Bulk Metal Forming

Abstract

A common definition of incremental forming is given in which an incremental forming process is characterized by regions of the workpiece experiencing more than one loading and unloading cycle due to the action of one set of tools within one production stage.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

4.1 Incremental Forming

• Bernd Kuhfuss &
• Eric Moumi 

4.1.1 Definition and Technologies

A common definition of incremental forming is given by Groche et al. [10, 11] in which an incremental forming process is characterized by regions of the workpiece experiencing more than one loading and unloading cycle due to the action of one set of tools within one production stage.

Incremental processes are divided into sheet metal and bulk metal forming. Well-established incremental bulk forming processes are open die forging, rotary swaging, flow forming and rolling processes (Fig. 4.1). In the following, only bulk metal forming with a focus on cold forming—namely rotary swaging—will be addressed.

Fig. 4.1

Incremental forming processes according to [11]. a Open die forging, b rotary swaging, c flow forming, d orbital rolling, e ring rolling, f cross rolling, g skewed rolling, h profile rolling, i planetary rolling

Parts that are produced by rotary swaging are used as macro scale products in different applications. An important area of application are components for the automotive industry, like axes, steering spindles and gear shafts, in particular those made from tubular blanks with the purpose of weight saving.

Basic research on rotary swaging started in the 1960s, and has continued since then on macro scale components. In the last few years the focus of scientific interest was extended to micro components [15, 16]. Fields of application for swaged micro parts are seen in e.g. medical mechatronic systems (minimal invasive diagnostics and surgery) and the automotive industry (miniature pumps and valves).

In order to get a better understanding of the relevant effects during micro forming, the process is in general considered as a system consisting of the components process, tools, material and machine respectively equipment [6]. In particular, the process characteristics and material conditions are referred to in the following sections. Special features of micro rotary swaging machines and tool design are addressed in 8.1.

4.1.2 Process and Tooling

According to DIN 8583, rotary swaging belongs to the processes of open die forging. The forming of the workpiece takes place in the swaging head in small steps by the radial oscillating movement of the tools (Fig. 4.2).

Fig. 4.2

Swaging head for the infeed process

The radial movement of the forming tools (die segments) (1) is generated by the rotation of the driven shaft (2) and the sine-shaped cam on top of the base jaws (3). With every pass of the cams along one of the cylinder rollers (4) the base jaws, and thereby the forming dies, are forced radially inwards producing an impact on the work (5). The rotation of the main shaft in relation to the workpiece leads to uniform forming over the circumference.

Rotary swaging can be sub-divided into two process variants with respect to the direction of feed. Figure 4.3 shows schematically the infeed swaging (left) and the plunge swaging (right) processes.

Fig. 4.3

Process variations of rotary swaging

In infeed swaging the workpiece (1) is axially fed into the swaging head and reduction takes place over the whole feed length in the reducing zone of the dies (2). In plunge swaging the oscillating movement of the dies is superimposed by a radial feed rate due to the axial movement of additional wedges (3), so that a local reduction in the workpiece occurs according to the die geometry. The components undergo a 2-dimensional pressure stress state in the reducing zone. This acts favorably on the formability, especially in micro forming.

Due to the geometrical scaling effect, i.e. ratio surface/volume increasing with smaller components, the vulnerability to tensile load also increases, see also 4.3.3. An example is given in Fig. 4.4 (top) showing a swaged sample of spray-formed AlSi10Zn13Cu4 without any detectable defects. The material cannot be formed by drawing due to cracking. Figure 4.4 (bottom) demonstrates the forming capabilities of rotary swaging. The component is a prototype for a micro valve housing formed over a mandrel. With this a calibrated inner diameter can be achieved, which would require an additional machining operation for other forming technologies like cold forging (extrusion).

Fig. 4.4

Swaged micro components

Important parameters that influence rotary swaging are the total and incremental deformation degrees, the forces acting on the system and the kinematics of the process that indicate the productivity. Figure 4.5 shows the geometrical relationships.

Fig. 4.5

Geometrical conditions on the workpiece during infeed rotary swaging

The deformation degree in the radial direction is described as follows (4.1):

$$ \varphi \; = \,\ln \frac{{d_{1}^{2} }}{{d_{0}^{2} }} $$

(4.1)

The incremental deformation degree due to each stroke can be defined as:

$$ \varphi_{st} = \;\ln \left( {\frac{{d_{0} - 2h_{st} }}{{d_{0} }}} \right)^{2} $$

(4.2)

h_st is the effective stroke:

$$ h_{st} \; = \;\frac{{v_{f} }}{{f_{st} }}\,\tan \alpha_{T} $$

(4.3)

with $v_{f}$ being the axial feed rate, $\alpha_{\text{T}}$ the die angle, and $f_{st}$ the stroke frequency.

