Tribological characterisation of additively manufactured hot forming steels

(1)

Tribological characterisation of additively

manufactured hot forming steels

Anna Vikhareva

Materials Engineering, master's level (120 credits) 2020

Luleå University of Technology

(2)

Preface

The work presented in this master thesis has been carried out at the Department of Engineering and Mathematics, Division of Machine Elements, Luleå University of Technology. This work has been a part of the FAMTool project, in cooperation with Uddeholm AB and Gestamp Hardtech AB, which is funded by the strategic innovation programme Metallic Materials, a joint investment by Sweden’s Innovation Agency (Vinnova), the Swedish Energy Agency and Formas, under grant number 2019-02941.

I would like to express my deep gratitude to my supervisors Associate Professor Jens Hardell and Dr. Leonardo Pelcastre for the continuous support of my M.Sc. thesis, for the encouragement and opportunity to be a part of this project. I am particularly grateful for the training with all used equipment and willingness to give his time so generously to Leonardo Pelcastre. I would like to thank Justine Decrozant-Triquenaux who always listened to me and supported during vital fika breaks. During my stay in Sweden, I’ve learned a lot working in high temperature tribology research team. I am vastly grateful for this experience and I had the great pleasure of working with you.

I also wish to thank the EUSMAT team for assistance during the whole master program and the huge work that you are doing to provide as much valuable information for students as possible.

My sincere thanks also go to AMASE students whom I meet during the master. Having friends all over the world is a stunning privilege. Thank you all for making this time a pleasant journey.

I am extremely grateful to my family for their profound beliefs in my abilities and the opportunity to stay in Germany. This decision wasn’t easy but with your support, I knew I can do it. Thank you for always being by my side regardless of the kilometers between us.

Many thanks to my friends for listening, supporting, discussing, laughing, and crying with me when it was necessary.

Special thanks to Oleg Brauer for criticism which motivated and empowered me to move on. Thank you for believing in me and providing unending inspiration and love.

(3)

Abstract

Over the last decade, the application of ultra-high strength steel as safety components and structural reinforcements in automobile applications has increased due to their favourable high-strength-to-weight ratio. The complex shaped components are widely produced using hot stamping. However, this process encounters problems such as galling and increased wear of the tools due to harsh operating conditions associated to the elevated temperatures. Moreover, quenching is a critical step that affects the hot formed components. Slow cooling rates results in inhomogeneous mechanical properties and increased cycle time. Therefore, fast and homogeneous quenching of the formed components in combination with reduction of wear rates during hot forming are important targets to ensure the quality and efficiency of the process. The use of additive manufacturing (AM) technologies opens up potential solutions for novel tooling concepts. The manufacturing of complex shape cooling channels and integration of high-performance alloys at the surface could benefit the tribological performance in the forming operation. However, the research into high temperature tribological behaviour of AM materials in hot forming applications is very limited.

The aim of this work is to study the tribological performance of additively manufactured materials. Two steels were used – a maraging steel and modified H13 tool steel. The hot work tool steel H13 is commonly applied for dies in metal forming processes. In this thesis it was used to study additive manufacturing as the processing route instead of conventional casting. The choice of a maraging steel is motivated by a possible application of high-performance alloys as a top layer on dies. The materials were post-machined and studied in milled, ground and shot-blasted conditions. The different post-machining operations were applied to study the effect of surface finish on the tribological behaviour and also to evaluate different methods of post-machining an AM surface. As fabricated dies are usually manufactured with milled surface. During its use, the dies undergo refurbishment after certain number of cycles and the surface condition is changed to a ground surface. These surface finishes are commonly tested for hot forming applications. The shot blasted operation was chosen as alternative surface finish. The process allows to prepare large sized tools easily and the surface has beneficial compressive stresses. The tribological behaviour of AM steels was studied using a hot strip drawing tribometer during sliding against a conventional Al-Si coated 22MnB5 steel. The workpiece temperature during the tests was 600 and 700°C. The results of the tribological performance of AM materials were compared to conventionally cast tool steel QRO90.

(4)

1. Introduction

This section presents tribology, and specifically high-temperature tribology, it also discusses hot sheet metal forming and additive manufacturing and the role of tribology in these processes and summarises the main knowledge gaps.

1.1 Tribology

The science of friction, wear and lubrication between surfaces in relative motion is called

tribology. This very broad subject has been studied by scientists and applied by engineers for many

decades. It studies how two interacting solid surfaces and other tribo-elements (lubricants and ambient media) behave in relative motion, under load in natural and artificial systems. The term tribology was implemented in the 1960s by Peter Jost and the word is derived from Greek (tribos and logos/logia) as «the science of rubbing» [1].

To describe friction it is common to consider the normal load Fn which is applied to a solid body on a horizontal plane. To set the body into motion and change its speed a parallel force F is required. The result of this displacement is an opposing friction force Ft in the plane, in the reverse direction to the sliding motion (Figure 1). The coefficient of friction µ is defined by

µ= Ft/ Fn

The coefficient of static friction (µs with Ft=Ft(s)) differs from the coefficient of dynamic friction (µd with Ft=Ft(d)) because Ft(s) is the maximum force to be applied to set the solid into motion; Ft(d) is the force applied to maintain this motion [2].

Figure 1. Definition of the tangential force Ft Figure 2. Evolution of the friction force through time

with sliding without stick-slip (left) and stick-slip sliding (right)

(6)

Material degradation occurs due to microstructural or compositional changes, deformation and damage accumulation, environmental attack, effects of elevated temperatures, or the combination of all of these effects. One of the degradation models of metals is wear. When a hard material is sliding against a softer surface, wear is typically seen as material loss of the softer material [3]. Wear and friction are caused by any physical or chemical phenomenon that can occur between interfacing surfaces, i.e.: mechanical deformation, thermal fracture, chemical reactions, brittle fracture, etc. Friction and wear are closely related but are distinct phenomena and should not be considered as material properties. Both terms highly depend on the tribological system Figure 3. It includes four main elements, controlling the tribological behavior of the system. Element one and two are interacting surfaces of tribo-elements, situated in contact. The third element represents the interfacial medium, usually a lubricant, in liquid, gas or solid state. The element number four is ambient medium – the surrounding environment of air, gases, fluids or vacuum. The combination of these elements determines the friction and wear characteristics of the particular system.

Figure 3.Tribological system Figure 4. An enlarged view of the cross-sections

of two surfaces

The real surface states: they are never clean and smooth. Figure 4 shows as an example, the cross-sections of die and metal in the flat rolling process [2]. The costs associated with wear in the metal rolling industry account for nearly 10% of total production costs. In the Jost report [4] (1966) was estimated, that the correct application of the basic principles of tribology would save approximately £515 million per annum [5]. Tribology is considerably important because it benefits in terms of reliability, performance and durability of moving machine components in all machinery domains.

