ROLL-BONDING METHOD

Field of the Invention

This invention relates to forming ultrafine grained materials. Specifically, the invention relates to a method of forming ultrafine grained metals by a method of roll- bonding .

Background of the Invention

Ultrafine grained (UFG) materials are defined as polycrystals having very small grains with an average grain size in the range of 100 nm to lμm. They are also called submicrostructured or nano-structured materials.

Compared with conventional materials, UFG materials show some excellent properties without the need to change the chemical composition. These properties include ultrahigh strength, low temperature and/or high strain rate superplasticity, enhanced fatigue behavior and superior corrosion. For these reasons, UFG materials have attracted significant interest in recent years.

Severe plastic deformation (SPD) is an extremely effective process for producing bulk UFG materials. Several SPD processing techniques have been developed to obtain UFG structures in both bulk and sheet materials, such as equal-channel angular pressing (ECAP) , high pressure torsion (HPT) , multi-axial compression/forging (MAC/F) and accumulative roll bonding (ARB) .

The ECAP, HPT, MAC/F processes, however, have two main drawbacks. First, forming machines with large load capacities and expensive dies are necessary. Second, productivity is relatively low and the amount of material produced is very limited. These processes are thought to

be inappropriate for practical application of the most commonly used product - sheet. Although some improvements have been proposed, at present the above drawbacks have not been overcome .

Accumulative roll bonding (ARB) imposes severe plastic strain to metallic materials by a rolling process. ARB is one of the most promising methods for industrial-scale production of UFG sheet materials. A schematic illustration of the ARB process is shown in Fig. 1. Two layers of the original sheet are stacked to form a thick specimen. In order to obtain one-body solid material, the rolling employed in the ARB is not only a deformation process but also a bonding process (roll bonding) .

To achieve good bonding, the sheet surfaces are subjected to surface treatments, such as degreasing and wire- brushing, before stacking. The two layers of material are joined together by rolling. The length of rolled material is then sectioned into two halves. The sectioned sheets are again surface-treated, stacked and roll -bonded. Recently, much attention has been focused on the development of ultrafine grained materials in the form of sheet via the ARB process because of its potential cost effectiveness in terms of commercial production, compared to other SPD processes.

The ARB process has been used to obtain ultrafine grained structure in various materials, including aluminium alloy, copper alloy and steels.

An important factor that affects strength of an ARB product is interfacial bond strength. This refers to interfaσial shear bonding strength between two layers of the ARB processed sheet. Surface treatment prior to roll bonding has a significant effect on the bond strength. Current treatments mainly comprise degreasing and wire-

brushing for removing oxide surface layers that inhibit interfacial bonding. However, degreasing and wire-brushing do not enable sound bonding at relative low processing temperatures around 0.35T _m and below, where T _m is the melting temperature of the material from which the sheets are formed.

Summary of the Invention

It is an object of the present invention to provide an improved method of forming ultrafine grained materials by accumulative roll bonding.

The invention provides a method of forming a multi-layer bonded product, the method comprising the steps of:

(a) providing two bodies of a material, each body having a surface for bonding to a surface of the other body;

(b) forming a distribution of fine particles on one or both surfaces, the particles being embeddable in the surfaces by forcing the surfaces together;

(c) arranging the bodies with the two surfaces facing each other so the particles are located between the two surfaces; and

(d) causing a thickness reduction to the two arranged bodies for bonding the surfaces to each other such that the particles are embedded in one or both surfaces, thereby bonding the bodies together.

The term "fine particles" will be understood to mean particles having a major dimension in the range of lnm to 50nm.

Roll bonding two sheets with particles between the surfaces causes the particles to become embedded in one or both surfaces of the sheets. Experimental work carried out by the applicant demonstrates that products roll-bonded in accordance with the invention have higher bond strength than products bonded without particles. The mechanisms by which bond strength increases are thought to be (a) dislocation pinning and (b) improved bonding due to disruption of an oxide surface layer caused by embedding of particles in the surface of the sheets.

