Free Academic Seminars And Projects Reports

SYNOPSIS

Aluminum auto body sheets produce appreciable amounts of slivers during trimming operations when trimmed with dies conventionally designed for steel sheets. The slivers can be carried through downstream processes and cause damage to the surface of formed parts resulting in costly manual repair. This paper reports a systematic experimental study at both the macro- and micro-scale of the sliver generation problem with relation to cutting angle, clearance, and blade sharpness. It is concluded that slivers can be reduced or even eliminated by modifying the current trimming tools designed for steel sheets. The most striking finding is that the conventional wisdom of 0-degree cutting conditions generates the largest amount of slivers for aluminum alloy sheets. With proper cutting angles, trimming aluminum auto body sheets can actually be more robust than that for steel sheets the clearances can be less restrictive and tools may require much less frequent sharpening.

[attachment=8027]

INTRODUCTION
Recent years have seen a vast increase in the use of aluminum alloys in the automotive components due to requirement for fuel efficiency and high performance. Aluminum alloys are natural choice over steel for light-weighting, corrosion resistance and easy recycling. However aluminum alloys as new material to automotive industry pose many challenges to manufacturing processes that have traditionally dealt with steels.

One such challenge occurs in the process of press forming (termed stamping) of aluminum alloy sheet material in to body panels (e.g. hoods, doors, fenders , etc). A typical manufacturing flow-path of press forming process consist of stretch/draw, trim and pierce, flange/pre-hem and final hem. Subsequently the panels go to assembly. The problem is Aluminum alloy sheets produce appreciable amount of sliver using trimming operations when trimmed with dies conventionally designed for Steel sheets.

Sliver is the general term for debris and small aluminum pieces produced in trimming operations. Because of high production volume in the automotive industry (as many as 1000 parts from a transfer stamping press), Sliver generated during each stroke of trimming operation quickly accumulate and spread all over trimming dies. The Slivers are then carried through downstream processes (eg. Flange/Pre-hem, Final hem and assembly) and cause damage to the surface of formed parts. Figure 1 shows Sliver damaged Aluminum alloy hood outer panels produced in an automotive manufacturing plant. The manufacturing cost increase due to Sliver problem for surface sensitive outer panels can range from 5-15%, making it one of the single largest factor that increases manufacturing cost. The Sliver problem is largely responsible for the typical stamping rate for aluminum panels to be 85% of stamping rate for steel panels. The sliver problem has been recognized as a critical concern by automotive manufacturers around the world.

Trimming is essentially a shearing process. Because of the widely spread use of shearing processes in industries ranging from primary metal, automotive, aerospace, and electronics, great advances has been made steadily. Generally many parameters in trimming operation can affect Sliver generation. Those normally considered include clearance, blade sharpness, trim line geometry/curvature, blade geometry, blade material and coating, blade surface finish, lubricants, press speed, temperature, clamping pad pressure, depth of entry, sheet thickness, sheet material etc. It is worth emphasizing that conventional wisdom suggests keeping the cutting angle as close as possible to 0 degree, i.e. blade travels vertically to sheet plane.

One of the most important findings of this work is that an operation window exists for trimming aluminum alloy sheets in terms of cutting angle, clearance and blade sharpness. The operation window is bounded by Sliver generation on one side and by Burr and Cut-surface quality on other.

Figure 1: Sliver damaged Aluminum alloy hood outer panels

EXPERIMENTAL PROCEDURE

Figure 2: Clearance, Blade sharpness and Cutting angle are the parameters considered

As schematically shown in figure 2, the Clearance is the horizontal distance between the cutting blade and the lower die. The percentage clearance is expressed as the ratio of this distance to sheet thickness. The blade sharpness is quantified by a blade edge radius. A smaller edge radius represents sharper blade while a larger edge radius indicates a duller blade. For convenience, the Cutting angle is defined as the complimentary angle of the blade travel direction with respect to the sheet plane. Therefore, a zero degree cutting angle represents the cutting condition that the blade travels vertically to sheet plane.

