WO2004096460A1

WO2004096460A1 - A punch for extrusion processes and a product produced by use of such punch

Info

Publication number: WO2004096460A1
Application number: PCT/DK2003/000274
Authority: WO
Inventors: Joachim Danckert
Original assignee: Univ Aalborg; Joachim Danckert
Priority date: 2003-04-25
Filing date: 2003-04-25
Publication date: 2004-11-11
Also published as: AU2003226957A1

Abstract

The invention relates to extrusion processes incorporating a punch design having a non-cylindrical punch land. With the new design, a possible small tilt of the punch only gives rise to minor changes in the contact between can and punch. With the new design, great improvements can be obtained, as follows: 1) A small tilt will only give rise to minor differences or variations in can height. 2) The punch does not drift, which means that a can with more even wall thickness in the height direction will be produced, 3) As the can quality is nearly unaffected by small changes in the tilt angle, it will be much easier to keep the production under control. 4) The punch may be easier to manufacture.

Description

A PUNCH FOR EXTRUSION PROCESSES AND A PRODUCT PRODUCED BY USE OF SUCH PUNCH

BACKGROUND OF THE INVENTION

Extrusion is a widely used process to manufacture a great variety of metal products. A number of different examples can be seen in e.g. /1, 2/. In Figure 1 is shown various basic extrusion processes a) forward rod, b) forward tube, c) forward can, d) backward rod, e) backward tube, f) backward can, g) sideward rod, h) sideward tube, i) sideward can. /2/. In Figure 2 is shown a schematic outline of the backward can extrusion process, 1. punch, 2. workpiece, 3. container, 4. ejector (after/2/). In the embodiment shown, the punch has a circular cylindrical cross-section seen in a direction parallel with a punch line, said punch line shown by an arrow. The punch has a punch land forming an outer surface. The punch land extends from a lower circumference of the punch to an upper circumference of the punch, said lower and upper circumference both exhibiting the largest diameter of the punch, and said upper and lower circumference delimiting the punch land.

GB 1,199,085 describes a punch having an outer convex tip. The convexity is rather large with a ratio from 53 % to 66 % between the height of the convex part and the diameter of the convex part. The large convexity leads to a high rate of friction and to wear of the punch. The purpose of this punch is to eliminate what is called structure separation.

The extrusion processes have been the subject of investigation in numerous books and papers. Despite the large amount of literature describing various aspects of the extrusion process, the inventor has not found any literature describing in detail the influence, which the punch land and/or die land has on the extrusion process and on the quality of the parts produced. In Figure 3 is shown a sketch of the ironing process, and in Figure 4 is shown a sketch of the ironing process in which the die land is slightly tilted /4/.

The die land in the conventional design of ironing dies is made cylindrical as shown in Figure 3. The punch and the die used in extrusion are commonly also made with a cylindrical punch land and a cylindrical die land. In /3,4/ it is shown that the die land can have a significant influence on the stability of the ironing process, and that a slight tilt of the die land can cause the ironing process to become instable. This tilt can be cause by e.g. inaccurate machining, inaccurate mounting of the tools or elastic deformation of the tools and/or press.

If the die land is slightly tilted, as sketched in Figure 4, contact between the cup wall and the die land is completely lost on side A, whereas the contact conditions on the opposite side B remains nearly unaffected by the tilt. This difference in contact conditions on side A and side B has the effect that the radial force on the die on side A is reduced compared to the radial force on the die on side B. This difference in radial force will try to move the die to the left (or the punch to the right) with the effect that the reduction ratio increases on side A and decreases on side B. In the opinion of the inventor, a slight tilt of the die land and/or punch land in the extrusion process will have the same effect, and the purpose of this work is to analyse the effect of a slight tilt of the punch land in case of can extrusion.

