Creep forming of metallic components by which a component such as an aluminium alloy plate is laid on a former and heated while the plate slowly takes up the form of the former is well known.

This technique suffers from the disadvantage that forming can take a long time, the tooling can be complex in shape to allow the correct profile to be formed and factors such as springback must be taken into account, requiring further processing steps and therefore can be uneconomic. Also the accuracy of formation can sometimes be inadequate, leading, for example, to the inability to produce large components of complex shape such as aluminium alloy wingskin panels, where errors in the shape formed accumulate over the length of the component to unacceptable levels above the required tolerances.

It is an object of the invention to provide an improved method of creep forming metallic components.

According to the invention there is provided a method of creep forming a metallic component including the steps of applying a static loading and a cyclic loading to the component during the creep forming thereof. It is preferred that the magnitude of the cyclic loading is much smaller than the magnitude of the static loading. The magnitude of the cyclic loading maybe less than or equal t 10% of the magnitude of the static loading, more preferably it may be less than 5%. In the experiments reported in this specification the magnitude of the cyclic loading is less than 2% of the magnitude of the static loading. Indeed it is less than 1%. A portion of the cyclic loading may be vibration.

The application of cyclic loading to components, artifacts, or structures as an additional energy source during creep forming is believed to accelerate permanent deformation experienced by the components and thus reduce springback. This is seen as a substantial advancement on the prior art.

Allowing large components to be creep formed without an unacceptable accumulation of error on the formed shape, tooling to be kept relatively simple and forming times to be reduced, thus making the process more economical.

Excitation may be applied both globally and locally to components, artifacts, or structures in either static or dynamic (adaptive) processing.

Methods of Application Application of cyclic loading/vibration may be made either to a localised component area, or to a whole component, depending upon size and specific forming requirements. Incremental application of the technique across a component will enable components of any dimension to be treated.

Components such as aircraft wing skins, stringers, spars, fuselage frames, fuselage panels etc. may be formed using this technique.

The technique is envisaged to be useful at any frequency over one cycle/hour.

Preferably frequencies of 20Hz-10, OOOHz are used.

In principal this technique may be used regardless of component material for example with steels, titanium or aluminium and titanium and aluminium alloys.

For some materials including the 2000,6000, 7000, and 8000 aluminium alloys cyclic loading can be applied as a supplement to conventional heating sources to increase strain retention during creep forming.

For materials where vibration/cyclic loading can be applied as a supplement to conventional heating sources to increase strain retention during creep forming this opens the exciting possibility of forming these materials to the end required shape without further precipitation hardening and thus negating the need for further heat treatment. Thus allowing the forming process to take place during the last stage of heat treatment.

There are six major advantages of this technique:- 1. An increase in strain retention during conventional or adaptive processing at normal forming temperatures over conventional heating when used alone.

2. The ability to creep form materials, including 2000, 6000, 7000 and 8000 aluminium alloys in the final heat treatment stage for example, 2024 aluminium alloy with a combination of conventional heating to relatively low temperatures (up to 100°C) with an additional vibration/cyclic loading.

3. Improved accuracy in final component shape as a direct result of the control over the vibration wave form within the artifact or component being processed.

4. Improved process speed at normal forming temperatures.

5. The ability to exercise forming zone control as a result of the local application of vibration/cyclic loading in thickened areas such as local reinforcements or integrally stiffened areas of components.

6. The ability to time curvature on large components such as large aircraft wing skins and stringers.

Equipment Portable excitation equipment is preferably used in the case of local application of the technique to a discrete area of a component. Otherwise bespoke equipment may be used for large components, for example aircraft wing skin panels.

The invention will now be further explained by way of example only with reference to the accompanying drawings of which:- Figure 1 is a graphical representation of the Stress Relaxation with Ageing Time results from displacement control tests carried out on 2024 T351 aluminium alloy for ten hours at 155°C plus or minus 5°C.

Figure 2 is a graphical representation of the Stress Relaxation with Ageing Time results from displacement control tests carried out on 7150 W51 aluminium alloy for ten hours at 155°C plus or minus 5°C.

Figure 3 is a graphical representation of the Creep Displacement with Ageing Time results from displacement control tests carried out on 7150 W51 aluminium alloy for ten hours at 155°C plus or minus 5°C.

