The University of Nottingham School of Mechanical, Materials and manufacturing Engineering

Transcription

1 The University of Nottingham School of Mechanical, Materials and manufacturing Engineering Project Title: Introduction of a segmented Blank Holder to press forming of Carbon Fibre composites Student: Simon Adams ( ) Session Spring 2005 Supervisor Dr P Harrison and Prof A C Long Individual Project Report submitted for the Degree of Master of Engineering in Mechanical Engineering 0

2 Contents Page 1.0 Aims and objectives Existing Blank holder Design of segmented blank holder Experimental procedure Experimental results Experiment Experiment Experiment Conclusion Further work Acknowledgements Appendix 1. General arrangement CAD drawing Detailed component drawings Material specification Appendix 2. Progress report 1

3 1.0 Aims and objectives. The aims and objectives of this experimental project detailed in this report are defined below. To design and manufacture a new segmented blank holder. This will further develop the existing experimental work done in forming experimental carbon fibre hemispheres. The blank holder will be capable of producing asymmetric boundary conditions during forming. To produce asymmetric hemisphere moulds. To investigate the effect of tool temperature in the forming of hemispheres To explore the accuracy of simulated hemispheres compared to experimental results. 2

4 2.0 Existing Blank Holder The existing blank holder used for all previous experiments is shown below (Fig [2.1]) in the hydraulic rig (Fig [2.2]) used to form the hemispheres. It is constructed of two halves of sheet steel, which are clamped together with 4 springs, 1 spring in each corner. Fig [2.1] Existing blank holder, left, lower half, right, the upper half of blank holder. Fig [2.2] Shows hydraulic rig, without the blank holder in place. Note male tool above, female tool below. Fig [2.3] Existing Blank holder assembled, Note springs mounted in the corners. The carbon fibre blank is clamped, and held in position by both the weight of the upper half of the blank holder and also the force exerted by the springs. The clamping force exerted by these springs can be adjusted by tightening the wing nuts by hand. This causes the spring to compress and exert a greater force. Since; F = kx Where, k is the spring constant x is the displacement of the spring length 3

5 The normal force exerted on the carbon fibre blank by both the weight of the upper half of the blank holder and the force of the clamping springs creates a clamping pressure on the material. When the blank is press formed, this material is drawn into the female mould as the two moulds are pressed together. As the material is drawn in, the friction generated by the clamping force resists this movement, since; Friction = µr Where µ is the coefficient of friction r is the normal reaction force The clamping pressure exerted by the blank holder is uniform, i.e. the boundary conditions for hemisphere forming are symmetrical. Furthermore, the sheet steel of the upper half of the blank holder is thick enough to be considered to be effectively rigid, so that there is no localisation of pressure on the material by the force of the springs. 3.0 Design of a segmented Blank Holder In order to control and vary the boundary conditions under which hemispheres are formed then a new blank holder must be designed. Figure [3.1] below shows the layout of the design. 4 Hydraulic posts form the rig constrain overall size of blank holder 8 adjustable pads Fig [3.1] Plan view diagram of the 8-pad design concept Tool forms hemisphere here Force distribution plate The concept of the design is that the 8 pressure pads can be set up in a variety of combinations, e.g. number of pads, pressure at each pad and orientation of pads relative to one-another. The pads do not come in contact with the blank itself, they exert pressure on a thin sheet of aluminium. This sheet is labelled in Fig [3.1] as the force distribution plate. It is sufficiently thin to localise the force of each pad experienced by the blank. However, it is necessary to prevent the pads themselves from snagging and catching the carbon fibre as it is drawn into the tool. 4

6 Each of the 8 pressure pads is designed to be easily changed in the blank holder configuration. Figure [3.2] below shows the details of the pressure pads. Bolt moved up and down to adjust the force of spring Pad Force distribution plate Blank Base plate Fig [3.2] Left, External view diagram of pressure pad. Right, Section view diagram of pressure pad. Figures [3.1] and [3.2] are taken form the complete General Arrangement AutoCAD drawing. This can be found in the appendix of this report as well as detailed drawings for each component. The design was then manufactured to this specification as shown in figures [3.3] and [3.4], below. All 8 pressure pads in position, i.e. fully uniform pressure all around the blank. Male tool Left, Fig [3.3] Blank holder on bench, with blank clamped ready inside. Right, Fig [3.4] Blank holder installed in the hydraulic rig. 5

