH=11.5) ... 41
5 Discussion and conclusions ... 43
5.1 Discussion of method ... 43 5.1.1 Porosity measurement ... 43 5.1.2 Heat treatment ... 43 5.1.3 Polarization ... 43 5.2 Discussion of findings ... 44 5.2.1 Metallography ... 44 5.2.2 Polarization ... 45 5.3 Conclusion ... 46 5.4 Future work ... 47 6 REFERENCE ... 48
4
3D printing is an additive manufacturing method based on digital model data. Compared with traditional manufacturing techniques, the 3D printing method provides faster production and less waste of material. Selective laser melting (SLM) is one of the 3D printing methods. Maraging steel is one of the materials that could be applied to this method. This material provides high strength and good ductility, so it is widely used in slat tracks, landing gear and helicopter undercarriages [1]
Together with mechanical properties and hardness, corrosion behavior is one of the most important properties of maraging steel to be focused, due to the high cost of corrosion.
Corrosion resistance can be affected by many parameters, such as microstructure and porosity. So it is necessary to understand the relationship between the microstructure/porosity and corrosion behavior to know the way to improve the corrosion resistance.
The microstructure and porosity could be affected by the manufacturing process or post heat treatment. Therefore, it is important to find out how the manufacturing process and post heat treatment would affect the microstructure by using potentiodynamic polarization test in this thesis.
1.2 Purpose and research questions
The purpose of this thesis is to understand the effect of microstructure and porosity from the manufacturing process and post-heat treatment, and the relationship between microstructure/porosity and the corrosion behavior. It is also important to find out how SLM affects the corrosion behavior of maraging steel. So the research questions are:
• What is the effect of different heat treatment methods on microstructure? • What is the relationship between the microstructure and the corrosion
behavior of maraging steel?
• What is the effect of SLM on the corrosion behavior compared with the conventional manufacturing method?
1.3 Delimitations
In this thesis, all conventional and SLM samples were made with a similar composition. The conditions of post-treatment were the same for both SLM and conventional samples. Besides, when choosing the tested samples, it is important to remove the
5
samples with cracks because cracks lead to a large effect on corrosion behavior. Also, porosity and surface roughness could be the factors that affect the corrosion behavior. But due to the limitation of the experimental method and implementation, the affection from porosity and surface roughness are not considered in this thesis. At last, only the optical microscope was used to observe the microstructure, so some parameters would affect the result, such as intermetallic, grain orientation, or phases, which are not considered in this thesis. The results were only used to find the relationship between porosity, microstructure and corrosion behavior.
1.4 Outline
The thesis starts with the theoretical background, including the introduction of 3D printing techniques (including materials, different methods, advantages, and challenges), followed by the selective laser melting and maraging steel, which is used in this project. Besides, corrosion of 3D printed metals and alloys and especially the corrosion of 3D printed maraging steel will be introduced.
In the next chapter, the experimental methods and materials that were used in this thesis will be introduced. Then the result of all the tests will be shown in chapter 4, including the porosity, optical microstructure, and the polarization curves. And the results will be compared and analyzed. In chapter 5, the results will be discussed. They will be concluded finally.
6
Additive manufacturing (AM), generally known as 3D printing, is the technique that produces products with 3D model data [1]. In 1987, the first commercial use of additive manufacturing occurred [2]. Since then, additive manufacturing got rapidly developed. As the name ‘additive manufacturing’, this method is to add material during production, usual layer by layer, which is different from the traditional methods, which are usually made by removing extra parts from the production. So additive manufacturing has some advantages compared with traditional techniques. Firstly, it increases material utilization with less waste of material. Secondly, compared with traditional techniques, it can produce complex shapes [3]. Besides, additive manufacturing allows personal customization.
There are several different additive manufacturing techniques, Fused deposition modeling (FDM), Stereolithography (SLA), Inkjet printing and contour crafting, selective laser melting (SLM), and many other techniques [3]. FDM usually used for thermoplastic polymer material; it is a low-cost method with a fast process [3]. SLA is one of the earliest methods used for resin material, this method provides extremely high properties, but the process is slow and expensive [4]. Inkjet printing used for ceramic materials. It is a fast and efficient method and could apply for a complex design [3].
Besides, there are several different types of materials have been used in 3D printing techniques, including metals and alloys, ceramics, concrete, polymers, and composites [3]. 3D printing materials types are not always the same. It could be powder, filament, or other types depends on the 3D printing techniques [5].
Although additive manufacturing has a lot of advantages, there are some challenges that need to be overcome.Additive manufacturing leads to porosity, which is produced between the layers and affected by the method and material [5]. Porosity has a strong influence on mechanical properties, such as ductility, tensile strength [6], corrosion resistance [8], room temperature impact strength and elongation [6], so it is important to find the most suitable method and parameters.
Also, the surface of additive manufacturing products is usually rough because of the manufacturing parameters [7], layer thickness [8] and the quality of the manufactured parts [9].
2.2 Selective laser melting
7
products. This technique uses metal powder as a material. Fig.1 shows the entire process of selective laser melting. The metal powder is distributed on the substrate plate, and only the selective area is melted by the laser. When the first layer finished, the metal powder would be placed on the previous layer and then repeat the step to get a new complete layer. Eventually, the extra powder will be removed, and the product will be revealed [10].
Fig.1. Concept of SLM process. (i) The high-power laser melts selective areas of the powder bed. (ii) The process is repeated for successive layers. (iii) Loose powder removed and
finished part revealed [10].
SLM technique is most used for metal and alloy powder, such as Titanium alloys [11] [12], Aluminum alloys [13], [14], [15], Nickel-based alloys [16], CoCr alloys [17] and Steel [18], [19], [20].
As shown in Fig.2, the research of steel and titanium-based materials for SLM techniques are more than half of overall research publications from 1999-2014 [10]. Steel and iron-based alloys account for more than 30%. So that steel and iron-based alloys are the most popular materials for SLM techniques as for today, such as stainless steel [21] and tool steel [22]. In this thesis, maraging steel will be printed by the SLM technique.
Fig.2. Research publications on SLM of various materials. Data are based on research publications on SLM, LaserCusing, and DMLS indexed by Web of Science and ScienceDirect
8
widely used in aerospace and military due to its superior strength and good ductility [24] [25] [26].
2.3.1 Heat treatment
Heat treatment is the process (by heating, heat preservation, or cooling) to make the materials achieve the expected microstructure and properties (including mechanical properties and other properties). There are a lot of heat treatment methods; some of them are commonly used for maraging steel. Such as austenitizing, quenching, aging and tempering [27]. These heat treatment methods will be introduced in the following sections.
