Abstract

By directed energy deposition (DED), a flexible design of cooling channels in forming tools, e.g., hot stamping, with a variety of sizes and a high positioning flexibility compared to machining processes is possible. The subsequent ball burnishing of the tool surfaces, in combination with a variation of the DED process parameters, enables control of the tool surface properties and friction behavior. Parameters such as the ball-burnishing pressure or the path overlapping in the DED process are investigated to quantify their effects on roughness, hardness, friction, residual stresses, and heat transfer coefficient of generic tool surfaces. The friction coefficient at elevated temperatures depends strongly on the surface roughness of the tool steel surfaces generated by DED and ball burnishing. The latter process improves the surface integrity: the roughness peaks are leveled by up to 75%, and the hardness and the residual stresses are enhanced by up to 20% and 70%, respectively. However, the roughness of the tool surfaces is determined mainly by the path overlapping of the welded beads in the DED process. Despite the higher surface roughness, the heat transfer coefficient is in the range of conventionally manufactured tool surfaces of up to 2700 W/m2K for contact pressures up to 40 MPa. First hot stamping experiments demonstrate that the tools manufactured by the novel process combination are able to manufacture 22MnB5 hat profiles with an increased and more homogeneous hardness, as well as more homogeneous thickness distribution, compared to conventionally manufactured tools.

1 Introduction

Hot stamping of automobile parts is used frequently when high-strength steel parts with low specific weight are required. This process combining forming and quenching has the advantages of high formability due to high temperatures and high strength of the part (Fig. 1). The most common materials used in hot stamping are manganese-boron steels such as 22MnB5. This is owing to their transformation from a ferritic-pearlitic to a martensitic microstructure in the process of combined forming and quenching, which results in high strengths of up to 1500 MPa of the formed part [1]. In order to accomplish this transformation, the manganese-boron steel blank has to be austenitized first. For the simultaneous forming and quenching, hot stamping tools with integrated cooling channels are employed. The cooling of the tool has to fulfill the requirement of a minimum cooling rate of 27 K/s for the martensite transformation of the material 22MnB5.

Fig. 1
Process chain of direct hot stamping [1]
Fig. 1
Process chain of direct hot stamping [1]
Close modal

Hot stamping tools have to meet various requirements in order to ensure appropriate forming and quenching of the workpiece. The requirements include, among others, a high compression strength to prevent plastic deformation, high hardness (>50 HRC) at elevated temperatures for the reduction of die wear, and high thermal conductivity (>20 W/mK) with regard to fast quenching of the formed part [2]. Materials meeting these requirements are hot working steels, such as 1.2343 (H11) or 1.2344 (H13). Hot stamping tools are rough machined and often hardened and/or tempered. The friction behavior between the hot stamping tool and the AlSi-coated metal sheet in the hot stamping process is influenced by several factors, such as the contact pressure, the relative sliding velocity, the interface temperature, and the surface roughness of both tool and blank [3]. In this context, the temperature is the most important factor influencing the friction coefficient, whereas the sliding velocity plays a minor role. Typical friction coefficients determined for AlSi-coated blanks at elevated temperatures (>500 °C) exhibit values between 0.4 and 0.6 [3]. Ghiotti et al. [4] evaluated the friction behavior of AlSi-coated 22MnB5 blank material, finding that the friction coefficient decreases with an increasing temperature (from 500 °C to 800 °C) as well as with an increasing contact pressure (from 5 MPa to 25 MPa at 500 °C). It is concluded that this is due to the dependence of the friction coefficient on the shear strength of the coating: With higher contact pressures (and higher temperatures), the surface asperities of the softer material are deformed, and the topography of the surface is flattened because of the lower shear strength. Pelcastre et al. [5] examined the friction behavior of AlSi-coated ultra high strength steel (UHSS) blanks and conventionally manufactured tool steel surfaces. The tool steel specimens exhibit a varying surface roughness resulting from grinding (Ra = 1.20–2.74 µm), milling (Ra = 2.31–2.44 µm), and polishing (Ra = 2.31–2.44 nm). At a temperature of 900 °C and a contact pressure of 10 MPa, the surface roughness of the tool steels influences the friction coefficient in the way that the ground surface with the highest surface roughness results in the highest friction coefficient compared to the milled and polished surfaces. This interdependency of surface roughness and friction coefficient dissolves with higher contact pressures (20 MPa) due to the flattening of rough asperities. In the case of the lower contact pressure, more material is transferred from the blank surface by the rough tool surfaces with the effect of a galling process.

The conventional manufacturing route of the cooling channels in hot stamping tools comprises the segmentation of the tool and the hole drilling. This procedure leads to the disadvantages of possible leakages and low flexibility regarding the size and location of the channels because of the limited accessibility. The additive manufacturing of hot stamping tools provides the advantage of increased flexibility regarding the geometry and the proximity of the cooling channels to the tool surface (e.g., in order to avoid hot spots). The creation of parts in the process of directed energy deposition (DED) is executed by melting materials as they are being deposited. For this, a laser beam is used to melt metals and to direct energy into a concentrated region, where the substrate and the deposited powder melt. By means of DED, thin layers of dense and wear-resistant metals can be deposited on components (e.g., forming tools), resulting in an enhanced performance and lifetime [6]. In the DED process, different deposition strategies influence the final properties of the workpiece: Ribeiro et al. [7] found for the manufacture of a stainless steel 316L cube that the toolpath strategy and the stepover (i.e., the path overlapping) affect the final part geometry, the density, the surface roughness, and the hardness of the part. Regarding the investigated toolpath strategies, the meander strategy results in the lowest microhardness measured, which is due to the heat-affected zones and heat transfer rates, respectively, alternating with the path strategy. The evaluation of two different stepovers shows that a smaller stepover (i.e., a greater path overlapping) leads to lower surface roughness values (e.g., for zigzag and meander strategy). A general drawback of the DED process is the rather poor resolution and surface finish with a roughness of more than 25 µm in most cases [6]. This fact indicates that a post-processing of tool surfaces manufactured by DED is inevitable.

A promising procedure within this scope is the ball burnishing of tool surfaces: this method reduces the surface roughness, as it has been shown for application on deep drawing dies with thermally sprayed hard metal coatings. Ball burnishing generates a tool surface with roughness on a comparable level to a ground tool surface. The resulting roughness can be predicted by an analytical model with the help of the original roughness and the hardness as input parameters. The deviation of the model to experimental results amounts to 10% on average. Furthermore, ball burnishing allows for influencing the material flow during the forming process by inducing defined textures to the tool surface. This local control of the friction coefficient results in an improvement of the geometry of deep-drawn parts by a compensation of anisotropy [8].

The surface finish achieved by ball burnishing of hardened steels (64 HRC) depends on the process parameters. By means of hard-turned samples, it has been shown that the burnishing pressure and the original surface roughness have a significant influence on the surface finish, i.e., the highest investigated pressure (38 MPa) and a low original roughness lead to the lowest final surface roughness [9]. In specific cases, especially for rather ductile materials (e.g., Inconel 718), the best surface roughness is not achieved by the highest possible burnishing pressure but by a defined lower pressure [10]. For very ductile materials, high burnishing pressures can lead to such a deformation of the surface that it results in geometric inaccuracies. Besides the surface roughness, the ball-burnishing parameters, e.g., burnishing pressure, influence the residual stresses in the workpiece as well. According to Chomienne et al. [11], the burnishing pressure is the process parameter with the most significant influence on the compressive residual stresses, which are induced in the burnishing process. An increase in the pressure leads not only to higher residual stresses but also to a greater depth of the residual stresses in the investigated material (martensitic stainless steel 15–5 PH with a Brinell hardness of HB 350). Residual stresses are already introduced to the workpiece by the additive manufacturing process, e.g., DED. In this context, compressive residual stresses are generated in the center of the part, whereas near to the surface tensile, stresses are dominant [12]. The residual stress pattern after DED is affected by the characteristic thermal cycle, which consists of rapid heating of the material because of the high energy intensity that is followed by a fast solidification due to the rather small fraction of the melt pool. Furthermore, a re-melting of previously solidified areas takes place. In the cooling phase, compressive residual stresses are generated due to the shrinkage of the material. However, tensile stresses occur due to the re-melting and re-solidification phases so that a rather complex residual stress field is induced [12].

In this paper, the novel process combination of DED and subsequent ball burnishing is suggested to manufacture tools for hot stamping. Through this unique combination, it is aimed to increase the geometric complexity of cooling channels and, at the same time, control the frictional and thermal conditions at the tool surface. The suggested process combination is explored essentially for generic tool surfaces by investigating the relationship between the process parameters of the DED (path strategy, path overlapping, number of layers) and the ball burnishing (burnishing pressure, sidestep) on the properties of the tools, including their hardness, their surface topology (important for friction and heat transfer), and the residual stresses. The combination of these processes offers to avoid the disadvantages of additively manufactured hot tamping tools, such as tensile residual stresses at the surface or high roughness values while benefitting from its flexibility. The next section describes the used materials and methods, followed by the results and their discussion, as well as the conclusions.