A double action laboratory die was specifically designed for this investigation. It closely simulates the actual trimming operation under production conditions. The die was placed in a 20 ton commercial two-step press with a punch speed of 0.05 m/s. The pad pressure was designed to be 1Mpa uniformly applied on the sheet. The clamping pad was one sheet thickness behind the cutting edge of the bottom die edge as in production trimming operations. The die is capable of trimming sheet metal at 0 , 5 , 15 , 20 and 25 cutting angle. The clearances were precisely controlled to be 5%, 10%, 15%, 20% and 25% of the sheet thickness by using thin metal shims.

In total, six blade sharpnesses were evaluated. Blades of five different sharpnesses with edge radii 25.4 m, 50.8 m, 127 m, 254 m and 0.5mm were precisely machined and measured using an optical comparator. The sharpest blade had an edge radius of 2.5 to 10 m.

The sheet metals tested were AA6111-T4 and AA6022-T4 which are the most commonly used aluminum sheet materials in North America for autobody panels, especially outer panels. Both alloys used in this experiment were 1mm thick, a typical thickness in stamping production in North America (European manufacturers generally use 1.1mm thick aluminum sheet materials). The dimensions of sheet test specimens were 38.1mm X 127mm. The effects of sheet orientation were also evaluated with specimens prepared parallel and transverse to the rolling direction. All specimens were tested under unlubricated conditions using straight blades. A total of 130 cutting conditions were evaluated which included different combinations of cutting angles, clearances, and blade sharpnesses. Five repeated tests were performed for each test conditions. If there was Sliver generation under a particular test condition, 10 repeated tests were conducted for this test condition. Over 1500 specimens for each alloy were tested. For comparison, selected cutting conditions were tested for 0.79mm AKDQ(Aluminum Killed Draw Quality) Steel sheets and 0.74mm DQIF(Draw Quality Interstitial Free) Steel sheets. The experiment matrix is summarized in table 1.


--
Table 1:
-- Cutting Angle 0,5,10,15,20,25 6
Clearance 5,10,15,20,25 5
Blade edge radius (in m) 2.5-10,25.4,50.8,127,254,0.5mm 6
Alloy 6111-T4, 6022-T4 2
Repeatability 5
Total specimen tested >3600
Steel sheets AKDQ, DQIF
--

As there were three independent parameters evaluated in this experimental investigation, it is not an easy task to present the huge amount of experimental data. Some of the general observations are briefly summarized and the focus is on the trends describing the effects of each parameter.

GENERAL OBSERVATIONS
First of all it was observed that there was no measurable difference between the results of sliver generation(as well as Burr and cut surface quality) for 6111-T4 and 6022-T4. Cut surface quality is defined as roughness of cut surface. Therefore these two alloys will not be distinguished.

With the blade having an edge radii of 2.5 to 10 m, the condition of 5% clearance and 0 cutting angle generated significant amounts of hair like sliver. As the cutting angle increased, the amount of slivers decreased. When the cutting angle was more than 10 degrees, no slivers were generated for all clearances tested. Similarly the amount of sliver also decreased as clearance increased. However the Burr height increased and the cut surface quality decreased as clearance increased. Sliver generated by the cuts along the rolling direction of sheet appeared longer than those in transverse direction. There was no distinguishable difference in Burr height and cut surface quality between the cuts in rolling and transverse directions.

The blades having an edge radius of 25.4 m and 50.8 m produce similar results. The 50.8 m radius blade generated slightly wider pieces of hair like slivers and slightly higher burrs than 25.4 m radius blade. The appearance of slivers began to change characteristically when blade edge radius reached 127 m. The hair like slivers were replaced by slivers of much wider and thicker long metal pieces and particles and flakes. The cut surface quality declined compared to cuts by blades with sharpnesses 2.5 to 10 m, 25.4 m and 50.8 m. With the combinations of blade sharpnesses, small cutting angles (0 to 10 degrees) and large clearances (15-20%), the roughness of cut surface increased significantly as clearance increased.