The design of conventional punches used in can extrusion has been dealt with in a number of text books and papers. The main focus has been the shape of the punch face with regards to lubrication and with regard to the extrusion force (punch force). To the knowledge of the inventor the effect, which the punch land may have on the stability of the extrusion process, has not been investigated (the position of the punch land on the punch can be seen from Figure 7, the punch land is denoted h)

In Figure 5 is shown different punch-face shapes: (a) Flat, (b) Conical, (c) Flat and conical, (d) and (e) Spherical, and in Figure 6 is shown the effect of the punch-face shape on the maximum punch pressure in backward can extrusion /5/. All the punch-face shapes shown in Figure 5, besides from (e), has a cylindrical punch land. In Figure 1 is shown the effect of the punch-face shape on maximum punch pressure in backward can extrusion. /5/

In Figure 7 is shown the punch design recommended by ICFG (International Cold Forging Group) for backward can extrusion of steel /5,6/. From Figure 7 it can be seen that ICFG, for punches for backward extrusion of cylindrical cans made from steel, recommends a cylindrical punch land with length h = (0.3 - 0.7) d , where d is in diameter of the punch land.

Radial pressure on the punch land in can extrusion:

The radial pressure on the punch land in can extrusion has been determined using the slab method. If the thickness of the can wall is thin compared to the radius of the punch, the deformation of the material between the punch land and the container wall takes place under nearly plane strain conditions. The calculation of the specific pressure on the punch land has been carried out assuming plane strain conditions.

In Figure 8 is shown a sketch of the region near the punch land. The cylindrical punch land has the length h and the can wall thickness is t. To the right in Figure 8 is shown the slab with the stresses and pressures action on the slab. If the friction conditions between can wall and container are same as between the punch land and the can wall (same friction coefficient or same friction factor), and it is assumed that the vertical stress in the can wall above the punch land is zero (the same as assuming zero friction between can wall and container above the punch land), the average specific pressure p on the punch land can easily be calculated to be (see e.g. Ill)

or (1) mh

2k\ 1 + ^~2t^~ where the first expression is when Coulomb friction is assumed with the friction coefficient equal to μ (τ = μp). The second expression is obtained if constant friction is assumed with the friction factor equal to m (τ = mk).

The total radial force P (per unit width in circumferential direction) is thus

(2)

k is in both (1) and (2) the shear yield stress of the can wall material.

Equation (1) and (2) can easily be modified to take into account different friction conditions in the punch land - can wall interface and in the can wall - container interface. From (2) it can be seen that

- the radial force is proportional to the yield stress of the can material

- the radial force increases with the length of the punch land.

- If friction is neglected the radial force is proportional to the length of the punch land - the radial force increases if the last term in (2) increases; that is

- the radial force increases if a) the friction increases and if b) h/t increases

The influence from a slight tilt of the punch land: If the punch land is slightly tilted as shown in the sketch in Figure 9, contact between can wall and punch land will be lost on one side of the punch land (side A in Figure 9), whereas a slight tilt will only have a marginal effect on the contact conditions between can wall and punch land on the opposite side (side B in Figure 9). A slight tilt of the punch land may be cause by e.g. inaccurate machining, inaccurate mounting of the tools in the press or elastic deformation of the press and/or tools during the extrusion.

In the region where contact between can wall and punch land is lost due to the tilt (side A in Figure 9), the radial force on the punch land drops to zero. On the opposite side (side B in Figure 9), the contact conditions remains nearly unaffected and the radial force on the punch land per unit width (in circumferential direction) is given by equation (2). This difference in radial force on the punch land on side A and side B will deflect the punch elastically to the right leading to an increased reduction ratio in region A and a decreased reduction ratio in region B.

If one only looks at the total horizontal force on the punch land, a "perfect" can extrusion process is in a state of unstable equilibrium when carried out with a punch with a cylindrical punch land (a slight disturbance causing a slight tilt of the punch land will create a net horizontal force on the punch land, which will cause a deflection of the punch)

DETAILED DESCRIPTION OF THE INVENTION

It is an object of the present invention to eliminate some or all of the aforementioned problems. This object is obtained by a punch for extrusion processes, where the punch has a cross-section, seen in a direction perpendicular to a punch line of the punch, having a circumference shaping an outer surface of the punch, and where between an upper circumference of the punch and a lower circumference of the punch a punch land is formed running along the outer surface of the punch and extending from a punch land start at the lower circumference to a punch land end at the upper circumference along a punch land trajectory, and where the punch land trajectory is non-cylindrical.