The maximum axial feed rate is obtained from the kinematics as:

$$ V_{f,\hbox{max} } = \frac{{h_{T} }}{{\left( {t_{cy} - t_{0} } \right)\tan \alpha_{T} }} $$

(4.4)

with t_cy being the period (=1/f_st) and t₀ the time segments with dies closed, and h_T the stroke of the base jaws.

Equation (4.4) gives an estimate for the processing time t_pr and thus the productivity. Calculating t_pr for the micro valve part of Fig. 4.4 would give 50 ms, the achievable production rate in practice is lowered due to handling times and the reduced feed rates because of process limits.

Figure 4.6 illustrates the forces during infeed swaging. Additionally 3 different sections are marked within the dies: reducing zone (I), calibration zone (II) and die outlet (III). The total radial forming force F_F is divided into components in zones I and II, and the related friction components are given as F_RI and F_RII. The axial forces are characterized by F_f and F_rej (rejection force). The neutral surface NS separates the material flow in and against the feed direction.

Fig. 4.6

Process forces in infeed swaging

The rejection force widely influences the process window for rotary swaging. In macro swaging technologies the reducing zone I is roughened by hardfacing. In micro rotary swaging rejection limits the achievable axial feed rate, see Fig. 4.7.

Fig. 4.7

Maximum feed rate in infeed swaging

The maximum applicable feed rates that were experimentally found for wire material are about 10 % of the theoretical value given by Eq. (4.4). The findings indicate a strong dependency on the Young’s modulus E, which is true for all degrees of deformation. Considering the ratio of part diameter to part length, which is less compared to the macro range, the limiting factor in infeed swaging is Euler buckling.

The buckling force F_K with the axial forces of Fig. 4.5 during the closing time of the dies is (Fig. 4.8):

$$ \begin{array}{*{20}l} {F_{K} = F_{f} + F_{rej} = 2\pi^{2} \frac{EI}{{l^{2} }}} & {\left( {\text{a}} \right)} \\ {F_{K} = F_{f} + F_{rej} = 4\pi^{2} \frac{EI}{{l^{2} }}} & {\left( {\text{b}} \right)} \\ \end{array} $$

(4.5)

Fig. 4.8

Euler buckling of micro parts

The smallest producible part diameter is limited by the osculation ratio of the tools and workpiece, the opening stroke of the tools during forming and the yield point of the material. The osculation ratio directly affects the achievable final geometry (diameter and roundness) and is difficult to predict analytically. The relation of working stroke to the final diameter is limited. Beyond this limit the workpiece can enter the gap between the tools (Fig. 4.9).

Fig. 4.9

Contact relationship between tool and workpiece

In a 4-segmented swaging head the gap g between the dies in the completely open status is 1.4 times the radial stroke. Design and process-related measures can be taken for the enlargement of the stroke with respect to the work diameter, see also Sect. 8.1.

The smallest stroke under which plastic deformation will occur is determined by the yield point. The remaining elastic deformation (spring back) is considered by a smaller contour in the dies than the final work diameter desired.

Besides buckling due to the minor relative axial stiffness of micro parts, other process failures like surface defects, torsion and so-called wing forming (material flow into the gaps of the dies when opened) limit the applicable process windows [15].

4.1.3 Material Effects

The material influence on the result is both similar and different to some aspects in comparison with the macro range.

Figure 4.10 illustrates the part diameter over part length when varying the feed rate. Measurements begin at the initially formed part end.

Fig. 4.10

Final diameters over part length for 1.4301

At the swaged end, a cone-like shape is formed backward to the forming direction, it seems like material flows over the core or the core is less deformed. The smallest diameter is found in this area. As can be seen, the final diameter steadily increases in the first 25 mm of swaged parts from 1.4301. This indicates that near this edge the missing supporting effect of adjacent material reduces the forming resistance. Another effect is obvious, i.e. the final diameter increases with feed rate. This can be explained by the declining number of impacts per volume element and an increase in spring back.

The Martens hardness distribution on steel 1.4301 and aluminum Al 99.5 is shown in Figs. 4.11 and 4.12 respectively.

Fig. 4.11

Hardness distributions of 1.4301 steel

Fig. 4.12

Hardness distributions of Al 99.5

As expected, hardness after cold forming was higher than in the initial materials, where the hardness is relatively homogenous, with values between 2,400 and 2,750 N/mm². Feed rates of wire have no significant influence on the absolute value of hardness, e.g. hardness values are similar at feed rates of 1.34 and 10.75 mm/s. The only difference is that the distribution is more homogenous at higher feed rates. An increase in hardness with deformation degree is obvious (Fig. 4.11). For deformation degrees of 1.8 and 2.4 the hardness is lower near the surface and becomes higher beneath it. An explanation for this characteristic is the Bauschinger effect, namely the consideration of friction acting on the surface.

Material Al 99.5 shows a different behavior. At the studied deformation degree there is also a noticeable increase in hardness from the blank status (420 N/mm²). The increase reaches it maximum values at the surface or very close to it.