2 Materials and Methods

2.1 Tool Materials.

The investigated materials comprise two different tool steel powders suitable for processing with directed energy deposition: UTP PLASweld™ Ferro 55 and UTP PLASweld™ Ferro 702 (voestalpine, Hamm, Germany). Given their distinct chemical composition (Table 1), the metal powders exhibit diverging properties regarding weldability (according to the content of carbon) and thermal conductivity. As the material Ferro 55 is comparable to the tool steel 1.2344 (H13, X40CrMoV5-1), it is assumed that the thermal conductivity amounts to 20–25 W/mK in a temperature range of 100–600 °C. The material Ferro 702 is equal to the maraging tool steel 1.2709 (X3NiCoMoTi18-9-5), which results in a thermal conductivity of 15–20 W/mK in the same temperature range as mentioned above.

Table 1

Chemical composition of the used metal powders in weight % according to the manufacturer [13,14]

Ferro 5
CSiMnCrMoFeothers
0.350.31.17.02.2rest
Ferro 702
CMoNiCoTiFeothers
0.034.818.09.51.0rest<0.5
Ferro 5
CSiMnCrMoFeothers
0.350.31.17.02.2rest
Ferro 702
CMoNiCoTiFeothers
0.034.818.09.51.0rest<0.5

The weld deposit exhibits a hardness of 53–58 HRC for the Ferro 55 and 32–37 HRC for the Ferro 702. For the latter, the hardness is increased to 48–53 HRC by subsequent tempering (which is not applied in the present case). Both metal powders hold a particle size in the range of 50–150 µm.

2.2 Directed Energy Deposition and Ball Burnishing.

The metal powders are processed in the hybrid machine Lasertec 65 3D (DMG MORI, Bielefeld, Germany) using an S235 steel basis with the dimensions of 100 × 100 × 10 mm, which is kept at room temperature at the beginning of the process. The machine includes a laser unit for DED as well as a five-axis milling unit. The DED procedure (Fig. 2(a)) is carried out by using a nozzle with a diameter of 3 mm and a laser spot diameter of 2.96 mm, respectively, for the deposition of the metal powder. This results in a working diameter of 2.7 mm. Further parameters are defined as follows: a feed rate (scanning speed) of 1000 mm/min, a powder rate of 14 g/min, and an initial laser power of 2000 W for one layer. The laser power is decreased by 200 W per layer (until 800 W) in order to avoid an extensive reheating of the previously deposited layer. The well-investigated input parameters, such as laser power, feed, and powder rate, for the DED process are based on already tested ones originating from the machine manufacturer. The subsequent ball burnishing (Fig. 2(b)) of the additively manufactured surfaces is carried out on a five-axis milling machine DMU 50 (DMG MORI). The utilized hydrostatic ball-burnishing tool HG13 (ECOROLL, Celle, Germany) has a ball diameter of 12.7 mm. By means of an external high-pressure pump, the burnishing pressure is adjusted to the targeted value at the burnishing ball. The ball-burnishing experiments are carried out with a constant feed of 33 mm/s.

Fig. 2
Working principle of (a) the directed energy deposition process and (b) the ball-burnishing process
Fig. 2
Working principle of (a) the directed energy deposition process and (b) the ball-burnishing process
Close modal

In the DED process, two building strategies for rectangular parts are applied: the zigzag strategy and the meander strategy (Fig. 3(a)). Along with these strategies, a path overlapping of 20% or 50% is used in order to analyze its influences on the surface properties with regard to the manufacturing time. The number of layers is varied in the range of one to three, with a thickness of 0.9 mm per layer, which results in different thicknesses of the additively manufactured surfaces. The different layers always change their primary direction after each deposited layer in the way the welding beads of one layer are deposited perpendicular to the welding beads of the next layer.

Fig. 3
(a) DED strategies applied for specimens (cross section: 60 × 60 mm) and (b) ball-burnishing parameters
Fig. 3
(a) DED strategies applied for specimens (cross section: 60 × 60 mm) and (b) ball-burnishing parameters
Close modal
The subsequent ball burnishing is performed perpendicular to the direction of the DED paths (completely for the zigzag strategy, but partially parallel for the meander strategy, Fig. 3(a)) with different burnishing pressures pball at the burnishing ball from 11.7 MPa to 23.4 MPa. The burnishing pressure is calculated as follows:
pball=FballAball
(1)
With the burnishing force Fball and diameter dball (Fig. 3(b)), the cross section of the burnishing ball Aball is given as
Aball=πdball24
(2)

The sidestep a or path distance is alternated between 0.05 mm and 0.5 mm (Fig. 3(b)). The ball-burnishing pressure and sidestep are varied in order to investigate in which limits the additively manufactured surface profile can be influenced by ball burnishing, e.g., to which extent the roughness can be reduced.

2.3 Surface Roughness Measurements and Strip Drawing Tests.

After each of the aforementioned process steps, the roughness of the surfaces of the samples is measured by a surface profiler (MarSurf XR20, Mahr, Goettingen, Germany). The measurements are carried out according to the German standard DIN EN ISO 4288 [15], with a measuring length of 40 mm consisting of five single measuring sections (with a length of 8 mm each). The direction of the roughness measurements is defined as perpendicular to the direction of the DED process (Fig. 3(a)). This is done in order to have a cross-sectional representation of the additively generated peaks and valleys with a broader analysis of the surface. By contrast, the investigation of the surface roughness in the longitudinal direction would be reasonable for the analysis of rather localized phenomena, e.g., Ref. [16]. For the evaluation of the surface roughness, the peak-to-valley height Rz, representing the greatest difference between the highest roughness peak and the lowest valley of the surface, is chosen. This is due to the fact that it represents the most significant roughness value giving the difference between the maximum and the minimum measuring value. By contrast, the mean roughness value Ra represents only a mean deviation of the profile from the midline, so that it does not allow any information about how high individual roughness peaks are. Additionally, the investigations of Hiegemann [17] give a thorough base for the evaluation of ball-burnished surfaces by using the parameter Rz.

Additionally, the hardness of the samples is measured in Brinell hardness tests according to the German standard DIN EN ISO 6506-1 [18] by using a macro hardness testing machine (ZHU 750 Top, Zwick/Roell, Ulm, Germany). The diameter of the testing ball is 5 mm, and the testing force is 7355 N. It is focused on macro hardness (rather than on micro or nano hardness) in this case because a rather big surface area has to be taken into account to cover the properties of a tool surface in a comprehensive way. The Brinell procedure for the hardness tests is chosen because of the resemblance of the test body with the geometry of the ball-burnishing tool based on the investigations by Hiegemann [17]. In order to prevent biasing, the indentations of the hardness tests were placed in the middle (or the peaks, respectively) of the welding beads. Furthermore, the diameter of the indentations does not exceed the width of the welding beads and is aligned with the direction of ball burnishing. In this context, it is worth mentioning that the ball burnishing did not cause grooves due to the sidestep of the burnishing of 0.1 mm. The hardness testing results in a diameter of the indentation of around 1.2 mm while the width of a welding bead is 2.7 mm, which is reduced at the additively manufactured surface according to the path overlapping. In this way, the indentations do not cross multiple beads (only burnishing paths).

The friction coefficient μ is determined in a strip drawing test at room temperature and elevated temperatures (Fig. 4) with additively manufactured and ball-burnished friction jaws and sheet metal strips (22MnB5 with an aluminum-silicon coating) with a thickness of 1.5 mm, a length of 300 mm, and a width of 50 mm for the tests at room temperature. For the metal strips, the manganese-boron steel 22MnB5 with an aluminum-silicon (AlSi) coating is used because it is applied commonly in hot stamping. The friction jaws are prepared by using an additively manufactured and ball-burnished surface on an S235 steel basis. The surface area of the jaws exhibits a width of 40 mm and a length of 34 mm.

Fig. 4
Strip drawing machine and principle
Fig. 4
Strip drawing machine and principle
Close modal

For the measurements with elevated temperatures, the blank is heated up inductively with the help of a high-frequency generator (Hüttinger Axio 10/450), with a power of 10 kW. Due to the size of the inductor (with a width of 26 mm and a length of 134 mm), the blanks have a width of 20 mm in this case in order to achieve uniform heating of the blank. The current temperature is measured by a near-infrared pyrometer (Sensortherm Metis M318). Beforehand, the emission coefficient of the heated blank has been determined by comparing the temperature measured by the pyrometer with the one measured by a thermocouple (as reference) and adjusting the value of the emission coefficient in the measuring software until both measuring devices exhibit the same temperature. On the basis of the relevant temperatures 500 °C, 650 °C, and 800 °C, the emission coefficient is defined as 0.54. In the procedure of the strip drawing tests, the metal strips are heated up slowly (within 30 s in order to attain uniform heating) to the austenitization temperature for 22MnB5 of 950 °C. This temperature is maintained for 30 s and then reduced to 800 °C, 650 °C, or 500 °C, respectively. After these temperatures are reached, they are held for 60 s, and the strip drawing test is conducted with a velocity of 30 mm/s, which is chosen on the basis of the real process conditions in hot stamping (Table 2). The three different temperatures are chosen according to representative temperatures, which occur during a common direct hot stamping process, i.e., the temperature at the beginning of forming (800 °C), at the end of forming (650 °C), and in the quenching phase (500 °C). Due to the limitations in the test setup, the investigated contact pressures are lower than the ones that appear in the real hot stamping process. However, the pressure is varied by three different values to represent the different pressure levels occurring in hot stamping.

