2014 Volume 54 Issue 12 Pages 2854-2859
The surfaces of dies used for flat sliding tests of galvannealed steel sheet (GA) have been investigated using several electron microscopic techniques in order to clarify the adhesion mechanism. Two kinds of adhesive materials were identified on the die surface. One consists of an Al oxide – Fe–Zn alloy composite and tightly bonds to ridges of the die surface. This resembles the built-up edges formed on tool surfaces during metal-cutting operations. Another adhesive material is composed of Fe–Zn alloys located in the hollows of the die surface which do not bond or only loosely bond to the die surface. A new adhesion model consisting three steps is proposed; (1) the Al oxide – Fe–Zn alloy composite layers are formed on the ridges of the die surface. (2) The Fe–Zn intermetallics on the GA surfaces are cut by this layer and (3) accumulate in the hollow of the die surface.
Zn-based coated steel is widely used in the world for automotive body parts because of their high corrosion resistance properties. One of the most important issues for coated steel has been formability which is significantly affected by the frictional characteristics of the coatings.1) The friction behavior of the Zn-based coated steel has been widely investigated from the view point of the coated materials, die materials, lubrications, and sliding-test conditions.1,2,3,4,5,6,7,8) The adhesion between the coated materials and die material occurring on the contact areas has been considered to be the main cause of friction resistance. From the macroscopic point of view, the adhesion is evaluated by the stick-slip behavior on the friction – stroke (distance) curve recorded during sliding tests.7) However, the mechanism of adhesion of Zn-based coating has been rarely discussed and has not been fully understood especially from the microscopic point of view.
Galvaneealed steel sheet (GA) is the most common coated steel for automotive application in Japan. Coatings and surface modification techniques have been studied and developed to reduce friction resistance and to improve the press-formability of GA.9,10,11) These operations have become more important in sheet metal forming high-strength steel sheet12) due to their increasing use for automotive parts. It is important to understand microscopic friction mechanisms in order to design GA surfaces with superior formability.
Many researchers have observed surfaces of Zn-based coated steel sheets after sliding tests using optical microscopy or scanning electron microscopy (SEM).2,3,5,6,7,8,9) In these papers the real contact surface areas or pits have been evaluated and discussed in the terms of sliding conditions and friction properties. For example, the friction forces of the contact area were correlated to the adhesion force6) and the concave pits were considered to play a role in supplying the lubricant to the sliding interface.3,5) These studies however have provided no answer as to what the adhesion of coated steel on the die surface is. One of the valuable approaches is to investigate die surfaces after the sliding test as well as the worked steel surfaces. Schey4) studied the bead surfaces after drawbead-simulation tests for Zn-coated steel sheets. In the case of electrogalvanized sheets (EG), the transfer layer forms on the bead surfaces by material-transfer from the worked surfaces and changes depending on the bead surface conditions. The bead surface is covered by zinc and the friction coefficient drops to the value typical for zinc sliding on zinc when nitriding steel beads were used for EG. This result shows that the transfer layer is an important controlling factor in determining friction stability as pointed out in Schey’s paper. For GA, the author mentioned that Fe–Zn debris formed a loosely attached coating on the beads and prevented any damage to bead and sheet.4) However interactions between the die surfaces and the surface of GA are still unclear.
We have carried out research focusing on the microscopic structure of the die surface in order to clarify the interactions between the die surfaces and the GA surfaces. The die surfaces after flat sliding tests of GA have been investigated using low-voltage SEM and transmission microscopy (TEM). In this paper the adhesion mechanism of GA is discussed from the plan view and cross-sectional imaging of the die surfaces and micro-probe elemental analysis data.
A typical GA sheet for automotive application is used in this study. The mechanical properties and fundamental coating characteristics of this steel sheet are listed in Table 1. The main intermetallic phase of the coating surface is Zn–Fe δ1, which was confirmed by X-ray diffraction analysis. The flat sliding tests were carried out using a flat sliding tester mentioned in Refs. 5) and 9). This testing instrument has the advantage that the friction property of the surface layer can be investigated without plastic deformation of the steel substrate. The sliding-test conditions are shown in Table 2 as well as the sizes of the die used in this study. The die shown in Fig. 1 was made of SKD11(JIS) grade tool steel and the sliding surface was finished by polishing with #2000 sand paper along the direction perpendicular to the sliding direction. A commercial washing oil actually used in autobody panel forming was applied to the GA sheets as a lubricant. The oil consists of mineral oil and some additives such as fatty acid salt and petroleum sulfonate. The sliding length was 130 mm for each sheet. The dies were removed from the instrument after one or two GA sheets were tested and their surfaces were investigated using electron microscopic techniques.
Photograph of the tool used in this study. The arrow shows sliding direction for steel sheet. (Online version in color.)