Higher bond strength provides ARB products formed by particle-embedding with improved mechanical properties without the cost associated with other superplastic deformation techniques.

Preferably, the distribution of particles on one or both surfaces of the two bodies is uniform.

Preferably, the particles have a hardness that is greater than the hardness of the material from which the two sheets are formed.

Preferably, the particles have a major dimension that is less than 1 micron.

Preferably, the particles have a major dimension that is in the range of 5 to 500nm. More preferably, the particles have a major dimension in the range of 5 to lOOnm and even more preferably, the particles have a major dimension in the range of 5 to 50nm.

Preferably, the particles are SiO ₂ , Al ₂ O ₃ , SiC, ZrO ₂ , TiO ₂ Ni, PeO, Fe ₂ O ₃ , Pe ₃ O ₄ , or SnO ₂ .

Preferably, the particles are applied to the surfaces in a concentration in the range of 0.1 to 3 mg/itim ² , and more preferably in the range of 0.4 to 1.8 mg/min ² .

Preferably, step (d) involves passing the two sheets through a pinch between rollers. Preferably, the thickness reduction is over 50%. More preferably, the thickness reduction is in the range of 50-65%.

Preferably, step (d) is carried out with the two bodies having a temperature in the range of room temperature to 0.45T _m .

Preferably, the method includes a surface pre-treatment step for removing oxides formed on the surfaces.

Preferably, the surface pre-treatment step involves wire- brushing or grinding the surfaces to remove oxides formed on the surfaces .

Preferably, the surface pre-treatment step involves a step of degreasing the surfaces prior to step (b) .

Preferably, the two bodies provided in step (a) comprise bodies previously formed in accordance with steps (a) to (d) .

Preferably, the surface pre-treatment step involves a step of pre-heating the bodies prior to step (d) . Preferably, the pre-heating step involves heating the arranged bodies at a temperature in the range of 0.25T _m to 0.4T _m for 1 to 10 minutes. Preferably, the arranged bodies are heated at 0.3T _m to 0.35T _m for up to 7.5 minutes.

Preferably, the bodies are subjected to step (d) before substantial re-oxidisation of the surfaces after the pre- treating step.

Preferably, the bodies are subjected to step (d) within 60 seconds of the pre-heating step ceasing.

Preferably, the material is aluminium alloy, copper alloy, titanium alloy or steel.

Preferably, the material comprises aluminium alloy and the particles comprise aluminium oxide.

Preferably, the aluminium oxide particles have a size in the range of 50nm to lpm.

Preferably, the material comprises steel and the particles are Ni-σontaining particles.

Preferably, the steel is IF (interstice free) steel or ultra low carbon steel.

Preferably, the Ni-containing particles have a size in the range of 50nm to lμm.

Preferably, the material comprises titanium allow and the particles comprise silicon carbide.

Preferably, the surfaces of steel bodies are prepared by grinding. Preferably, the grinding is across the width of the body.

The invention also provides a roll bonded product formed in accordance with the method outlined above.

Brief Description of the Drawings

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings, in which:

Figure 1 is a schematic representation of an accumulated roll bonding process.

Figure 2 is a graph of a true stress-strain curve for samples prepared in accordance with the invention and, for comparison, a sample comprising the same material without ARB in accordance with the invention.

Figure 3 is a graph of stress and strain as a function of layers in ARB samples prepare in accordance with the invention.

Figure 4 is a graph of average rolling pressure versus average true strain.

Figure 5 is a photograph of two samples (one comprising original material and the other comprising an ARB product with 4 layers) subjected to tensile testing.

Figure 6 is an optical micrograph of a cross -section of a sample prepared in accordance with the invention and having 32 layers.

Figure 7 is a graph of bonding strength and ratio of bonding strength to yield strength as a function of the number of layers in samples prepared in accordance with the invention.

Figure 8 is a graph of ratio of bonding strength to yield strength as a function of the number of layers in 5050 Al samples prepared in accordance with the invention.