Mechanism of friction

(7)

friction coefficient can be obtained and referred to as hydrodynamic lubrication. For example, the metal pressing is lubricated by grease and oil, will obtain friction, partially controlled by viscous drag [3]. The major mechanisms of friction are presented in Figure 5.

Figure 5. Three major mechanisms of friction [5]

Mechanism of wear

Wear is the damage to a solid surface, involving the progressive loss of the material, because of the relative motion between the surface and a contacting material or substance [6]. High friction coefficient and wear rates are frequently caused by adhesive wear (Figure 5). If the protective films between surfaces have been removed in a sliding contact, strong adhesive bonds form across the interfaces. In effect, the opposing surfaces tear fragments from each other during their interaction. This type of wear is very severe, and it is commonly occurred in most metal working operations such as cutting, pressing and extrusion. The presence of sufficient levels of contact stresses and sliding speeds during processes provokes the adhesive wear. Another mechanism is abrasive wear (Figure 6). This type of wear is the most significant in terms of magnitude and occurrence. The earlier stages of abrasive wear is characterised by abrasive grits acting as small cutting tools which produce and remove the wear particles between interacting surfaces. Abrasive wear is progressing within the time due to repeated deformations or fatigue of worn materials. When cohesion between grains is weak, this wear occurs by intergranular separation, i.e. entire grains are pulled out by the abrasive grits. Abrasive wear is a rapid form of wear but it can be easily restrained by improving the hardness of the material [3].

(8)

Generally, there are defined phenomena, occurring between two contacting surfaces that control the wear process, regardless of the initiated mechanisms. These are the chemical and physical interactions of surfaces with lubricants and other constituents of the environment; the transmission of forces at the interface through asperities and loose wear particles; the response of a given pair of solid materials to the forces at the surface [5].

Lubrication

Friction absorbs a lot of work and wasting power, which is mainly converted to the heat at the sliding surfaces. To minimize frictional forces, the surfaces should slide easily over one another. The obvious way is to «contaminate» the asperity tips, preventing the atom-to-atom contact. Polymers and soft metals can do this, but a sufficient reduction of friction is provided by lubricants [6]. Thin low shear strength layers of lubrication are used to improve the smoothness of movement of one surface over another and to prevent damage. These layers are often difficult to observe and their thickness is varied in a range from sub-micron to a few hundred µm. There are no restrictions on the type of material required to form a lubricating film because it does not influence the limits of film effectiveness [7]. Plain oil creates a liquid film, solid lubricant such as molybdenum disulphide function as lamellar crystalline materials which allow easy sliding. Moreover, solid lubricants are essential for high temperature applications where conventional liquid lubricants may decompose. Tribological coatings are also considered as a way of solid lubrication.

(9)

1.2 High-temperature tribology

The friction and wear behaviour at elevated temperatures of metals and alloys is the key point in the effective performance of adjoined parts in internal combustion engines, cutting tools, bearings in aerospace propulsion systems and metalworking processes. The interfacial temperature between moving parts derives from an external source of heat or frictional heating at the contact as in vehicle brakes. If the system cannot be lubricated by conventional lubricants, the application can be regarded as high temperature from a tribological point of view. At about 300°C, lubricants such as oil and grease are decomposed. At high temperatures, changes occur in mechanical and thermophysical properties of the bulk as well as in surface reactivity. The wear response of metals and their mechanical properties are tied to oxidation, sulfidation and other processes that alter the chemistry of the tribo–contact [8].

When metals are exposed to air, water or acid, ionic chemical reactions between oxygen and metal molecules take place. Oxidation is a form of metal corrosion and leads to the creation of an oxide layer. Metallic oxides have a strong influence on the tribological system. Oxidation can affect wear and friction differently: protectively or detrimentally. The effect of oxidation is rapidly increased at elevated temperatures. Archard an Hirst [9] introduced to the research field of tribology the terms “mild” and “severe” wear under unlubricated conditions in 1956. Researchers have distinguished, based on observations, mild (oxidational) wear from severe (adhesive and ploughing) wear. To characterise the mild wear, Quinn made a comparison of the wear debris of steel from the results of static oxidational experiments with the published studies. He found out that contact temperature could account for the formation of oxidised debris; he observed that the rapid cooling of the surfaces once they are separated does not enable oxidation. Thus, the state of mild oxidative wear surface proved that metallic fragments were oxidized during interaction of the surfaces [9]. He also developed a model - the general theory of oxidational wear. At elevated temperatures on metals exposed to air, an oxide layer is formed and during the sliding, these oxides tend to break and generate wear debris that influences the tribological response. The types of oxides are governed by the metals and its alloying elements, the temperature and the concentration of oxygen in the atmosphere. For example, the main constituent of steel is iron (Fe) and can form: hematite (Fe2O3) which is abrasive and hard and can result in increased wear; wüstite (FeO) and

magnetite (Fe3O4) are more ductile and show better resistance to wear.

Not only the natural oxides influence the tribological contact. The formation of tribolayes and the presence of in-situ oxidation of wear debris play an important role. Different models of how these tribolayers are formed have been proposed. The wear debris could be either trapped within the formed grooves caused by wear or removed from the tribological system [10]. Pauschitz et al. [11] proposed that the tribolayers can be discerned according to chemical composition, mechanical properties, physical appearance and failure mechanism. Moreover, four different possible scenarios were stated in a sliding contact between metals. Near the ambient temperature and when one of the materials is significantly harder than the other, either No Layer formation (NL) or Transfer Layer formation (TL) can occur. At high temperature the Mechanically Mixed Layer (MML) could be formed (when materials in contact have similar hardness), as well as the Composite Layer (CL), which is formed and characterised as a hard, brittle layer with a high concentration of oxygen.

(10)

(11)

1.3 Hot sheet metal forming process

The need to manufacture automobile safety and structural components from ultra-high strength steel a few decades ago was apparent. It was motivated by the increasing demands in reduction of weight of the vehicles; improved safety, and crashworthiness qualities. The hot stamping (Figure 7) was developed and patented by a Swedish company Plannja in 1977. In the beginning, the process was used for manufacturing of saw blades and lawn mower blades. The first vehicle manufacturer Saab Automobile AB adopted and produced the hardened boron steel component for the Saab 9000 in 1984. The rapid increase in productivity was estimated from 3 million parts/year to 8 million parts/ year by 1997. The further development resulted in more hot stamped parts in the cars, up to 107 million parts/year in 2007. The main utilisation of the process in the automotive industry is the production of the car body-in-white components, such as A-pillar, B-pillar, bumper, tunnel, roof rail and rocker rail [13]. The parts are characterised by outstanding mechanical properties and a high strength-to-mass ratio. The sheet metal forming at elevated temperatures improves the formability of high-strength steels due to the increased material ductility at higher temperatures; higher mechanical properties can also be obtained thanks to rapid quenching within the dies after forming. In comparison to sheet metal forming at room temperature, hot stamping delivers a higher geometrical accuracy of the manufactured parts, because of the reduced elastic springback at elevated temperatures. The hot stamping is a complex technology and requires an understanding of both the press forming process and microstructural changes occurring during the cooling. The forming temperature, the deformation degree, the heating and the cooling rates, the austenitisation temperature and the hardening time are critical for the final properties of the product [14].