4.2 Upsetting

• Heiko Brüning 

4.2.1 Basics in Upsetting

Upsetting is one of the most important forming techniques belonging to the group of bulk metal forming processes [3]. As stated in DIN 8583, upsetting is a subcategory of free forming and is further divided into flat coining and partial upsetting. This paragraph is addressed to (partial) upsetting as this is the process which is more often applied and has greater practical relevance.

The workpiece is placed between two coplanar surfaces, called tools, which move toward each other reducing their relative distance, thus forming the workpiece and decrementing its height. Due to constant density of the workpiece material, the width of the workpiece increases so that the total volume remains constant. Figure 4.13 schematically shows an upsetting process with friction coefficient $\mu \, > \,0$ between the tools and workpiece.

Fig. 4.13

Upsetting of workpiece with $\mu \, \ne \,0$

Depending on the value of friction coefficient $\mu$ different behavior of formability is observed. If upsetting is carried out with no or negligible friction between the workpiece and tools, the shape of the workpiece remains as is, that means that the lateral surface stays flat. This can be achieved by either using surfaces well-lubricated with paraffin [18] or by using specially-formed specimens with concave surfaces, as proposed by [26]. In this case, a homogeneous strain of the workpiece and its grain structure is assured thus leading to a uniform allocation of hardness. For friction coefficients $\mu \, > \,0$ between the workpiece and tools, the lateral surface adopts a convex shape so that the strain is inhomogeneous. In this case the workpiece does not have the same overall hardness, if the hardness is dependent on the absolute value of strain such as for work hardening materials. Figure 4.14 shows a schematical cross-sectional view of a cylindrical specimen which has been processed by upsetting.

Fig. 4.14

Effect of upsetting of work hardening material on hardness. Area I largest strain, highest hardness. Area II medium strain, medium hardness. Area III lowest strain, lowest hardness

The upsetting process is relevant for both theoretical investigations and practical applications. On the one hand, upsetting is a model process to carry out compression tests (see 3.3.1.) in order to determine flow curves, and on the other hand upsetting is used daily a multitude of times in many process chains to produce screws, nuts, bolts, rivets and nails. In contrast with machining operations, upsetting benefits from the economic utilization of materials, low cycle times and, one factor which is very important, a homogeneous grain structure making goods mechanically more resistant [3].

There are two main parameters to characterize an upsetting process and its limits:

• 1. The maximum natural strain $\varphi .$ This value reaches its maximum, called major strain, parallel to the tool movement direction:

$$\varphi _{{max}} = \varphi _{h} = ln\frac{{h_{1} }}{{h_{0} }}. $$

(4.6)

As soon as a certain value of $\varphi$ is exceeded, the limit of formability of the workpiece is reached and cracks start to develop.

• 2. The upset ratio u as a limit for buckling. Buckling will occur when upset ratios greater than $u_{{max}}$ are processed

$$ u = \frac{{h_{0} }}{{d_{0} }}. $$

(4.7)

Exact values determined by experiments for both u and $\varphi$ can be found in paragraphs 4.2.2 and 4.2.3.

Fig. 4.15

Models of stress curve during upsetting

Cold forming of metallic materials generally does not lead to a significant change in density. In conjunction with the fact that material cohesion is not reduced by forming operations, the volume of the workpiece stays constant. This relation can be stated as follows:

$$ \varphi_{h} + \varphi_{r} + \varphi_{t} = 0 $$

(4.8)

which describes that the sum of the strains in each main direction necessarily equals zero. Among others, such as for compression tests, axially symmetrical cylindrical specimens are used with the consequence that, with equal boundary conditions applied, radial strain equals tangential strain so that:

$$ \varphi_{r} = \varphi_{h} $$

and

$$ \varphi_{h} + 2\varphi_{r} = 0. $$

(4.9)

A constant strain is achieved as soon as the stress $\sigma$ in one direction reaches the flow stress $k_{f} .$ The maximum strain is realized parallel to the direction of the tool movement so that the stress state reaches its major value in the z-direction, called $\sigma_{z} ,$ at the contact areas between the tool and workpiece. $\sigma_{z}$ is dependent on flow stress $k_{f} ,$ the radial distance r to the neutral surface, the upset ratio u of the workpiece and the friction coefficient $\mu$ between the workpiece and tools [3]. If a homogeneous grain structure is assumed and the average grain size $L_{K}$ is much smaller than the workpiece diameter, $\sigma_{z}$ can be calculated analytically by an exponential function [27] as follows:

$$ \sigma_{z} = - k_{f} *exp\left[ {\frac{2\mu }{h}\left( {\frac{d}{2} - r} \right)} \right]. $$

(4.10)

For practical applications a simplification of 4.5 is often used. The exponential fraction can be approximated by a series expansion, neglecting terms of higher order as shown in (4.11).

$$ \sigma_{z} = - k_{f} \left[ {1 + \frac{2\mu }{h}\left( {\frac{d}{2} - r} \right)} \right]. $$

(4.11)

Calculating the maximum tension using (4.10) or (4.11), it is obvious that $\sigma_{z} = - k_{f}$ is constant and independent of r if the friction coefficient $\mu = 0.$ Figure 4.15 shows all three stress states: upsetting without friction, simplified model with friction, and analytical model with friction.