Figure 9 is a graph of ratio of ultimate tensile strength as a function of the number of layers in 5050 Al samples prepared in accordance with the invention.

Detailed Description of Embodiments

An experimental procedure carried out in accordance with one embodiment of the invention is described below in reference for forming accumulative roll-bonded sheets of aluminium alloy 6060. It will be appreciated however that the invention is applicable to forming accumulative roll- bonded products of other materials, including copper, titanium, magnesium and steel .

• Rolling Mill

A Hille 100 two-high experimental rolling mill was used to conduct experiments. The mill was driven by an infinitely variable speed motor of 56kW. The maximum rolling load, torque and speed were 150OkN, 13kNm and 70rpm respectively. The rolls were designed with 250mm diameter and 254mm barrel length, made of Bδhler W302 tool steel. The roll surface was treated by the nitride hardening and its hardness is 65 to 70HRc within a depth of 0.2-0.3mm. The roll surfaces have been ground and the roughness is Ra=O .30μm along the circumferential direction and Ra=O.35μm along the axial direction. The roll gap could be set by a mechanical screwdown system and two hydraulic capsules.

• Instrumentation

Two strain gauge load cells were installed between top roll chocks and screws on operating side and drive side to measure the rolling force. Two strain gauge torque transducers were placed in the drive spindles to measure the rolling torques of both top and bottom roll. The signals of the rolling force and torque were recorded by a digital data acquisition system including a computer, an amplifier and a data acquisition broad. A position

transducer, LVDT, was mounted on the roll chocks to monitor the roll gap.

• Workpieces

Rolled samples were prepared from aluminium alloy 6060. The samples were cut parallel to the original rolling direction, having the dimension of 1.66mm x 100mm x 500mm (thickness x width x length) . The initial surface roughness of the samples was Ra= 0.5δμm along both the rolling direction and the transverse direction. All the samples were annealed at 400 ⁰ C for 4 hours. Aluminium alloy 5050 samples were also tested. The dimension of AA5050 samples was 1.2mm x 100mm x 250mm.

• Procedure

Prior to each rolling pass, the roll was cleaned with acetone and the roll gap and speed were set to the required setting. Two pieces of the original samples were degreased by acetone and wire-brushed. The surface roughness after wire-brushing was Ra= 2.80μm along the rolling direction and Ra= 2.66μm along the transverse direction. The particles were uniformly spread between the two pieces. The samples were then stacked and heated in a furnace. The layered sample was then rolled under dry condition. The rolled samples were cut into two halves and the edges were trimmed to avoid propagation of edge cracks. Above procedure proceeded for several passes.

Two groups of experiments have been conducted. In the first group, the sheet material is AA6060 and the nana- particle is SiO2 with the size of 10nm-30nm. The nominal thickness reduction is 50%. The heating temperature and heating time are 290 ⁰ C and 3min. The details for each pass are shown in Table 1.

Table 1 Process parameters for the AA6060 ARB experiments

In the second group of experiment, as -received AA5050 samples were ARB-processed to compare the effects of particle conditions. Summary of the particle conditions are shown in Table 2.

Up to four rolling cycles were carried out in order to achieve high strength material and strong bonds. Nominal thickness reduction for each cycle was 50%. The preheating temperatures were about 300 ⁰ C and the heating time was in the range of 5min to 7.5min. Some samples were wire brushed and others were left with the surface oxide buildup to investigate the importance of the nano-particles being in direct contact with a bare metal surface. Light (0.5 mg/mm ² ) and heavy (1.5 mg/mm ² ) amounts of the nano particles (20-40 micron SiC and 14nm SiO ₂ ) were distributed on the bonding surfaces to determine their influence on bond strength. However, it will be appreciated that particle distributions in the range of 0.1 to 3 mg/mm ² may be used and more specifically particle distributions in the range of 0.4 to 1.8 mg/mm ² may be used.