7 4.0 Experimental Procedure The basic procedure used for the production of each hemisphere is shown below. However, the experimental matrix in the appendix of this report must be read to fully realise the all parameters affecting the forming of each hemisphere. 1. Chemlease, a release agent is applied to the surface of both the male and female tool. This is done carefully with a wetted rag of cloth, using safety gloves. Several layers of Chemlease must be applied with time between layers for it to dry fully. It is important that the tooling in contact with the carbon fibres is as smooth and clean as possible. The Chemlease prevents the carbon fibre form sticking to the tool. 2. Remove the carbon fibre from the freezer in which it is stored (the prepreg carbon fibre begins slowly to cure at room temperature) and cut out a rough square, larger than the blank required. The material comes in roll form and care must be taken at this stage to ensure that the tows of the material are orthogonal to one another. This is in fact rarely the case and so the material must be sheared by hand back to 90º/0º. Fig [4.1] shows how this is done with the help of a paint marker pen, following the tows of the material carefully by eye. Fig [4.1] Left, material as it first arrives from the supplier, tows are not straight (indicated by the white paint dots-following the tows), need to be sheared back into orthogonal configuration, Right, shows corrected blank. 3. Once the blank is cut to shape (290mm square) then a 10mm square grid is drawn on with a 2.3mm white paint pen. This grid is drawn so that the lines follow the direction of the fibres. This is for later analysis as to how the material distorts as it forms. 4. The blank is then placed into the base of the blank holder (grid lines facing downwards so that distorted grid is on the outside of formed hemisphere) as shown below in Fig [4.2]. Fig [4.2] Blank placed into base of blank holder. 6

8 5. The force distribution plate is then placed on top of the blank, as shown below Fig [4.3]. Fig [4.3] Force distribution plate in position on top of the blank. 6. The top half of the blank holder is then placed in position and bolted securely in place into the 4 upright posts. As a result of this the pressure pads should now be in contact with the force distribution plate. This should now be checked to make sure that each pressure pad is in fact exerting force on the plate beneath it. Shown below, Fig [4.4]. Fig [4.4] The blank holder, ready to be lifted into the rig. 7. The blank holder is now lifted into the rig, the lower half of the holder is designed to locate accurately on the female tool. Fig [4.5] Blank now ready to be formed, female tool is driven hydraulically up, taking the blank holder with it, to meet the male tool above. 7

9 8. The two tools are brought together slowly until they meet and the blank is pressed into the mould. This is the start of the cure time. The cure time is one of the variables listed in the experimental matrix in the appendix of this report. The cure temperature is also a variable, several experiments were carried out using tools at room temperature, so that the blank was cold pressed. This means that the heaters installed in the tools were switched on only after the tools are brought together. 9. After the period of curing the hydraulic ram is turned off and the blank holder is lowered using the air retracts attached. The hemisphere formed may need a little persuasion to release form the male mould, care must be taken, and heatproof gauntlets must be used, as the blank holder will be very hot. Fig [4.6] shows the blank holder just having been removed form the rig. Fig [4.6] Left, The formed hemisphere, form the inside. Right, The same hemisphere, form the outside, note the distorted grid. 10. The tooling must be thoroughly checked and cleaned with the solvent Acetone and then several coats of Chemlease to maintain the smooth surface finish required for the next hemisphere. 11. The completed hemisphere can then be released form the blank holder, by undoing the 4 bolts. Care must be taken since the metal surfaces will remain hot to the touch for approx. 30 minutes. Fig [4.7] shows a completed hemisphere. Fig [4.7] Completed hemisphere mould 8

10 5.0 Experimental results In total, 18 hemisphere moulds were formed during the course of the project, under various conditions and criteria. Not all of the hemispheres are discussed in the main body of this report, however details of the forming conditions of all experiments are included in the experimental matrix of this report. There are 3 main experiments discussed in this report, they are as follows; 1. Creating asymmetric hemisphere moulds. 2. Investigate the effects of cold tool forming. 3. To explore the accuracy of simulated hemispheres compared to experimental results. Each hemisphere formed was numbered in sequence as the forming conditions were recorded, the numbering starts with H0 (Hemisphere 0), H1, H2, H3, etc until H19. See logbook in appendix for details. Full specification of the material used for all experiments can be found in the manufacturer s data sheet in the appendix of this report. 9