The required punch force $F_{z}$ can be calculated by the integration of (4.11) across the surface A of the workpiece. For cylindrical workpieces $F_{z}$ results in:

$$ F_{z} = - A*k_{f} \left[ {1 + \frac{1}{3}\mu \frac{d}{h}} \right]. $$

(4.12)

4.2.2 Buckling in the Upsetting Process

Buckling is one sort of defect that might occur during the upsetting process. As soon as buckling appears the grain structure is defective, which is highly visible in cross-sectional polishes with the result that mechanical properties are shortened. Experiments have been carried out by Messner [19] with specimens in the micro range as well as macro range to determine the characteristics of the buckling effect in workpieces with major dimensions smaller than 1 mm. Specimens of brass (CuZn15) as well as an austenitic chromium stainless steel (X4CrNi18-10) have been used. The grain size $L_{K}$ is adjusted by heat treatment so that its effect on buckling could also be determined. It is to be stated that there is an increasing influence of grain size on the evenness of the lateral surface after the upsetting process for decreasing ratios of sample diameter to grain size ${{d_{0} } \mathord{\left/ {\vphantom {{d_{0} } {L_{K} }}} \right. \kern-0pt} {L_{K} }}.$ Specimens with an initial diameter of $d_{0} = 0. 5\,{\text{mm}}$ and a grain size $L_{K} = 6 5\,\mu {\text{m}}$ showed an inhomogeneous lateral surface after the upsetting process, because some crystal grains extruded the surface. Inhomogeneities are not only limited to grains corresponding to the surface but also within the specimen. There are areas whose deformation is larger than others, even though friction between the workpiece and tools is close to zero. The strength of this effect increases with the decreasing dimensions of the workpiece.

Buckling of the specimen occurs as soon as an asymmetrical movement of the material flow perpendicular to the tool movement direction is observed. The value of buckling ∆x is measured as shown in Fig. 4.16.

Fig. 4.16

Measurement method for buckling ∆x

The size of relative buckling ∆x/d ₀ is investigated by means of the size of the specimen, the upset ratio and the grain size. As shown in Fig. 4.17, a decreasing size of specimen inevitably leads to an increase in relative buckling, for example an upset ratio u = 2.4 for specimens d ₀  = 4.8 mm causes a relative buckling of 0.05 whereas, specimens with d ₀  = 0.5 mm experience relative buckling of 0.25. Figure 4.17 only shows these effects for an absolute value of natural strain $\left| \varphi \right| = 0. 8 ,$ but it is to be noted that similar tendencies are recognized for $\left| \varphi \right| = 0.2$ and $\left| \varphi \right| = 0.4.$ Overall there is no influence of the grain size L _K on relative buckling, not even if the cross-section of the specimen consists of an average of five grains. This effect is valid for both the macro and micro range, and is supported by finite element simulations by Mori et al. [21] for specimens with conventional dimensions. Vollertsen [28] states that the statistical spread increases with decreasing sample size. This can also be seen in Fig. 4.17, as the spread in size of relative buckling is larger for samples with $d_{0} = 0. 5\,{\text{mm}}$ compared to samples with $d_{0} = 4. 8\,{\text{mm}} .$

Fig. 4.17

Influence of upset ratio u on relative buckling during the non-lubricated upsetting of cylindrical specimens CuZn15 according to [19]

As the micro structure, represented by the grain size, has no significant influence on relative buckling, this size effect can only be based on the shape inaccuracy of specimens due to the machining process. This means that buckling only occurs if the cylindrical specimen or the tool has at least minor deviations in shape compared to its ideal geometry, or if the tool movement is not exactly perpendicular to the workpiece surface. In Fig. 4.18 possible causes for buckling while upsetting are shown. Due to the fact that impurities in the ambient atmosphere such as dust belong to the same order of magnitude as the specimens being upset, it is advisable to carry out micro cold forming operations such as upsetting in clean room conditions.

Fig. 4.18

Causes for buckling during the upsetting process

In the macro range a limit of upset ratio $u = 2. 3$ is generally recognized [18]. Due to the fact that the value of relative buckling is a steady function of upset ratio, it is also clear that for ratios $u\, < \, 2. 3$ upsetting operations lead to buckling and thus to deviations in the shape of specimens. The value of this deviation is accepted for industrial applications. If the value of relative buckling in the macro range is also applied in the micro range, the maximum upset ratio u is smaller, because relative buckling increases with decreasing sample diameter for constant upset ratios. Thus in the micro range the tolerable upset ratio u is far below 2 [19].