Table 2

Strip drawing test parameters compared to hot stamping conditions

Strip drawing testsHot stamping (based on Ref. [19])
Blank temperature800 °C—500 °C800 °C (beginning of forming)— ≈ 550 °C (end of forming)
Relative velocity30 mm/s30 mm/s—10 mm/s
Contact pressure2.5 MPa—7.5 MPa (due to the limitations of the test setup)variable up to 40 MPa (depending on the forming stage and the geometry)
Strip drawing testsHot stamping (based on Ref. [19])
Blank temperature800 °C—500 °C800 °C (beginning of forming)— ≈ 550 °C (end of forming)
Relative velocity30 mm/s30 mm/s—10 mm/s
Contact pressure2.5 MPa—7.5 MPa (due to the limitations of the test setup)variable up to 40 MPa (depending on the forming stage and the geometry)

As a preparation before the conduction of the strip drawing tests, the blank is treated thermally in a furnace for 4 min at a heat of 950 °C (so that the blank experiences a total austenitization time of 5 min). This is done in order to generate a diffusion of the AlSi coating beforehand for avoiding an uneven distribution of the coating caused by the influence of the electromagnetic field during inductive heating, as it has been observed by Veit et al. [20].

2.4 Analytical Modeling of the Surface Roughness.

For the prediction of the surface roughness after ball burnishing, an analytical model that is based on the model for thermally sprayed surfaces is applied [21]. According to this model, the roughness after ball burnishing is dependent on the roughness before ball burnishing, the flow stress at a roughness peak σy, and the contact pressure under the burnishing ball pcontactarea(z) (Fig. 5) conforming to [22]. Thus, input parameters for the model include the original roughness (after the DED process), the Brinell hardness, and the flow stress. Whereas the first two parameters are measured, the last one is determined analytically from the ball-burnishing test by using the width of a burnishing path (del,pl) and the burnishing force Fball by means of
σy=83πFballdel,pl2φ2
(3)
Fig. 5
Contact pressure conditions in the ball-burnishing (also referred to as rolling) process [20]
Fig. 5
Contact pressure conditions in the ball-burnishing (also referred to as rolling) process [20]
Close modal
The surface roughness after ball burnishing can be calculated as follows:
Rz,burnished=Rz,untreated(1(23pnominalcontactarea(z)σy))
(4)
In this context, the averaged contact pressure pnc,ave is given as the quotient of the force of the burnishing ball Fball and the elastic-plastic contact area Ael,pl. According to Ref. [23], a fully plastic contact begins in the case of pnc,ave = HBW (hardness according to Brinell, tungsten carbide ball), i.e., the contact pressure is equal to the hardness value, and it does not exceed its value. Consequently, for the calculation of Rz,burnished, the following boundary conditions are used:
pnc,max={32FballAel,pl(1a2del,pl2),forpnc,aveHBW32HBW(1a2del,pl2),forpnc,ave>HBW
(5)

In contrast to the model of Ref. [21], the geometry of the surface to be ball burnished shows differences in terms of a periodical elliptical structure due to the deposited welding beads. This is taken into account for the determination of the width of a burnishing path with regard to the contact conditions of the burnishing ball and the elliptical surface.

2.5 Determination of Residual Stresses.

The determination of the residual stresses at the surface of the tool steel samples is carried out by X-ray diffraction (XStress 3000, Stresstech, Rennerod, Germany) at which the sin2ψ-method is used for the mathematical solution. The measurements are done at three different positions in the middle of each sample, with defined measuring parameters (Table 3).

Table 3

Measuring parameters used for X-ray diffraction

X-ray tube anodeWavelength (K-α)Diffraction angle Tilt angle ω (five iterations){hkl} plane
Chromium K-α radiation0.229 nm156.4±42 deg{211}
X-ray tube anodeWavelength (K-α)Diffraction angle Tilt angle ω (five iterations){hkl} plane
Chromium K-α radiation0.229 nm156.4±42 deg{211}

Each measurement is repeated three times. Residual stresses in both 0-deg and 90-deg directions are displayed. In this context, the 0-deg direction is equal to the direction of ball burnishing (and perpendicular to the direction of the DED process), whereas the 90-deg direction corresponds to the DED direction (Fig. 6). Since the data are taken from three different positions on the sample, the variation across the part is taken into account by the scatter resulting from these positions. Because the measurements are taken in a depth on a µm scale near to tool surface, it can be assured that it is taken in the area of the deformation depth of ball burnishing, which is on a µm scale due to the high hardness of the tool steels as well.

Fig. 6
Additively manufactured and ball-burnished specimen with measuring directions and points for the determination of residual stresses
Fig. 6
Additively manufactured and ball-burnished specimen with measuring directions and points for the determination of residual stresses
Close modal

2.6 Determination of the Heat Transfer Coefficient.

For the determination of the heat transfer coefficient between the additively manufactured and ball-burnished tool surfaces and the AlSi-coated 22MnB5 blank (with a thickness of 1.5 mm), experiments with a plate tool on the basis of hot stamping conditions are conducted. The tests are carried out on a universal testing machine (Zwick Z250, Ulm, Germany) on which a plate die is mounted (Fig. 7). The parts of the plate die being in contact with the blank are the actual contact plates made of the tool steel H13 (1.2344). They are coated with a 3-mm thick layer of the powder material Ferro 55 by means of DED. Type K thermocouples—located perpendicular to the contact surface with a distance to the surface of 1 and 3 mm—are used to measure the temperature of the contact plates. Before the test, the sheet is first preheated in a chamber furnace (Nabertherm N161) at 950 °C for 4 min. This is done according to the reason it is done for the strip drawing tests in order to have a prediffused state of the AlSi coating (see above). Afterward, it is clamped in the plate die by means of clamping jaws so that it is located centrally between the contact plates. The clamping jaws move with the blank so that bending of the blank is prevented. In order to heat the blank conductively, current is passed through it via the clamping jaws.

Fig. 7
Test stand for the determination of the temperatures relevant for the heat transfer coefficient: (a) front view and (b) side view
Fig. 7
Test stand for the determination of the temperatures relevant for the heat transfer coefficient: (a) front view and (b) side view
Close modal

The current for conductive heating is supplied by a DC source, which produces a maximum of 12 V at a maximum of 2000 A. The temperature of the sheet is measured by a pyrometer (Sensortherm Metis M318, Steinbach, Germany). The sheet is heated to 950 °C within 30 s and held at this temperature for 30 s. The ramp time of 30 s reduces the deformation of the sheet during heating; the 30-s holding time was chosen to ensure complete austenitization of the sheet. This is to produce the same conditions as in the furnace. After 60 s, the power supply is interrupted, and the plate tool is closed at 10 mm/s until a defined force corresponding to the desired contact pressure, which is varied in the range of 2.5–40 MPa, is reached. After 60 s, the plate tool is opened again. During the entire test, a thermal imaging camera (MicroEpsilon ThermoImager, Ortenburg, Germany), which is directed frontally at the center of the sheet, records the temperature when the tool is closed.

The heat transfer coefficient α is determined from Newton’s law of cooling as
α=ρcpVAtln(Tbl,0TtlTbl(t)Ttl)
(6)

Within the model, the temperature of the tool Ttl is calibrated with the help of the temperatures measured by the thermocouples, whereas the data from the thermal imaging camera is used for the temperature of the blank Tbl until 450 °C. Starting with an average temperature of 850 °C, a mean value of the heat transfer coefficient is determined in this temperature range using the temperature data recorded by the thermal imaging camera.

2.7 Hot Stamping Experiments.

For the application of additively manufactured cooling channels in a hot stamping tool, the geometry of a hat profile as a workpiece is chosen. In the setup, the upper part of the punch and its cooling channels are manufactured by DED using a path overlapping of 50% and the tool steel powder Ferro 55 (Fig. 8). The punch surface manufactured by DED is ball burnished with a burnishing pressure of 23.4 MPa and a sidestep of 0.1 mm. The cooled upper part of the punch is deposited on a premachined base material (1.2367) in order to save time and costs (Fig. 8(c)). Moreover, the cooling channels are intended to be located very close to the tool surface in the uppermost layers of the tool so that the functionality designed by the use of DED is limited to that area. Preliminary analyses regarding these aspects and detailed information about the cooling channel design can be found in Ref. [24].

Fig. 8
(a) Test setup for hot stamping experiments, (b) micrograph of a cooling channel in the additively manufactured punch, and (c) premachined punch basis
Fig. 8
(a) Test setup for hot stamping experiments, (b) micrograph of a cooling channel in the additively manufactured punch, and (c) premachined punch basis
Close modal

During the hot stamping tests, the temperature of the punch was measured by thermocouples located 2 mm beneath the surface. The following cases are compared:

  • Conventionally manufactured punch (tool steel 1.2367) with two drilled cooling channels with a diameter of 8 mm and a distance to the tool surface of approx. 15 mm (due to restrictions of the deep-hole drilling process).