Figure 4 is a graphical representation of the total Creep Ageing Displacement left after the tests represented in Figures 1-3.

Experimental Data There follows a description of the testing of bent-beam cyclic creep forming specimens for two aluminium alloys ; 2024 T351 and 7150 W51 and the subsequent results.

Experimental Method The tests involved the quantitative stressing of beam specimens by application of a four point bending stress using a servohydraulic cyclic Instron machine.

The applied stress was determined from the size of the specimen and the bending deflection. The stressed specimens were then exposed to a test temperature and a cyclic load of small amplitude applied. Displacements along the length of the beam specimens were then measured and reported.

The stresses and displacements were calculated using the following formula : amax=12xExtXyl (3xH2-4xA2) (1) Where: o : maximum tensile stress E: modulus of elasticity = 73 GPa t: thickness of specimen = 0.0035 m y: maximum displacement H: distance between outer supports = 0.180 m A: distance between inner and outer supports = 0.045 m

Displacement Controlled Tests The displacements shown in below in Tables 1 and 3 were kept constant during the tests. The loads were measured at given intervals. The permanent displacement left after the tests was measured along the length of the specimens.

2024 T351 Aluminium Allov The displacements applied to the specimens of 2024 T351 aluminium alloy for the stresses selected, as calculated by equation (1), were as follows : Table 1: Applied Displacements Stress (MPa) Max. Displacement (m) Cyclic Displacement +/_ m 230 0. 006684 232.5 (230+/-2. 5) 0. 006756 0. 000072 235 (230+/-5) 0. 0068300. 000146

The ladings applied were: <BR> <BR> 3xAxW<BR> #max = (2)<BR> b x d² Where: W: Load (MN) A: Distance between inner and outer supports b: Specimen width (0.05 m) d: Specimen thickness (0.0035 m) Table 2: Applied loads Stress (MPa) W load (KN) Cyclic load +/- (KN) 230 1. 044 232. 5 1. 055 0. 011 2351. 0660. 022

7150 W51 Aluminium Allov The Displacements applied to the specimens of 7150 W51 aluminium alloy and for the stresses selected, as calculated by equation (1), were as follows : Table 3: Applied Displacements Stress (MPa) Max. Displacements (m) Cyclic Displacement _ +I- (m) 350 0. 01017 352.5 (350+/-2. 5) 0. 010244 0. 000074 355 (350+/-5) 0. 0103165 0.0001465

The ladings applied as calculated by equation (2) were: Table 4: Applied loads Stress (MPa) W load (KN) Cyclic load +/- (KN) 350 1. 588 352. 5 (350+/-2. 5) 1. 6 0. 012 255 (350+/-5) 1. 611 0. 023 Forming time is defined as the time from the inception of the test until the required time has elapsed. The above tests began when the stressed specimen achieved the required temperature.

The Forming time was 10 hours (15 minutes). The temperature was 155°C 5°C.

Load Controlled Tests (7150 W51 Aluminium Allov) The load was maintained at 350 MPa in the static test and 350 +/-2.5 MPa or +/-5MPa in the tests with a small cyclic load. The specimens were then left to creep. The Forming time was 10 hours (15 minutes). The temperature was 155°C + 5°C.

RESULTS Displacement Controlled Tests 2024 T351 ALLOY Test 1-Static Load Only 10 hrs at 155 °C Table 5: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 105 0.142 0.16 0.102 After Test 0 0. 295 0.412 0.34 0.05

Table 6 Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1. 04 230 1 0. 75 165 2 0. 67 148 3 0. 64 141 4 0. 62 137 5 0. 62 137 6 0. 62 137

Test 2-Static Load plus +/-2. 5 MPa cyclic load-25 Hz 10hrsat155°C Table 7: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0.06 0.072 0.09 0.04 After Test 0 0. 285 0.405 0.34 0. 05 Table 8: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1. 04 230 1 0.78 172 2 0. 67 148 3 0. 64 141 4 0. 64 141 0.64 141 5 0.64 141 6 0.64 141

Test 3-Static Load plus +/-2. 5 MPa cyclic load-50 Hz 10 hrs at 155 °C Table 9: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 135 0. 155 0. 107-0. 05 After Test 0 0.36 0.463 0.32-0. 055 Table 10: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1. 043 230 1 0. 78 172 2 0.69 152 3 0.66 146 4 0.65 143 5 0.65 143 6 0. 65 143