Figure 7. Direct hot stamping

Process and materials

(12)

in hot stamping are illustrated in Figure 9. The blank material with low mechanical properties and ferritic-pearlitic microstructure is heated in a furnace above the austenitisation temperature around 950 °C. After a certain soaking time the formation of a homogeneous austenitic microstructure is finished and the blank is transferred into dies with the temperature around 750 °C. The blank is formed and quenched simultaneously by the water cooled dies for 5-10s. The optimal cooling rate around 27 K/s induces the diffusionless martensitic transformation at 400°C. Fast and homogeneous cooling results in favourable mechanical properties of the produced components and shortened cycle times during the hot forming. The described process mainly applied at the industrial scale is referred to as direct hot stamping. It is used for components with a relatively small drawing depth or low degree of forming. The second variant is the indirect hot stamping, which is characterised by longer process steps and includes a cold forming step before heating of the blank. Usually, the parts with very complex shapes undergo indirect sheet metal forming [16].

Figure 8. Dies and final structural components in

hot stamping

Figure 9. Temperature vs time diagram of the

hot stamping

(13)

Table 1. Alloying elements in tool steels and obtained properties

Characteristics Elements (a)

Hot hardness W, Mo, Co (with Mo or W), V, Cr, Mn

Wear resistance V, W, Mo, Cr, Mn

Deep hardening Mn, Mo, Cr, Si, Ni, V

Minimum distortion Mo (with Cr), Cr, Mn

Toughening by grain refinement V, W, Mo, Mn, Cr

(a) Elements are arranged roughly in order of decreasing potency when added in usual amounts for the desired characteristics

During austenitisation, the formation of oxide layer occurs immediately when the steel is exposed to air. In hot stamping, to prevent the excessive oxidation also known as scaling, the sheet metal blanks are usually pre-coated with an Al-Si layer using a continuous hot-dip galvanizing process. The chemical composition of the layer includes 10% of silicon, 3% of iron and 87% of aluminium in weight percentage. The coating gives good corrosion resistance and has a good paintability and weldability. During the heating stage in hot stamping, the interdiffusion between steel and coating interfaces is thermally activated and determines the formation of Al-Fe-Si ternary phases [13]. Despite the small difference in chemical composition diverse intermetallic phases are formed increasing the hardness of the coating and additionally it becomes brittle. The most common intermetallic phases observed after the heating of Al-Si coated boron steel are illustrated in Figure 10 (b). The multi-layered structure from bottom to the top consists of: a diffusion layer between the substrate and the coating; a layer with FeAl2 and/or Fe2Al5; an intermediate layer of

Fe2Al2Si; and an unstable layer of Fe2Al7-8Si. It is clear that the tribological response from tool

steel – coated metal blank tribopair is affected by the phases in the coating as they have different physical and mechanical properties [20].

Figure 10. Intermetallic phases on the Al-Si coating: a) as-delivered and b) after heating [20]

Tribological performance during hot stamping

(14)

parts. However, it also gives rise to some problems such as transfer and build-up of the coating material onto the tool surface. The formed transferred layers influence the quality and assurance of final products. The necessity of frequent refurbishing of the tools results in reduced productivity in hot stamping. Adhesion is identified as the main acting wear mechanism. However, the abrasive wear is also frequently found [17]. Two- and three-body abrasive wear is caused by the hard Al-Si intermetallic phases in the coating or by the oxidised wear debris. It can act as a third body or get embedded to the surface and abrade the counter body.

Hernandez et al. [21] investigated the high temperature three body abrasive wear. It was observed that wear behaviour is temperature dependant. Wear mechanisms vary for each material and range of temperatures. Generally, a transition from microploughing to a combination of microcutting and microploughing has been observed. For elevated temperatures, the wear rate is increasing. Venema et al. [22] pointed out larger abrasive areas and severe form of adhesive wear - galling at high temperatures of the tools as major problems in hot stamping. Pelcastre et al. [23] identified that galling onto the uncoated tool steel surface occurred by direct adhesion (adhesive

galling) and by accumulation and compaction of wear debris (compaction galling). When more

severe compaction galling occurs, wear debris is accumulated within valleys in the surface or surface defects. During sliding, the entrapped debris starts to form compacts of intermetallics of Al-Si with oxidised debris from the tool. Adhesive galling appears to progress at the high contact pressure points such as tips of asperities or through adhesion and formation of the transfer layer.

Figure 11. Compaction galling and abrasive wear after drawing tests [22]

Figure 11 presents a Scanning Electron Microscope (SEM) analysis in Back Scattered Electron (BSE) mode of the cross sections of tool surfaces after 10 drawing tests. Compaction galling is evident at both temperatures at 500 °C (left) and 700 °C (right). The nature of the layers is distinct. At 700 °C the trapped particles are larger and therefore the fracture in the coating is temperature dependent. At high temperatures, the wear mechanisms are characterised by larger abrasive areas and occurrence of more severe compaction galling. [22].

The surface roughness of the tool steel [24] has an influence on the occurrence and severity of the transfer of Al-Si coating onto the tool steel. Additionally, the orientation of the surface lay with respect to the sliding direction significantly influences the galling behaviour. The parallel orientation of the surface lay to the sliding tends to reduce the galling tendency.

(15)

forces. The resulting friction is affected by oxidation and the strong adhesion between workpiece coating and tool.

(16)

1.4 Additive manufacturing

Additive manufacturing (AM) is one of the most rapidly developing technologies which offers new opportunities to engineers and materials scientists to fulfil a diverse range of properties of the products [26]. Three dimensional parts can be produced directly from a CAD-model and several process steps that are required to obtain the final product, e.g. casting, forming, machining etc., can be reduced. AM became a very important manufacturing method for future industrial applications. All additive manufacturing technologies are based on slicing a solid model into layers with a defined thickness. The building of parts is performed in a layer-by-layer approach. A typical AM process starts with the selection of the parts and determination of the main requirements for it. The design process begins with the creation of a computer aided design (CAD) model. The optimisation steps of design can include changes in topology for weight reduction or an application of multi-materials for improved functionality of the parts. The building layout decides the printing strategy including the part orientation and design of support structures. Then the computer-aided-manufacturing (CAM) toolpath is established and process parameters are chosen concerning applied build materials. The part is printed layer by layer following the one-layer-at the time methodology. At the end of the process the parts are cleaned, stress relieved or heat treated. The support structures are removed and the final machining is performed [27].

The ASTM classifies AM technologies into different categories: Binder jetting (BJ), Directed energy deposition (DED) and Powder bed fusion (PBF) technologies. These operate within metal processing. The PBF technology enables to build complex features, hollow cooling channels and high-precision parts. But the process is limited by single material per build and building envelope. The DED overcomes these limitations and provides a higher deposition rate but is limited by a finer geometry. BJ could be applied for high feature resolution but requires a post-process sintering to remove binders and infiltrate the pores with liquid metal. The most widely used technologies for metal 3D printing are Laser-PBF(L-PBF) and DED.