  • Additively manufactured punch with approx. 4.5 mm distance between the tip of the drop-shaped cooling channels and the top surface. The width of the drop-shaped channels is 6 mm, and the distance between each is 8 mm.

The procedure of the hot stamping tests using AlSi-coated 22MnB5 blanks consists of the austenitization of the blanks in a furnace at 950 °C for 5 min, the direct transfer to the press, the forming operation, and the quenching with a holding time of 15 s.

3 Results and Discussion

In this section, the results from the roughness measurements, hardness tests, strip drawing tests, and the determination of the residual stresses as well as the heat transfer coefficient are presented and discussed as a function of the various parameter combinations of the DED and the ball-burnishing process (Table 4).

Table 4

Varied parameters in the DED and the ball-burnishing process and successive surface measurements

Directed energy depositionPath strategyPath overlappingNumber of layersSurface measurements
Zigzag/meander20%/50%1/2/3Roughness/hardness/residual stresses
Ball burnishingPath strategySidestepBurnishing pressureSurface measurements
Zigzag0.05 mm/0.1 mm/0.5 mm11.7 MPa/19.5 MPa/23.4 MPaRoughness/hardness/residual stresses/friction coefficient/heat transfer coefficient
Directed energy depositionPath strategyPath overlappingNumber of layersSurface measurements
Zigzag/meander20%/50%1/2/3Roughness/hardness/residual stresses
Ball burnishingPath strategySidestepBurnishing pressureSurface measurements
Zigzag0.05 mm/0.1 mm/0.5 mm11.7 MPa/19.5 MPa/23.4 MPaRoughness/hardness/residual stresses/friction coefficient/heat transfer coefficient

The variations of the parameters in the two processes are investigated against the background of influencing the material flow and the heat transfer in hot stamping by modified tool surfaces. In order to investigate whether the surface roughness has an impact on the material flow and the heat transfer, the path strategy and the path overlapping are altered. The two different path strategies are chosen on the basis of their suitability for rectangular workpieces. Like the path strategies, the path overlapping is varied to examine the influence, especially on the roughness. It is assumed that a higher path overlapping leads to a lower roughness because of the smaller gap between the welding beads (Fig. 9). The higher path overlapping of 50% is selected for having a balance between a low waviness of the surface and an efficient manufacturing process without consuming a large amount of metal powder.

Fig. 9
Schematic depiction of different path overlappings with (a) a higher value of overlapping (such as 50%) and (b) a smaller value of overlapping (such as 20%)
Fig. 9
Schematic depiction of different path overlappings with (a) a higher value of overlapping (such as 50%) and (b) a smaller value of overlapping (such as 20%)
Close modal

The number of layers is varied between one and three due to the need of depositing several layers of metal powder in the additive manufacturing process (e.g., because of the geometry of the part). By this means, their influence on the mechanical properties is examined. In this context, the properties of one layer are examined as a part of the referential analyses. In the ball-burnishing process, the parameters sidestep and burnishing pressure are varied because they are likely to affect the resulting roughness, as it has been shown by Hiegemann [17], for example. On this basis, a reduction of the surface roughness is expected for a small sidestep and a higher burnishing pressure (with the limitation of a formation of waviness). Furthermore, the hardness, which is relevant for the wear resistance of the tool surfaces, is expected to affect the roughness. The residual stresses play a role in the application in forming tools concerning the failure limit. The experimental plan shows which parameter variations were applied to the two different materials (Table 5). The plan does not aim at a full factorial design but rather at a representation of the most important surface effects achievable by the process combination of directed energy deposition and ball burnishing.

Table 5

Experimental plan with variations applied to the investigated materials

Measurements
RoughnessHardnessFrictionResidual stressesHeat transfer coefficient
DED ParametersPath strategyFerro 55/Ferro 702Ferro 55
Path overlappingFerro 55/Ferro 702Ferro 55/Ferro 702Ferro 55Ferro 55
LayersFerro 55/Ferro 702Ferro 55/Ferro 702
Ball-burnishing parametersBurnishing pressureFerro 55/Ferro 702Ferro 55/Ferro 702Ferro 55Ferro 55/Ferro 702
SidestepFerro 702Ferro 702
Measurements
RoughnessHardnessFrictionResidual stressesHeat transfer coefficient
DED ParametersPath strategyFerro 55/Ferro 702Ferro 55
Path overlappingFerro 55/Ferro 702Ferro 55/Ferro 702Ferro 55Ferro 55
LayersFerro 55/Ferro 702Ferro 55/Ferro 702
Ball-burnishing parametersBurnishing pressureFerro 55/Ferro 702Ferro 55/Ferro 702Ferro 55Ferro 55/Ferro 702
SidestepFerro 702Ferro 702

3.1 Analysis of the Roughness of Additively Manufactured and Ball-Burnished Tool Steel Surfaces.

The additively manufactured layers influence the roughness Rz in the way that it is accumulated with a higher number of layers and increased by 16% for three layers, especially in the case of a path overlapping of 20% (Fig. 10). Regarding the overlapping of the paths, the roughness is reduced by 38% when using overlapping of 50% compared to an overlapping of 20%. The meander path strategy is showing higher variations (by 10%) compared to the zigzag strategy.

Fig. 10
Measured roughness of the additively manufactured tool surfaces with a measuring length of 40 mm consisting of five single sections
Fig. 10
Measured roughness of the additively manufactured tool surfaces with a measuring length of 40 mm consisting of five single sections
Close modal

The comparison of the surface roughness of the material Ferro 55 with the material Ferro 702 displays that the surface roughness can be reduced by up to 45% (for one layer) with the use of Ferro 702 (Fig. 11).

Fig. 11
Measured roughness of the additively manufactured tool surfaces with different materials
Fig. 11
Measured roughness of the additively manufactured tool surfaces with different materials
Close modal

The characteristics of the additively manufactured tool steel surfaces are dependent on the process parameters applied in the two manufacturing processes. The surfaces manufactured in the DED process exhibit a high surface roughness, which increases with a higher number of layers (Fig. 10). In this way, an accumulation of the waviness of the subjacent layers takes place. The decrease of the roughness due to the higher path overlapping originates from the filling of the gaps that is more dominant in the case of an overlapping of 50% (Fig. 12). For the latter, a wider melting pool (including the adjacent welding bead), which leads to a stronger absorption of powder, is generated and leads to a flattening of the surface [7].

Fig. 12
(a) DED manufactured sample with a path overlapping of 50% (left) and 20% (right), micrograph of cross section of two additively manufactured layers with (b) 50% and (c) 20% path overlapping
Fig. 12
(a) DED manufactured sample with a path overlapping of 50% (left) and 20% (right), micrograph of cross section of two additively manufactured layers with (b) 50% and (c) 20% path overlapping
Close modal

The meander strategy causes higher variations of the surface roughness, resulting in an increase in the roughness compared to the zigzag strategy. In this context, the heat distribution in the meander strategy is considered as a probable reason: the heat is mainly concentrated in the middle of the sample where the welding beads are located closer together. When coming to the outer part of the sample, the beads begin to cool down before the adjacent welding bead is placed (i.e., rapid cooling and heating are alternating). These findings indicate that the generated properties of additively manufactured parts depend on the process parameters because they influence the formation of the macro- and microstructure due to the varying heat distribution and cooling rate in the DED process. The comparison of the surface roughness of the two materials demonstrates that the roughness of the samples manufactured with the material Ferro 702 exhibits a lower roughness level. This is caused mainly by the lower carbon content of Ferro 702, which leads to better weldability because of the prevention of martensite formation affecting the hardness evolution in the heat-affected zone (in contrast to the Ferro 55, which is hardened in the DED process). Additionally, the Ferro 702 has a lower thermal conductivity, which leads to slower cooling in the DED process so that a more homogeneous bonding of the powder is achieved.

The ball-burnished surfaces with a path overlapping of 50% (DED) show for both materials a reduction of the surface roughness with higher burnishing pressures pball defined in Eq. (1) (Fig. 13). In the case of the Ferro 702, the reduction of the surface roughness adds up to 75%, whereas the reduction in the case of the Ferro 55 amounts to 35%.

Fig. 13
Measured roughness of the additively manufactured tool surfaces after ball burnishing with varied burnishing pressure
Fig. 13
Measured roughness of the additively manufactured tool surfaces after ball burnishing with varied burnishing pressure
Close modal

The sidestep in the burnishing process influences the surface roughness of the tool steel in the manner of a reduction of the roughness with smaller sidesteps (Fig. 14). By this means, a reduction of up to 70% in comparison to the additively manufactured surface is achieved.