The blades with cutting edge radii of 254 m and 0.5mm are considered very dull for cutting sheets of 1mm thick material. With these blades the slivers were in form of either large thick metal pieces or large flakes and particles. However, with 15-25 cutting angle and 5-10% clearance no slivers in any form were generated and the burr was minimal for blades with edge radii upto 0.5mm. The quality of cut surface was also very good.

It was observed that majority of slivers occurred during the down stroke. The press used for the testing is a two step press. It could be paused after down stroke and then restarted for return stroke. The conclusion that the majority of slivers occurred during the down stroke was also confirmed by extensive metallographic investigations to trace the origin of sliver.

No sliver were generated for both the AKDQ and DQIF steel samples for all evaluated cutting conditions and grain orientations. Approximately 150 steel specimens were tested using selected combinations of cutting angle, clearance, and blade sharpness. The restrictions placed on trimming angles, clearances and blade sharpness for cutting steel sheets is not attributable for concerns, for slivers, but rather the ability to actually trim the metal with acceptable burr and goods cut surfaces.

As there were three independent parameters evaluated in this experimental investigation, it is not an easy task to present the huge amount of experimental data. Some of the general observations are briefly summarized and the focus is on the trends describing the effects of each parameter.

TRENDS OF PARAMETRIC EFFECTS
One of the most important findings of this work is that the trimming operation windows for aluminum alloy sheet in terms of cutting angle, clearance, and blade sharpness are bounded by sliver generation at one side and by burr and cut surface quality on the other. This finding can be illustrated schematically by the trends of effect for each individual parameter.
CUTTING ANGLE
Figure 3 schematically shows the trends of effects of cutting angle for both aluminum and steel sheets.


The left vertical axis scales the severity of sliver generation, the right vertical axis scales the Burr height. The horizontal axis represents the cutting angle. Correspondingly the trimming operation window is bounded by the left curve for sliver generation and by right curve for burr height.

As observed in laboratory testing as well as in production stamping lines slivers are not a problem for steel sheets. There fore the operation window for steel sheets is just bounded by the right curve , the Burr height. A zero degree trim angle provides the best condition to cut steel sheets and produce good cut surface quality and smaller Burr. There-fore trimming tools conventionally designed for steel sheets require cutting angles as close as possible to zero degrees. In practice automotive manufacturers generally try to avoid cutting angles greater than 15 .

The cutting angles required for aluminum sheets differ dramatically from that for steel sheets. The best condition for steel cutting provides largest amount of sliver for aluminum sheets. As the cutting angle increases the amount of slivers generated gradually decrease. When the cutting angle is greater than 15 slivers were significantly reduced or completely eliminated even with blade radius as dull as 0.5mm for a wide range of clearances. However if cutting angle increases beyond a certain value the Burr height and quality of cut surface become intolerable even though no slivers were generated.

CLEARANCE
Under the optimal cutting angles for aluminum and steel sheets the effects of clearance is schematically shown in figure 4.


For steel auto body sheets the optimum 0 cutting angle it is usually desirable to keep clearance below 10 of metal thickness. Above this value the Burr height and cut surface quality become unacceptable and it may even be impossible to cut off sheet metal. Slivers usually are not a problem.

For 6111-T4 and 6022-T4 with optimum cutting angles the clearance can be 15 of sheet thickness with blade being as dull as 0.5mm and almost no slivers plus a tolerable Burr height and acceptable cut surface quality are observed. It clearly indicates that under optimal cutting angle conditions the clearance for aluminum alloys can be less restrictive than that for steels.

BLADE SHARPNESS
Again under optimum cutting angle for aluminum and steel sheets figure 5 schematically show the effect of blade sharpness.

It is well known that maintaining sharp blades is critical to successful trimming of steel sheets. Under the optimal cutting angle for steel sheets when the radius of cutting blade exceeds 125 m and clearance exceeds 10 large and sharp Burrs and unacceptable cut surface quality results for steel sheets.