4.2.3 Formability in the Upsetting Process

Formability is a value given to describe the capability of enduring the true strain of a workpiece. The limit of formability is reached as soon as parts of the workpiece lose their cohesion. An indication of this is the appearance of cracks, which can either occur along the outer surface of the workpiece or inside. Cracks inside the workpiece are easily detectable by analyzing cross-sectional polishes, but also during the upsetting process a discontinuous increase in punch force can be an indication of material failure due to excess formability [12].

In meso and macro scale, upsetting without any externally detectable defects is possible up to absolute values of natural strain of $\left| \varphi \right| = 2.0$ [19]. This value is independent of sample diameter to grain size ratios for at least ${{d_{0} } \mathord{\left/ {\vphantom {{d_{0} } {L_{K} }}} \right. \kern-0pt} {L_{K} }}\, \le \, 70.$ Nevertheless, a correlation between grain size, surface roughness and natural strain has been reported: surface roughness increases with both increasing absolute values of natural strain and grain size. As shown in Fig. 4.14, the shape of the cylindrical specimen is dependent on the friction coefficient $\mu$ between the tool and workpiece surface. As soon as the friction coefficient $\mu \, > \,0$ is obtained, barreling of the specimen takes place. In the macro range, plastic material flow forces grains which used to form the lateral surface of the specimen to move radially and axially, thus forming the top and bottom surface and coming in direct contact with the tool surface.

In the micro range different behavior of specimens with regard to formability is reported. Upsetting is only possible up to $\left| \varphi \right| = 2.0$ if ${{d_{0} } \mathord{\left/ {\vphantom {{d_{0} } {L_{K} }}} \right. \kern-0pt} {L_{K} }}\, \approx \, 30$ or above, which means that a fine microstructure is required. But also in this case, the fillet from the bottom and top surfaces to the lateral surface is very rough, leading to stress peaks. The fact that still no cracks appear indicates that the stress peaks are diminished by locally extended plastic material flow. If the ratio of the sample diameter to grain size ${{d_{0} } \mathord{\left/ {\vphantom {{d_{0} } {L_{K} }}} \right. \kern-0pt} {L_{K} }}$ decreases, the maximum natural strain until no cracks appear also decreases. For ${{d_{0} } \mathord{\left/ {\vphantom {{d_{0} } {L_{K} }}} \right. \kern-0pt} {L_{K} }}\, \approx \, 8$ and a sample diameter $d_{0} = 0. 5\, {\text{mm}},$ at the lateral surface as well as at the bottom and top surface cracks tend to partially develop, starting at $\left| \varphi \right| = 1. 5.$ As soon as $\left| \varphi \right|\, \ge \, 1. 7$ is reached only defective workpieces with cracks mainly located at the top and bottom surfaces can be produced by upsetting. These cracks are oriented tangentially, leading to the conclusion that the maximum formability of the material is exceeded, because neither at the bottom nor at the top surface do radially tensile stresses occur if $\mu \, > \,0.$ A further difference between upsetting in the macro and micro range is the fact that the movement of grains which formed the lateral surface before the upsetting operation only come in contact with the tools for larger values of absolute natural strain compared to the macro range.

4.2.4 Influence of Miniaturization on Friction Coefficient in the Upsetting Process

The upsetting process reduces the height of the workpiece. Due to its constant volume the width increases so that there is relative movement between the tools and workpiece perpendicular to the tool movement direction. The value of relative movement is mainly influenced by the friction coefficient $\mu$ between the tools and workpiece, as friction elevates the resistance against relative movement [29]. Figure 4.19 shows a schematic cross-sectional view of a specimen and the resulting stress state under applied load. The compression stress $\sigma_{n}$ is caused by the punch force. The relative movement between the top surface of the specimen and tool surface is restricted in the radial direction by friction, thus leading to tangential stresses $\tau_{F}$ so that a multiaxial stress state exists. The force available for reducing the height of the specimen is reduced because a fraction is needed to transcend friction. So the greater the friction coefficient $\mu$, the greater are both the barreling effect and punch force.

Fig. 4.19

Schematic cross-sectional view of specimen with applied load and resulting stress state

When a forming operation is carried out with rigid tools, the surface deformation of tools can be assumed to be neglected as the surface of the tools has a greater hardness than the workpiece. Due to the fact that the friction coefficient is a consequence of the coaction of the topography of the interacting surfaces, the friction coefficient changes with the increasing strain of the workpiece and its surface [29]. The surface roughness decreases with increasing absolute values of natural strain, because a flattening effect takes place which minimizes the size of the peaks on the surface.

The amount of friction can be influenced by lubricating the interacting surfaces. This can either be achieved by applying suitable fluent lubricant such as oil or paraffin or by coating the tool surfaces with e.g. PTFE, as proposed by [19]. For non-lubricated surfaces there is no size effect reported for the upsetting of work hardening steel X4-CrNi18-10 concerning the size of the specimen or grain size $L_{K}$, which means that neither the grain size nor the absolute size of the specimen has an influence on the average value of friction coefficient. But, as also stated by [19], the mean variation in friction coefficient increases with decreasing sample size. For other materials these influences are still under investigation, and inconsistent results are reported. For upsetting with lubricated interacting surfaces, the influence of grain size on friction coefficient is also at the most very small. In contrast with dry surfaces, a correlation between sample size and friction coefficient exists: the smaller the specimen, the larger the friction coefficient. This phenomenon is based on the fact that with decreasing sample size the fraction of closed lubricant pockets is lowered (see 2.2.1), thus increasing the total true contact area between the tool and surface of the specimen, so that the contact forces grow.