Test 4-Static Load plus +/-5 MPa cyclic load-25 Hz 10 hrs at 155 °C Table 11 : Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 0125 0.0175-0. 03-0.04 After Test 0 0. 19 0.276 0.162-0. 06 Table 12: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1. 043 230 1 0.79 174 2 0.69 152 3 0. 64 141 4 0.63 139 5 0.63 139 6 0. 63 139

Test 5-Static Load plus +/-5 MPa cyclic load-50 Hz 10hrsat155°C Table 13 : Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 0225 0.02 0. 00-0. 05 After Test 0 0. 305 0.401 0.3-0. 05 Table 14: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1.042 230 1 0.83 183 2 0. 71 156 3 0.67 148 4 0. 65 143 5 0. 65 143 6 0. 65 143

REPEATED TEST Test 4A-Static Load plus +/-5 MPa cyclic load-25 Hz 10hrsat155°C Table 15 Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 03 0. 025 0. 04 0.02 After Test 0 0. 210 0.285 0.220 0.05 Table 16: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1. 046 230 1 0.81 179 2 0. 69 152 3 0.66 145 4 0. 64 141 5 0.64 141 6 0. 63 139 The Stress Relaxation with Ageing Time results from displacement control tests carried out on 2024 T351 aluminium alloy for ten hours at 155°C plus or minus 5°C are shown in a graphical representation in Figure 1.

7150 W51 ALLOY Test 1-Static Load Only 10hrsat155°C Table 17: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0-0. 015 0 0. 03 0.05 After Test 0 1. 465 2. 17 1. 645 0.16 Table 18: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1. 585 350 1 1.453 320 2 1.361 300 3 1.333 294 4 1.302 287 5 1. 288 284 6 1.287 284

Test 2-Static Load plus +/-2. 5 MPa cyclic load-25 Hz 10hrsat155°C Table 19: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 005 0-0. 02-0. 04 After Test 0 1.43 2.09 1. 55 0. 025 Table 20: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1.585 350 1 1.270 280 2 1. 155 255 3 1.097 242 4 1. 079 238 5 1.071 236 6 1. 060 234 7 1. 053 232 8 1. 047 231

Test 3-Static Load plus +/-5 MPa cyclic load-25 Hz 10hrsat155°C Table 21: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 13 0. 1 0. 05-0. 075 After Test 0 1. 40 2. 03 1. 485 0.01 Table 22: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1.589 350 1 1.166 257 2 0.972 214 3 0.912 201 4 0.872 192 5 0.851 188 6 0.837 184 7 0. 828 182 8 0. 825 182

Test 4-Static Load plus +/-2. 5 MPa cyclic load-50 Hz 10hrsat155°C Table 23: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 145 0.155 0.04-0. 085 After Test 0 1. 640 2. 32 1. 58-0. 05 Table 24: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1.554 350 1 1. 302 287 2 1. 141 251 3 1.110 245 4 1.088 240 5 1.070 236 6 1.058 233 7 1.050 231 8 1.045 230

Test 5-Static Load plus |/-5 MPa cyclic load-50 Hz 10 hrs at 155 °C Table 25: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 2 0. 28 0. 2 0 After Test 0 1. 22 1.81 1.35 0.08 Table 26: Load Relaxation Time (hrs) Load (KN) Stress (MPa) 0 1.587 350 1 1.116 246 2 1.027 226 3 0. 987 218 4 0.968 213 5 0. 962 212 6 0.958 211 7 0.948 209 8 0.942 208 The Stress Relaxation with Ageing Time results from displacement control tests carried out on 7150 W51 aluminium alloy for ten hours at 155°C plus or minus 5°C are shown in a graphical representation in Figure 2.

7150 W51 Aluminium Alloy Load Controlled Tests 7150 W51 ALLOY Load Control (Load 1.588 KN) Test 1-Static Load Only 10 hrs at 155 °C Table 27: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 42 0. 66 0. 57 0.06 After Test 0 2.95 4.14 3.19 0.26 Table 28: Displacement vs Time Time (hrs) Actuator Displacement position (mm) (mm) 0 1. 127 0 1-0. 504 1.631 1.5-0. 936 2.063 2-1.225 2.352 3-1.4012 2.528 4-1. 631 2.758 5-1. 852 2.979 6-2. 043 3.17 7-2.181 3.308 Note This specimen was initially loaded for 15 min at 155 °C to a stress of 230 MPa. The specimen was then unloaded, cooled to room temperature and the test re-started as above. Hence the initial curvature of 0.66 mm at the centre.