For both 6111-T4 and 6022T4 with optimum cutting angles the blade edge radius can be as dull as0.5mm while the clearance can be as large as15 and the trimming still produces almost no slivers, fairly small burr and good cut surface quality.

Based on the general observations and the trends identified for each parameter manufacturing tool design guidelines were generated to specify the requirements for clearance and blade edge radius for a sequential range of cutting angles.

Therefore it is concluded that with optimal cutting angles trimming aluminum sheets can be more robust than trimming steel sheets. For aluminum alloys the clearance can be less restrictive so that time to fabricate trimming dies as well as the time for line set up and maintenance is less. Blades will require much less re sharpening because the cutting is less sensitive to blade edge radius. There fore much less tool life can be achieved.

ORIGIN AND MECHANISM OF SLIVER GENERATION
The morphologies of the slivers generated under cutting conditions were statically categorized. The slivers generated at all cutting conditions were collected using adhesive cellophane tapes. Micrographs were taken of many of the slivers. Based on the analysis of many of the micrographs, the slivers were categorized into three basic types.
The first type which statically represents 80% of the total sliver generated in all the cutting conditions is shown in fig.6
Fig.6 Optical micrograph of first type of sliver

The fine hair like slivers observed with unaided eyes are in this category. Under magnification in an optical microscope, the long slivers have a smooth and shiny surface and a rough surface on the other side. Scanning electron microscope revealed that the rough surface is a fractured surface dominated by shear deformation but with a slight amount of tensile deformation.

The lab generated slivers were similar to those generated in production process. The width of this type of sliver increases as the blade edge radius increases. SEM observations indicate that these slivers appear to be thin and curled when viewed in their longitudinal direction. To trace the origin of this type of sliver, many of the trimmed sheet pieces were examined carefully under optical microscope. As schematically shown in figure 8, the portion being trimmed off is the scrap and the portion remaining between the hold-down pad and the die is the part. Occasionally the slivers had not completely dropped off from the trimmed sheet pieces, providing the evidence of the origin of slivers.
Fig.8 Sketch illustrating the part and scrap of a trimmed sheet


Fig.9 shows the optical micrograph of trimmed edge of scrap piece viewed from the bottom surface. When observed from this direction, the surfaces of the slivers hanging on the edge of scrap are rough.

Fig 9


Fig.10 given below shows an optical micrograph of a trimmed edge of a scrap piece as viewed from top surface. The surfaces of sliver observed from this direction are smooth and shiny.

Fig.10

The above observation was confirmed numerous times to assure it is not an isolated incident. As a separate validation of the finding scrap pieces with slivers still attached were examined using SEM. The rough surface of the attached sliver has the same fracture characteristics as the rough fracture surface of slivers collected with cellophane tapes.

The above observations indicate that the first type of slivers are generated at the upper edge contact with the scrap during the downstroke of trimming operation. This conclusion is also supported by an investigation of an interrupted test. The blade travel was stopped midway during the cutting process and the specimen was then recovered. The cross section surface of this still collected part and scrap were polished. Fig.11 shows an optical micrograph of the blade contact portion of the specimen.

The figure clearly shows that a C shaped crack has developed. The piece above the C shaped crack will break away along the crack to form the sliver of first type. This observation also explains why the sliver of first type has smooth surface on one side (contacted with the blade) and rough surface on the other (fracture surface). The thickness of the sliver is defined by the depth of C shaped crack. The sliver is certainly thin and curled validating the SEM observations on individual slivers.

The fact that this type of sliver develop during downstroke is also confirmed by macroscopic experiments. The press used is a double-action press which can be paused after downstroke and then restarted for return stroke. It was observed that the slivers of this type had already been generated after the downstroke. The upstroke did not increase the amount of this type of slivers, but did increase other types of slivers.

It is also noted in the experimentation that cuts along the rolling direction of sheet produced longer pieces of slivers than the cuts perpendicular to the rolling direction. This is most likely due to grain geometry difference in the two directions. The grains are much longer along the rolling direction. The sliver generated by the cuts along the rolling direction contain fewer numbers of grain boundaries. Therefore, there is less chance for the long slivers to separate into smaller pieces because grain boundaries are often the sites of fracture initiation.