The distribution of elements on the die surfaces were measured by an electron probe microanalyser (EPMA), JXA-8600MX (JEOL, Japan) with an incident electron beam with an accelerating voltage of 8 kV. The diameter and step-size of the scanning electron beam were set to 30 μm. The die surfaces were observed using low-voltage scanning electron microscopic techniques.13,14,15) A model LEO1530 (LEO, now Carl Zeiss, Germany) was used in this study with accelerating voltages of 1 kV and 5 kV. Cross-sectional specimens of the die surfaces were made with a focused ion beam (FIB) instrument, FB2000A (Hitachi, Japan) for observations using the SEM and a TEM. The thinned specimens were made using the micro-sampling technique in the FIB instrument and investigated by a TEM, model CM20FEG (FEI, USA).
Friction coefficients (drawing force/normal load) are shown in Fig. 2 as a function of sliding distances for drawing tests of two GA sheets. The friction coefficients are leveled after sliding about 40 mm. There was a good reproducibility on the friction coefficients for the GA used in this study.
Friction coefficient curves as a function of sliding length. (a) the first GA sheet drawing, (b) the second GA sheet drawing. (Online version in color.)
The distributions of O-K, Al-K, Fe-L and Zn-L emission intensities measured by EPMA are shown in Fig. 3 for the die surfaces after 130 mm (one GA sheet) and 260 mm (two GA sheets) drawings. O, Al, and Zn are clearly detected especially on the right side which corresponds to the sheet incoming side. The distributions of these elements spread to the left side which corresponds to the sheet outgoing side after the two-GA sliding tests. The average x-ray counts of O-K, Al-K, Fe-L, and Zn-L emissions are denoted under each figure. The average counts of the x-ray from O, Al, and Zn are higher and those of Fe emissions are lower on the die surface after two-GA sheet drawings than those on the die surface after one-GA sheet drawing. These results and the distribution of Al and Zn in Fig. 3 demonstrate that Al and Zn were transferred from the GA surfaces to the die surface during sliding and their quantity increases with sliding distance. A Fe–Zn intermetallic (FeZn13) was detected by x-ray diffraction on the die surfaces after the sliding tests. This shows that Fe also moves to the die surface from the GA surface through the sliding, while the EPMA data show that the average intensity of Fe-L decreases through the sliding (Fig. 3). This latter result is explained by the fact that the increases of Zn, Al and O prevent Fe-L emission from being excited on the under side of the die matrix or escaping from where and these effects are larger than the increase of Fe-L emission from the intermetallics on the die surface. It should be noted that oxygen also increases by the sliding of GA on the die surfaces.
EPMA intensity mapping of Zn, O, Fe, and Al on the die surfaces after sliding tests. Sliding lengths: (a) 130 mm [one GA sheet drawing], (b) 260 mm [two GA sheet drawings] Average intensities (counts) are shown under the each mapping. (Online version in color.)
Figure 4 shows SEM images observed from a 45 degree direction for the die surface after the sliding test of one GA sheet. The sliding direction of the GA sheet is indicated by the arrow in the figure. The observed position is 0.5 mm from the edge of the steel-sheet-incoming side and the center of the width of the die. There are adhesive materials on the die surface along with the grooves caused by the polishing. The materials seem to be located on the ridge of the grooves and on the incoming side of the ridges. The die surface seems to be partially covered by the materials. These characteristics were seen on the die surfaces where O, Al, and Zn were detected by EPMA (Fig. 3) and the results for the die surface after two-GA sheets drawing were closely similar to these situations.
SEM images of die surface after sliding test with sliding length of 130 mm. The image was obtained by a 45 degree directional observation from the die surface with the primary electron energy is 1 keV. An arrow shows the sliding direction for steel sheet.
The die surfaces were observed by the SEM with high magnification from the direction of the surface normal. Figures 5(a) and 5(b) are the SEM images of the adhesive materials observed with the primary electron energy of 1 keV using an Everhart-Thornley detector (In-Chamber detector) and an In-Lens detector, respectively. The former image highlights the topography of the top surfaces.15) The adhesive materials have shapes stretching to the sliding direction. The image obtained by the In-Lens detector (Fig. 5(b)) visualizes the differences of the materials with thin thickness down to few-tens nanometers on the surfaces.15) The upper regions (position 1; sheet outgoing side) of the adhesive materials show remarkably darker contrast than the other regions (position 2). This result shows that the adhesive products on the die surface consist of two parts and suggests that the products showing darker contrast have lower electron conductivity than the other ones.15) Figure 5(c) shows a SEM image with the same magnification as Fig. 5 (a) and (b) for the flattened area of the GA surface after the sliding test. Microscopic grooves along the sliding direction are observed on the whole flat area. This suggests that the surface of the GA was abraded during the sliding test. We will discuss this feature below.