Load Control (Load 1.588 KN) Test 2-Static Load plus +/-2. 5 MPa cyclic load-25 Hz 10 hrs at 155 °C Table 29: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 33 0. 43 0. 25-0.03 After Test 0 2. 22 3. 14 2. 145-0. 04 Table 30: Displacement vs Time Time (hrs) Actuator Displacement position (mm) (mm) 0 1. 454 0 1 0. 416 1.038 2-0. 0673 1.521 3-0. 2665 1.72 4-0. 3625 1.816 5-0. 4431 1.897 6-0. 4413 1.895 7-0.5544 2.001

Load Control (Load 1.588 KN) Test 3-Static Load plus +/-5 MPa cyclic load-25 Hz 10 hrs at 155 °C Table 31: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 145 0. 2 0. 125 0.03 After Test _ 0 1. 65 2.305 1. 685 0. 09 Table 32: Displacement vs Time Time (hrs) Actuator Displacement position (mm) (mm) 0 3.331 0 1 2. 184 1.147 2 2.122 1.209 3 1.953 1.378 4 1.883 1.448 5 1. 809 1.552 6 1. 734 1.597 7 1. 690 1.641 8 1.673 1. 658 9 1.645 1. 686

Load Control (Load 1.588 KN) Test 4 (Repeat of Test 2) -Static Load plus +/-2. 5 MPa cyclic load-25 Hz<BR> 10 hrs at 155 °C Table 33: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 022 0.0725 0.095 0.095 After Test 0 1. 59 2. 29 1. 62 0. 115 Table 34: Displacement vs Time Time (hrs) Actuator Displacement position (mm) (mm) 0 2. 357 0 1 0.6209 1.736 2 0.086 2. 271 3-0.3268 2. 684 4-0.2934 2.65 5-0. 3097 2. 667 6-0. 4321 2. 789 7-0. 5198 2. 877

Load Control (Load 1.588 KN) Test 5 (Repeat of Test 3) -Static Load plus +/-5 MPa cyclic load-25 Hz 10hrsat155°C Table 35 : Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 055 0.085 0.095 0.075 After Test01. 00 1. 435 1.075 0. 05 Table 36: Displacement vs Time Time (hrs) Actuator Displacement position (mm) (mm) 0 2.475 0 1 0.8531 1.622 2 0.6984 1. 777 3 0.3650 2. 11 4-0. 3936 Note: The machine shut down after 4 hrs testing

Load Control (Load 1.588 KN) Test 6-Static Load plus l/-2. 5 MPa cyclic load-40 Hz 10 hrs at 155 °C Table 37: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 13 0. 17 0. 13-0.01 After Test 0 1. 565 2. 17 1. 55 0.03 Table 38: Displacement vs Time Time (hrs) Actuator Displacement position (mm) (mm) 0 2.061 0 1 0.7958 1.265 2 0.3590 1. 702 3 0. 1265 1. 935 4 0. 7396 1. 321 5 0. 6557 1. 405 6 0. 6336 1. 427 7 0. 6373 1. 424

Note: The frequency was 40 Hz (instead of 50 Hz) due to instability on the signal at 50 Hz.

Load Control (Load 1.588 KN) Test 7-Static Load plus +/-5 MPa cyclic load-40 Hz 10hrsat155°C Table 39: Specimen Displacements (mm) Distance from edge 45 90 125 170 215 (mm) (centre) Before Test 0 0. 07 0. 08 0. 025-0. 105 After Test 0 2. 7 3. 88 2. 83 0. 04

Table 40: Displacement vs Time Time (hrs) Actuator Displacement position (mm) (mm) 0 1.962 0 1-3. 094 5.056 2-3.247 5.209 3-3. 580 5.542 4-3.629 5.591 5-3.764 5. 726 6-3. 892 5. 854 7-3. 900 5. 862

Note: The frequency was 40 Hz (instead of 50 Hz) due to instability on the signal at 50 Hz.