2.3.1.1 Austenitizing
Austenite is usually used as an initial microstructure in phase transformation. So steels usually are heated to achieve austenite state and then cooled down to have expected microstructure. From the iron-carbon phase diagram shown in Fig.3, it can be seen that the austenitizing temperature is related to the carbon weight percent, from around 750°C to around 1500°C.
9
Fig.3. Iron-carbon phase diagram [28].
The different austenitizing temperatures will lead to a difference in mechanical properties [29]. The high austenitizing temperature is not always lead to good properties, and previous studies mentioned that high austenitizing temperature might cause worse mechanical properties than a suitable temperature [30] [31]. The extremely high temperature during austenitizing is called overheated, which leads to the redistribution of sulfur, and the decrease of ductility and fracture toughness [32]. There have been some researches mentioned that the optimized austenitizing temperature is generally around 900-950°C [33] [34].
2.3.1.2 Quenching
Quenching is the process that fast cooling the heated steel from austenite region, to avoid phase transformation and produce martensite [35]. The cooling rate leads to different microstructure formed.
CCT (continuous cooling transformation) diagram is widely used in determining the cooling rate during heat treatment. It could help to find the relationship between phase transformation and cooling rate [36]. As shown in Fig.4, it is clear that a higher cooling rate leads to martensite, and a lower cooling rate may lead to bainite, pearlite, or ferrite [37] [38].
10
Fig.4. CCT diagram of steel [39].
2.3.1.3 Aging
Aging, also known as age hardening, is a heat treatment method to increase some mechanical properties such as hardness and strength but may cause ductility decrease. The fundamental of age hardening is that dislocation slip obstructed by secondary phase precipitation, increasing hardness and strength of the material. [40].
The temperature and time are the main factors that affect the properties of steel. Some research has mentioned that the relationship between aging time and hardness [41] [42] [43]. They found that with the increase of aging time, the hardness increases first and then decreases after reached the peak value, this is because of the grain growth/grain recrystallization and coarsening of the precepted particles, which leads to lower hardness [42]. Besides, the relationship between aging temperature and hardness shows a similar trend, as shown in Fig.5 [42], This is because the lower aging temperature leads to under-aging, which causes μ-, S- and X- intermetallic phases precipitation. And the higher temperature leads to over-aging, which causes the austenite phase formed [42].
11
Fig.5. The effect of aging time and aging temperature on the hardness of 18Ni-300 steel [42].
The aging temperature of maraging steel is usually produced in the range 440 °C to 650 °C depending on the composition [44] [41], in order to have the precipitates (Ni3X (X = Ti, Al, Mo)) to increase the hardness of maraging steel [45].
2.3.1.4 Tempering
Martensite is known as its high hardness and strength, and it is brittle as well. Tempering is the process that helps martensite transforming into a more ductile microstructure [46]. Generally, tempering reheats the quenched steel to the temperature below A1, keeps for a certain time and then cools down [47]. A1 temperature refers to the eutectoid temperature; if the tempering temperature is higher than A1, the microstructure will transform back to austenite [48].
2.3.2 Martensite transformation
As mentioned above, the martensite transformed from austenite by quenching with a high cooling rate. As soon as the temperature achieves martensite start temperature (Ms), the martensite occurs without an incubation period. And the transformation stops when the temperature reaches martensite finish temperature (Mf) [49].
To have the maraging steel with proper mechanical properties (hardness and strength), the retaining austenite needs to be reduced [49]. The way is to control the martensite transformation temperature Ms and Mf and control the composition of the material as well. Formula (1) is one of the methods to calculate the Ms temperature [49].
Ms⁄(°C) = 550 -330wt%(C) -35wt%(Mn) -17wt%(Ni) -12wt%(Cr) -21wt%(Mo)
-10wt%(Cu) -5wt%(N) -10wt%(Si) +10wt%(Co) +30wt%(Al) (1)
2.3.3 Microstructure
The austenite has a face-centered cubic (FCC) structure, and it changes to a body-centered tetragonal (BCT) structure after the austenite transformed to martensite, which is supersaturated with carbon [50]. As mentioned before, to have a martensite microstructure, the cooling rate should be fast In this case, the carbon atom of austenite moves to the void position between iron atoms, which is smaller than the volume of the carbon atom. So the crystal matrix transforms into BCT structure [50]. Schematic diagram of austenite and martensite shows in Fig.6.
12
Fig.6. Schematic diagram of (A) FCC austenite and (B) BCT martensite [50].
It has been reported that steel with low carbon content (less than 0.6%), the shape of martensite is lath, named as lath martensite, between 0.6% and 1%, the martensite formed as a mixture of lath and plate, and the plate martensite formed when the carbon content is more than 1% [51] [52]. The typical microstructure of lath martensite and plate martensite shown in Fig.7.
Fig.7. A typical microstructure of (A) lath martensite and (B) plate martensite [53].
2.4 Corrosion
Corrosion is caused by the electrochemical reaction between the material and the environment. All metals may corrode, but with different corrosion rate. Material with higher corrosion resistance will corrode slower.
Some parameters could affect the corrosion resistance, such as surface roughness, porosity, grain size, and some other factors.
Surface roughness is one of them. The smoother surface, which means lower surface roughness, would have a lower corrosion rate [54]. Since the rough surface would cause, more electrolyte stays in the concaved area and increase the corrosion [55].
Porosity is considered as another parameter which would affect corrosion resistance. High porosity may cause corrosion attacks easier [56], this may because the local acidification occurs in the pore and leads to an anodic area in the pore and the cathodic
13 area out of the pore [57].
The grain boundary is another factor that influences corrosion resistance. The high amount of grain boundaries brings more anodic areas for corrosion occurring and leads to poor corrosion resistance [58].
Besides, the refined grain size will affect the corrosion resistance. Research has reported that the fine grain size is beneficial to corrosion fatigue of SLM 316L [59]. And some other researches have reported that the finer grain size is beneficial to corrosion resistance because it has less segregation [60] [61]. While some other researchers hold the opposite point, it has been reported that the grain refinement leads to the increasing of anodic areas, and the corrosion current of steel [58] and aluminum alloys [62] increases as well.
2.4.1 Corrosion types
Corrosion can be categorized into two major categories, general or uniform corrosion and localized corrosion. And localized corrosion includes crevice corrosion, intergranular corrosion, and pitting, galvanic corrosion [63].