Fig. 14
Measured roughness of the additively manufactured tool surfaces after ball burnishing with a variation of the sidestep value
Fig. 14
Measured roughness of the additively manufactured tool surfaces after ball burnishing with a variation of the sidestep value
Close modal

The ball-burnishing process leads to a reduction of the surface roughness with specific differences regarding the processed material and the ball-burnishing pressure. An increase in the burnishing pressure leads to lower roughness for both materials (Fig. 13). In this context, the reduction of the roughness of the Ferro 702 is considerably higher compared to the roughness of the Ferro 55. In combination with the lower hardness of the Ferro 702, it can be reasoned that with higher burnishing pressures, plastic deformation of the material occurs. In the ball-burnishing process, a second parameter, the sidestep (or path distance), influences the resulting surface roughness in the manner that a smaller sidestep results in a lower roughness. Consequently, a higher burnishing pressure is not required necessarily for achieving a low roughness level because this can be achieved by a smaller sidestep as well. However, a decrease in the sidestep leads to an increase in the processing time, which represents a drawback for the manufacturing of large workpieces. Altogether, the path overlapping in the DED process shows the most significant influence on the surface roughness, i.e., by means of the path overlapping, the roughness can be adjusted in the widest range compared to the other parameters (Fig. 15). In the same regard, ball burnishing has a significant influence on the roughness of the tool surface manufactured by DED, but in a less wide range regarding the roughness values compared to the ones generated by DED. In this context, the effect of the number of layers exhibits corresponding trends with the path overlapping so that these two factors are interdependent. However, in the most practical cases, a number of more than two layers will be deposited on the workpiece so that the case of one layer serves as a reference experiment. The ball-burnishing parameters have a less significant effect regarding the range in which the roughness can be regulated. Nevertheless, the ball-burnishing process enables a reduction in the surface roughness after the DED process, as shown in the presented investigations.

Fig. 15
Summary of the influencing parameters on the surface roughness
Fig. 15
Summary of the influencing parameters on the surface roughness
Close modal

3.2 Analytical Prediction of the Surface Roughness After Ball Burnishing.

For the analytical prediction of the surface roughness after ball burnishing, the above-described model is applied targeting an avoidance of preparatory experiments for the investigation of the resulting roughness. The comparison of the analytical and the experimental values of the surface roughness for two different ball-burnishing pressures shows a deviation of 6% on average (Fig. 16). By this means, extensive experimental preliminary works can be avoided. The analytical results demonstrate that an increase in the burnishing pressure leads to a lower surface roughness for both path overlappings of 20% and 50%. However, in the case of the experimental results, there is a tendency for the path overlapping of 20% to result in a slightly higher surface roughness after ball burnishing with a pressure of 19.5 MPa compared to 11.7 MPa. In this context, there is not only a flattening of the roughness peaks taking place but also an intensification of the valleys between the welding beads, especially for the cases of one and two layers due to the lack of a filling of the valleys between the welding beads in these states. This effect is not displayed by the analytical model and leads to an underestimation of the roughness after ball burnishing for a path overlapping of 20%. Consequently, a higher path overlapping, e.g., 50%, leads to a more reliable prediction of the surface roughness after ball burnishing by the model since such effects like the ones for smaller overlappings, e.g., 20%, do not occur.

Fig. 16
Analytical prediction of the roughness of additively manufactured surfaces ball burnished with a pressure of (a) 11.7 MPa and (b) 19.5 MPa
Fig. 16
Analytical prediction of the roughness of additively manufactured surfaces ball burnished with a pressure of (a) 11.7 MPa and (b) 19.5 MPa
Close modal

3.3 Analysis of the Hardness of Additively Manufactured and Ball-Burnished Tool Steel Surfaces.

The hardness of the additively manufactured surfaces increases with the number of layers by 30% for the comparison of three layers to one layer manufactured with the zigzag strategy (Fig. 17). This can be (for one layer) due to the influence of the softer substrate material (S235) that is diminished with a higher number of deposited layers. Additionally, the hardness of the Ferro 55 is increased with a higher number of layers due to its thermo-mechanical properties, which causes a hardening of the material during the DED process. There is also an increase in the hardness by 16% for the path overlapping of 50% compared to the path overlapping of 20%. The meander strategy shows higher variations analog to the results of the roughness measurements. Hence, the path strategy does not change the hardness. Rather, there is a slightly higher scatter in the case of the meander strategy.

Fig. 17
Brinell hardness of the additively manufactured tool steel surfaces
Fig. 17
Brinell hardness of the additively manufactured tool steel surfaces
Close modal

After ball burnishing, the hardness of the additively manufactured tool surfaces increases with the burnishing pressure by up to 15% (Fig. 18). However, this is limited to the samples manufactured with one layer in the DED process—there is no significant increase in hardness for the cases with three layers.

Fig. 18
Brinell hardness of the additively manufactured tool surfaces after ball burnishing for the material Ferro 55
Fig. 18
Brinell hardness of the additively manufactured tool surfaces after ball burnishing for the material Ferro 55
Close modal

The hardness of the material Ferro 702 exhibits a lower level (by up to 60%) compared to the material Ferro 55 after the DED process. In contrast to the Ferro 55, the Ferro 702 does not change its hardness values significantly during DED. After ball burnishing, the hardness increases by up to 20% (Fig. 19). As described for the material Ferro 55, this increase is less significant for the samples with three DED layers.

Fig. 19
Brinell hardness of the additively manufactured tool surfaces after ball burnishing for the material Ferro 702
Fig. 19
Brinell hardness of the additively manufactured tool surfaces after ball burnishing for the material Ferro 702
Close modal

The hardness of the materials is relevant concerning the wear properties for use in hot stamping tools and the potential of leveling the welding beads originating from the DED process in the ball-burnishing process. For the material Ferro 55, the resulting hardness depends mainly on the number of layers and the path overlapping (Fig. 17). The hardness increases with the number of layers and the path overlapping. In both cases, the fraction of material involved in the deposition is extended. In combination with the thermal conductivity of the material, the heat generated during the DED process dissipates faster, i.e., the high thermal conductivity causes a high cooling rate and thus a self-quench hardening of the material. After ball burnishing, the hardness exhibits an increase of up to 15%, which indicates that the surface is work hardened during the ball-burnishing process. Since tool steel is investigated, there are no phase changes expected due to forming (i.e., ball burnishing), but rather work hardening effects with respect to dislocations. In accordance with the lower hardness level of the material Ferro 702, the increase of the hardness by ball burnishing is higher (by up to 20%) compared to the Ferro 55. However, the increase of the hardness saturates for most cases. It can be assumed that with higher burnishing pressures, the hardness will decrease, as it has been observed by Srinivasa Rao [25]. This is due to a flaking of the surface layers resulting from the intense work hardening. Such flaking is witnessed for the investigated samples burnished with a pressure of 23.4 MPa as well. The summarized comparison of the hardness achieved after the DED process demonstrates that the material Ferro 55 enables to generate a wider range of hardness values compared to the material Ferro 702 (Fig. 20).

Fig. 20
Summary of the influencing parameters on the hardness
Fig. 20
Summary of the influencing parameters on the hardness
Close modal

3.4 Analysis of the Friction Behavior of Additively Manufactured and Ball-Burnished Surfaces.

The investigation of the friction behavior of the friction pair consisting of the additively manufactured and ball-burnished friction jaws and the 22MnB5 sheet metal strips shows a decreasing friction coefficient with higher contact pressures (Fig. 21).

Fig. 21
Friction behavior resulting from strip drawing tests with 22MnB5 strips at room temperature for the materials (a) Ferro 55 and (b) Ferro 702
Fig. 21
Friction behavior resulting from strip drawing tests with 22MnB5 strips at room temperature for the materials (a) Ferro 55 and (b) Ferro 702
Close modal

The friction coefficient decreases with higher path overlapping (DED) in most cases as well. Depending on the material, the friction coefficient is decreased in the case of Ferro 702 as the material of the fiction jaw surface.

In the strip drawing tests at room temperature, a drawing velocity of 10 mm/s, i.e., slower compared to the tests at elevated temperatures, is applied. This is done to avoid an impact of the velocity on the other parameters so that a high accuracy of the results can be ensured. The friction behavior at elevated temperatures indicates that the friction coefficient decreases with an increase in the contact pressure, as found in the strip drawing tests at room temperature (Fig. 22(a)). The friction coefficient is at a higher level for elevated temperatures (compared to room temperature) in all examined cases.

Fig. 22
Friction behavior resulting from strip drawing tests with 22MnB5 strips at elevated temperatures with a variation of (a) the contact pressure and (b) the path overlapping (DED)
Fig. 22
Friction behavior resulting from strip drawing tests with 22MnB5 strips at elevated temperatures with a variation of (a) the contact pressure and (b) the path overlapping (DED)
Close modal

The DED process influences the friction behavior in a similar way as with room temperature: larger path overlapping leads to lower friction coefficients (Fig. 22(b)). The friction coefficient can be reduced slightly by manufacturing the friction jaws with higher burnishing pressures (Fig. 23(a)). Additionally, a reduction of the friction coefficient is obtained when using the material Ferro 702 for the surface of the friction jaws (Fig. 23(b)). Regarding the three investigated elevated temperatures, the friction coefficient at 500 °C is on a higher level compared to the temperatures of 650 °C and 800 °C for most parameter combinations.