11 5.1 Experiment 1. Asymmetric boundary conditions. In order to discover how the new blank holder would work, the first series of experiments were carried out in the form of control experiments, i.e. Attempt to create a symmetrical hemisphere similar to those created by the previous blank holder. This control hemisphere can then be compared with asymmetric cases later. The segmented blank holder was set up as shown in Fig [5.11]. All 8 pressure pads were used. Forming conditions of Hemisphere 5 (H5) Fig [5.11] Plan view diagram of the blank holder set up with uniform boundary conditions to create a symmetric hemisphere. Pressure pad Blank Force distribution plate H5 Forming conditions: Tool forming temp (ºC) 120 Curing temp (ºC) 120 Curing time (mins) 30 Fig [5.12] Shows H5. Left, Plan view of formed hemisphere, note the equal sizing of the 4 quadrants, indicating good symmetry. Right, Fig [5.13] Shows H5, Note high shear angle at the base of the hemisphere, lessening towards the tip. 10

12 Graphical analysis of Hemisphere 5 Before each mould is pressed, a 10mm square grid is drawn onto the blank, this grid is aligned with the tows of the material. As the material distorts during pressing, so does the grid, this is used to determine the shear characteristics of the material. Shown below in Fig [] is a graphical representation of the shear behaviour of H5. Graph to show how grid shear angle varies with distance from the apex of Hemisphere 5 (H5). Shear angle (Degrees) Distance fom the Apex (Number of 10mm grid squares) Fig [5.14] shows the shear behaviour of This graph is made from the careful measurement of the grid squares by hand and then converted into a shear angle. A fully orthogonal grid square has a shear angle of 0. It can be seen from the above graph that the shear angle builds steadily from the apex of the hemisphere, and then falls away at the edge of the mould. This can be seen clearly in the photo Fig [5.13]. 11

13 Since H5 shows that it is possible to form a good, symmetric hemisphere by setting up uniform boundary conditions, the following experiments seek to explore the possibility of a highly asymmetric hemisphere mould. Fig [5.13] shows the set-up of the blank holder used to do this. Forming conditions of Hemisphere 9 (H9) Fig [5.15] Plan view diagram of the blank holder set up with highly asymmetric boundary conditions. Only 2 pads are used, in opposing CHANGE corners of the blank holder. Pressure pads H9 Forming conditions: Tool forming temp (ºC) 120 Curing temp (ºC) 120 Curing time (mins) 30 Note the high shear angle (Diamond shape) of the grid squares at the hemisphere apex. Fig [5.16] Plan view of H9, Note massively asymmetric quadrants 2 and 4, due to the positioning of the pressure pads as shown in Fig [5.13] above. Fig [5.17] Side view of H9, showing tow waviness and minor wrinkling between quadrants. 12

14 Graphical analysis of Hemisphere 9 Since H9 is highly asymmetric, it cannot be analysed like that H5, since the quadrants are so different. Therefore, 2 quadrants of H9 are analysed below. Quadrant 1 Quadrant 2 Figure [5.18] The two quadrants, 1 and 2 analysed in graphical form below. 60 Graph to show how grid shear angle varies with distance from the apex of Quadrant 1 of Hemisphere 9 (H9) Shear angle (Degrees) Distance from the apex (Number of 10mm grid squares) Fig [5.19] Shear angle analysis of quadrant 1, hemisphere 9. Graph to show how grid shear angle varies with distance form the apex of Quadrant 2 of Hemisphere 9 (H9) 60 Grid shear angle (Degrees) Distance from the apex (Numberof 10mm grid squares) Fig [5.19a] Shear angle analysis of quadrant 2, hemisphere 9 13

