Influence of deformation on metal structure in the forward microextrusion process

Comparing microforming to traditional forming process it can be observed, that while going to a microscale, parameters such as grain size or surface structure are not changed [1, 2]. Relationships between the dimensions of the manufactured items and morfometric parameters of their microstructure and surface stereometry, both items and tools are different in macro and microscale. This leads to the size effect formation, which presence does not permit direct application in microforming of metals available technological knowledge relating to conventional forming methods [3, 4]. Microforming is a young technology, but more and more research centers in the world deal with it. Currently works refer to experimental diagnosis of the impact of size effect on metal forming processes and of the products mechanical properties. In parallel, together with the reports of these works results, attempts to take into account and apply the identified phenomena in all sorts of software simulating the metal deformation processes are undertaken eg. [5, 6]. Experimental studies of metal microextrusion have been presented in a thematic literature for two decades [7]. Influence of the containers and dies surfaces smoothness and a material grain size on the mechanical properties of the product have been investigated – size effect. Similarly, as in the case of macroscale, in microscale there are conducted: a forward microextrusion eg. [8, 9], a backward microextrusion eg. [10] and an angled one [11, 12]. These processes are carried out in dry conditions, with lubrication, at ambient and elevated temperatures. These parameters appear to have different, often heightened influence on mechanics of deformation in the microscale. Miniaturization of these processes introduces need to further understand the peculiarities with it emerging. In microscale material can no longer be treated as a homogeneous, because in the area of deformation it can be just a few grains. It should also be taken into account the influence of container surface roughness, which significantly affects on deformation processes at the microscale level.


Introduction
Comparing microforming to traditional forming process it can be observed, that while going to a microscale, parameters such as grain size or surface structure are not changed [1,2].Relationships between the dimensions of the manufactured items and morfometric parameters of their microstructure and surface stereometry, both items and tools are different in macro and microscale.This leads to the size effect formation, which presence does not permit direct application in microforming of metals available technological knowledge relating to conventional forming methods [3,4].
Microforming is a young technology, but more and more research centers in the world deal with it.Currently works refer to experimental diagnosis of the impact of size effect on metal forming processes and of the products mechanical properties.In parallel, together with the reports of these works results, attempts to take into account and apply the identified phenomena in all sorts of software simulating the metal deformation processes are undertaken eg.[5,6].
Experimental studies of metal microextrusion have been presented in a thematic literature for two decades [7].Influence of the containers and dies surfaces smoothness and a material grain size on the mechanical properties of the product have been investigatedsize effect.Similarly, as in the case of macroscale, in microscale there are conducted: a forward microextrusion eg.[8,9], a backward microextrusion eg.[10] and an angled one [11,12].These processes are carried out in dry conditions, with lubrication, at ambient and elevated temperatures.These parameters appear to have different, often heightened influence on mechanics of deformation in the microscale.Miniaturization of these processes introduces need to further understand the peculiarities with it emerging.In microscale material can no longer be treated as a homogeneous, because in the area of deformation it can be just a few grains.It should also be taken into account the influence of container surface roughness, which significantly affects on deformation processes at the microscale level.

Experimental investigations
In order to determine the influence of tool roughness on metal plastic flow in the forward microextrusion process conducted in dry conditions, the toolkit author's copyright project was designed and manufactured [6].There are two containers and two rectangular dies of the same size, differing only in the degree of containers roughness, hollowed in the halves of the bodies (Fig. 1).The diameter of the containers is D p = 1.8 mm, while the dies D m = 0.9 mm.Toolkit includes also the piston that presses on the "floating" over his face punches.Body and the piston are made of X210Cr12 steel, and punches of bearing steel 100Cr6.Samples made of aluminum and copper were extruded (Fig. 2).In order to illustrate the material structures, samples made of annealed electrolytic copper Cu99.99E have been used.Fig. 1 View of the toolkit for forward microextrusion [6] Fig. 2 Extruded samples [6] Containers and dies were drilled with the use of the EDM method.In order to obtain diverse wall roughness of containers, they were drilled with different degrees of energy pulsed electric discharge.In order to determine the degree of containers roughness, the roughness profiles were designated using laser microscope LEXT OLS4000 3D, and average values of parameter Ra were determined for each.As a result of the measurements, following values were obtained: Ra = 1.2 μm and Ra = 3.5 μm.Fragments of received profiles of the tested containers are shown in Figs. 3 and 4. Extruded samples were sunk in epoxy resin and their metallographic sections were prepared.The microscopy images at the longitudinal sections of samples before and after extrusion were obtained.Width measurements of the strips with highly deformed structure caused by containers roughness in the places where these bands are the widest were realized.Measurements were conducted using MultiScan software, that allows morfometric analysis of objects in the scaled photographs.
Fig. 5 shows the material structure in the longitudinal section of the sample before extruding.At the edge (Fig. 5, photo a) and inside (Fig. 5, photo b) of the sample a homogeneous grains distribution appears.