The punch is constituted by an oblong metal member, the one end being fastened to a hydraulic press or the like force-acting machine, and the other end being the actual punch being forced into the metallic blank constituting the output blank for the manufacturing process. The punch has an outer surface, part of which is the "active" part during the punching process. The "active" part of the outer surface of the punch is delimited by a lower circumference near a tip at the other end of the punch and an upper circumference near the one end of the punch. Between the lower circumference and the upper circumference, trajectories run along the outer surface, thus forming the "active" outer surface of the punch.

The reference to the lower circumference and upper circumference is based on the fact that the punching process may take place from above and downwards. However, the reference is not to be construed as limiting the scope of protection. Thus, despite the reference to lower circumference and upper circumference, the punching process may take place from beneath and upwards or from the side and sideways. The new punch design:

As described above, a slight tilt of the punch land may have a significant influence on the total radial force on the punch in horizontal direction when a conventional punch design with a cylindrical punch land is employed. To reduce the effect from a slight tilt of the punch land on the stability of the can extrusion process, it is suggested to make the punch land slightly curved as shown in Figure 10, a sketch of a punch nose with a circular profiled punch land. If the punch land is curved e.g. with a circular profile as shown in Figure 10, a slight tilt of the punch land will only change the contact conditions between can wall and punch land slightly.

A slight tilt of the punch land will thus only give rise to a small total radial force on the punch land and the elastic deflection of the punch will thus be small compared to the deflection of the punch if this was made with a cylindrical punch land. It is thus believed that by making the punch with a curved punch land the extrusion process becomes more stable and it becomes possible to produce cans with less variation in can wall thickness as compared to when a conventional punch design with a cylindrical punch land is employed.

The effect of a slight tilt of the punch land has been investigated using FEM-simulations. The simulations show that a slight tilt of a cylindrical punch land can give rise to a significant total radial force on the punch, whereas a slight tilt only give rise to a small total radial force when the punch land has a circular profile as shown in Figure 10.

Experimental verification of the proposed punch design:

The effect of using a punch with a circular profiled punch land in backward can extrusion has been verified experimentally using a backward can extrusion process running in a Danish company. The material in all the experiments was soft aluminium (Al 99.5%).The dimensions of the slug was ø72.9*15 mm and the final dimensions of the can was 073*250 mm with a nominal wall thickness of 0.915 mm. In Figure 11 are shown the slug to the left and the extruded can. The backward can extrusion process is shown schematically in Figure 2.

Two different designs of the punch nose were used, a) a punch nose with a cylindrical punch land and b) a punch nose with a circular profiled punch land. In Figure 12 is shown the dimension of the conventional punch nose with the cylindrical punch land and in Figure 13 is shown the punch nose with the circular profiled punch land. Besides from the circular profiled punch land the punch nose was identical to the conventional punch with the cylindrical punch land shown in Figure 12 (same dimensions, same material and manufactured to the same specifications).

The punch nose was mounted screwed on to a long punch (length of punch shaft 220 mm) mounted in the press. The two punch noses used were measured with a co-ordinate measuring machine. The punch profile was also measured using a very accurate profile measuring machine (actually a roughness measuring machine which also can be used as a profile measuring machine in the range +- 0.75 mm).

Below is listed the main measured data:

Data measured with the co-ordinate measuring machine:

Punch with cylindrical die land: cylindricity: 0.0079 μm flatness of upper surface: 0.0058 μm parallelism of upper and lower surface: 0.0082 μm diameter of punch land: 071.1728 μm Punch with circular profile punch land: flatness of upper surface: 0.0013 μm parallelism between upper and lower surface: 0.0021 μm max. diameter of punch land: 071.1486 μm

Data measured on the profile measuring machine:

Punch with circular profile punch land: parallelism of punch land in relation to punch shaft: < 1 μm

Punch with circular profile punch land: the dimensions measured are those (besides from the diameters) shown in Figure 13. 10 cans produced with the conventional punch (Figure 12) were picked from the actual production. After the production run of the cans using the conventional punch nose, had been finalised, the punch nose with the conventional punch design (Figure 12) was replaced with the punch nose with the circular profiled punch land (Figure 13) and 12 cans were produced.