15 Further asymmetry experiments-varying the thickness of the force distribution plate The asymmetry seen in H9 (Fig [5.14]) was developed by the use of localised pressure, the following experiment H17, explores the extent to which the force of the pressure pads can be localised. This is done by doubling the thickness of the force distribution plate and in doing so making it more rigid. This will result in less bending and a more uniform distribution of pressure on the blank. Forming conditions of Hemisphere 17 (H17) The experimental setup of the blank holder is the same as that detailed in Fig [5.13] in the forming of H9, 2 pads, each in opposite corners of the blank. However, the only difference being the thickness of the force distribution plate which is doubled from 1mm to 2mm thickness. Fig [5.19b] H17, Note that compared to H9 in Fig [5.16], H17 is less asymmetrical. The quadrants are, more equal in size than those of H9. The grid squares at the apex have a lower average shear angle than H9. Fig [5.19c] Side view of H17, shows waviness in formation of tows, but no wrinkling compared to H9 in Fig [5.17] 14

16 5.2 Experiment 2. Forming with cold tools The following 2 experimental hemisphere moulds were produced using cold tools. This means that the tools were brought together at room temperature (19ºC). Then the heaters installed in both tools are switched on and the hemisphere is then cured under constant temperature and pressure. Forming conditions of Hemisphere 14 (H14) H14 was formed with the same uniform boundary conditions as that of H5, detailed in Fig [5.11]. The blank was left to cure for 60 minutes, longer than previous experiments, since more time is required to bring the tools up to heat from room temperature. Fig [5.21] Plan view of H14. Hemisphere has good symmetry in 4 quadrants. Apex grid shows orthogonal tows. Some fibre pull-out at edges. Fig [5.22] H14 compared to the control hemisphere H5, below Fig [5.23]. H14 shows high degrees of tow waviness compared to smooth tow path of H5. H14 also shows fibre pull-out at the edges between the quadrants. Fig [5.23] Side view of control hemisphere H5. 15

17 Forming with cold tools and asymmetric boundary conditions The following experimental hemisphere is a repeat of the asymmetric boundary conditions used in experiment [1. (H9)] described in diagram Fig [5.13]. The following Photos compare the two asymmetric hemispheres H9 and H13. H9 formed with hot tools (120ºC) and H13 formed under the same boundary conditions, but with cold tools (19ºC, room temperature). Tow waviness across apex Fig [5.24] H13 formed with cold tools, although highly asymmetric, H13 has not got the quadrant elongation of H9 (Right), instead, severe wrinkling in areas of low in-plane tension-quadrants 1 and 3. Note also the waviness of the tows across the apex. Fig [5.25] H9 formed with hot tools. Compare with the same boundary conditions as H13 (Left). See experiment 1. Fig [5.26] H13, shows severe wrinkling of quadrant 1. Note that material has completely overlapped in places. Some fibre pull-out at edges. Fig [5.27] H13, further detail of wrinkling from the inside of the hemisphere. 16

18 5.3 Experiment 3. Exploring the accuracy of simulated hemispheres compared to experimental results. It is possible to use computer simulations to develop an understanding of how a material is going to behave when it is formed in real life. One such simulation is DrapeIt, an energy based mapping code, developed at The University of Nottingham. By using this program it is possible to determine data such as the shape of an ideal blank shape. The concept of an ideal blank shape is one that, when cut will form the hemisphere perfectly and use a minimum of material in the process, i.e. minimal wastage. Below, Fig [5.31] shows the modelled hemisphere. Fig [5.31] Computer simulation DrapeIt, showing modelled hemisphere. Red and orange grid squares are areas of highest shear angle, blue corresponds to areas of least shear. Using this simulation, it is possible to unwrap the hemisphere simulated above and so determine its net shape. This net is then the shape of the ideal blank. Model unwrapped to find blank shape. Blank outline Fig [5.32] Left: overhead view of simulated hemisphere. Right: Unwrapped blank, note 10mm square grid used, as in all experiments 17

19 Once the blank outline was known, it was possible to export this outline as a dxf. AutoCAD file, this could then be plotted onto a paper template, and this used as shown in Fig [5.33] below. Fig [5.33] Ideal blank shape marked out on a 10mm square grid. Forming conditions of Hemisphere 18 (H18) The blank shown above was used in the blank holder under fully uniform boundary conditions detailed in Fig [5.11]. The hemisphere was formed with cold tools (18ºC, room temperature). The result is shown below in Fig [5.34] and Fig [5.35] Fig [5.34] H18, formed from ideal blank shape, very square apex grid. No obvious waviness. Footprint slightly ovalised. Fig [5.35] H18, hemisphere appears to be closely matched to simulated shape shown in Fig [5.31]. 18