Advantages of additive manufacturing compared to conventional fabrication includes: – material efficiency – by reused powder in the process;

– resource efficiency – due to reduced auxiliary resources such as cutting tools, jigs, coolants; – part flexibility – through the fabrication of complex features in a single piece;

– production flexibility – AM machines do not require costly setups and hence small batch production is economical.

Regarding the disadvantages of AM technology:

– size limitation – AM unable to produce large sized objects due to lack of material strength; – imperfections – manufactured by AM parts usually have a rough and ribbed surface; – cost – AM equipment is expensive and the process required the cost of operation materials,

(17)

The PBF enables to processes a variety of materials such as maraging steels, 316L/304 stainless steels, Ni alloys, Co, Ti, Al and Cu alloys. The limitations include formed residual stresses that can cause distortions; repair of damaged parts is difficult; multiple materials on a single build are not available commercially [27]. A subcategory of PBF – Selective Laser melting (SLM) in Figure 12 is performed by selectively melting of powder feedstock by a high-power laser beam. The liquid melt pool cools down rapidly and forms a solid track. It offers a set of unique advantages over conventional manufacturing: full density of parts, flexibility in geometric design, rapid production of parts and high spatial resolution (cellular lightweight structures). The customisation of products with acceptable costs reduces the consumption of raw material [29]. The AM steels used in this master thesis were manufactured by SLM.

Figure 12. Schematics of powder bed fusion (PBF) technology SLM [27]

Additive manufacturing is believed by many experts as a “next generation” technology. The fields of application of AM are broad and include aerospace, military, automotive, dental, and medical industries. In tooling industry AM is applied for production of functional tool components in the small batch. For example, the manufacturing of die cast tooling considered as a promising application field [30]. Another promising field of the applications of SLM process is manufacturing of hot stamping tools with improved cooling efficiency [18]. Geometrical freedom of AM allows fabricating complex shaped cooling channels close to the surface. Hence, the heat removal from dies can be increased improving the efficiency of the process [31]. To overcome some of the limitations of conventional production of forming dies, the AM technology has a huge potential to contribute to the fabrication of the tools. It allows due to higher cooling rates during printing to obtain much more refined microstructure. Moreover, the production of functionally graded materials (FGMs) [32] in L-PBF can be used to enhance the surface properties of the dies, e.g. improve wear/corrosion resistance [18]. High performance alloys could be implemented as a top surface layer on the dies with inexpensive alloys in the bulk. Previous research has established a deposition of maraging steel on H13 tool steel by the L-PBF technique indicating success for the bimetals approach performed by S. Shakerin et al. [33] .

Additively manufactured tool steels

(18)

microstructure, soft and ductile martensite. Through the ageing treatment, its strength is enhanced by precipitations of intermetallic phases. In fusion AM processes, the cooling rates are high and implement the martensite formation in both steels. The maraging steel is can be easily processed thanks to ductile martensite while carbon-bearing steels with a brittle matrix tend to crack as a result of induced thermal stresses during the building process. Therefore, the optimisation of process parameters for tool steels to obtain crack free and dense products are crucial. For maraging steels the main focus is at the optimisation of microstructure, mechanical properties and post-process heat treatments.

The maraging steels are widely used in aerospace, tooling sectors and recently attracted interest for SLM applications. The most utilised alloy for maraging steels in AM is 18Ni-300 (1.2709, X3NiCoMoTi 18-9-5). In the literature, only a few other alloys used in AM are mentioned, such as 18Ni-250, 14Ni-200. The 18Ni-300 grade can be produced with a density above 99%. Conventionally produced (wrought) maraging steel usually results in a fully martensitic microstructure. The obtained microstructure for AM maraging steels is cellular/dendritic solidification microstructure with a significant amount of retained austenite (6-11% depending on process parameters). Regarding mechanical properties, steel produced by PBF show higher yield and ultimate tensile strength (UTS) compared to conventionally produced materials. Generally, the microstructure of AM maraging steels is finer. After ageing the hardness is significantly increased from 381 to 645 HV [34].

The most known examples of tool steels for AM are high speed steels M2 (1.3343, HS 6-5-2 C) and M3.2 (1.3294, HS 6-5-3-8). Other used steels are hot working tool steels H11 (1.2343, X38CrMoV5-1) and H13 (1.2344, X40CrMoV5-1). To obtain crack free samples for AM H13 steel the base plate is heated to at least 100°C during PBF. The microstructure is similar to maraging steels and consists of solidification dendrites with retained austenite. In as-built microstructure of AM H13 carbide precipitates can be already present. The hardness values for L-PBF produced tool steel lies in a range from 570 to 680 HV which is higher in comparison to conventionally wrought H13 material. The yield strength and UTS of these steels after tempering are similar to conventionally manufactured material [34].

Wear performance of AM steels

The application of AM steels in hot forming application implies the need for the materials to provide excellent wear resistance. Some studies on tribological performance of maraging and hot work tool steel was mainly done in pin on disc configuration. Yin et. al. [29] studied the tribological behaviour of SLM maraging steel to evaluate the difference in wear mechanisms before and after ageing. It was observed that wear mechanisms of SLM samples changed from abrasion in as-fabricated state to adhesion tribofilm in aged steel (by optimal ageing conditions), probably due to increased hardness after treatment. The coefficient of friction (COF) of the aged samples was always lower (0.586±0.071) than that of as-fabricated (0.629±0.061) because of the lubricating effect of the tribofilms.

(19)

Kumar and Kruth [36] characterised the wear behaviour of AM concept tool steel CL50WS by wear fretting test. The wear zones of tool steel showed scratching lines stating that abrasive wear is the primary cause of degradation. The authors pointed out that the wear resistance of steel-based SLM materials is not dependent on their hardness, but it is affected by the density of printed steels. Moreover, the composition of the material is more important rather than the type of processing for developing a wear-resistance material.

The work by Casati [37] was focused on tribological performance of hot work tool steel H11 and a leaner version with 30% less carbon (L-H11). The wear resistance of AM manufactured samples was evaluated by ball-on-disk tribometer (CSM Instruments) under dry sliding conditions with an alumina pin. The quenched and tempered H11 specimens showed the best wear behaviour whereas L-H11 alloy exhibited lower wear performance which was related to lower hardness values.

Several studies have also investigated the ceramic reinforcement of metals to prepare a metal matrix composites (MMCs) using AM to improve the tribological performance of materials. The MMCs exhibited an optimum combination of metallic matrix and stiffer and stronger ceramic reinforcements. It was found that the wear rate of SLS manufactured SiC/Fe MMCs shows ~ 66·7% fold decrease compared to the unreinforced Fe [30].