The cans produced, both with the conventional punch design and with the new punch design with the curved punch land have been produced with

- the same material - the same slug geometry

- the same lubrication and lubrication procedure

- the same tool (besides from punch nose) and same press

- same tools set-up (the punch was not dismounted from the press, when the conventional punch nose was replaced with the punch nose with the circular profiled punch land) - the same pressing speed

Main emphasis was placed on keeping everything the same (besides from the punch nose) when producing the cans using the conventional punch design and using the proposed punch design with the circular profiled punch land. Any difference in the can wall thickness in the cans produced with the punch with the conventional punch design (Figure 12) and with the punch with the proposed punch design (Figure 13) can thus be attributed to the difference in the geometry of the punch nose. In Figure 14 are shown the geometry of the punch lands of the two punches drawn together. From Figure 14 it can be seen that there is only a slight difference between the punch nose with the cylindrical punch land and the punch with the circular profiled punch land. But as shown below, this slight difference has a significant influence on the quality of the cans produced.

Experimental results:

All the cans looked very much the same and had nearly the same height. The can rim was on all the cans fairly even; the maximum difference between the highest and lowest point on the can rim was less than 2 mm for all the cans.

The thickness of the can wall of all the cans (10 cans made with the punch with the cylindrical punch land and 12 cans made with the punch with the circular profiled punch land) was measured on a coordinate measuring machine. The can was aligned with the cylinder axis of the can in the z-direction. The wall thickness was measured in the height direction in 0, 90, 180 and 270 degree. The zero degree direction was placed at the lowest point on the rim of the can. z was set equal to zero just below the lowest point on the can rim. The wall thickness was measured for every 2 mm from z = 0 (just below the can rim) to z = 205 mm below the can rim. The wall thickness was also measured as function of the angle (for every 5 degree) for z = 0, 25, 50, 75, 100, 125, 150, 175 and 200 mm below the can rim.

The measured wall thickness distributions were fairly much the same in the cans produced with same punch. In the following are thus, for each punch nose, only shown experimental data obtained from the can showing the smallest and the can showing the largest thickness variation respectively.

In Figure 15 is shown the can wall thickness in 0, 90, 180 and 270 degree as function of the distance from the can rim for the cans produced with the conventional punch with cylindrical punch land and in Figure 16 are shown the wall thickness as function of the angle with the distance from the can rim as parameter. In Figure 15 and figure 16, the wall thickness is shown as function of the distance from the can rim measured in four different directions.

The upper figure is for the can showing the smallest variation in wall thickness and the lower figure is for the can showing the largest thickness variation. In Figure 16, the wall thickness as function of angle with the distance from the can rim as parameter. The upper figure is for the can showing the smallest variation in wall thickness and the lower figure is for the can showing the largest thickness variation.

In Figure 17 and 18 are shown the corresponding thickness distributions for the two cans produced with the punch with circular profiled punch land. In Figure 17, the wall thickness as function of the distance from the can rim measured in four different directions. The upper figure is for the can showing the smallest variation in wall thickness and the lower figure is for the can showing the largest thickness variation. In Figure 18, the wall thickness as function of angle with the distance from the can rim as parameter. The upper figure is for the can showing the smallest variation in wall thickness and the lower figure is for the can showing the largest thickness variation.

The minimum wall thickness of all 10 cans made with the conventional punch design was 0.702 mm and the maximum wall thickness 0.978 mm (max - min wall thickness = 0.278 mm). The minimum wall thickness of all the 12 cans made with the punch with the circular profiled punch land was 0.795 mm and the maximum wall thickness 0.925 mm (max - min wall thickness = 0.130 mm).

Discussions:

From Figure 15 it can be seen that when the conventional punch design is used, the punch has been deflected elastically off centre, increasing the reduction ratio one side and decreasing it on the opposite side. The maximum deflection occurs around 100 mm below the can rim. It is not fully understood why the deflection of the punch changes at this point, but it is believed to be connected with a change in lubrication condition between the can wall and the container wall. It can be seen from Figure 11 that there is a change in the appearance of the outside surface of the can. The upper app. 100 mm is shining bright, whereas the surface below is dull. One explanation for this change in surface (and probably also change in friction condition) is that the surface becomes dull when material/lubricant, which initially has been at the bottom of the slug, comes into contact with the container wall. However further investigations are required to investigate the exact reason for the change in wall thickness and change in surface appearance.