Fig. 5 Initial copper grain structure
At the section of extruded in a container with the roughness Ra = 1.2 μm sample a narrow strip of deformed structure at the outer edges of its upper part is visible (Fig. 6, photo d and e).Core of the sample remained unchanged and grains in this area have an original, uniform layout (Fig. 6, photo a and b).In the areas of the upper outflow and die a fragmented structure appears (Fig. 6, photo c and f), caused by intensification of deformation.Lower part of the sample shows also significant fragmentation of the structure, especially at its outer edge (Fig. 6, photo g).
Sample extruded in the container with roughness Ra = 3.5 μm have a broader range of deformed grains at the outer edge of its upper part (Fig. 7, photo b, c and d).Core of this part, as in the previous case, has an uniform grain structure (Fig. 7, photo e and f).

Microextrusion process modeling conditions
Axially symmetric extrusion process allowed to carry out simulations in a half sections what reduced time of calculation.Simulations have been performed using software DEFORM based on the theory of plasticity and using FEM.Tool elements were treated as a rigid bodies.Simulations have been carried out with keeping constant punch velocity equal v 0 = 0.01 mm/s and workpiece temperature T = 20°C.Workpiece material was considered as a homogeneous, plastic with isotropic hardening, described by the stress-strain curve defined for the experimentally investigated one.Its model was introduced to the software library of curves.
The container was D p = 1 mm in diameter, while the die D m = 0.5 mm (Fig. 10).Initial length of the workpiece model was 2.5 mm.Modeled meshes have been designed with the greatest possible number of elements, what was intended to take into account the variable geometry of the containers surfaces.These meshes had respectively: 7 and 9 thousand of elements.Workpieces were designed in the form of rectangles, connected with the peeks of containers roughness.While extruding the material filled triangles, which was the signal, that the mesh sensitivity is sufficient and takes into account the surface In order to illustrate the influence of container roughness on metal deformation process during microextruding the numerical models of roughness were designed and introduced into simulations (Fig. 11).At the toolworkpiece interface a zero friction factor has been given.The friction was the resistance of the material movement caused by a container roughness wave.Forward microextrusion in container with the roughness wave numerical simulations following effective stress  ̅ distributions revealed (Fig. 13).Effective stress in the material is characterized by specific distribution in the vicinity of the wave. ̅ increases with increasing motive parameters and with the height of the container.

Results of the microextrusion numerical analyses
Fig. 14 shows the deformation of the material using a stratified strain effective images ε̅ .In the vicinity of the wave there is a significant change in the conditions of deformation.With increasing motives depth increases strain value in the workpiece section.
Particles velocities distributions (Fig. 15) reveal the slow of movement of the layers in the container roughness vicinity.The larger motives dimensions the bigger volume of the material with reduced speeddead zone.This dependency compared to images of the strain fields suggests, that in the initial phase of material flow a signifi-cant deformation of the material and the flow suppression is followed.As a result, a conical surface area rises, in respect of which an intensive shearing is followed by.

Conclusions
Analyzed microextrusion processes were characterized by a large ratio of the container roughness value Ra to its diameter D p .In experimental part of investigations, microextrusion cases with the above relation Ra / D p equa-ling 0.00067 and 0.00194 have been examined.This gives respectively 11 and 39% of participation of the fragmented by the containers roughness grain structure layers in the inputs cross sections.Volume of the fine grained structure depending on the roughness of a container is the source of size effect in this case.This phenomenon has been confirmed experimentally and numerically.Reduction of the degree of roughness or increasing in the dimensions of the input will result in the disappearance of this occurrence.
Subsequent studies should be focused on identifying and recognizing the value of the parameter, which allows to specify the appearance of the size effect resulting from the impact of containers roughness.This parameter could be the ratio of a surface roughness Ra to a container diameter D p : Ra / Dp.Line forecasting of this ratio value, based on the obtained data suggests, that when it is equal to 0.0002, the participation of fine grained structure at the input cross section will be negligibly small.It will be, therefore, the size effect emergence value during forward extruding of the tested metal.This estimate should be verified experimentally.Precise determination of this start parameter value would allow to determine the container diameter and its degree of roughness, to which modeling of container surface roughness wave in forward microextrusion numerical simulations would be appropriate.

Fig. 10 Fig. 11
Fig. 10 Adopted for calculations model for forward micro-extrusion roughness of the containers.Motives method according to the PN-EN ISO 12085:1999/AC:2009 divides a roughness profile into socalled motives.A motive is a part of the original profile between the highest points of two local, not necessarily neighboring, peaks of the profile.Using the motives method, the average length AR and average depth R of the motives for both roughness profiles have been determined, with a given 95% confidence interval using the normal distribution: 1. R = 4.89 ± 1.09 μm AR = 16.66 ± 2.05 μm, 2. R = 13.87 ± 2.30 μm AR = 26.27± 3.41 μm.

Fig. 12 Fig. 12
Fig.12shows half sections of the forwardly extruded workpieces in containers with top layers characterized by triangular profiles.Finite element meshes are dense enough to take into account variability of the material geometry in the vicinity of the containers roughness.