One possibility to decrease both the friction coefficient $\mu$ and flow stress $k_{f} ,$ is to apply high-frequency vibration to the specimen. The flow stress is lowered by acoustic softening and stress superposition. The advantages are that on the one hand there are lower mechanical forces on the tools and on the other hand there is an improvement in surface quality of the surface of the workpiece into which the vibration is conveyed [31]. In this case the very top of the surface is heated due to friction by the incremental longitudinal movement of the punch, thus enabling a plastic deformation of the roughness peaks so that the surface quality significantly increases.

4.3 Cold Forging

• Volker Piwek 

4.3.1 Methods of Cold Forging

Cold forging is a process which is not assisted by any heat from the outside, and characterized by the fact that the process starts at room temperature (20 °C) [13]. All methods of cold forging are characterized by high productivity and are therefore suitable for bulk production. Among upsetting and tube drawing, a large area of cold forging is covered by the various extrusion processes where the material is formed due to a relative movement between the punch and the extrusion tool (sleeve) [17]. Figure 4.20 shows the classification of the different variants of extrusion methods with respect to the specific material flow.

Fig. 4.20

Classification of extrusion processes

The movement of the punch and the material flow by direct extrusion are generally oriented in the same direction. In contrast, in backward extrusion the material is usually deflected by 180°. With the lateral extrusion process the material flow performs a change of direction of around 90°. The 3 methods described are based on a direct transmission of the punch movement to the face of the workpiece blank. With hydrostatic forward extrusion processes however there is no direct contact between the punch and the workpiece. The workpiece is partially embedded in a fluid which is also used for force transmission. In Fig. 4.21 the characteristic geometries for extruded parts are presented according to the above-described process variants respective of their combinations or different process chains [24].

Fig. 4.21

Examples of producible geometries and process combinations

Basically, all ductile materials are formable by extrusion. The main advantages of this resource-efficient process are:

• optimum use of material with uninterrupted grain flow,

• very high productivity and short cycle times,

• high dimensional and shape accuracy,

• high surface quality,

• process-specific cold work hardening.

The convenient material flow (see also Fig. 4.22) and process-specific cold work hardening offers a sufficient strength and stiffness of the extruded workpiece. A near net-shape geometry is very important because of the complicated handling of micro parts. Thus a finishing operation in following process steps can hardly be realized. Here cold extrusion conforms to the high requirements in view of the shape accuracy and surface quality, and is very cost-effective due to its short cycle times. On the other hand, it has to be considered that distortion can occur due to the influence of the friction conditions and the applied material characteristics (e.g. grain structure). Furthermore, the design of the micro extrusion tools has a significant influence on process stability.

Fig. 4.22

Flow of material (left) and schematic diagram of a direct-forward extrusion process (right)

4.3.1.1 Fundamentals of Extrusion

Basically forging processes can be stationary or transient. Due to the coincidence of flow lines and trajectories, extrusion processes can be determined as stationary (Fig. 4.22). For a stationary process the machine work W _m is the product of the punch force F _p and the punch stroke h _p .

$$ W_{m} = \int\limits_{0}^{{s_{p} }} {F_{p} } ds\, \approx \,F_{p} \cdot h_{p} $$

(4.13)

The medium flow stress k _fm can be obtained by the integration over the deformation degree $\varphi$ in the limits from the start (Index 0) to the end of the process (Index 1).

$$ k_{fm} = \frac{1}{\varphi }\int\limits_{{\varphi_{0} }}^{{\varphi_{1} }} {k_{f} } d\varphi $$

(4.14)

For ideal work W _id, considering only the formed volume V of the material and the variance in the effective strain $\Updelta \varphi_{v} ,$ follows:

$$ W_{id} = k_{fm} \cdot \Updelta \varphi_{v} \cdot V $$

(4.15)

In practice, the provided punch force F _p has to exceed the ideal work W _id due to friction and shift work. These additional losses during the deformation process are detected by the efficiency of deformation $\eta_{F}$. For the required work of the plunger W _pl :

$$ W_{pl} = \frac{1}{{\eta_{F} }} \cdot W_{id} $$

(4.16)

Substituting Eq. 4.15 into Eq. 4.16 and equating this term with Eq. 4.13 results in:

$$ \begin{aligned} F_{p} & = \frac{1}{{\eta_{F} }}k_{fm} \cdot \Updelta \varphi_{v} \cdot V \cdot \frac{1}{{h_{p} }} \\ & = \frac{1}{{\eta_{F} }}k_{fm} \cdot \Updelta \varphi_{v} \cdot A_{1} \\ \end{aligned} $$

(4.17)

In this expression, A ₁ is the cross-sectional area after forming (A ₀ before forming). The forming forces F _p are increasing with a higher deformation degree $\varphi$ = ln (A ₁/A ₀) with respect to the opening angle $2\alpha$ of the extrusion tool and with a higher material strength and hardness. Figure 4.22 shows schematically the steady flow of material and the process parameters for direct-forward extrusion.