(20)

1.5 Surface post-machining for additive manufactured steels

It is known that machining operations provide components not only with the final shape but also alter their surface and subsurface properties. Specimens produced by L-PBF technologies have high roughness surfaces with the occurrence of defects such as partially sintered/melted powders, porosity, lack of fusion or delamination, making components susceptible to early mechanical failure and/or corrosion. Moreover, the surface finish has an influence on tribological performance of components, especially of dies in forming operations. Hence, the post-machining of additively manufactured components is a necessary step to obtain a functional part [38].

Bagehorn et al. [39] studied the effectiveness of surface finishing processes such as milling, blasting, vibratory, micro-machining on AM Ti-6Al-4V alloy. They concluded that milled specimens exhibit better results with fewer amount of surface defects and a significant reduction in roughness (Ra = 0.3 µm). The other finishing techniques did not perform as good but showed promising results: blasting (Ra = 10 µm), vibratory (Ra = 1 µm), micro-machining (Ra = 0.4 µm)

Eyzat et al. [40] observed the influence of Surface Mechanical Attrition Treatment (SMAT) on SLM manufactured titanium alloys. The results showed that SMAT can strongly modify the surface characteristics and also enhance mechanical properties. The surface roughness of as-built specimens could be reduced by 80% with an increase of mechanical properties by 10 to 15%.

Effect of milling and/or deep rolling on SLM fabricated H13 steel was studied by Breidenstein et al. [41]. The results showed that deep rolling reduced roughness values by a factor of two, however a side effect of this process is that the partially melted powder particles are pressed into the surface which could lead to rapid damage initiation of the materials under loading. Surfaces obtained by milling showed roughness values (Rz = 1.4 - 2.0 µm) similar to values of conventionally casted or forged steels. The authors pointed out that the combination of milling and deep rolling leads to best obtained Rz = 0.6 – 0.8 µm.

The comparison between grinding, blasting, electropolishing and plasma polishing was studied on SLM 316L steel [42]. The results showed that with grinding, the surface roughness was reduced to 0.3 µm representing 14% reduction of surface roughness in comparison to as fabricated samples.. With shot blasting, the roughness value was decreased by 56% to as-built samples with the roughness for sand blasted samples being around 8 µm. Electro and plasma polishing reduced the initial roughness to 55-60% of the surface roughness of 9 µm. The combination of mechanical surface treatments with polishing resulted in the best roughness values of 0.12- 3.6 µm with no machining marks on the surfaces.

(21)

1.6 Existing knowledge gaps

The conventional production of dies is performed by casting or through conventional power metallurgy technologies that have some production limitations on the design of the tools. The requirements for hot work tool steels in hot stamping such as resistance to wear, high strength, and toughness at elevated temperatures, are dependent on the steel making process. Moreover, the tribological behaviour of the die-workpiece tribopair are affected by the manufacturing process of the dies.

Galling and uniform quenching are considered as the main critical problems in hot stamping of Al-Si coated UHSS. By improving the tribological performance in the process and cooling channels design – lifetime of the tool, repair cost and quality of the final products could be improved for manufacturers. This can be realised through AM processing of hot stamping dies but the tribological behaviour of AM materials has not been extensively studied yet at high temperatures. Research studies in AM focus mainly on materials processability, optimisation of process parameters, effects of post-process heat treatments and mechanical properties of the alloys. The application of AM process for manufacturing of dies for hot stamping is also not studied and considered as a novel approach.

(22)

2. Aim and Objectives

AM opens up possibilities to improve the tribological behaviour and efficiency of hot stamping by modifying the cooling channel design and use high performance alloys in the die surface. The overall aim of this MSc thesis work is to evaluate the tribological performance of additively manufactured steels for hot stamping applications.

The specific objectives are to:

✓ characterise the microstructure and mechanical properties of AM steels pre/post test ✓ assess the tribological behaviour of AM steels at high temperatures, i.e. assess the galling

and abrasive wear

✓ evaluate the influence of surface topography on the tribological behaviour

(23)

3. Experimental work

The following section describes studied materials and specimens, as well as the applied experimental procedure and techniques such as metallographic sample preparation, microscopic analysis, microhardness and topography measurements.

3.1 Materials and specimens

The materials studied in this master thesis included additively manufactured Maraging steel (MAR) and modified tool steel H13. The nominal composition of the steels are presented in Table 2. The hot work tool steel H13 is commonly applied for dies manufacturing. In this thesis it was used to study how the application of additive manufacturing compared to conventional casting. The choice of a maraging steel is motivated by a possible application of high-performance alloys as a top layer on dies. The specimens were manufactured in the form of pins with flat surfaces and a radius at the leading and trailing edges. A schematic representation of the test specimens and their dimensions are shown in Figure 13. After printing, the samples were heat treated and machined by electro discharge machining (EDM). To evaluate the influence of surface topography on the tribological behaviour, the surfaces of maraging steel pins were prepared by milling, grinding and shot blasting. The post machining operations were chosen to evaluate different post-machining processes for AM steels. As fabricated dies are usually manufactured with milled surface. During use, the dies undergo refurbishment after certain cycles and the surface condition is changed to a ground surface. The milled and ground surfaces are widely tested conditions for evaluation of tribological performance in hot forming applications. The shot blasted surfaces were evaluated as an alternative surface finish. The process allows easy processing of large sized tools and the surface contains beneficial compressive stresses. The H13 was tested in milled surface finish as conventionally manufactured tools for hot forming applications. The tribological behaviour of AM materials was compared to conventionally cast tool steel QRO 90® Supreme with a ground surface finish, which was reported by Deng et al. [43] and was used in this work as a reference. All pins were tested against commercially available Al-Si coated UHSS boron steel 22MnB5. The workpiece material was supplied by Gestamp HardTech in the form of strips with 15 mm width, 1000 mm length and 1.5 mm thickness.

Table 2. The chemical composition of AM steels and the reference tool steel, weight %. Fe makes up the balance.

Material C Si Mn Cr Mo Co Cu Ni V

Maraging 0.03 0.35 0.40 5.0 8.0 12.0 2.0 2.0 -

H13 mod 0.35 0.2 0.5 5.0 2.3 - - - 0.6

(24)

Figure 13. Surface condition of the tested AM materials and reference tool steel, as well as the schematic

(25)

3.2 Experimental procedure and techniques

Tribological studies

The tribological tests were carried out using a Ducom TR-20M-47 Hot Strip Drawing Tribometer, which enables the utilization of process-like conditions relevant to hot stamping. The schematic representation of the test-set up and the main elements of the tribometer are presented in Figure 14.