From Figure 17 it can be seen that when the punch with the circular profiled punch land is used the deflection of the punch is significantly less than when the conventional punch is used (Figure 15). It can be seen that there also is a change in can wall thickness around 100 mm below the can rim. The reason for this change may be a suggested above. By comparing Figure 15, 16, 17 and 18 it is obvious that the cans produced with the punch with the circular profiled punch land have a significantly smaller variation in can wall thickness compared to the cans produced with the conventional punch with the cylindrical punch land. With the conventional punch the wall thickness (for all the 10 cups) is in the interval 0.702 - 0.978 mm, whereas the wall thickness (for all the 12 cups) produced with the circular profiled punch land is in the interval 0.795- 0.925 mm. The use of the punch with the circular profiled punch land has in the backward can extrusion investigated experimentally had a very significant influence on the stability of the extrusion process; the process has become much more stable with regard to variations in the can wall thickness.

From Figure 14 it can be seen that the there is only a slight difference between the geometry of the punch with the cylindrical punch land and the punch with the circular profiled punch land. The experimental results thus show that even very small changes in the geometry of the punch land may have a significant influence on the stability of the extrusion process. Conclusions:

The punches used in can extrusion are commonly made with a cylindrical punch land. It is shown that a slight tilt of the punch will give rise to a total horizontal force on the punch land, which will deflect the punch elastically; the deflection will decrease the reduction ratio on one side and increase it on the opposite side. A simple slab analysis shows that the total radial force on the punch land due to a slight tilt of the punch land increases with the length of the punch land and that the ratio (length of punch land)/(can wall thickness) may have a significant influence on the total radial force. Based on the slab analysis, problems with elastic deflection of the punch due to a slight tilt of the punch land will be most pronounced when cans with a small can wall thickness are extruded.

A new punch design has been proposed. The new punch design encompasses a curved punch land in place of the cylindrical punch land. With a curved punch land, a slight tilt of the die will only give rise to minor changes in contact conditions between punch land and can wall, and the total radial force on the punch due to a slight tilt of the punch land will thus be small. It is thus believed that by employing a punch with a curved punch land (e.g. a punch land with a circular profile as used in the experiments) it becomes possible to extrude cans with less variation in can wall thickness compared to when a punch with a cylindrical punch land is employed.

The backward can extrusion of a thin walled aluminium can has been investigated experimentally. The extrusion was carried out with a conventional punch with a cylindrical punch land and with a punch having a circular profiled die land. In the experiments main emphasis was placed on keeping everything the same besides from the geometry of the punch. The measured differences in wall thickness in cans made with the conventional punch and with the punch with the circular profiled punch land is thus attributed to the difference in punch geometry. The experimental results show that the variation in can wall thickness is reduced significantly when the punch with the circular profiled punch land is employed compared to when the conventional punch design is used. Both the theoretical and the experimental results strongly indicate that the extrusion process can be made more stable by employing a punch with a curve punch land in place of the conventional cylindrical punch land as recommended e.g. by ICFG.

The experiments also show that a very small change in the geometry of the punch land can have a significant influence on the stability of the can extrusion process. The difference between the punch with the cylindrical punch land and the punch with the circular profiled punch land is very small as seen in Figure 14, but the difference in the cans produced with the two punches are significant. The maximum wall thickness is in all the cans found approximately 100 mm below the can rim. On the outside surface of the cans a distinct change in surface appearance can be seen at approximately 100 mm below the can rim; the surface of the upper part of the can is shiny whereas the lower part is dull. A likely explanation for the change in surface appearance is a change in friction conditions caused when material/lubricant initially located at the bottom of the slug comes into contact with the container wall.

The present investigation has focused on the effect of a slight tilt of the punch land in the can extrusion process. It is however the belief of the inventor that a similar approach can be used with all metal forming processes, where a cylindrical punch and/or die land is used. When a cylindrical punch and/or die land is used, a slight tilt of the land will have a significant influence on interface forces between the material being plastically deformed and the land. If the land is made with a curve profile, e.g. by making the land with a circular profile as in this investigation, or conical the change in contact forces due to a slight tilt of the land can be reduced significantly and the processes can thus be made more stable.