Creep Displacement with Ageing Time results from displacement control tests carried out on 7150 W51 aluminium alloy for ten hours at 1550C plus or minus 50C are shown in a graphical representation in Figure 3.

The data showing the permanent displacement left at the end of all of the tests is shown graphically in Fig. 4.

Observations From the results of the displacement controlled tests outlined above and it was observed that there was very little influence of the small amplitude cyclic loading is observed on the age-forming of 2024 T351 Aluminium Alloy. This can be seen clearly from the graphical representation at Figure 1. The specimen subjected to static loading alone creep-aged at approximately the same rate as those that also included a small amplitude cyclic loading.

Conversely, the 7150 W51 Aluminium Alloy displacement controlled tests results showed a definite effect of the small cyclic loading on the creep-aged rate as illustrated in Figure 2. For example, the stress relaxed to a low value of 180 MPa from an initial stress of 350 MPa when the loading included small amplitude cycling (+/-5 MPa (25Hz)) while, with static loading alone, it relaxed from an initial stress of 350 MPa to only 284 MPa in the same time span.

From these results it also seems possible that the magnitude of the stress relaxation is dominated by the range of the small amplitude cycles, i. e. the +/-5 MPa tests showed a larger stress relaxation than the +/-2.5 MPa.

In the Load Controlled Tests indicate results showed that the specimen tested at +/-5 MPa and 40 Hz frequency showed a substantially higher creep elongation than the other specimens tested. Thus a combination of cyclic amplitude and high frequency seems to produce higher creep elongation.

The graphical representation of the permanent deflection left at the end of all of the tests as shown in Figure. 4, clearly shows that the 7150 W51 Aluminium Alloy retained a much higher permanent deflection than the 2024 T351 Aluminium Alloy after the Displacement Controlled tests.

The data from the load controlled tests showed a much higher scatter than that from the displacement controlled tests. It should be noted that the very low amplitude cyclic loading was very difficult to control because they were of the same magnitude as the noise, this can explain the larger scatter in the results of in this type of test.

Discussion 7150 Aluminium Alloy seems to age-creeps more readily than 2024 Aluminium Alloy. This may be the result of two factors. Firstly, the stresses applied to the 7150 W51 Aluminium Alloy were higher than those applied to the 2024 T351 Aluminium Alloy (350MPa against 230 MPa). Obviously, 7150 Aluminium Alloy being a stronger material than the 2024 Aluminum Alloy can be subjected to higher stresses. For example, the ratio of the yield stresses when fully aged are 1.63 while the ratios of the applied stresses in the tests was 1.52. Secondly, the 2024 Aluminium Alloy was already aged to a temper T351 while the 7150 Aluminium Alloy was not artificially aged prior to testing. This means that the 7150 Aluminium alloy material tested was initially softer than the 2024 Aluminium Alloy fully aged material tested, which may result in a higher creep rate than may be anticipated by the ratio of 1.52 mentioned above.

From these results the potential for a method of creep forming metallic components including a step of applying cyclic loading/vibration is very good.

Such a technique will enable large components to be creep formed economically, whilst maintaining the necessary accuracy and thus keeping the component within the required tolerance.

CLAIMS 1. A method of creep forming a metallic component including the steps of applying a static loading and a cyclic loading to the component during the creep forming thereof.

2. A method of creep forming a metallic component as in claim 1 wherein the magnitude of the cyclic loading is less than or equal to 10% of the magnitude of the static loading.

3. A method of creep forming a metallic component as in claim 2 wherein the magnitude of the cyclic loading is less than or equal to 5% of the magnitude of the static loading.

4. A method of creep forming a metallic component as in claim 3 wherein the magnitude of the cyclic loading is less than or equal to 2% of the magnitude of the static loading.

5. A method of creep forming a metallic component as in claim 4 wherein the magnitude of the cyclic loading is less than or equal to 1 % of the magnitude of the static loading.

6. A method of creep forming a metallic component as in any preceeding claim wherein a portion of the cyclic loading is vibration.

7. A method of creep forming a metallic component as in any preceding claim in which the cyclic loading has a frequency of 1 Hz to 1, OOOHz.

8. A method of creep forming a metallic component as in claim 7 wherein the cyclic loading has a frequency of 1 OHz to 100Hz.

9. A method of creep forming a metallic component as in claim 8 wherein the cyclic loading has a frequency of 20Hz to 50Hz.