Figure 14. Simplified schematic of the

tribological test

Figure 15. Main elements in tribometer [43]

Before the test, the pin specimens were cleaned in an ultra-sonic cleaner USC-TH with Heptane for 3 min at 25°C, thereafter rinsed with ethanol and dried in air. Every single test included 600 mm sliding distance that is corresponded to 3 strips in total, i.e. 200 mm per strip. The workpiece strip specimens were wiped clean using ethanol. The test procedure was defined as follows:

– The pins were mounted in holders and fixed within the tool assembly (Figure 15), followed by alignment of the pins using pressure sensitive paper and shims. This is a necessary step to ensure a homogeneous distribution of the applied load to perform a full contact between pins during sliding;

– Then strip specimen was mounted and a pretension was applied to prevent deformation of the strip during heating and sliding;

– Resistive heating was then applied on the strip by means of an electrical current passed through it (Joule effect). Due to the high heating rated obtained with resistive heating, a step-wise application of voltage was used. The main goal of this approach is to obtain a similar microstructure of the Al-Si coating as in the real application. The process included heating to ~600°C for 4 min, then 2 min to ~750°C and 30 s to ~920°C to reach austenitisation temperature. Afterwards, the voltage was decreased depending on desired test temperature, i.e. sliding at 600°C or at 700°C. The temperature was measured by means of a pyrometer.

(26)

– When the load reached the prearranged value, sliding was set for 200 mm, with a sliding speed 100 mm/s;

– Finally, upon completion of the test, quenching of the system was performed by pressurised air.

The sequence of actions was repeated for 2 other strips with the same mounted pair of pins to obtain the full 600 mm sliding distance. For consistent results, the tests were repeated two times for all temperatures and surface finishes of maraging steel. Milled H13 specimens were supplied in smaller batch and therefore only two tests were carried out per condition. All process parameters were selected based on conditions encountered in the forming operation.

Metallographic sample preparation

For microstructural analysis and analysis of cross-sections, the pins were cut with a Secotom-10 precision cutting machine. The following steps were performed for metallographic preparation:

– Grinding. This operation is mainly used to reduce the damage caused by the previous sectioning. Grinding was conducted using SiC abrasive paper, reducing the particle size (from P120 to P2500) to prepare for final polishing. The time for each abrasive paper defined by full removal of grinding marks from the previous step.

– Polishing. The purpose of this process is to smooth surfaces after grinding. Rough polishing was done using diamond abrasive reducing the particle size (from 6 µm to 1 µm) in each polishing step. A final polishing step was conducted using colloidal silica with a particle size of 0.05 µm.

– Etching. The etching is necessary to reveal microstructural features of the material such as grain boundaries, size, deformed grains. The used etching agent was Nital 3% for 30s. For as-built specimens, Nital 10% was applied for 3-4 min but did not work. Further optimisation of etching of as-built samples is required.

Microscopic analysis

– Optical Microscopy (OM). The samples were analysed with OM to reveal the proper time for etching and microstructure defects using Nikon Eclipse MA200.

– Scanning Electron Microscopy (SEM). This technique was used for observation of specimen surfaces, nature of transferred layer on the worn surface of the tool, microstructures, and deformed layers near the sliding surface. The utilised microscopes were a Jeol and a Zeiss Merlin. The used voltage for SEM and EDS (Electron Dispersive Spectroscopy) are 10 kV in Jeol and 15 kV in Zeiss Merlin.

Mechanical characterisation and topography

– Microhardness. Hardness profiles were measured on mirror polished sides of specimens using a Matsuzawa MXT-CX microhardness tester. The used load was 25g and the indentation time was 15s. The tests were done taking five points per depth step, starting at 20 µm from the edge of the pin; the highest and lowest values were neglected and then the average of the remaining three was considered.

(27)

(28)

4. Results and discussion

This section presents the characterisation of as-delivered AM steels in terms of surface topography, hardness and microstructure. After the tribological tests it discusses the resulted frictional behaviour of additively manufactured steels, occurred wear mechanisms and the mechanisms behind material removal on AM steels.

4.1 Characterisation of as-delivered additively manufactured steels

4.1.1 Surface topography

Surface characterisation before test allows to evaluate the surface roughness, defects of surfaces after the machining operations, and later the progression of galling on the tested pins. The presented defects, perpendicular surface lay to sliding direction, and surface imperfections are undesirable and affect the tribological behaviour, in particular galling development. In Figure 16 the surfaces and the Sa values of the as-delivered pins are shown.

Figure 16. Surface characterisation of as delivered pins

(29)

pins followed the direction of rotation milling and had a curved surface lay. For the tribological tests, the surface lay of the pins was oriented aiming for consistent perpendicular surface lay to sliding direction. The surface lay of the ground specimens was tested parallel to the sliding direction. Shot blasting with sand and glass resulted in irregular surfaces with a variation of the height of peaks and depth of valleys/cavities.

Figure 17. Photographs of as delivered pin surfaces (top) and SEM of the same (below)

4.1.2 Hardness

The hardness for both steels showed no significant differences in measurements along their depth as shown in the profiles given in Figure 18. This confirmed the homogeneous microstructure of the pins printed by SLM after their heat treatment. The average hardness of MAR was 500 ± 12 HV0.025 and H13 was 497 ± 9 HV0.025. The measured hardness of specimens after tests showed

similar values: the hardness of MAR tested at 600°C was 544 ± 11 HV0.025 and at 700°C 543

± 11 HV0.025. Therefore, the hardness of AM materials was unaffected by the test conditions.

(30)

4.1.3 Microstructure

The microstructures of the vertically printed specimens after heat treatments are shown in Figure 19. As can be seen, the microstructure of AM materials is very fine and homogeneous. The SLM process resulted in a density of more than 99%, and no defects or voids were found in the observed cross sections. After etching, multiple phases were revealed on maraging steel (MAR) as there are different contrast regions. Both steels usually present a martensitic microstructure.

Figure 19. Cross sections of steels after etching

With SEM at higher magnification, more features of the microstructures can be seen (Figure 20). MAR is characterised by a very high density of precipitates due to the ageing process. Different cell shapes were observed within the steel matrix. Coarse particles were primary at cell boundaries and triple points. For H13, a fine cellular microstructure was defined. The existence of coarse grains is due to the influence of a complex temperature gradient during printing.

(31)

4.2 Frictional behaviour of AM steels

The frictional behaviour from tests with Maraging steel is shown in Figure 21. The results indicated a good reproducibility of tests at 600°C. There was a slight difference in the coefficient of friction (COF) at the beginning of the tests between the different surface topographies, which is correlated to the running-in behavior of tribological tests. For milled and ground surfaces, COF started at around 0.6 and tends to decrease during the first 200 mm. After 200 mm sliding distance, COF is stabilised around 0.4 for both surfaces. The coefficient of friction for shot blasted pins was slightly lower at the beginning of sliding and showed a consistent value of 0.4 almost during the full sliding distance.

Figure 21. Frictional behavior of AM maraging steel at 600°C and 700°C (repeat tests for each type of

surface) sliding against Al-Si coated 22MnB5 steel.

