Stretch Forming Analysis for Skin of Aircraft, Model. B717-200 Paper Reference Number: 2001-89 Myung-Hyun. Jung Aerospace Div. Korean Air co, Korea. Tel: 82-01-051-970-5639 E-mail: jumyhy@dreamx.net Abstract This paper represents three-dimensional finite element simulation of stretch forming process in manufacturing of skins of aircraft model B717-200, using MSC.Marc program. The aim of the analysis is to determine the possibility of the stretch forming of which stretching direction is toward longitude due to the limit of the capacity of the stretch forming machine in our plant. In addition to, the aim of it lies in the determination of the optimized stretch-forming process avoiding a failure and minimizing an amount of a spring-back. The analysis shows where the highest stress is concentrated on, how much the adjustment of the amount of the stretching force and the moving die and the alternation of the order between those have influenced on formability, and gives the history of changing stress over whole process,. Results demonstrate that the finite element simulation is capable of predicting the formability in longitudinal stretch-forming. Through the stretch forming analysis using MSC.Marc program, not only the practical stretch forming process can be developed in detail but also the trial and error can be reduced. 1
1 Introduction Most of the aircraft s skins have a long length and a smooth contour with large radius. These kinds of parts are formed by the roll forming or the stretch forming process. Generally skins having simple curvature are formed by the roll forming process but those having compound curvature are formed by the stretch forming process. The analysis of the stretch forming on this paper is concerned with the case of the longitudinal stretch forming of which direction is turned 90 degrees from transverse stretch forming direction which is the normal stretching direction. Due to a limit of width of stretch forming machine s jaw, some skins of aircraft which exceed the capacity of stretch forming machine cannot be stretched by the normal transverse-stretching method. For this reason, longitudinal stretch-forming method can be considered to form the workpiece of which width is exceeds the size of the machine s jaw but its longitudial width is not. In order to find out the possibility of the stretch forming for the skin of aircraft model b717-200 when longitudinal stretch-forming method is applied, the stretch forming analysis using MSC.Marc program is conducted before the above longi tudinal stretch-forming method in practice is progressed 2 Problem definition If a workpiece having large curvature is formed by one step simple bending, a large amount of springback is occurred due to material elasticity. To get rid of springback and reach to range of plasticity, as one of the forming technique among various kinds of those for it, a stretch-forming method can be considered. Stretch-forming is a forming process that edges of workpiece are clamped by a jaw and stretched by moving die upward and pulling it. Stretch-forming machine is shown on Fig.3. As you see, the forming capability of the machine depends on the size of clamping jaw. The maximum width of workpiece which can be formed by the machine in korean air company is 72 inch identical with the width of the jaw. Width of a part which exceeds the size cannot be stretched by our own machine with the normal transverse stretch-forming method. 5910343-19 Skin and 5911409-239 Skin shown on this paper are parts in the such case. The transversal length of these Skin exceeds the capacity of the stretch forming machine in our plant, We tried to seek a certain stretching technique for these parts using our own stretch forming machine in order to not only save production cost but also utilize our facilities. Generally, normal stretching direction is toward transverse of the part, but in order to avoid the limit of width of the jaw, the longitudinal stretching direction can be considered. In this case, the longitudinal stretching method requires greater amount of stretching than it of transverse, because the workpiece is stretched in worse condition in which the contour of workpiece cannot be formed easily. Therefore, the longitudinal stretching method may be apt to cause fracture, in a while, if, to avoid fracture, the amount of stretching is lesser, it will bring about a large spring back. Under this situation, it is difficult to decide whether the longitudinal stretch-forming can be of success or not without actual forming test. There are several problems which makes the progress of actual forming test difficult. The price of raw material for 5910343-19 skin and 5911409-239 skin is higher than it of others because it is special size which is larger than standard production size of raw material. Moreo ver, the stretch forming die would be very huge. If the longitudinal stretch-forming method is failed, the loss of economic cost will be severe. For this reason, finite element analysis of longitudinal stretch-forming is eagerly needed. Fig.1 typical stretch forming(left), a shape of aircraft s skin(right) 2