Plan view SEM images of die surface ((a) (b)) and GA surface (c) after the sliding test with sliding length of 130 mm. Primary electron energy is 1 keV. Arrows show the sliding direction for steel sheet. ET detector image (a) and In-lens detector image (b) were taken from the identical area.
The x-ray microanalysis was carried out using an energy dispersive x-ray spectrometer (EDS) with the primary electron energy of 5 keV and the results are shown in Fig. 6. Typical EDS spectra were measured for the adhesive products with dark contrast (position 1), brighter contrast (position 2) and die surface where the adhesive product can not be seen (position 3). The O, Al, S, Fe, and Zn are detected from position 1 whereas Fe and Zn are mainly detected from position 2. The latter datum suggests that the material at position 2 is the Fe–Zn intermetallics detected by the X-ray diffraction measurements mentioned above. Only Fe is detected except for a small C peak for the EDS spectrum measured at position 3. This demonstrates that the adhesive materials do not cover whole area of the die surface and exist on the dispersive positions along the sliding direction as suggested in Fig. 4. Adhesive characteristics clarified in this study continuously covered the whole die surface as the sliding length increased as shown by Al mapping in Fig. 3.
SEM image and EDS spectra measured for the areas shown in the image for die surface after the sliding test with sliding length of 130 mm. Primary electron energy is 1 keV for the imaging and 5 keV for the EDS measurements. The arrow in the image shows the sliding direction for steel sheet.
The cross sections of the die surface were fabricated for the SEM and TEM analysis using the FIB technique in order to specify the positions and avoid top surface damage induced by conventional polishing methods. A typical SEM image of the 45 degrees-box fabrication for the SEM observation is shown in Fig. 7(a). The cross sections were made along the sliding direction. Figure 7(b) shows backscattering electron (BSE) images of the cross section of the die surface. The images were observed from the direction normal to the cross sectional surface, 45 degrees from the die surface, therefore the vertical length is √2 times larger than the real one along the depth direction. The adhesive materials on the die surface consist of two parts, which is in good agreement with the plan view SEM observations (Figs. 5 and 6). One showing dark contrast in the BSE image is formed on the ridges of the die material (“1” in Fig. 7(b)). This part corresponds to the materials with dark contrast in Fig. 4(b). O, Al, Fe, and Zn and small amount of S were detected from this part by the EDS analysis. Other materials showing brighter contrast in the BSE image are located in the hollow parts of the die surface (“2” in Fig. 7(b)). Zn and Fe were detected from this part by the EDS analysis, which corresponds to the materials with brighter contrast in Fig. 4 (b) and is thought to be Fe–Zn alloys as discussed before. These characteristics were observed on the whole cross section of the die surface in Fig. 7(a). The following points should be noted for the adhesive materials “2”. These materials (1) seem not to be bonded to the die surface directly, but (2) connect to the former adhesive material “1” at the GA-incoming side of the material “1”. (3) Some lines which spread from the connecting point are observed on the cross section of the materials as illustrated in Fig. 7(b). These characteristics suggest that the materials were cut by the materials “1” and accumulated intermittently.
SEM images of the cross section for die surface after the sliding test with sliding length of 260 mm. The cross section was fabricated by FIB. Arrows show the sliding direction for steel sheet. (a) SE image observed by ET detector at low magnification including the fabricated area. (b) Back scattering electron image at high magnification with the primary electron energy of 5 keV. Lines shown in the adhesive material “2” are illustrated in Fig (b).
TEM observations were carried out for the cross section which was picked up from the die surface using the micro-sampling technique and the results are shown in Fig. 8. The TEM bright field image in Fig. 8(a) clearly shows the two types of the adhesive materials and their positions as mentioned in the previous section. The brighter regions “1” on the ridges correspond to the adhesive materials with darker contrasts Figs. 4(b) and 7(b) and the darker regions “2” in the hollows correspond to the materials with brighter contrast in Figs. 4(b) and 7(b). An high-magnification TEM bright field image and a line-analysis result using scanning transmission microscopy (STEM) mode for the adhesive layer on a ridge of the die surface are shown in Figs. 8(b) and 8(c), respectively. This adhesive layer has a very fine structure with heterogeneous contrast and contains O, Al, Fe and Zn as the main elements. We consider the layer is a fine Al oxides and Fe–Zn alloy composite. Some minor elements S, Ca, and Ba are also detected from the layer. This result suggests that ingredients in the oil used as a lubricant were mixed in with this layer. The TEM and STEM data also show that these adhesive layers are strongly bounded to the die surface at the ridge without any interfacial layers and are hard compared to Fe–Zn intermetallics. Another type of adhesive material at the hollow regions was confirmed to be Fe–Zn intermetallics by EDS analysis. Figure 8(a) clearly shows that this material is not adhering to the die surface on the large areas as mentioned previously.