2.4.1.1 General or uniform corrosion
A uniform attack is a common form of corrosion. The corrosion shows similar behavior in most areas of the surface [64]. The corrosion happens in the whole areas without preferential part on the sample for corrosion. One typical sample is that the steel exposed to the atmosphere. In general, it is not thought of as a dangerous corrosion type because the thickness reduction could be predicted, and the protection is available as well [65].
2.4.1.2 Pitting
Pitting is localized corrosion, which causes a hole on the surface of the metal. Pitting is one of the most dangerous corrosion forms because it loses little weight and difficult to be detected [66]. Pitting corrosion is usually caused by chloride, bromide, and iodide [67]. These ions may cause the corrosion to start at the inclusions or grain boundaries, and the reaction between the dissolved metal and water may cause further corrosion [68]. The mechanism of pitting shows in Fig.8.
14
Fig.8. Schematic of the chemical reactions that occur during the pitting corrosion process [69].
Once the pit occurs, the oxygen in the pit starts to be consumed, which causes the cathodic reaction to move to the surface. The cell becomes separated with all anodic reaction in the pit and cathodic on the surface and negative chlorine ions move to anodic because of the positive ions there (Mez+, H+). The anodic reaction concentration
in the pit causes the increase of acidity and leads to further corrosion [70].
2.4.1.3 Crevice corrosion
Crevice corrosion is localized corrosion on metal materials [71]. As shown in Fig.9, in the crevice region, the metal ions increase, and the dissolved oxygen decreases. Besides, the chloride ions from the environment enrich in this region. The corrosion happens concentrated on the crevice region due to the same rate of the cathodic reaction and the anodic reaction [72].
This kind of corrosion could happen in the metals or alloys with good corrosion resistance such as magnesium alloys [73], stainless steel [74] and titanium alloys [75], so it could be the most dangerous corrosion form [76].
15
2.4.1.4 Intergranular corrosion
Fig.10. (A) Schematic of intergranular corrosion. (B) Intergranular corrosion below a metal surface [78].
Intergranular corrosion is the corrosion that happens at the grain boundaries because of elemental segregation and the precipitation of the secondary phase. Shown in Fig.10, the corrosion occurs through the grain boundaries and does not affect the core of the grains [78], this is because the difference of local composition or the difference of element concentration between core and grain boundary, so that the core of the grains act as cathodic and the grain boundary act as anodic [78], in a corrosive atmosphere or corrosive fluid, a galvanic cell occurs [79].
2.4.1.5 Galvanic corrosion
Galvanic corrosion is an electrochemical process. The galvanic corrosion needs different metals, closed circuits and electrolytes, the potential exists when the metals contact. In this case, the metal with higher corrosion resistance acts as the cathode, and the lower corrosion resistance metal acts as anode [80]. Fig.11 shows a sample of Al-Cu, in this case, Al acts as anode and Cu is the cathode, with the closed circuit and the electrolyte, the aluminum will be corroded [81].
16
Fig.11. Galvanic corrosion of aluminum-copper [81].
2.4.2 Passivation
Passivation is an important technique to protect the material by using of coat, it can be applied by chemical reaction or spontaneous reaction [82]. Passivation would happen with some metals or alloys in some particular environmental conditions. Usually, iron, nickel, titanium and their alloys have passivity, while zinc, tin and their alloys have limited passivity in typical conditions [83].
2.4.3 Corrosion test
Several methods are used to test the corrosion resistance, such as salt spray test [84], immersion test [85], electrochemical impedance spectroscopy (EIS) test [86], potentiodynamic polarization test and the low cycle corrosion fatigue (LCCF) tests [87].
2.4.3.1 Potentiodynamic polarization test
The potentiodynamic polarization test is popular for laboratory corrosion testing. This method could help to find the result of corrosion rate, passivity, and the cathodic and anodic behavior [88].
To activate the electrochemical process, anode, cathode, ions, and the path which could transfer electrons are needed. Oxidation occurs at the anode and could be presented as [88]:
M → Mx+ + xe- (4)
2OH- → ½O2 + H2O + 2e- (5)
and the reduction takes place at the cathode, as shown below [88]:
Mx(OH)y + ye- → xM+ + yOH (6)
H2O + e- → H + OH- (7)
17
Cathodic Scan, in this thesis, the Anodic Scan will be used. Anodic Scan is the method to measure the corrosion-potential relationship during metal dissolution. With the results of the potential and current density, the polarization curve could be drawn and analyzed [88].
Fig.12. Theoretical polarization curve [88].
Polarization curve usually used to show the corrosion behavior of the sample at a different potential. Fig.12 shows a theoretical polarization curve. Some important parameters could help to study the corrosion behavior, such as critical current density, passivation current density, passivation and trans-passivation potential.
The scan starts from point 1 to point 2 along the positive potential direction. A is open circuit potential, which means the anodic reaction has the same reaction rate with a cathodic reaction, and the measured current will be zero.
At point C is the primary passive potential and the critical current density, the critical current density is an important parameter to determine if the passivation could happen or not. The previous study mentioned that the passivation would occur after the critical current density has been exceeded [89]. The corrosion rate starts decreasing after this point and reaches the stable current density due to the oxidation layer or passivation layer covers on the surface.
In region E-F, the passivation current density can be seen. The passivation current density is the current density when passivation is stable, between passivation potential
18
Corrosion resistance is influenced by the microstructure of materials. The microstructure of 3D printed samples will be strongly affected by the parameters such as the input energy density and the cooling rate. 3D printed samples usually have a dendritic microstructure, which is not good for both mechanical properties and corrosion resistance [55].
But some other researches show the opposite point of view. It has been reported that the 3D printed titanium alloy and cobalt-chromium alloy samples have better corrosion resistance because of the finer grain size, which leads to shallow corrosion [91]. Another literature also reported that the 3D printed aluminum alloy sample has less corrosion rate both in air and 3.5% NaCl solution compared with casting samples [92]. The reason why 3D printed aluminum alloy has high corrosion resistance is that more homogenous microstructure and better dissolution of metal elements due to the higher solidification rate [92].
For 3D printed samples, some main factors could affect corrosion resistance. One is porosity, compared with the samples with lower porosity, the samples with higher porosity have higher pitting tend [93]. Daniel et al. [94] mentioned the same result. Another factor is chemical inhomogeneities. 3D printed samples are usually not homogeneous in both scales and lengths due to the process, the elements with higher melting temperature will solidify first and concentrate on the grains, and the elements with lower melting temperature will concentrate on grain boundaries, which leads to localized corrosion [95].