4.3.1.2 Technology for Cold Forging of Micro Parts

The previously mentioned basic principles and relationships are generally independent of workpiece size. The extrusion process is applied for the bulk production of high quality products in the domains of automotive and machine parts, especially for connecting elements, and in electrical industries [8]. Due to the increasing miniaturization of microelectronic and mechatronic assemblies and the complicated application of other manufacturing methods, e.g. metal cutting at once the requirement to realize even those micro components by extrusion operations is given. While the masses of workpieces in the macro scale vary from a few grams to several kilograms, micro parts have only up to some 100 mg of material that has to be shaped.

With the manufacture of micro parts by extrusion, the friction effects due to the small workpiece dimensions (surface-volume ratio) and the grain size and its distribution exert a significant influence on the entire forming process [6]. The friction can be up to 20 times higher than in the macro area [4]. On the one hand this influence is reflected in a decreasing efficiency of deformation $\eta_{F}$ and on the other an enlarged deviation of the formed part’s geometry appears [7]. The most important test for detecting friction conditions during extrusion is the Double-Cup-Extrusion-Test (DCE), wherein a cup is formed in the forward direction whist at the same time being generated in the reverse direction (Fig. 4.23).

Fig. 4.23

Schematic diagram of the Double-Cup-Extrusion-Test (left), and various sizes of specimen [6] (right)

In this test, the ratio of upper to lower cup height h_u/h_l is a measure of friction during the forming process. For the ideal situation, without any friction (m = 0), two cups of equal height (h_u = h_l) will result, while at maximum friction (m = 1) only a cup geometry in the reverse direction of the punch feed will be formed. Therefore the analysis of the tribological effects of design processes for extruding metallic micro parts has a significant relevance [9].

Experimental results of the DCE-Test on CuZn15 specimens with diameters from 4 mm down to 0.5 mm show a significant influence of the specimen size on the friction coefficient [5]. Furthermore, there are other important methods to determine the friction conditions within the field of micro forming. The micro-ring-compression-test, for example was used, by [2] to determine the friction on aluminum alloy 6061 specimens in a diameter range between 2 and 3 mm. Diameter dimensions below 2 mm down to less than 1 mm were explored for copper alloy CuZn30 by [14] using the pin-length-test. In addition to the experimental investigations, analytical or numerical methods, e.g. finite element method (FEM), can also be applied to determine the friction coefficient. Here the determination of the forming forces is based on the Upper-Bound-Method and the slice-model. The displayed results of a FE-analysis are based on a coarse grained structured material with an average grain size of 211 μm. The determined friction coefficients by the different methods and size ranges of the specimen are arranged in Fig. 4.24. Generally the tendency can be recognized that friction increases with the smaller dimensions of the formed test piece.

Fig. 4.24

Determined friction coefficients by different methods

On the one hand, the FE approach shows similar results to the Double-Cup-Extrusion-Test in that the length of the formed workpiece section decreases with increasing friction. On the other hand it was found that the friction coefficient again decreases with an increasing workpiece diameter, but distortion effects become significant [20]. These effects can be primarily explained with the grain size effect in the material and not with the friction conditions between the contact surface of the punch and the surface of the specimen. Generally, to reduce friction the use of lubricants is suitable. Furthermore, coatings of the tool surfaces made of silicon-based Diamond-Like Carbon(DLC), chromium nitride (CrN) and titanium nitride (TiN) can be considered.

4.3.1.3 Component Applications in the Micro Range

Figure 4.25 shows examples of extruded metallic micro parts for industrial applications (a, c) and the field of research (b).

Fig. 4.25

Examples of extruded metallic micro parts

Workpieces (a) are produced by direct-forward cup extrusion on an industrial extruding press. Component (b) has geometry similar to the shaft of a micro motor and was manufactured by direct-forward in combination with cup-backward extrusion with a shaft diameter of 0.5 mm and a wall thickness of 50 μm. Copper pin (c) was manufactured by the same process combination with a shaft of 0.5 mm and a nominal inner cup diameter of about 1 mm.