Fig. 23
Friction behavior resulting from strip drawing tests with 22MnB5 strips at elevated temperatures with a variation of (a) the ball-burnishing pressure and (b) the material of the tool surface
Fig. 23
Friction behavior resulting from strip drawing tests with 22MnB5 strips at elevated temperatures with a variation of (a) the ball-burnishing pressure and (b) the material of the tool surface
Close modal

The same result is given for the strip drawing tests at elevated temperatures: The friction coefficient exhibits the lowest values for the highest contact pressure because roughness peaks are leveled in this way. As for room temperature, the friction coefficient is on a lower level (by 7%) for a higher path overlapping of the material Ferro 55 because the surface roughness in this case is lower (by 50%), too (Fig. 24(a)). This reflects the same tendency as the values of the surface roughness, which are also decreasing with a higher path overlapping. The material Ferro 702 does not show the described tendencies regarding the influence of the path overlapping in the same manner: A smaller path overlapping does not (always) lead to a higher friction coefficient, although the surface roughness increases (Figs. 12 and 21). This indicates that the lower hardness of the Ferro 702 contributes to a leveling of the roughness peaks of the tool surfaces in the strip drawing test. For the elevated temperatures, the friction coefficient is at a higher level compared to the friction coefficient at room temperature, as expected. In this context, the AlSi coating of the metal sheets plays an important role in the tribological behavior of the tool surface. When comparing the three different elevated temperatures, the friction coefficient is at the highest level for the temperature of 500 °C and decreases for the temperatures of 650 °C and 800 °C (Figs. 22 and 23). This discontinuous development is owing to the AlSi coating, as it has been observed by Schwingenschlögl [26]: At temperatures around 500 °C, the coating is still in a state of a high strength so that roughness peaks are leveled only slightly. With higher temperatures over 600 °C, the shear strength of the coating is lowered and leveling of the surface occurs. Additionally, loose material particles working as a solid lubricant are generated by cleavage of the coating and fill roughness lows of both friction partners [26]. Furthermore, the friction behavior with regard to the AlSi coating on the tool surface plays a critical role in the forming stage (for strains starting from 0.017): In this context, it is likely that the coating experiences micro cracks, leading to increased abrasive wear at higher temperatures [27].

Fig. 24
(a) Comparison of roughness and friction coefficient for different manufacturing parameters of the friction jaw surface and (b) summary of the influencing parameters on the friction coefficient
Fig. 24
(a) Comparison of roughness and friction coefficient for different manufacturing parameters of the friction jaw surface and (b) summary of the influencing parameters on the friction coefficient
Close modal

The burnishing pressure affects the friction coefficient as well: a higher burnishing pressure leads to a slight decrease of the friction coefficient of 5% in the case of the material Ferro 55. This slight influence corresponds to the development of the surface roughness with increasing burnishing pressures: The roughness decreases slightly by 9% as well (Fig. 24(a)). The summary of the influencing parameters on the friction coefficient shows that the path overlapping of the welded beads enables the most significant influence on the friction coefficient regarding the tool surface conditions (Fig. 24(b)). The strip drawing conditions like blank temperature and contact pressure influence the friction coefficient as well. These two parameters are interdependent and are determined by the hot stamping process itself.

3.5 Analysis of Residual Stresses of Additively Manufactured and Ball-Burnished Surfaces.

The residual stresses are determined for the material Ferro 55 (Fig. 25) under consideration of varying path overlapping (DED) and ball-burnishing pressure. In all cases, compressive residual stresses are determined. Due to the asperities of the tool surfaces manufactured by DED, the X-ray measurements exhibit a relatively high scatter, especially in the case of small path overlappings (≤20%). Because of the fact that these asperities accumulate with a higher number of layers, only samples with one layer have been taken into account in order to avoid a considerable influence of the asperities in the measured residual stresses.

Fig. 25
Measured residual stresses for the material Ferro 55 with a path overlapping of (a) 20% and (b) 50%
Fig. 25
Measured residual stresses for the material Ferro 55 with a path overlapping of (a) 20% and (b) 50%
Close modal

For the path overlapping of 50%, the residual stresses are mainly on a lower level compared to the ones with a path overlapping of 20%. After the ball-burnishing process, the compressive residual stresses reach higher values compared to the state after the DED process. The trends of the residual stresses also match with the corresponding hardness values, i.e., the increase of the hardness with higher burnishing pressure corresponds to the increase of the residual stresses.

However, for the highest burnishing pressure, there is a tendency for a decrease in the residual stresses introduced by ball burnishing. Similar tendencies can be seen for the residual stresses of the material Ferro 702 (Fig. 26): The determined residual stresses are compressive stresses, which are enhanced by ball burnishing. Additionally, in this case, the sidestep during ball burnishing is varied for the ball-burnishing pressure of 19.5 MPa, with a path overlapping of 50% (DED). The results of this variation demonstrate that a bigger sidestep leads to lower residual stresses, i.e., an increase of the sidestep from 0.1 mm to 0.5 mm results in a reduction of the residual stresses of 24% for the direction of 0 deg and 74% for the direction of 90 deg, respectively.

Fig. 26
Measured residual stresses for the material Ferro 702 for the directions of (a) 0 deg and (b) 90 deg
Fig. 26
Measured residual stresses for the material Ferro 702 for the directions of (a) 0 deg and (b) 90 deg
Close modal

The compressive residual stresses occur due to the high cooling rate in the process of typically 103–105 °C/s [6], which leads to a shrinkage of the material. In this context, it is likely that tensile residual stresses appear as well due to the repetitive re-melting and re-solidification in the DED process, which can lead to a complex residual stress field. The compressive stresses are increased by the ball burnishing as a function of the burnishing pressure. The latter represents the most important ball-burnishing parameter regarding the influence on the residual stresses [11]. The decrease of the residual stresses at the highest burnishing pressure can be explained by the occurrence of flaking of the work-hardened surface resulting in stress relief. The stepover in the ball-burnishing process influences the residual stresses in a less significant way because the stepovers of 0.1 mm and 0.05 mm are almost on the same level, whereas the stepover of 0.5 mm leads to a degradation of the residual stresses due to the smaller contact area of the burnishing tool with the tool steel surface. In combination with the lower surface roughness quality of the biggest stepover, it is not considered as beneficial for the surface properties of the tool surfaces. The comparison of the two path overlappings and the resulting residual stresses indicate that the higher hardness level of the surfaces manufactured with a path overlapping of 50% leads to a minor influence of the subsequent ball burnishing regarding the level of the compressive residual stresses. The latter also differs depending on the direction of the residual stresses: The residual stresses exhibit higher values for the direction of ball burnishing (0-deg direction). Thus, the plastic deformation does not have the same intensity in both 0-deg and 90-deg directions, leading to a directionality of the residual stresses. The observed variations for the residual stresses of the material Ferro 55 result from the high surface roughness, which leads to a less deep and unsteady penetration depth of the X-ray beams [28]. In general, compressive residual stresses are beneficial for tool surfaces, e.g., regarding their endurance limit. In this context, the ball-burnishing process contributes to the surface integrity of parts, which comprises the decrease in roughness, the introduction of compressive residual stresses, and the increase in hardness. In summary, the residual stresses are influenced by the burnishing pressure and the sidestep in the ball-burnishing process most significantly (Fig. 27(a)).

Fig. 27
(a) Summary of the influencing parameters on the residual stresses and (b) residual stresses after preheating
Fig. 27
(a) Summary of the influencing parameters on the residual stresses and (b) residual stresses after preheating
Close modal

In the context of a tool used with heated workpieces, it has to be considered that the residual stresses can be influenced by high temperatures. This is investigated by exemplary samples heated up to 100 °C and 200 °C, respectively, letting them cool down completely before the conduction of the measurements (Fig. 27(b)). By heating the tool specimen in a furnace, the compressive residual stresses are reduced by up to 35% in the given example when comparing the nonheated with the preheated specimen at 200 °C. In this sense, it can be reasoned that the residual stresses will be lowered during the use of the tool for hot-forming operations.

In the framework of these investigations, it is worth mentioning that deeper analyses of the residual stress evolution by DED and ball burnishing of tool steels can serve a better understanding of the mechanisms taking place. This is especially true with regard to the corresponding microstructure evolution and the interdependency of the (depth of the) residual stresses and the plastic deformation zone in ball burnishing. The existence of those effects has been shown by Sunny et al. [29] and Courbon et al. [30], for example. However, here the process combinations used—deviating from the presented case—are powder bed fusion with interlayer roller burnishing and laser cladding with ball burnishing for stainless steel, respectively.

3.6 Analysis of the Heat Transfer Coefficient of Additively Manufactured and Ball-Burnished Tool Surfaces and 22MnB5 Blank.

The analyzed configuration for the tool surface is chosen because it exhibits the lowest roughness compared to other investigated manufacturing routes (see results above) so that the contact conditions are as close as possible to conventionally manufactured tool surfaces. For the case of additively manufactured tool surfaces with a path overlapping of 50% that is ball burnished subsequently with a burnishing pressure of 23.4 MPa, the average heat transfer coefficient shows the trend of an increase with increasing contact pressures (Fig. 28). With an increasing contact pressure, the real contact area between the tool surface and the blank is increased because there is a plastic deformation of the (softer) blank material taking place. Consequently, at low pressures, the real contact area is smaller, wherefore the blank is cooling down slower. By contrast, the contact area increases with increasing pressure, leading to a faster cooling through an increase of the heat transfer.