20 6.0 Conclusions The following section refers to the findings of the experimental work described in the previous section of this report. Objective 1. To design and manufacture a new segmented blank holder. This will further develop the existing experimental work done in forming experimental carbon fibre hemispheres. The blank holder will be capable of producing asymmetric boundary conditions during forming. The blank holder developed and described in this report (see appendix for full technical drawings) was a success. It completed the design criteria set out at the beginning. It produced symmetric hemispheres when the boundary conditions were set to be symmetric, this provide the project with a control experiment with which to compare the subsequent experiments. From the results of this project, it is evident that the blank holder has created a variety of highly asymmetric hemisphere moulds, thus fulfilling the objective it was designed for. Furthermore, the blank holder has a great deal of potential for further experiments in the future. Objective 2. To produce asymmetric hemisphere moulds. The blank holder has produced asymmetric hemispheres, H9 and H17 are detailed in this report. H17 supports the hypothesis that an increase in the thickness of the force distribution plate (i.e. less localised pressure concentration) results in a lessening of asymmetry in the hemisphere mould produced. H17 has lower shear angles on the grid squares of the apex compared to those of H9 and fewer signs of the material wrinkling between quadrants at the edges. This is because a thicker force distribution plate provides a more consistent level of clamping pressure on the blank. This results in a more even distribution of in-plane tension in the fibres of the material, meaning that the material deforms in a more symmetrical way when moulded by the tools. Objective 3. To investigate the effect of tool temperature in the forming of hemispheres In experiment 2., the effects of cold tooling are investigated, with both uniform and asymmetric boundary conditions. Hemisphere 14 (H14) was formed under uniform boundary conditions, the result was minor wrinkling at the base of each quadrant as well as tow waviness. Furthermore, there was evidence of fibre pull-out at the edges of the blank. Hemisphere 13 (H13) suggests that when forming with cold tools under asymmetric boundary conditions that asymmetry is reduced. This is because the effective shear modulus of the prepreg material is much greater at lower temperatures, when the tooling is cold, the effectiveness of the clamping force of the blank holder is reduced. This is because the tool in contact with the material does not lower the shear modulus of the material it is in contact with due to the fact it is cold, it does not reduce the viscosity of the matrix, therefore, instead of stretching this material, the material remains relatively resilient and is instead simply drawn into the mould itself. This results on the blank being drawn into the mould with reasonable symmetry, as shown in the photo (Fig [5.24]). 19

21 Objective 4. To explore the accuracy of simulated hemispheres compared to experimental results. In experiment 4., the simulated hemisphere shape gave a good approximation to the resulting experimental hemisphere. This was particularly successful in the understanding the correct shape to which the blank should be cut too in order to produce a round shaped footprint, which is a key factor in the minimisation of material wastage in forming. Indeed further detailed analysis with the software package Camsys would form part of further work in this area. 7.0 Further work New segmented blank holder provides considerable scope for further experimental work in the area of asymmetric boundary conditions. The use of pressure pads and a thin sheet of material like the force distribution plate in this project in more complex mouldings than a simple hemisphere, could uncover exciting new phenomena. For example moulding more complex shapes such as a helicopter pilot s helmet, where the use of a varied clamping pressure may lead to a betterfinished item with better use of material properties. Further analysis of formed hemispheres using the software package Camsys. Although time consuming, this will enable better, quantifiable analysis of each hemisphere by computer analysis of the grid drawn on the surface of the mould. Further the work detailed in Hemisphere forming experiment report J. Wang, November By using the segmented tool to experiment with changing the blank holder force used. Undergo work with the segmented blank holder using a different type of material to that used in this report to observe phenomena and attempt to optimise is formation as well as assessing computer based models of the material under various forming conditions. 8.0 Acknowledgements The author acknowledges the help and enthusiasm of Dr Phil Harrison the project supervisor, as well as Prof A.C. Long, the project coordinator. The help from senior research fellow Jo Wiggers in the use of DrapeIt was gratefully received. Thanks also for the endless help and cooperation of Lab technicians Paul, Roger and Geoff, who assisted with all elements of practical work, also Dave Oakland for the manufacture of the segmented blank holder. The work was funded by the University of Nottingham, for which the author is most grateful. 20