4.3.2 Materials: Process Windows and Limits

Steel as well as nonferrous metals are formable by extrusion. While primarily unalloyed and low alloyed steels are used for macro components, such as screws, piston pins and piston rods [23], aluminum and copper are in particular applied in the production of micro components in the field of electrical engineering. The most important aluminum material for extrusion forming operations is pure aluminum Al 99.5 and for higher strength requirements self-hardening (thermosetting) alloys, e.g. AlMg5, AlMgSi and AlCuMg. Furthermore, titanium-based light metals alloys, especially titanium-aluminum, are suitable for the extrusion process. Pure copper is mainly used in the electrical industry, such as E-Cu, Cu and Se-Sf-Cu, in the production of contact pins for example. In the group of copper alloys, brass materials ($\alpha - {\text{brass}}$) with a tin content of 28 % up to a maximum of 37 % (CuZn28 … CuZn37) are particularly suitable. Also bronze materials, especially tin- and silicon-bronzes, e.g. SnBz1 and SnBz2Mn, are used.

For forming operations in the macro range, guidelines for accessible material-related deformation degrees are available. Due to the significant increase in the friction conditions in the micro range, the relevance of these values is limited. From the macro range it is known that the highest deformation degrees can be reached for direct-forward extrusion of pure aluminum with values up to 4. This is also applicable with some restrictions to hollow-forward extrusion. The deformation degrees for copper and bronze materials range from 1.2 to 1.6. However, with heat treatable or stainless steels, in general only deformation degrees of less than 1 can be realized.

Especially for micro forming, the achievable shape accuracy is influenced by the grain size and its distribution (anisotropy) in the blank that is to be extruded, and the friction conditions between the contact surfaces of the workpiece and tools [4]. Extended investigations have shown that the occurring distortion of extruded micro parts is caused by irrational local deformation, spring back effects and interfacial friction stress [2]. These features are size-dependent so that reduced shape accuracy is expected with a decreasing size of specimen. Research on extruded brass pins indicated an influence of the grain size on the distortion achieved. Thus, a material with coarse grains shows a significant straightness deviation, while a fine-grained structure is less affected [14].

4.3.3 Extrusion Tools for Micro Forming

Figure 4.26 illustrates the basic design of a tool for a direct-forward extrusion process with special features for micro forming. The tool consists of an upper, vertically movable section, where the punch is supported on the upper base plate and can be adjusted in the vertical direction by a set screw.

Fig. 4.26

Tool for a direct-forward extrusion process

The geometry of the workpiece is produced by the extrusion tool that consists of an inner sleeve, which is strengthened by ring reinforcement. In the vertical direction, the tool is supported by the lower base plate, in which the ejector is also linearly mounted. This element works against the punch direction and enables the detachment of the workpiece after the forming operation. In extrusion tools for micro forming the matrices have correspondingly small holes in a comparatively large sleeve, which is generally not limited in its external dimensions. The punches, or mandrels, are usually very slim structures, where buckling resistance is one of the fundamental design criteria. Due to this requirement in the micro range, the demand for wear resistance for the stationary as well as the relocatable tool components in compression molding increases significantly. The general requirements on extrusion tools are e.g.

• conical design of the interfaces between several parts

• axis symmetrical design of the parts to balance out non-coaxialities.

For design in the micro range there are specified guidelines, such as

• a minimum number of tool elements to minimize the tolerances of the entire tool assembly

• floating respectively self-adjusting positioning of punch and pin in the horizontal direction

• adjustable punching tool respectively reinforcement.

The axially symmetrical and self-adjusting designs of the extrusion sleeve and a coaxial alignment to the punch are both important to fulfill the requirements of close fittings, especially for micro extrusion processes. The conical design of the interfaces eliminates the clearance that is needed for mounting. Furthermore, there are approaches to optimize the accuracy and the ejection process of the workpiece, which is complicated by the narrow clearances in the micro range, by using a piezo-driven prestress and expandable punching tool [25].

In addition to the basic requirements of high shape accuracy and durability of the tools, in the micro range there are approaches for flexible tool systems to cover an enlarged spectrum of workpiece geometries [30]. The goal of such systems is to provide tools with modular-like structures for the various methods of extrusion or combination thereof [22]. There are essentially three types of tool component: basic-, process- and workpiece-specific components. To reduce the time for changing of the tool (machine setup time), quick-change tooling systems can be used, such as a clip device for the attachment of the punch. Furthermore, the opportunity is given to separately renew or exchange specific areas with increased tool wear due to friction.

Another important feature of the tools is the material used. A suitable material for an extrusion tool has to fulfill on the one hand the requirements of its own manufacturing process as well as for its subsequent application in forming operations. Figure 4.27 shows the main requirements of such materials divided in two groups with focus on the tool manufacturing and the forming process [13].

Fig. 4.27

General demands on the materials for extrusion tools

Generally hardened and tempered tool steels and carbide metals are used, which provide hardness from 50 up to 67 HRC and a bending and burst strength from 2.400 to 2.800 MPa and 3.000 to 4.100 MPa respectively. Furthermore, powder metallurgical steel is used to achieve a high material homogeneity and a fine carbide grain structure (grain size from 2 to 6 μm), while in conventional tool steel the carbide size is between 30 and 50 μm [1]. In particular, a high bending strength is necessary for punching of micro extrusion tools to achieve a sufficient buckling resistance, due to the small dimensions and the thickness ratio required.