Sliding against the workpiece at 700°C resulted in unstable friction behaviour. Higher fluctuations were observed for all surfaces compared to the tests at 600°C. Earlier failure of tests was found for the shot blasted surface. The ground surfaces showed a tendency to fail after 200 mm sliding distance whereas milled lasted up to 400 mm. Galling is associated with instabilities in frictional behaviour. Friction peaks indicate severe material transfer onto the tool surface that leads to failure of a test. It was observed that in some cases, sudden formation and removal of lumps took place, which was seen as isolated friction peaks during the test. In Figure 22, the worn strip after test at 700°C for milled pin is shown. The dashed lines indicate the friction peak and the corresponding area on the worn strip. When a lump on the pin surface was formed (Figure 23), it started to scratch the strip and results in an increasing COF. After a short time friction decreases and sliding continues with full contact until complete failure of test occurs at the end.

(32)

Figure 22. Worn strip in comparison to COF curve after the test at 700°C

Photographs in Figure 23 present worn surfaces of milled pins. At both temperatures the worn zones was observed on the leading edges due to higher contact pressure concentration in this area during sliding. The pin surface tested at 600°C showed a distributed worn surface with affected surface lay but still distinguishable marks after machining. Whereas at 700°C, due to increased severity of wear rates, the surface lay on the leading edge was removed. The built-up material (lump) is usually formed at the leading edge radius of the pins. The difference in color on pin’s surfaces indicates oxidation during tribological testing.

(33)

The friction behaviour of additively manufactured H13 is presented in Figure 24. The curves showed similar behavior to MAR where at 600°C, the test lasted for a longer sliding distance, but the COF showed lower values. At 700°C the first test (blue line) failed before 200 mm sliding distance. This behaviour was unusual for milled surfaces. With observation under SEM on cross section of H13 (Figure 25) presence of Zn was found in the material transfer. This indicates that these tests were carried out on a different strip material, as Zn is not an alloying element in the Al-Si coated boron steel, the results from EDS confirmed changed composition in the coating of boron steel. This was due to a mix-up of the material when the strip material was resupplied. Therefore, the results of test at 600°C are not directly comparable to tests of maraging steel. Since H13 pins were tested against a different material, the main comparison in this project of H13 and MAR is based on the results of the second test at 700°C (orange line), where the conventional Al-Si coated boron steel was used. The formation of lumps and removal was also observed for this test, and furthermore, the test lasted the 600 mm sliding distance. The COF of H13 has similar behaviour to MAR at higher temperature. However, for an adequate comparison, repeat tests are required.

Figure 24. The COF of AM H13

(34)

Comparing the COF between AM materials (Figure 26) and the conventional QRO 90 tool steel it has been observed that friction values stabilised for MAR, H13 and QRO 90 around 0.4. At lower workpiece temperature 600°C the full sliding distance was completed for all materials and surfaces. At 700°C, the QRO 90 showed a stable behaviour similar to that at 600°C, whereas for AM materials the COF between the two temperatures was significantly different.

Figure 26. Comparison of COF for AM steels and casted steel

(35)

4.3 Wear behaviour

4.3.1 Surface characterisation after tribological tests

AM maraging steel surfaces were characterised with optical profilometer after tests. For comparison, the most representative pins are shown in Figure 27. Evaluation of surfaces was done for all repeated tests. All results can be found in Appendix - A. The 2D topography images showed that galling and abrasion occurred at both temperatures. With the increase of temperature, the severity of both wear mechanisms increased. Abrasion was presented in form of deeper grooves that were spread on surfaces towards the trailing edge. The formation of localised pits on surfaces was also observed (Figure 28). At 600°C abrasion was localised on the leading edge for all surface finishes. Galling formation was affected by surface topographies.

On milled pins, material transfer was mainly developed on protruding parts of the surface due to the waviness of the pins. Material transfer on ground samples was milder and consisted of isolated particles dispersed across the pin surface. Although the higher surface roughness tends to facilitate galling, on the shot-blasted surfaces (Sa= 1.94 µm), formation of a thick transfer layer was not found. On the surfaces from both temperatures, only a thin layer was observed. The surface characterisation showed that galling on milled and ground surfaces prevails in comparison to the shot-blasted finish. At 700°C accumulated material transfer with wear debris from steel grows faster in compact layers and form lumps.

(36)

Figure 28. Abrasion as pits on worn maraging steel pins

On AM H13 the same wear mechanisms were found (Figure 29). Compared to MAR, less severity at tested temperatures was observed. The formation mechanism of galling was similar to QRO 90 at 600°C. Adhered material was concentrated on the leading edge and accumulated following the surface lay. It is important to note that the obtained results are affected by the presence of Zn in the coating for the tests at 600°C. At 700°C abrasive wear was characterised by deep grooves along the surface of H13. In case of QRO 90, galling covered the whole surface whereas in H13 the transferred material presented in a smaller scale on the leading edge.

The surface characterisation after tribological tests showed occurrence of galling and abrasion on AM steels. Higher severity of wear occurred during tests at 700°C. The surface waviness promotes growth of thicker transfer layers. The high initial Sa on shot blasted specimens did not promote higher rates of galling since specimens with this finish had the smallest material transfer level, but the sliding distance to failure was shorter. On milled and ground surfaces, galling was observed as formation and accumulation of transferred material particles. In comparison to the conventional tool steel, the AM materials underwent more severe abrasion.

(37)

4.3.2. Galling

Ground and milled surfaces

The important damage mechanism encountered in hot stamping is material transfer or galling. This type of wear is originated from severe adhesion. High adhesive forces between tool and workpiece lead to micro welding of the asperities of the two surfaces. During sliding, particles of one surface brake away and adhere to the counter surface. Under SEM, galling appears as a compact layer on the tool surface (Figure 30 (a)). This transferred layer can be easily identified with the BSE detector, as a difference in elemental contrasts exist between the transferred material (dark grey) and the tool steel (light grey), as shown in Figure 30 (c). Galling can be characterised by existence of different forms. Mainly, material transfer generates in the form of localised lumps or in more spread out layers on the tool pins.

For the AM materials with ground and milled surfaces, galling occurred similarly at both temperatures. The earlier stages of galling can be seen in Figure 30 (b). The adhesion of transferred material started with thin layers on the tool surface. With the time, material transfer developed and formed a spread out transfer layer as presented in Figure 30 (a). The wear debris (Figure 30 (d)) from coating and steel tends to accumulate (Figure 30 (c)) in surface valley regions of the milled or ground surfaces, which provide sites for particle entrapment. Trapped wear debris gets compacted due to the load and sliding. The parallel surface lay to sliding direction on ground specimens allowed for wear debris to travel along the grooves. The entrapped particles were dispersed and therefore the formation of transfer layers was initiated across the surface. On the milled pins, due to perpendicular direction of surface lay and waviness, thicker layers were formed near the leading edge and on protruding parts of pins. The severity of galling increased with increasing workpiece temperature.

Figure 30. Galling on AM milled and ground surfaces (a) formation of compact layer, (b) initial

(38)

For characterisation of galling, BSE mode in SEM is widely applied. With this observation, the “sandwich” structure of galling was observed as shown in Figure 31 (c). Figure 31 (a) evidently showed the different layers in this structure. With higher magnification and EDS, the elements of mixed regions were investigated. As shown in Figure 31 (d) the presence of Fe, Al and Si confirmed the mixture of wear debris from the steel and coating.