Fig. 28
Heat transfer coefficient determined by experiments and analytical method
Fig. 28
Heat transfer coefficient determined by experiments and analytical method
Close modal

The heat transfer coefficient determined analytically for the cases of 20% and 50% path overlapping shows only slight differences, e.g., in the case of a contact pressure of 2.5 MPa, the heat transfer coefficient is 25% higher for 50% path overlapping. For higher contact pressures, the value for the heat transfer coefficient of both examined cases does not show significant differences. Generally, the determined values for the heat transfer coefficient are, on average, slightly below those of comparable experiments in literature (Table 6). Although the tool surfaces exhibit high surface roughness, especially for the case of 20% path overlapping, the heat transfer coefficient is less significantly affected hereby compared to the more considerable differences in surface roughness for the two path overlappings. Here, it has to be taken into account that the heat conductance of the material Ferro 55 is higher than the one of the material 1.2379 so that the higher roughness is (at least partially) compensated.

Table 6

Determined heat transfer coefficients with comparison to literature values of 22MnB5 and conventionally manufactured tool steel surface (1.2379)

Contact pressure in MPaDetermined analyticallyDetermined inversely (by finite element method)
Heat transfer coefficient in W/m2K (50% path overlapping)Heat transfer coefficient according to [31] in W/m2KHeat transfer coefficient according to [32] in W/m2KHeat transfer coefficient according to [33] in W/m2K
2.510561400
51201153910002000
7.516231678
102015181815002200
202660237417002700
302546293025003100
40272234863500
Contact pressure in MPaDetermined analyticallyDetermined inversely (by finite element method)
Heat transfer coefficient in W/m2K (50% path overlapping)Heat transfer coefficient according to [31] in W/m2KHeat transfer coefficient according to [32] in W/m2KHeat transfer coefficient according to [33] in W/m2K
2.510561400
51201153910002000
7.516231678
102015181815002200
202660237417002700
302546293025003100
40272234863500

3.7 Hot Stamping of 22MnB5 Hat Profiles.

Based on the aforementioned analyses of the tool surfaces manufactured with the novel process route, a hot stamping punch is manufactured. The comparison of the punch temperatures measured by thermocouples located 2 mm below the tool surface shows that the additively manufactured cooling channels not only reduce the temperature in the hot stamping process by up to 50%. It also keeps the temperature constant during consecutive forming cycles, whereas the temperature of the conventionally manufactured punch increases (Fig. 29).

Fig. 29
Punch temperatures during five hot stamping cycles with conventionally manufactured punch and punch manufactured by DED and ball burnishing
Fig. 29
Punch temperatures during five hot stamping cycles with conventionally manufactured punch and punch manufactured by DED and ball burnishing
Close modal

After the hot stamping tests, hardness measurements are performed on the hat profiles. The results indicate that the hat profile exhibits a higher hardness when using near-surface cooling channels manufactured by DED (Fig. 30(a)). The hardness distribution across all areas is more homogeneous compared to conventionally manufactured parts. Those parts exhibit a sharp drop in the hardness from the top to the wall area. Similar observations can be made regarding the sheet thickness and its distribution: The part has a higher thickness across all areas, and the distribution is more homogeneous (Fig. 30(b)). For the conventionally hot-stamped parts, the drop in hardness in the wall correlates with the lowest sheet thickness observed in the same region. Based on these findings, the effects of the additively manufactured and ball-burnished tool surfaces on the hot-stamped hat profile are now qualitatively known.

Fig. 30
(a) Hardness measurements performed on the hot-stamped hat profile and (b) evaluation of the sheet thicknesses of the hat profile after hot stamping
Fig. 30
(a) Hardness measurements performed on the hot-stamped hat profile and (b) evaluation of the sheet thicknesses of the hat profile after hot stamping
Close modal

4 Conclusions

The presented investigations characterize the influence of varied parameters in the DED and the ball-burnishing process on the surface quality of tool steels, e.g., for application in hot stamping. Regarding the presented processes of DED and ball burnishing, it has been shown that these procedures are feasible for the manufacturing of tool surfaces. The ball-burnishing process is able to achieve an enhancement of the surface integrity of additively manufactured tool steel surfaces. By ball burnishing, the surface roughness resulting from the DED process can be reduced by up to 75% in the case of Ferro 702 and by up to 35% for Ferro 55—depending mainly on the path overlapping (DED) and the ball-burnishing pressure. The prediction of the roughness after ball burnishing is possible by applying the principle of the analytical model of Ref. [21]. On the basis of the average deviation from the experimental results of 6%, the analytical roughness prediction provides the advantage that an experimental characterization of the surface roughness after ball burnishing becomes avoidable. The friction behavior of the examined tool surfaces in combination with AlSi-coated 22MnB5 blanks shows a correlation with the surface roughness of the tools for the material Ferro 55, i.e., the friction coefficient increases by 12% with an increase of the surface roughness by 55%. In this context, the most significant influence is found for the path overlapping from the DED process. Additionally, the ball burnishing of the additively manufactured tool surfaces induces compressive residual stresses in the direction of ball burnishing and perpendicular to it, leading to a higher endurance strength of the tools. In an outline, the manufactured tool surfaces exhibit the following properties:

  • Higher roughness compared to conventionally manufactured tools

  • Hardened tool surface after first manufacturing step (depending on material)

  • Friction behavior comparable to known hot stamping conditions

    • Heat transfer coefficient values slightly lower compared to conventional tool surfaces

  • Ability to manufacture hot-stamped 22MnB5 parts with high hardness and low thickness variation in combination with keeping the tool temperature at a lower and more constant level compared to conventionally manufactured tools.

With regard to the intended field of application, the two investigated materials Ferro 55 and Ferro 702 differ in their qualification for the manufacturing of (hot stamping) tools exhibiting a great range of adjustable parameters (Fig. 31).

Fig. 31
Summary of the properties adjustable by DED and ball burnishing with corresponding limits
Fig. 31
Summary of the properties adjustable by DED and ball burnishing with corresponding limits
Close modal

Regarding a possible field of application of the investigated tool surfaces for hot stamping, it has to be analyzed how the determined heat transfer coefficients influence the properties of the formed part, e.g., whether a sufficient cooling rate and a fully martensitic microstructure is achieved. This depends not only on the surface properties of the tool (e.g., the roughness) but also on the geometry, positioning, and structure of the cooling channels. However, a lower heat transfer coefficient can also be used to achieve a retarded cooling of the blank, e.g., in the area of the blank holder. By this means, the material flow into the forming area can be improved due to the avoidance of high temperature gradients at the beginning of the forming stage. The first results of hot stamping tests conducted with a punch manufactured by DED and ball burnishing show that it is possible to produce hardened sheet metal parts with the investigated tool surfaces generated by the presented novel process combination. Moreover, an increased overall and more homogeneous hardness as well as more homogeneous thickness distribution as compared to conventionally manufactured tools is achieved.

The potentials of the novel process combination of DED and ball burnishing for the manufacturing of tool surfaces are now determined. The combination of the described methods allows for manufacturing tool surfaces, e.g., for hot forming applications. By the presented variations of the manufacturing parameters, a global or local adjustment of the surface properties of the tools is possible. However, the possibility to actively adjust the material flow and the heat transfer is limited. To enable more significant effects in this field, additional use of textured tool surfaces combined with near-surface cooling channels is desirable. Further investigations within the presented scope comprise the analysis of different cooling channel configurations in combination with texturing with respect to the heat balance in the (hot stamping) tool.

Acknowledgment

The authors thank the German Research Foundation (DFG) for the financial support of the project “Functionalization of additively manufactured hot stamping tools by ball burnishing” (project number 417202720). The authors also thank Dr. Ricker Meya and Dr. Till Clausmeyer for their helpful comments on this paper.

Conflict of Interest

There are no conflicts of interest.

Data Availability Statement

The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.

Nomenclature

t =

time

A =

sidestep

V =

volume

cp =

heat capacity

dball =

diameter of burnishing ball

pball =

ball burnishing pressure

wburnishing path =

width of ball burnishing path

Aball =

cross section (of ball burnishing ball)