Table 2

Particle conditions for the ARB experiments of 5050

Aluminium

• Testing

Tensile samples were machined from the ARB processed samples with the tensile axis parallel to the rolling direction. Tensile tests were carried out on a servo- controlled Instron tensile tester at a speed of lmm/s to failure. The true strain-true stress curves were then calculated. Tests were carried out in accordance with the methods disclosed in Krallics and Lenard (J. Mater. Proc. Tech., 152 (2004) 154) .

Two narrow slots were milled to the middle plane along the thickness direction of the samples with the width of 2mm. Tensile tests were then performed on an Instron tensile tester. The bond strength was then determined.

The tensile properties of the ARB deformed AA6060 samples are shown in Fig.2. It is clear that the strength increases and the elongation (ductility) decreases with increasing numbers layers in the samples (or rolling passes) due to work-hardening. The annealed original sample exhibits the good ductility and low strength. The strength rapidly increases once the samples are processed by ARB. The sample with 16 layers has the maximum the strength. As the ARB rolling pass further increases to five passes (32 layers) the strength is compressed. The

ARB process dramatically reduces the ductility as commonly seen in severe plastic deformation processes.

The ultimate tensile strength, yield strength and elongation at peak are collected from Fig.2 and plotted against the layers in the samples in Fig.3. The ultimate tensile strength is the maximum strength on the true strain-true stress curve. The yield strength is the maximum strength of the linear part of the true strain- true stress curve. The elongation at peak, which is the strain corresponding to the ultimate tensile strength, is adopted to describe the ductility property.

It is found in samples produced in accordance with the invention that the ultimate tensile strength and yield strength have the same trend. Both of them increase with increasing number of layers in the samples up to 16 layers and then decreases. The ultimate tensile strength and yield strength of the original sample are 108MPa and 39MPa respectively. They are significantly enhanced to 176MPa and 114MPa respectively after one ARB pass. The maximum values of the ultimate tensile strength and yield strength (258MPa and 178MPa) appear in the 16-layered sample. The ultimate tensile strength of the 16-layered sample is 2.4 times that of the original sample.

The strengths for the 32 -layered AA6060 sample are smaller than those of the 16-layered sample. This indicates that the strength decreases with further decreases the grain size when the grain size is reduced to a certain level.

Fig.3 shows that the elongation drops from 23.4% for the original sample to 3.7% when the sample is rolled by one ARB pass. The elongation seems to be independent of the number of the ARB pass. It always remains in a small range between 3.6% and 4% when the layers in the samples are increased from 2 layers to 32 layers.

The average rolling pressures versus the average true strains are shown in Fig.4. The average rolling pressure is calculated by F/ (L*W) , where F is the measured rolling force, L the length of the roll bite and W the width of the samples. The average true strain is determined by (εθ+ εl)/2, where εO is the accumulative true strain before the pass and εl is the accumulative true strain after the pass. The curve of the average rolling pressure against the average true strain has a same trend as the ultimate tensile strength and yield strength shown in Fig.3. The 16-layered sample has the maximum the average rolling pressure. In addition, the average rolling pressure is much larger than the ultimate tensile strength. This demonstrates that the coefficient of friction in the ARB processes of aluminium is quite large. The reason can be attributed to: (a) dry rolling condition and (b) the increased adhesion strength at the roll -sample interface caused by work-hardening.

Fig.5 shows the pictures of the fractured original sample and 4 -layered AA6060 sample after the tensile tests. It can be seen that the fractures of two samples exhibit the different appearances. The original sample has a normal necking fracture, which is perpendicular to the tensile direction. The fracture in the 4 -layered sample has no

obvious necking and locates at an angle of about 70° to the tensile direction. All the ARB processed samples have the similar fractures as the 4 -layered sample. This indicates that a certain rolling texture exists in the ARB processed samples.

For the purpose of comparing the above results, the ARB experiments without the nano-sized particles were also conducted. Only surface treatments of wire brushing and cleaning by acetone, were used. However, the bonding could not be soundly formed, even for the first pass, at the current process conditions without assistance of the nano- size particles. This indicates that the nano- sized SiO ₂ particles have a significant influence on the interfacial bonding.