When analysing the worn surfaces, different stages of galling could be identified. Closer to the leading edge a higher concentration of compacted layers could be detected as the sliding contact starts between workpiece and pins at this point. Towards the trailing edge adhered material showed thinner layers indicating the early stage of galling. Keeping this in mind, in regions where thin transfer layers are observed, the deformation of the steel surface was encountered (Figure 31 (b)). It is likely that the steel material is shifted in the direction of sliding, with subsequent accumulation of Al-Si wear debris.

Figure 31. Galling under BSE mode (a) layers in transferred material, (b) deformed steel in the direction

of sliding, (c) “sandwich” structure, (d) EDS on mixed region

(39)

Figure 32. Galling on AM milled H13 (a) sintered compact layers, (b) deformed steel in the direction of

sliding

Further analysis was conducted on cross sections of worn samples. As seen in Figure 33 (b), the compacted layers consist of debris that sinter together during sliding. The detachment of coarse fragments from this layer could cause abrasion on the interacting counter surface. Figure 33 (a) and (c) showed microstructures of cross section for AM maraging and H13 steels that were characterised by refined grains at the surface and unaffected bulk. Grains near the surface were deformed in the sliding direction. Compressed grains on the surface were probably plastically deformed by the applied load during the test. Smaller grains tend to decrease ductility meanwhile increasing the hardness and tensile strength of materials referring to strain hardening phenomenon. Such deformation induces anisotropic high strength properties in certain directions. However, the hardness measurements did not reveal differences on a surface compared to bulk. More precise measurements are necessary, for example with nanoindentation, to evaluate if this refinement results in surface hardening for AM materials.

Figure 33. Cross sections of MAR and H13 (a) compressed grains, (b) sintered material transfer on tool

surface, (c) deformed grains under compacted layer of galling

(40)

Figure 34. Galling on conventionally casted QRO 90 (a) formation of compacted layer, (b) abrasive wear

marks, (c) accumulation of adhered material

On AM milled and ground materials, the galling formation is similar to conventional tool steel and could be described in the following sequence. When sliding starts, fragments of Al-Si coating adhere to high asperities tips of the steel surface lay or embed into the surface. Once the initial material transfer has occurred, further agglomeration of coating particles continues to occur in those regions (Figure 35 (a)). Accumulation and entrapment of the wear debris in surface valleys lead to thickening/compaction of the transferred layer as can be seen on tilted pins in Figure 35 (b) for ground surface. If sliding progresses, thicker build-up of adhered material form and develops into lumps.

Figure 35. Tilted surface of ground pins (a) accumulation and entrapment of wear debris, (b) compaction

of transferred layer

Shot-blasted surfaces

(41)

Figure 36. Galling on shot blasted finish (a) formation of thin transferred layer, (b) surface flattening, (c)

tilted surface of the pin

On the shot-blasted pins, the formation of galling could be described in the following sequence. Harder phases in the workpiece Al-Si coating caused a grinding effect on the original rough topography as shown in Figure 37 (a). The initial accumulation of Al-Si particles occurred in surfaces cavities (Figure 37 (a)). As sliding progresses, a thin layer of material transfer developed on flattened parts of the maraging steel surface. Wear debris from steel and coating continue to accumulate and got entrapped in the remaining surfaces cavities (Figure 37 (b)). As presented in Figure 37 (c) the transferred layer continues to grow during sliding at the places were thin and flat layers have initiated. These differences in galling formation are driven by the significant differences between milled/ground and shot-blasted topographies after machining.

Figure 37. Galling on shot blasted AM steel (a) initial flattening and entrapment of wear debris in surface

cavities, (b) developed thin layer, (c) growth of material transfer

4.3.3 Abrasive wear

(42)

coating from workpiece in the contact zone. Subsequently, these debris accumulated together and got embedded into the steel surface, initiating micro-cutting of AM materials (Figure 38 (b)).

Figure 38. Abrasive mechanisms on AM maraging steel (a) smeared surface, (b) micro-cutting, (c)

grooves formation. These types of mechanisms were also observed for H13

In Figure 39 (b) the micro-cutting of material was clearly observed when analysing cross sections. EDS mapping confirmed the presence of Al in embedded wear debris. The wear of the maraging steel is characterised by removal of material taking place at the grain boundaries (Figure 39 (a)). The loading and sliding action promoted embedment of Al-Si particles into the surface and shearing of material. The generated stresses induced material fracture as seen in Figure 39 (c). More SEM micrographs of galling and abrasive wear on maraging steel can be found in Appendix - B.

Figure 39. Abrasive wear in AM materials, cross-sections view (a) delamination in steel, (b)

(43)

It was observed that the removed material could be entrapped at the asperities on the strip surface, resulting in transfer of maraging steel onto the counter surface (Figure 40 (c)). In Figure 40 (a), the Al-Si coating is shown after exposure to high temperatures outside the contact. It was reported in many studies that it undergoes morphological changes on the surface and is characterised by increased roughness, hardness and brittleness. As seen in Figure 40 (b), steel material had a slightly different contrast in comparison to the rest of the coating in a contact area. To define the composition of adhered patches in a contact zone further analysis was conducted by EDS mapping on strip surface (Figure 41). Mapping showed that surfaces contained: Al, Fe as these are alloying elements in coating (88wt%Al, 9%Si and 3%Fe) and oxygen due to formation of oxide layer on the surface. But the prevalence of Fe and O on smeared surfaces confirmed that these patches are transferred steel from the tool. The lack of Fe (dark area in Al map) also confirmed the presence steel in this region.

Figure 40. Worn Al-Si coated 22MnB5 surfaces after tribological tests (a) outside the contact, (b) contact

surface, (c) adhered maraging steel on strip

(44)

4.4 Mechanisms behind material removal onto AM steels

In comparison with conventional tool steels, the abrasion on AM materials was found to be more severe. The main findings in terms damage mechanisms in steel could be described with schematics as follows (Figure 42):

1) During early stages of galling, debris from the Al-Si coating started to adhere to the steel surface;

2) Embedded particles served as obstacles and therefore as initial points for debris accumulation. As sliding progresses, harder phases of material transfer and oxidised steel debris cut through the material generating more wear debris. It leads to compact layers with a “sandwich-like” structure;

3) Due to further sliding, this mixed feature causes shearing of the steel matrix and pushes the material in sliding direction. In the end, the deformed layers of material grow on the steel surfaces with further material adhering on top;

4) These deformed layers in steel could be removed with time.

The susceptibility of the AM steel to plastic deformation resulted in removal of the material by the initial adhesion/embedment of particles. The formation of observed pits on maraging steel could be explained by this process. This damage mechanism was observed for all tested surfaces for maraging and H13 steels.

Tribological characterisation of additively manufactured hot forming steels