2.4.5 Corrosion of maraging steel
A research mentioned that heat treatment could affect corrosion behavior. The research compared the corrosion behavior in phosphoric acid between aged and annealed maraging steel. Compared with fully annealed samples, fully aged samples have higher critical current density and passivation current density [96].
Besides, the amount of retained austenite also
affects the corrosion behavior. It has
been reported that the micro-segregation of retained austenite to a localized area causes
a bad corrosion resistance. So the low amount of retained austenite leads to a good
19
corrosion resistance [97].
Maraging steel tends to have micro galvanic corrosion because the intermetallic precipitation, such as Ni3Ti, Ni3Mo will act as cathodic while the α-Fe with low potential will act as anodic [98], so the corrosion will more happen on the grain. Another study mentioned the same conclusion due to the local strain fields of the precipitates of the martensite matrix [99].
But some others reported the opposite point, such as the corrosion occurs more on and around the grain boundaries of maraging steel; this may cause by the intermetallic, which will be produced at the grain boundary [96]. According to Dinesh’s research, with the grain size increases, the corrosion resistance increases as well. This may because the corrosion is easier happens in higher energy areas like grain boundaries, while with the larger size of the grain, the grain boundary area becomes less [100].
20
Fig.13. The test and experimental procedures.
3.1 Preparation of samples
3.1.1 Chemical composition
This thesis includes both SLM and conventional samples. It is important to have a similar chemical composition to avoid the effect of different compositions. The chemical composition of SLM and wrought maraging steel samples show in Table.1.
Table.1 The chemical composition of SLM and wrought samples.
3.1.2 Process and post-treatment
21
3D printed samples are produced by selective laser melting at the University of Pretoria. And the conventional maraging steel was bought from Matmatch GmbH company. The conventional samples were produced as below:
· Melted · Ingot casted · Re-melted
· Diffusion annealed
· Forged in two steps (with intermediate heating) · Solution annealed
· Quenched · Straightened
Fig.14 shows the process and post-treatment of the materials in this thesis. Both SLM samples and wrought samples were heat treated. The first step is austenitizing at 900 °C for 30 minutes and quenching in water. The second step is aging at 480 °C for 3 hours and cooling in air. The previous researches did not mention the relationship between heat treatment conditions and corrosion behavior, so the conditions (time and temperature) of heat treatment were selected by searching for and analyzing others’ research. Choosing conditions with better mechanical properties like hardness for heat treatment [101] [102] [103].
For the sake of reading, we will abbreviate the ‘As printed’ for SLM as ‘AP’, ‘Austenitized & quenched’ as ‘AQ’ for both SLM and wrought, and ‘Aged’ still as ‘Aged’ for both SLM and wrought samples in this thesis.
3.1.3 Schematic diagram of samples
The schematic diagram of the samples shows in Fig.15. The unit is mm. Different surfaces were named as ‘Top’, ‘Vertical’, and ‘Horizon’ for both SLM and wrought samples.
Fig.15. Schematic diagram of (A) wrought samples, (B) SLM samples.
3.2 Experiments
3.2.1 Porosity measurement
22
Fig.16. Density measurement machine.
The density is calculated by the formula (6) [105]. A is the weight of the sample in air, B is the weight of sample in measuring fluid and ρ0 is the density of measuring fluid,
which is 1 g/cm3. ρm represents the theoretical density of the samples, and it could be
measured by using the machine in Fig.16 [106]. The porosity could be calculated by the formula (7) [107].
ρ = A
A−Bρ0 (6)
Porosity = 100(1-ρ/ρm) (7)
3.2.2 Mounting
In this thesis, the volume of SLM samples is small and difficult to hold, so all SLM samples are embedded. The two different kinds of embedding had been used, as shown in Fig.17.
23
Fig.17. (A) Hot mounting samples, (B) cold mounting samples.
For the samples for the etching test, the hot mounting was used. The hot embedding was made by Struers CitoPress. The phenolic resin with a carbon filter was used for mounting. The parameters of the hot mounting shown in Table 2.
Table.2 Some parameters of hot mounting.
Heating Cooling
Temp. (°C) Time (min) Pressure (bar) Time (min) Rate
180 3.5 250 1.5 High
For the samples for the polarization test, the hot embedding could not be used due to the cable is needed on the backside of the sample. In this case, Epofix resin was used to do the cold mounting.
Fig.18. A mold of cold mounting.
The mold of cold mounting shows above in Fig.18. Firstly, the high vacuum grease should cover the inner surface of the mold so that the sample could be removed easily from the mold after solidification. The resin needs to mix with the hardener; the mix
24
affect the result. By using Presi Mecapol p310 VV, the samples were ground from the sandpaper with 80 grit till the previous surface has been removed completely. And then move to 120 grit, by changing direction with 90°, the scratch is vertical to the previous scratch. This step is finished when the new scratch covers the previous scratch completely. Then repeat this step with 220, 320, 500, 800, 1000, 2000, and 4000 grit sandpapers. The rotation speed was around 200-300 RPM, because high speed may cause the samples out of control, and the low speed needs much more time. And the water was needed during the grinding to cooling the samples’ surface and remove the particles. After the last step, the surface should look like a mirror without any scratch and then the sample needs to be cleaned and dried.
3.2.4 Polishing
The samples were polished by Struers Tegramin-30 before etching. The parameters for polishing are shown in Table 3.
Table.3 Parameters of polishing.
Surface Suspension Composition Particle size (μm) Time (min)
SiC Foil #800 22 5
SiC Foil #1000 18 5
Dur DiaP.Dac3 Unique diamond suspensions 3 15 Nap DiaP.Nap-B1 Unique diamond suspensions 1 15 Chem OP-U NonDry Non-drying colloidal silica suspension 0.04 3
Both polishing cloth and the holder should be clean completely before and after polishing. During the polishing, particles are removed and may stay on the surface of polishing cloth and the holder. It may cause scratch when it moves to the next step or after changing to another sample. The samples also need to be washed after every step and dried completely.
25
3.2.5 Etching
The etching solution is Fry’s reagent. Fry’s reagent is widely used in etching martensitic and precipitation-hardenable steel [108] [109]. The composition of Fry’s reagent used in this thesis shows in Table4. Before etching, the surface needs to be cleaned. Then immerse the surface into the solution vertically and shake to have homogeneous etching. Then the samples need to be cleaned by distilled water and dried completely. Then observe the surface, if the surface becomes matte, then observe under the microscope to see if the grain boundaries show up clearly.
Table.4 Composition of Fry’s reagent.
Composition Amount
CuCl2 5g
HCl 40ml
H2O 30ml
Ethanol 25ml
If the grain boundaries are not well observed, then repeat the steps until the microstructure could be seen clearly.