Fball =

ball-burnishing force

Ra =

average roughness value

Rz =

maximum roughness depth

Tbl,0 =

initial temperature of blank

Ttl =

temperature of tool

α =

heat transfer coefficient

μ =

friction coefficient

ρ =

density

σy =

flow stress

References

1.
Karbasian
,
H.
, and
Tekkaya
,
A. E.
,
2010
, “
A Review on Hot Stamping
,”
J. Mater. Process. Technol.
,
210
(
15
), pp.
2103
2118
.
2.
Billur
,
E.
, and
Senuma
,
T.
,
2019
, “Die Design and Manufacturing,”
Hot Stamping of Ultra-High Strength Steels—From a Technological and Business Perspective
,
E.
Billur
, ed.,
Springer Nature
,
Cham, Switzerland
, pp.
105
129
.
3.
Porzner
,
H.
, and
Billur
,
E.
,
2019
, “Computer Modeling of Hot Stamping,”
Hot Stamping of Ultra-High Strength Steels—From a Technological and Business Perspective
,
E.
Billur
, ed.,
Springer Nature
,
Cham, Switzerland
, pp.
203
223
.
4.
Ghiotti
,
A.
,
Bruschi
,
S.
, and
Borsetto
,
F.
,
2011
, “
Tribological Characteristics of High Strength Steel Sheets Under Hot Stamping Conditions
,”
J. Mater. Process. Technol.
,
211
(
11
), pp.
1694
1700
.
5.
Pelcastre
,
L.
,
Hardell
,
J.
, and
Prakash
,
B.
,
2011
, “
Investigations Into the Occurrence of Galling During Hot Forming of Al-Si-Coated High-Strength Steel
,”
Proc. Inst. Mech. Eng. Part J J. Eng. Tribol.
,
225
(
6
), pp.
487
498
.
6.
Gibson
,
I.
,
Rosen
,
D.
, and
Stucker
,
B.
,
2015
, “Directed Energy Deposition Processes,”
Additive Manufacturing Technologies
, 2nd ed.,
I.
Gibson
,
D.
Rosen
, and
B.
Stucker
, eds.,
Springer
,
New York
, pp.
245
268
.
7.
Ribeiro
,
K. S. B.
,
Mariani
,
F. E.
, and
Coelho
,
R. T.
,
2020
, “
A Study of Different Deposition Strategies in Direct Energy Deposition (DED) Processes
,”
Procedia Manuf.
,
48
, pp.
663
670
.
8.
Hiegemann
,
L.
,
Weddeling
,
C.
,
Ben Khalifa
,
N.
, and
Tekkaya
,
A. E.
,
2013
, “
Control of the Material Flow in Deep Drawing by the Use of Rolled Surface Textures
,”
6. Öffentliches Kolloquium SFB 708
,
W.
Tillmann
, and
I.
Baumann
, eds., pp.
33
45
.
9.
Luca
,
L.
,
Neagu-Ventzel
,
S.
, and
Marinescu
,
I.
,
2005
, “
Effects of Working Parameters on Surface Finish in Ball-Burnishing of Hardened Steels
,”
Precis. Eng.
,
29
(
2
), pp.
253
256
.
10.
López de Lacalle
,
L. N.
,
Lamikiz
,
A
,
Sánchez
,
J. A.
, and
Arana
,
J. L.
,
2007
, “
The Effect of Ball Burnishing on Heat-Treated Steel and Inconel 718 Milled Surfaces
,”
Int. J. Adv. Manuf. Technol.
,
32
(
9–10
), pp.
958
968
.
11.
Chomienne
,
V.
,
Valiorgue
,
F.
,
Rech
,
J.
, and
Verdu
,
C.
,
2016
, “
Influence of Ball Burnishing on Residual Stress Profile of a 15-5PH Stainless Steel
,”
CIRP J. Manuf. Sci. Technol.
,
13
, pp.
90
96
.
12.
Li
,
C.
,
Liu
,
Z. Y.
,
Fang
,
X. Y.
, and
Guo
,
Y. B.
,
2018
, “
Residual Stress in Metal Additive Manufacturing
,”
Procedia CIRP
,
71
, pp.
348
353
.
13.
voestalpine
,
2016
, “Material Data Sheet for UTP PLASweld Ferro 55,” www.voestalpine.com/welding.
14.
voestalpine
,
2017
, “Material Data Sheet for UTP PLASweld Ferro 702,” www.voestalpine.com/welding.
15.
DIN EN ISO 4288
,
1997
,
Geometrical Product Specifications (GPS)—Surface Texture: Profilemethod—Rules and Procedures for the Assessment of Surface Texture
,
Beuth GmbH
,
Berlin
.
16.
Jardon
,
Z.
,
Ertveldt
,
J.
,
Lecluyse
,
R.
,
Hinderdael
,
M.
, and
Pyl
,
L.
,
2022
, “
Directed Energy Deposition Roughness Mitigation Through Laser Remelting
,”
Procedia CIRP
,
111
, pp.
180
184
.
17.
Hiegemann
,
L.
,
2017
, “
Glattwalzen Beschichteter Oberflächen
,”
Ph.D. thesis
,
Institute of Forming Technology and Lightweight Components, TU Dortmund University
,
Shaker, Aachen, Germany
.
18.
DIN EN ISO 6506-1, 2014
,
Metallic Materials—Brinell Hardness Test—Part 1: Test Method
,
Beuth GmbH
,
Berlin
.
19.
Karbasian
,
H.
,
2010
, “
Formgenauigkeit und Mechanische Eigenschaften Pressgehärteter Bauteile
,”
Ph.D. thesis
,
Institute of Forming Technology and Lightweight Components, TU Dortmund University
,
Shaker, Aachen, Germany
.
20.
Veit
,
R.
,
Hofmann
,
H.
,
Kolleck
,
R.
,
Sikora
,
S.
,
Chinesta
,
F.
,
Chastel
,
Y.
, and
El Mansori
,
M.
,
2011
, “
Investigation of the Phase Formation of AlSi-Coatings for Hot Stamping of Boron Alloyed Steel
,”
AIP Conf. Proc.
,
1315
, pp.
769
774
.
21.
Hiegemann
,
L.
,
Weddeling
,
C.
, and
Tekkaya
,
A. E.
,
2016
, “
Analytical Contact Pressure Model for Predicting Roughness of Ball Burnished Surfaces
,”
J. Mater. Process. Technol.
,
232
, pp.
63
77
.
22.
Hertz
,
H.
,
1881
, “
Über die Berührung Fester Elastischer Körper
,”
J. Reine Angew. Math.
,
92
, pp.
156
171
.
23.
Zhao
,
Y.
,
Maietta
,
D. M.
, and
Chang
,
L.
,
2000
, “
An Asperity Microcontact Model Incorporating the Transition From Elastic Deformation to Fully Plastic Flow
,”
ASME J. Tribol.
,
22
(
1
), pp.
86
93
.
24.
Komodromos
,
A.
,
Kolpak
,
F.
, and
Tekkaya
,
A. E.
,
2022
, “
Manufacturing of Integrated Cooling Channels by Directed Energy Deposition for Hot Stamping Tools With Ball Burnished Surfaces
,”
Berg Huettenmaenn Monatsh
,
167
(
9
), pp.
428
434
.
25.
Srinivasa Rao
,
D.
,
Suresh Hebbar
,
H.
,
Komaraiah
,
M.
, and
Kempaiah
,
U. N.
,
2008
, “
Investigations on the Effect of Ball Burnishing Parameters on Surface Hardness and Wear Resistance of HSLA Dual-Phase Steels
,”
Mater. Manuf. Processes
,
23
(
3
), pp.
295
302
.
26.
Schwingenschlögl
,
P.
,
2020
, “
Erarbeitung Eines Prozessverständnisses zur Verbesserung der Tribologischen Bedingungen Beim Presshärten
,”
Ph.D. thesis
,
Lehrstuhl für Fertigungstechnologie, Friedrich-Alexander-University Erlangen-Nuremberg
.
27.
Venema
,
J.
,
Hazrati
,
J.
,
Zaman
,
S. B.
,
Matthews
,
D. T. A.
, and
Atzema
,
E.
,
2022
, “
Temperature and Strain Dependent Fracture of AlSi Coating in Hot Stamping of Press Hardening Steel
,”
Int. J. Mater. Form.
,
15
(
3
).
28.
Li
,
A.
,
Ji
,
V.
,
Lebrun
,
J. L.
, and
Ingelbert
,
G.
,
1995
, “
Surface Roughness Effects on Stress Determination by the X-Ray Diffraction Method
,”
Exp. Tech.
,
19
(
2
), pp.
9
11
.
29.
Sunny
,
S.
,
Mathews
,
R.
,
Yu
,
H.
, and
Malik
,
A.
,
2022
, “
Effects of Microstructure and Inherent Stress on Residual Stress Induced During Powder bed Fusion With Roller Burnishing
,”
Int. J. Mech. Sci.
,
219
.
30.
Courbon
,
C.
,
Sova
,
A.
,
Valiorgue
,
F.
,
Pascal
,
H.
,
Sijobert
,
J.
,
Kermouche
,
G.
,
Bertrand
,
P.
, and
Rech
,
J.
,
2019
, “
Near Surface Transformations of Stainless Steel Cold Spray and Laser Cladding Deposits After Turning and Ball-Burnishing
,”
Surf. Coat. Technol.
,
371
, pp.
235
244
.
31.
Lechler
,
J.
,
2009
, “
Beschreibung und Modellierung des Werkstoffverhaltens von Presshärtbaren Bor-Manganstählen
,”
Ph.D. thesis
,
Lehrstuhl für Fertigungstechnologie, Friedrich-Alexander-University Erlangen-Nuremberg, Meisenbach
,
Bamberg
.
32.
Bosetti
,
P.
,
Bruschi
,
S.
,
Stoehr
,
T.
,
Lechler
,
J.
, and
Merklein
,
M.
,
2010
, “
Interlaboratory Comparison for Heat Transfer Coefficient Identification in Hot Stamping of High Strength Steels
,”
In J. Mater. Form.
,
3
(
S1
), pp.
817
820
.
33.
Löbbe
,
C.
,
2019
, “
Temperaturunterstütztes Biegen und Wärmebehandeln in Mehrstufigen Werkzeugen
,”
Ph.D. thesis
,
Institute of Forming Technology and Lightweight Components, TU Dortmund University
,
Shaker, Aachen, Germany
.