In the opinion of the inventor a "perfect" can extrusion process carried out with a punch with a cylindrical punch land is in a state of unstable equilibrium (the process is not be self-centring). If and how the punch nose should be designed in order to make the extrusion process self-centring will be a topic of future research.

References

1. ICFG Document No. 1/77 "Production of Steel Parts by Cold Forging",

Portcullis Press. Ltd, Queensway, England, 1978 2. ICFG Document No. 13/02 "Cold Forging of Aluminium",

Meisenbach Verlag Bamberg, BRD (ISBN 3-87525-161-x), 2002

3. Danckert, J., Ironing of thin walled cans, Annals of the CIRP Vol. 50/1 (165 - 168), 2001

4. Danckert, 3., The Influence from the Die Land on the Stability of the Ironing Process, Numisheet2002 (Eds. Yang, D-Y., Oh, S.I., Huh, H., Kim, Y.H.), Vol. 1

(115 - 120), 2002

5. Lange, K., Lehrbuch der Umformtechnik, Band 2 Umformtechnik, Springer Verlag, 1974

6. ICFG Document No.6/82 "General Recommendations for Design, Manufacture and Operational Aspects of Cold Extrusion Tools for Steel Components".

(in International Cold Forging Group 1967 - 1992, Meisenbach Verlag, Bamberg (ISBN 3-87525-058-3), 1992

7. Hosford, W.F., Caddell, R.M., Metal Forming, Prentice Hall (ISBN 0-13-588526-4), 1993

Claims

1. A punch for extrusion processes, where the punch has a cross-section, seen in a direction parallel with a punch line of the punch, having a circumference shaping an outer surface of the punch, and where a punch land is formed running along the outer surface of the punch and extending from a punch land start to a punch land end along a punch land trajectory, and where the punch land trajectory is non-cylindrical.

2. A punch according to claim 1, where between an upper circumference of the punch and a lower circumference of the punch the punch land is formed, and where the punch land start is positioned at the lower circumference and the punch land end is positioned at the upper circumference.

3. A punch according to claim 1, where between an upper circumference of the punch and a lower circumference of the punch the punch land is formed, and where the punch land start is positioned at the upper circumference and the punch land end is positioned at the lower circumference.

4. A punch according to any of claims 1-3, where the punch land trajectory is partly circular.

5. A punch according to any of claims 1-3, where the punch land trajectory is partly elliptic.

6. A punch according to any of claims 1-3, where the punch land trajectory is partly conical.

7. A punch according to any of claims 1-3, where the punch land trajectory is part of a tractrix.

8. A punch according to any of the preceding claims, where the punch corner radius (Rpunch) is between 0.1 and 1 times, as example approximately 0.5 times, the maximum cross-sectional linear dimension of the punch land.

9. A punch according to any of the preceding claims, where a length (h) of the punch land is between 0.01 and 1 times the square root of maximum cross-sectional linear dimension of the punch land.

10. A product produced by use of a punch according to any of the preceding claims, said product being a final product or being an intermediate product, and said product constituting a can with a bottom and side walls extending from the bottom, and said side walls already after having been produced by a punch according to any of the preceding claims having a substantially uniform wall thickness along an entire longitudinal extension 5 of the side walls.

11. A product produced by use of a punch according to any of the preceding claims, said product being a final product or being an intermediate product, and said product constituting a can with a bottom and side walls extending from the bottom, and said side

10 walls already after having been produced by a punch according to any of the preceding claims having a substantially uniform height along an entire circumference of the side walls.

12. A product according to claim 10 or claim 11, where the can has a circular cylindrical 15 cross-section.

13. A product according to claim 10 or claim 11, where the can has a continuous non- circular cylindrical cross-section.

20 14. A product according to claim 10 or claim 11, where the can has a polygonal cylindrical cross-section.

15. A product according to any of claims 10-14, where the thickness/height ratio is between 0.1% and 10%, as example 0.5%. 25

30

35

40