But the immerse time is different for different type samples. Some samples need to be observed more often like every 2 seconds, while some other samples need much more time, these samples are observed every 10 seconds.
3.2.6 Microstructure observation
Microstructure observation includes two parts. One is after grinding to observe the amount of porosity on the surface. The other one is after etching to observe the microstructure of the samples. By using Olympus GX71F optical microscope shown in Fig.19, the microstructure could be seen and photographed easily from the computer software.
26
The connection between the samples and the equipment shown in Fig.20. As mentioned above, conventional samples have a different shape with SLM samples. And because conventional samples have a large size while only a small area needs to be tested, so the connection method is different. As shown in Fig.20 (A), the conventional sample was clamped tightly, and the SLM samples were immersed in the solution as shown in Fig.20 (B).
Fig.20. Connection of polarization test of (A) conventional samples, (B) SLM samples.
The polarization test includes two steps. The first step is to achieve a stable potential. The open-circuit potential (OCP) need to be stable before the corrosion test start. In this case, 1200s monitor time was set to keep the open circuit potential stable to avoid the effects on the measurement.
And the second step is to draw the polarization curve, and it needs an applied voltage. The applied voltage was from -0.05V to 2.5V. The scan rate is the rate of potential change during the measurement. The scan rate would affect the information and the result of the measurement. If the scan rate is too fast, it is not allowed to have enough time to reach the stable state, which leads to misinterpretation of polarization process.
27
If the scan rate is too slow, the interfacial structure may change [110]. So it set as 0.2mV/s. All the tests should be repeated at least twice to avoid the wrong result. If the results are different, then it needs to be repeated more until the results are similar.
28
it is important to find out the relationship of porosity between SLM and Convention samples.
To find the average porosity of all the different kind of samples, the porosity of all the samples were measured. But there are some samples with cracks on the surface or in the center, which may be produced during the manufacturing. These samples with an irregularly high amount of porosity were found and not going to use in any further experiment.
The porosity of AQ (SLM) and AQ (Conventional) series shown in table.5, the difference of porosity between SLM and conventional samples is clear. The porosity of SLM samples was around 29% and the porosity of conventional samples was around 11%. So the porosity of SLM is about 3 times more than the conventional one. But due to the large difference of porosity between the samples, the relationship between porosity and corrosion behavior cannot be studied in this thesis.
Table.5 Porosity measurement (%)
Conditions AQ (SLM) AQ (Conventional)
Porosity 28.87±15.90 10.76±12.61
4.2 Metallography after etching
By observing the microstructure after etching, the difference of grain size and the grain shapes could be compared easily between different samples.
4.2.1 Metallography of SLM samples
The microstructure of 3 kinds of SLM samples (As printed, AQ and Aged) shows in Fig.21, 22 and 23, and each fig shows 3 different planes to comparing the difference between these planes. It can be seen that there are black and white parts, in metallography, the difference between black and white areas caused by the difference of the height. As all the samples have been polished before etching, the reaction during the etching leads to the difference of height and the black and white areas.
29
Fig.21. Metallography of As printed samples: (A) Top, (B) Vertical, (C) Horizon, etched by Fry’s Reagent under a 20x optical microscope.
Fig.21. shows the microstructure of different planes of As printed samples. It can be found that the microstructure as printed are similar in different planes, and the grain boundary is difficult to be found.
30
Fig.22. Metallography of AQ (SLM) samples: (A) Top, (B) Vertical, (C) Horizon, etched by Fry’s Reagent under a 20x optical microscope.
Fig.22 and fig.23 show the microstructure of AQ(SLM) and Aged(SLM). It can be seen that these two series have homogeneous microstructure. By comparing the microstructure of different planes, it can be found that both AQ(SLM) and
Aged(SLM) samples have similar microstructure in different planes so all the planes could be used in the comparison between different kinds of samples.
31
Fig.23. Metallography of Aged (SLM) samples: (A) Top, (B) Vertical, (C) Horizon, etched by Fry’s Reagent under a 20x optical microscope.
Fig.21(A), fig22(A) and fig23(A) show the top planes of As printed, AQ (SLM), Aged (SLM) samples. It can be found that after austenitizing and quenching, the
microstructure of SLM samples becomes homogeneous, and the lath martensite microstructure shows up.
After aging treatment, the microstructure becomes more homogeneous, and the lath martensite can be seen clearly.
4.2.2 Metallography of Conventional samples
Fig.24 shows the schematic diagram of conventional samples, 9 different positions were chosen to be observed.
32 •
• Fig.24. Schematic diagram of Conventional samples.
•
Fig.25 shows the metallography of conventional (AQ) samples. When comparing the metallography in the same planes, it can be seen that the microstructure is different. The center (inner) part has a larger grain size than the outer part. This may cause by the temperature gradient. The solidification will start from the outer part to the inner part, so the solidification speed of the outer part is higher. The grain size is minimized with rapid solidification.
By comparing the different planes of Conventional (AQ) samples, a similar microstructure shows up.
33
Fig.25. Metallography of Conventional (AQ) samples in different positions: Vertical (A) inner, (B) middle, (C) outer; Horizon (D) inner, (E) middle, (F) outer and Top (G) inner, (H) middle, (I)
outer etched by Fry’s Reagent under a 20x optical microscope.
Fig.26 shows the metallography of conventional (Aged) samples. It shows a similar result with conventional (AQ) samples. The inner part has a larger grain size than the outer part, and different planes show a similar microstructure.
34
Fig.26. Metallography of Conventional (Aged) samples in different positions: Vertical (A) inner, (B) middle, (C) outer; Horizon (D) inner, (E) middle, (F) outer and Top (G) inner, (H) middle, (I)
outer etched by Fry’s Reagent under a 20x optical microscope.
As mentioned before, the microstructure of both Conventional (AQ) and Conventional (Aged) samples in different planes are similar, so the horizon plane could be typical samples to be compared. The grain size seems to decrease after aging.
4.3 Metallography before polarization
Fig.27 shows examples of surface for the polarization test. The surface to be tested should not have much porosity and without any scratches. As mentioned in the previous section, the different levels of porosity would cause a difference in corrosion behavior. Scratches could lead to the same problem. In this thesis, it is important to avoid the effects of these factors and compare the corrosion behavior in a similar porosity level by polarization test.
35
Fig.27. Samples of metallography before etching.
4.4 Polarization
In this thesis, 2 different kinds of solution were used in the polarization test, in order to find the corrosion behavior in a different environment.