The cross sections parallel to the rolling direction were observed by an optical microscope. Optical micrograph on the cross section of an 32 -layered AA6060 sample deformed by five ARB passes with assistance of the nano-size particles is displayed in Fig.6. The bonding interface (five pass interface) formed by the latest pass (five pass) is more obvious than other interfaces. This indicates that the latest pass interfacial bonding is weaker than the bonding interfaces formed by previous passes, and the interfacial bonding can be improved via the further deformation.

To quantitatively analyse the effect of the nano-sized SiO ₂ particles on the bonding property, the bond strength at the newly formed interfaces is measured in accordance with a method outlined in Krallis and Lenard.

Bond strength as a function of the number of layers in the AA6060 samples is shown in Fig.7. The bond strength increases as the layers in the samples increase from 2 layers to 8 layers, and then it decrease with the layers.

The ratios of the bond strength to the yield strength are also calculated and shown in Fig.7. All the ratios exceed 0.17. Krallis and Lenard report that the ratio of bond strength to the yield strength is about 0.1 for 4-layered and 8 -layered samples treated by wire-brushing and acetone cleaning only. The corresponding ratios in our samples are 0.29 and 0.25 respectively, which are more than 2.5 times the value determined by Krallis and Lenard.

The bond strength and the ultimate tensile strength for roller bonded AA5050 samples are shown in Pig.8 and Fig.9. The major conclusions are summarized as follow:

(i) There was much success in utilising nano particles to improve bonding in aluminium alloys. The nano-particle bonds were much stronger and durable compared to the non nano-particle bonded sheets. The strongest bond was obtained with those in the brushed AA5050 (sample 8) using the 50nm aluminium oxide at 140.15 MPa in shear strength.

(ii) The tensile strength of all AA5050 sheets increased to an average of 406 MPa in the fourth ARBed sheets being approximately 1.8 times the original strength. The AA5050 sheet with the highest tensile strength was manufactured using the 50nm aluminium oxide. This sheet had an ultimate tensile strength of 456.41 MPa being approximately twice the original strength of 226 MPa.

It can be concluded that the nano- sized particles, and particularly nano-sized particles, significantly enhance the bond strength in ARB.

One mechanism leading to improved interfacial bond strength and to improved tensile properties of ARB

materials is believed to be embedding the particles in the surface of layers being roll bonded.

The particles are harder than the materials being bonded, so the particles become embedded in the surfaces being bonded at the bonding interface. When shear strain is applied, dislocations move along the bond interface. The dislocations tend to pile up at a particle embedded in the surface, thereby pinning the dislocations and increasing the shear strength of the bonded product. However, if the size of the particle is greater than a particular size, such as in the micro-size scale, the surfaces of the sheets may not completely wrap around a particle and the surface may not completely contact. As a result gaps between the surfaces, ad between the surfaces and particle, form. Such gaps, or cracks, reduce shear strength because the surfaces are no longer bonded substantially continuously along their surfaces.

Another mechanism believed to improve interfacial bond strength is that the particles are thought to improve shear strength by improving bond adhesion between the two surfaces. Specifically, the particles are thought to break the oxides layer that exists at the surfaces of the sheets to be roll bonded. Wire brushing has a similar effect, but time lag between wire brushing and roll bonding enables the oxide layer to re-form on the surface to an extent. However, the particles cause fresh surfaces to be exposed and, without sufficient time to re-form an oxide layer, the surfaces are bonded more effectively. Additionally, it is believed that some diffusion, into the sheets, occurs from the particles, thereby causing solid solution strengthening in the areas of diffusion. This may also contribute to improvements in bonding strength.

Reference to any prior art in the specification is not, and should not be taken as, an acknowledgment or any form

of suggestion that this prior art forms part of the common general knowledge in Australia or any other country.

Many modifications may be made to the preferred embodiment of the present invention as described above without departing from the spirit and scope of the present invention.

It will be understood that the term "comprises" or its grammatical variants as used in this specification and claims is equivalent to the term "includes" and is not to be taken as excluding the presence of other features or elements .