TEM bright field images ((a), (b)) of the cross section for die surface after the sliding test with sliding length of 260 mm. The arrow in Fig (a) shows the sliding direction for steel sheet. Fig (b) is a high-magnification image of the area shown in Fig (a). (c) EDS line analysis results for a line indicated in Fig (a). Carbon was ignored in the quantitative calculation. (Online version in color.)
According to the results obtained in this study the adhesion model of GA on the die surface can be proposed as illustrated in Fig. 9. (1) Al oxide – Fe–Zn composite layers are formed at the ridge of the die surface. (2) Fe–Zn alloys are cut by the composite layer. (3) The cut alloys accumulate in the hollow of the die surface intermittently. This scenario is supported by the SEM image for the flattened GA surface (Fig. 5(c)) where the surface is covered by microscopic grooves along the sliding direction and interval of the grooves is of the same order to the interval of elongated adhesive materials (a few tens nanometers, Figs. 5(a) and 6).
Schematic drawing of die surface model during sliding test of GA. (Online version in color.)
It should be pointed out that the microscopic structure revealed in this study for the adhesive materials on the die surfaces is similar to the situation of metal cutting operations. The adhesive layer on the ridge of the die surface resembles the built-up edge growth on the tool during cutting. The adhesive materials located at the hollow side are worked materials cut by the built-up edge like adhesive layer. Although the formation mechanisms of the layer are still unclear, Al and O seem to play important roles in the formation process according to the microstructure of the layer (Fig. 8(b)). The SEM-EDS analysis with a primary energy of 5 keV was carried out for the GA surfaces before and after the sliding test. Al was not detected from the flattened GA surface after the test while the Al was detected from the GA surface before the test. Al may transfer to the die surface from the GA surface which is easily understandable because the surface oxide layer on GA contains Al. Our results show that the microscopic contacting area of the die surface is limited to the ridge regions. This means that real contact pressures at the contacting points are significantly higher than the expected ones from the normal load and the flattened GA surface area after the test. This high contact pressure may be another reason for the formation of the tied bonding adhesive layers. Wear mechanisms have been presented for Zn and Zn-based coating.16) In the case of 55%Al-Zn the particles are formed by mechanical milling and subsequent merging of smaller wear debris. Although the adhesion to the die was not discussed in this reference, the formation mechanism might be similar to this process.
The main factor of the friction resistance during the sliding has been usually considered to be shear force of adhesive materials. In the case of EG, the friction behavior is well explained by the shear force on Zn if the stable adhesive layer, or transfer layer, forms on the die surface.4) The scenario illustrated in Fig. 9 shows that the shear force of Fe–Zn intermetallics is not only a factor of friction resistance. Our model proposes that cutting or ploughing resistances of the Fe–Zn alloy on the GA surface by the build-up edge-like adhesive layers is the important factor contributing to the friction resistance. Both adhesive and abrasive (or ploughing) wear were reported for galvanized steel sheet (GI) from SEM observations for the GI surfaces after the sliding test.7,8) Miguel et al.7) mentioned that high apparatus pressure produces ploughing because of the grooves observed on the GI surfaces. The built-up edge-like adhesive layers possibly form on the die surfaces after sliding the GI in some sliding conditions because the GI also contains Al in the coating just as the GA does.
In the many previous researches on the friction behaviors of Zn-based coating steel, steel-sheet surfaces after the sliding tests have been mainly investigated although the importance of the transfer layer on the die surface was pointed out.4) Our study spotlights the die surfaces from different aspects using several microscopic techniques and shows new findings. The microscopic observations for the die surfaces should be carried out not for the steel-sheet surfaces only in order to understand the interaction between the die and other Zn-based coatings completely.
The latest microscopic investigations on die surfaces after a flat sliding test of GA provide new findings on the morphologies of adhesive materials. A model of the interaction between GA and die surfaces during the sliding was proposed from the results obtained in this study. Al oxides – Fe–Zn alloys composite layers form on the ridge of the die surface. The Fe–Zn intermetallics on the GA surfaces are cut by this layer and accumulate on the hollow of the die surface. This scenario is similar to metal cutting operations accompanied by the built-up edge. Our results also show that cutting resistances of the Fe–Zn alloy on the GA surface by the built-up edge-like adhesive layers are the important factor contributing to friction resistance. Further investigations are necessary to clarify the formation mechanism of the unique adhesion materials and their contribution to the friction properties of GA. It should be emphasized, finally, that our results point out again the importance of investigations on the die surfaces to clarify the fundamental phenomena occurring at the contact points.