4.4.1
Polarization in Na
2SO
4solution (pH=6.5)
As mentioned in 4.1, the porosity for SLM and conventional samples are different, but the polarization experiment will be tested on a surface of the samples, and all the tested samples were chosen carefully in 4.3. So the effects of porosity on corrosion behavior of the samples could be ignored.
Fig.28 shows the polarization curves of SLM samples in a 3.5% Na2SO4 solution. All
these three kinds of samples show similar behavior. With the potential increases, the current density decreases first and then increases until the critical current density.
Fig.28 (A), (B) and (C) show the polarization curves of three different planes of As printed, AQ, and Aged samples. It can be seen that different planes have similar corrosion behavior or important parameters.
36
37
(C)
Fig.28. Polarization curves of different planes in (A) As printed, (B) Austenitized & quenched and (C) Aged 3D printed samples in 0.5 mol Na2SO4 solution.
But when comparing between different conditions, the corrosion behavior is different. As printed samples have much higher passivation current density than Austenitized & quenched and Aged samples, although all the samples have similar critical current density. See in Fig.29.
38
(C)
Fig.29. Polarization curves of 3D printed with different post-process in (A) top, (B) vertical and (C) horizon planes in 0.5 mol Na2SO4 solution.
For Convention (AQ) and Conventional (Aged) samples, the polarization curves of different planes almost overlap, as shown in Fig.30. Also, as shown in Fig.31, for the same plane in different samples, the critical current density and the passivation current density are similar as well, which shows the same behavior with the 3D printed samples. Besides, there is secondary passivation happens in conventional samples which do not happen in 3D printed samples. But there are not enough studies on the mechanism of secondary passivation.
39
(A)
(B)
Fig.30. Polarization curves of different planes in (A) Austenitized & quenched and (B) Aged conventional samples in 0.5 mol Na2SO4 solution.
40
41
(C)
Fig.31. Polarization curves of conventional samples with different post-process in (A) top, (B) vertical and (C) horizon planes in 0.5 mol Na2SO4 solution.
The important parameters, as mentioned before, can be found in table 6. By comparing these parameters, it can be found that SLM samples have a similar value of critical current density than conventional samples. Besides, the corrosion current density pf SLM samples are lightly higher than conventional samples.
Besides, due to the high passivation potential, the passivation area only occurs when high energy is applied to the samples, all five kinds of samples do not have the passivation area under natural situations.
Table.6 Polarization data of maraging steel in Na2SO4 solution. SLM As printed SLM AQ SLM Aged Conventional AQ Conventional Aged Critical current denstiy (A/cm2) 0.065 0.055 0.069 0.020 0.018 Corrosion current density (A/cm2)
8.59E-03 1.63E-02 2.30E-02 5.10E-03 3.95E-03
Passivation potential (V)
0.436 0.766 0.843 0.593 0.702
4.4.2
Polarization in NaOH solution (pH=11.5)
NaOH solution was chosen for polarization to see whether the acidity would affect the corrosion behavior of maraging steel.
42
Fig.32. Polarization curves of SLM and conventional samples in NaOH solution.
As shown in fig.32, it is easy to be found that all 5 different kinds of samples have similar corrosion behavior in NaOH solution.
The curves are similar to the SLM samples in the Na2SO4 solution. But compared with SLM samples in Na2SO4 solution curves, as soon as it reaches the critical current density, it does not decrease, but start to have passivation.
Table.7 shows the data of maraging steel’s corrosion behavior in NaOH solution. The corrosion current density and passivation density are similar for all five kinds of samples. Besides, the passivation areas are similar as well. So the corrosion rate and corrosion behavior of all five kinds of samples are similar in NaOH
solution(pH=11.5).
Table.7 Polarization data of maraging steel in NaOH solution.
SLM As printed SLM AQ SLM Aged Conventional AQ Conventional Aged Passivation current density (A/cm2)
9.71E-07 9.03E-07 2.79E-06 9.64E-07 9.99E-07
Corrosion current density (A/cm2)
2.86E-08 2.33E-08 1.12E-07 2.96E-08 4.42E-08
Passivation potential (V) -0.115 -0.070 -0.129 -0.112 -0.133 Trans-passivation potential (V) 0.553 0.591 0.553 0.495 0.486 Passivation area 0.668 0.661 0.683 0.605 0.619
43
5
Discussion and conclusions
In this chapter, the methods and the results from previous sections will be discussed, and the conclusion and future work will be given based on the discussion.
5.1 Discussion of method
5.1.1
Porosity measurement
There are several methods for measuring porosity, such as the Archimedes’ method, a micrograph of a cross-section and X-ray scanning.
Archimedes’ method could help to measure the porosity accurately, the measurement of density for each kind of sample was repeated several times, the samples should be measured in the machine, and the density value should be recorded after a stable value shows up.
Then the porosity value of each sample is calculated and the mean value of porosity is calculated as well. The standard deviation was at a low level so that the results are reliable with low errors.
Due to the size of the machine, the samples need to be cut first. The density value shows in the machine was not a stable value, but changes in a region, so the recorded value could be higher or lower than the real value, which could affect the porosity value as well. In addition, due to the uneven distribution of pore, the pore may occur after grinding and affect the result.
5.1.2
Heat treatment
Heat treatment is one of the most important factors which could affect the microstructure. By controlling the time and temperature of heat treatment, the difference between different batches was controlled to be the smallest.
Due to the capacity of the heat treatment machine and a large amount of the samples, the heat treatment was divided into several times and two machines.
The different heating rate between the two machines leads to the small difference of microstructure. Besides, even for the same machine, different batches of samples may have a difference in heat conditions as well.
To avoid the effects of the heat treatment, it is better to do the heat treatment for all the samples together, but both the heat treatment machines do not have enough space for a large number of samples. In this case, it may be a good idea to do the heat treatment for the samples which need to be compared with each other.
5.1.3
Polarization
The polarization test was repeated several times due to the changes of the filler on the surface. At first, there was no cover on the gap between the sample and the mold as
44 electrode and the cables were changed as well.
To find the result, the polarization was tested for each sample at least twice. If the results were similar, then they are reliable. If not, then it needs to be tested for another time to see which result was correct.
Fig.33. Polarization sample.
5.2 Discussion of findings
5.2.1
Metallography
• Due to the lack of further experiment or data on metallography, it is difficult to say what is the black and white part.
45
As mentioned in the previous section, the metal elements (especially iron) may react with Fry’s reagent, but the carbide or carbon enrichment region has less reaction. So that the difference in height occurs, and according to previous research, the local carbon distribution is not homogeneous after quenching [111]. Fig.25 shows a similar result, the black and white part was not homogeneous in as printed and AQ samples, but homogeneous in Aged sample.