Fig.12 is an Optical micrograph of sliver of second type. It is generally smaller in width and shorter in length than the first type, however it is thicker in the third direction compared to the first type.


Fig.12 optical micrograph of second type of sliver a)generated in the lab, b)generated in production line. Note that the picture background is adhesive cellophane tape.

Fig.13 shows the second type of sliver generated by blade rubbing off burr area. The slivers were still attached on the part providing the evidence of their source.


The third type of sliver is generally in the form of flakes and particles.


Fig.14a shows optical micrograph of some of these particles and flakes. In general these particles and flakes result from frictional interaction between blade and cut surface of part.
Fig.14b is an optical micrograph of cut surface of part with particles and flakes still attached.
Tracing the origins of the slivers and understanding their generation mechanism help shed light on effects of trimming parameters evaluated. The first type of sliver is generated due to the formation of C-shaped crack. The C-shaped crack develops mainly due to shear deformation. The angled cut eliminates the condition which promotes the formation of C-shaped crack. Therefore the first type of slivers are eliminated or reduced. Geometrically an angled cut reduces the tendency and probability of blade friction interaction with cut surfaces of the part because zero degree cutting angle results in a curved S-shaped cut surface profile for the part while the optimal angled cut produces straight cut surface profile. In addition, an angled cut accelerates the fracture process thereby reducing the time that the blade interacts with the sheet material. Consequently, optimal angled cut reduces the tendency and probability of generating the second and third types of slivers. Increasing the clearance has similar effects to that of increasing cutting angle due to the additional bending deformation.

The blade sharpness determines the appearance of sliver. When the blade edge radius is small, the indentation deformation zone is small and the size of C-crack is small. As a result, the width of the first type is small. When the blade is sharper, naturally the second and third type of slivers rubbed off from the cut surfaces of the part will also be smaller.

The understanding of the origins of sliver suggest that the depth of entry has no effect on generation of first type of slivers. However, minimization of entry depth will reduce the opportunity to generate the second and third types of slivers by reducing frictional interaction of blade cut surface during the down as well as back stroke. For the same reason, the improvement of blade surface finish as well as application of lubrication will not reduce the generation of first type of sliver but will certainly reduce the generation of second and third types.

CONCLUSION
Aluminum has been perceived by the automotive industry to be a difficult-to-handle material. The manufacturing cost of making an automotive outer body panel from aluminum alloy are usually appreciably higher than making the same panel from steel. The purpose of this paper ultimately is to change this perspective.

Experiments show that with appropriate tool design, the sliver issues with aluminum alloys can be resolved and the trimming of aluminum autobody sheets can actually be more robust than that of steel sheets. The fact that the conventional zero degree cutting angle results in most sliver generation for aluminum alloys is a paradigm shift. One needs to think and treat aluminum conceptually different than steel, because of their inherent difference in microstructure and mechanical properties. It is so comforting that the clearances can be less restrictive and the tools require much less frequent resharpening (longer tool life) when trimming aluminum alloys. These translate into less time for die fabrication and line setup, much less line down-time and much lower maintenance cost.

This technology has been requested by almost every major automotive manufacturers around the world. It has been successfully implemented in several North American automotive stamping plants. It is recently learned that implementation in one manufacturing plant resulted in significant reduction (from 25% to just 4%) in the manual repair-rate for aluminum alloy outer hood panels. The production line also had five consecutive monthly record runs, as there was no need to clean up slivers in addition to significant increase in robustness of trimming process. In another case, sliver reduction guidelines were the driving force for a conceptually new design for an aluminum hood outer in a next generation mid-size car model which has been released to production.

REFERENCES :
Ming Li, Sliver Generation Reduction in Trimming of Aluminium Auto body Sheets ; ASME Journal, Feb.2003 Pg.128-137