• The grain size is different between different samples. For SLM samples, AQ samples have finer grain size and homogenous microstructure than As printed samples which caused by the heat treatment, because austenitizing could help various phases transform into homogenous austenite and transform to martensite by quenching. And Aged samples have finer grain size than AQ samples. As it has been reported that the grain can be refined with suitable aging conditions [112]. • For conventional samples, the grain size of the different areas is different.
Compared with SLM samples, conventional samples have a larger volume, so that the cooling rate is much more difference between the inner and outer area. The outer area has a higher cooling rate than the inner area, so the grain size is different. SLM samples are small and thin, so the grain size is similar in all areas. • The SLM samples have a finer grain size than conventional samples. This is
because the SLM method uses a high-temperature laser to melt, the grain cannot grow up due to the short time [113].
5.2.2
Polarization
• In this thesis, the corrosion behavior of SLM and conventional samples are similar in NaOH solution (pH=11.5), and the AQ and Aged SLM samples are much better than conventional samples in Na2SO4 solution (pH=6.5). These can be found by
the different current density of the passivation area.
• This could be related to the SLM method that can lead to a finer microstructure due to the higher speed of solidification, it also helps to reduce the amount of the precipitates. These can help increase corrosion resistance [114].
• By analyzing the polarization curves in NaOH solution, it can be found that the iCorr in NaOH (10-7 A/cm2) is at a low level, and the passivation current density in
NaOH is at a low level as well. This is because in the alkaline environment (pH=11.5), the main reaction shows in formula (6) [115].
Fe+2OH-=Fe(OH)2+2e- (8)
Due to the high concentration of OH- in NaOH solution, promoted the occurrence
of the reaction, leads to a protective film occurs on the surface of the samples. • Due to the pre-treatment and sample preparation, the effects of porosity and
surface roughness on corrosion behavior could be ignored.
• For conventional samples, all the values related to corrosion behavior are similar, and the microstructure is similar as well, but due to the lack of data on
46
combination of various factors such as grain structure, intermetallic, and orientation relationship.
• By comparing the polarization curves between SLM and conventional samples, it can be found that SLM samples have much lower current density, and the passivation potential and the corrosion rate are similar. But the SLM samples have higher critical current density, so this may be affected by some other factors such as intermetallic and orientation relationships as mentioned in this thesis.
5.3 Conclusion
In this thesis, both wrought and SLM were chosen to produce maraging steel samples. By study their appearance, porosity, microstructure, and corrosion behavior, the conclusion will be given by answering the research questions.
• What is the effect of different heat treatment methods on microstructure?
In this thesis, aging, austenitizing and quenching were used as heat treatment for both SLM and conventional samples. By comparing the microstructure, it can be found that the microstructure, especially the grain boundary, becomes clearer, and the grain size becomes finer. On the other hand, there is no evidence to provide the relationship between the heat treatment and corrosion behavior.
• What is the relationship between the microstructure and the corrosion behavior of maraging steel?
The microstructure shows that the grain size has differences between different samples, but by comparing the corrosion behavior and data of all the samples, it can be found that the difference between 3 conditions of SLM samples is small, and the difference between 2 conditions of conventional samples is small as well. So there are no related trends between the grain size and corrosion behavior.
• What is the effect of SLM on the corrosion behavior compared with the conventional manufacturing method?
In NaOH solution, all the data and corrosion behavior are similar. But in Na2SO4 solution, it can be found that the SLM samples have lightly higher corrosion current density than conventional samples. Besides, SLM samples have higher critical current
47
density than conventional samples (about 3 times). So in this thesis, the conventional samples have lightly better corrosion resistance than SLM samples.
5.4 Future work
This thesis only includes a part of the research on the SLM maraging steel, and more researches should be followed, in order to figure out more results. The following list could be thinking as future work.
• In this thesis, only two heat treatment methods were used, in order to get more data, other heat treatment conditions could be considered, such as different heating temperature, heating time and other methods.
• Porosity may be an important factor that affects the corrosion behavior of 3D printing maraging steel. The relationship between porosity and corrosion behavior need to be found in future work.
• In this thesis, only optical microscope was used to observe the microstructure. Further microstructure observing including SEM or EBSD can be used to figure out the phases, intermetallic and orientation relationship of maraging steel.
48
[3] A. K. G. I. K. T. N. D. H. Tuan D. Ngo, "Additive manufacturing (3D printing): A review of materials, methods, applications and challenges," Composites Part B , pp. 172-196, 2018.
[4] M. J. Z. Z. J. G. D. H. Xin Wang, "3D printing of polymer matrix composites: A review and prospective," Composites Part B: Engineering, pp. 442-458, February 2017.
[5] C. B. Annamalai Pandian, "A review of recent trends and challenges in 3D printing," 2016.
[6] G. G. N. N. M. L. N. a. B. B. N. Moscovitch, "The effect of porosity on the microstructure and mechanical properties of die cast Mg alloys".
[7] S. Saunders, "Investigating Effects of Process Parameters on Surface Roughness of 3D Printed Down-Facing Surfaces," 2 October 2018. [Online]. Available: https://3dprint.com/226342/3d-print-down-facing-surfaces/.
[8] N. A. Sukindar, "Optimization of the Parameters for Surface Quality of the Open-source 3D Printing," Journal of Mechanical Engineering, pp. 33-43, June 2017.
[9] G. M.-S. ,. A. G.-C. M. G. a. D. C. Mercedes Pérez 1, "Surface Quality Enhancement of Fused Deposition Modeling (FDM) Printed Samples Based on the Selection of Critical Printing Parameters," Materials, August 2018.
[10] C. Y. Yap, C. K. Chua, Z. L. Dong, Z. H. Liu, D. Q. Zhang, L. E. Loh and S. L. Sing, "Review of selective laser melting : materials and applications," Applied Physics Reviews 2, 2015. [11] H. A. Lai‐Chang Zhang, "Selective Laser Melting of Titanium Alloys and Titanium Matrix
Composites for Biomedical Applications: A Review," Advanced Engineering Materials,
October 2015.
[12] W. Y. F. E. W. Swee LeongSing, "Selective laser melting of titanium alloy with 50 wt% tantalum: Microstructure and mechanical properties," Journal of Alloys and Compounds, pp. 461-470, March 2016.
[13] U. H. V. H. D. B. Erhard Brandl, "Additive manufactured AlSi10Mg samples using Selective Laser Melting (SLM): Microstructure, high cycle fatigue, and fracture behavior,"
Materials & Design, pp. 159-169, February 2012.
[14] T. N. Takahiro Kimura, "Microstructures and mechanical properties of A356 (AlSi7Mg0.3) aluminum alloy fabricated by selective laser melting," Materials & Design, January 2016. [15] F. C. M. L. J. P. A. A. E. P. A. M. L. P. F. D. M. Francesco Trevisan, "On the Selective Laser
Melting (SLM) of the AlSi10Mg Alloy: Process, Microstructure, and Mechanical Properties," Materials, January 2017.
49
[16] M. T. U. G. Johannes Strößner, "Mechanical and Microstructural Investigation of Nickel‐ Based Superalloy IN718 Manufactured by Selective Laser Melting (SLM)," Advanced Engineering Materials, June 2015.
[17] X.-Z. X. J. C. B. W. Nan Xiang, "Metal–ceramic bond strength of Co–Cr alloy fabricated by selective laser melting," Journal of Dentistry, June 2012.
[18] A. M. K. A. Y. Sun, "Sliding Wear Characteristics and Corrosion Behaviour of Selective Laser Melted 316L Stainless Steel," Journal of Materials Engineering and Performance,
pp. 518-526, February 2014.
[19] S. N. A. G.Casalino, "Experimental investigation and statistical optimisation of the selective laser melting process of a maraging steel," Optics & Laser Technology, pp. 151-158, January 2015.
[20] J. K. S. Kumar, "Wear Performance of SLS/SLM Materials," Advanced Engineering Materials, July 2008.
[21] Z. Sun, X. Tan, S. B. Tor and W. Y. Yeong, "Selective laser melting of stainless steel 316L with low porosity and high build rates," Materials and Design, pp. 197-204, 2016. [22] M. L. M. M. M. a. B. Mazur, "SLM additive manufacture of H13 tool steel with conformal
cooling and structural lattices," Rapid Prototyping Journal, pp. 504-518, 2016.
[23] C. Masneri, "Microstructural and mechanical properties of maraging steel parts produced by selective laser melting," 2016.
[24] S. Hossein Nedjad, J. Teimouri, A. Tahmasebifar, H. Shirazi and M. Nili Ahmadabadi, "A new concept in further alloying of Fe–Ni–Mn maraging steels," Scripta Materialia, pp. 528-531, 2009.
[25] W. S. D. Z. Guo, "Quantification of phase transformation kinetics of 18 wt.% Ni C250 maraging steel," Materials Science and Engineering: A, pp. 10-20, May 2004.
[26] W. M. G. Jr., "Ultrahigh-strength steels for aerospace applications," The Journal of The Minerals, Metals & Materials Society, pp. 20-24, May 1990.
[27] A. P. Mouritz, "11 - Steels for aircraft structures," in Introduction to Aerospace Materials, 2012, pp. 232-250.
[28] M. 3. M. L. Procedures, "FeC," [Online]. Available: https://web.utk.edu/~prack/MSE%20300/FeC.pdf. [Accessed July 2019].
[29] S. K. H. S. M. h. G. Parsa Abbaszadeha, "Effect of Austenitizing Temperature on Mechanical Properties of the Mixed Bainite - Martensite Microstructure in CrMoV Steel,"
Materials Research, 27 November 2017.
[30] G. R. T. V. A.Rajasekhar, "Influence of austenitizing temperature on microstructure and mechanical properties of AISI 431 martensitic stainless steel electron beam welds,"
Materials & Design, pp. 1612-1624, May 2009.
[31] P. K. Susil KPutatunda, "Influence of austenitizing temperature on fracture toughness of a low manganese austempered ductile iron (ADI) with ferritic as cast structure,"
Materials Science and Engineering: A, pp. 15-31, 15 August 1999.
[32] R. W. C. .. P. V. K.H. Jürgen Buschow, Encyclopedia of Materials: Science and Technology (Second Edition), 2001.
50 (Eighth Edition), 2014, pp. 473-498.
[38] T.Inoue, "12.06 - Metallo-Thermo-Mechanical Coupling in Quenching," in
Comprehensive Materials Processing, 2014, pp. 177-251.
[39] A. M, "TTT Diagrams: Introduction and Limitations | Material Engineering," [Online]. Available: http://www.engineeringenotes.com/engineering/phase-diagram/ttt-diagrams-introduction-and-limitations-material-engineering/34522.
[40] H. Z. W. W.Sha, "11 - Phase transformations in maraging steels," in Phase Transformations in Steels, 2012, pp. 332-362.
[41] M. P. C. F. M. R. d. S. J. M. N. H. F. G. A. J. M. PardalS. S. M. Tavares, "Influence of temperature and aging time on hardness and magnetic properties of the maraging steel grade 300," Journal of Materials Science, pp. 2276-2281, April 2007.
[42] C. C. X. Y. X. F. R. J. P. O. M. L. H. L. R. L. Shuo Yin, "The influence of aging temperature and aging time on the mechanical and tribological properties of selective laser melted maraging 18Ni-300 steel," Additive Manufacturing, pp. 592-600, 2018.
[43] S. T. V. T. M. D. S. D. D. S. J.M. Pardal, "Modeling of precipitation hardening during the aging and overaging of 18Ni–Co–Mo–Ti maraging 300 steel," Journal of Alloys and Compounds, pp. 109-113, 2005.
[44] I. F. B. H. F. G. d. A. Venceslau Xavier Lima Filhoa, "Influence of Solution Annealing on Microstructure and Mechanical Properties of Maraging 300 Steel," Materials Research,
2016.
[45] K. Z. M. K. W. M. &. T. K. Chaolin Tan, "Microstructural characterization and properties of selective laser melted maraging steel with different build directions," Science and Technology of Advanced Materials, pp. 746-758, 2018.
[46] S. E.Hughes, "Chapter 4 - Materials and Their Weldability," in A Quick Guide to Welding and Weld Inspection , 2009, pp. 36-48.
[47] D. M. K. B.Edenhofer, "13 - Carburizing of steels," in Thermochemical Surface Engineering of Steels, 2015, pp. 485-553.
[48] Satyendra, "The Iron-Carbon Phase Diagram," 11 March 2013. [Online]. Available: https://www.ispatguru.com/the-iron-carbon-phase-diagram/.
[49] J. Yue, The new maraging stainless steel and its strength and toughness, 2016. [50] D. J. SCHAEFFLER, "A review of steel microstructures," 1 August 2016. [Online]. Available:
https://www.thefabricator.com/article/metalsmaterials/a-review-of-steel-microstructures.