Fig.2 Stretch forming machine(left), movement of die and jaw(right) 3 Analysis Stretch forming Analysis for 5910343-19 Skin In order to figure out the possibility of the longitudinal stretch forming for 5911409-239 Skin of aircraft model b717-200, the stretch forming analysis using MSC.Marc program is conducted before the longitudinal stretching method is progressed in actual stretch-forming. Prior to the FE analysis of the stretch forming for 5911409-239 Skin, the analysis for 5910343-19 Skin was performed in parallel with the practical stretch forming. Fig.3(left) shows the finite element model of which boundary conditions and contact bodies are established. Material properties are used those of 2024- T3 heat-treated aluminum alloy,.050 thickness and strain-stress curve of it shown on Fig.5(right). The surface of contact bodies is imported to Mentat as a format of IGES File from CATIA data file. The finite element is modeled only a quarter of the full size to reduce computation time. Fig. 3 FE modeling, contact bodies and boundary conditions(left), strain-stress curve 2024-T3(right) The stretch forming machine in our plant is capable of angling the jaw up to 18. The below Fig.4 shows the jaw angled 18. It seems that using a rounded jaw is more advantageous than using an angled jaw in the forming condition. To have the machine get a rounded jaw, a costly machine modification for it is needed. In order to figure out whether the modification of our machine is needed or not, the finite element simulation comparing with forming conditions of two cases is conducted. Results of the analysis of two cases represent that the contact condition using an angled jaw is not so good as that of a rounded jaw, but the highest stress value in angled jaw is lower than that of the rounded jaw. In the case of using an angled jaw, it has a margin of more forming a workpiece to make better the contact condition. Being stretched more in using an angled jaw so as to be the same the contact condition as it of the case using a rounded jaw, the highest stress in the case of using an angled jaw is equivalent to that of using a rounded jaw. Therefore, it is concluded that machine modification which have it get a rounded jaw is not needed. The following analysis is about forming result in the case of using the angled jaw. 3
Fig. 4 The shape of the angled jaw and the rounded jaw In the next, the effect of the alternation of the order of movement between jaw and die is investigated. It is remarkable that the highest stress value is being reduced gradually in proportion to the moving distance of jaw after die stopped going up. This means that, to stretch-form it to the utmost, a tension by moving jaw plays a more important role than by ascending die. The stress history plot of the node having the highest stress is shown on fig.5. The actual stretch forming test for 5910343-19 Skin using an angled jaw have been failed in several times before a workpiece was stretched enough to form, due to stress concentrated on the extreme ends of workpiece. As far as other stretch forming technique which can disperse the concentration of the stress on the extreme end is not there, it seems that the longitudinal stretch-forming method cannot be succeeded. Even though a rounded jaw is used, it is expected that the result would be the same as the above result because the highest stress value is higher than that of the angled jaw. As the result, it become known that other stretch -forming technique which can makes the highest stress value low is needed to form it without failure. Fig. 5 History of the highest stress in progress, (left)the case of moving die and tension simultaneously, (right) the case of applying only additional tension after moving die and tension simultaneously It seems that, in stretch-forming using a rectangular shaped raw material, it is difficult to stretch-form it completely without a failure because the stress on the extreme ends of workpiece exceeds the limit of its ultimate stress. As a means of the other methods, the shape of a tensile test specimen can be considered, that is, the both sides of a raw material is cut off and made it to the shape of a tensile test specimen shown in fig.6. The following Fig.7 is the analysis of the stretch forming with the workpiece shaped a tensile test specimen. The analysis of two cases that one is using raw material shaped rectangular, the other is using workpiece shaped tensile test specimen is conducted. Fig.7 shows a deformed shape of a workpiece shaped tensile test specimen. As you see on fig7, the highest von-mises stress value in the case of the 4
workpiece shaped a tensile test specimen is 73,610 psi. But, in the case of the workpiece shaped rectangular, its value is 82,260 psi. Fig.7 is shows the stress history plot of specified nodes. The highest von-mises stress value in the case of using the workpiece shaped a tensile test specimen is lower than that of using the workpiece shaped rectangular. Also, the analysis demonstrates that it is possible to make the higher stress value low to the utmost by adjusting the location of the cutout area. Fig. 6 The shape of material: (left) rectangular type, (right) tensile specimen type 1 Node 66 958 Fig. 7 Stress of the formed material shaped tensile specimen(left) and history of stress of specific nodes(right) In the analysis of t he stretch forming with the workpiece shaped a tensile test specimen, it is investigated how much the adjustment of an amount of the moving distance between jaw and die and the alternation of the order of those have influenced on formability. Below 4 cases is result of the analysis referred to their influence. If the contact condition in all cases come to be the same as the other s, the stress of each cases come to be different from each other. Table.1 shows a detail process of the analysis, including the number of increment and highest stress value in each cases at the position in which the contact condition is best. Results show that a sufficient amount of tension is more contributing to lowering the highest stress and make it possible to form further than that of moving die. Table.1 result of analysis of several cases which different process applied respectively Increment Detail process Inc of best Highest contact cond stress Case1 0~40 Create angle18 41~100 Up 6, stretch.6, simult Inc.92 74,000 psi 0~40 Create angle18 Case2 41~60 Stretch.2 Inc.96 63,000 psi 61~120 Up 6, stretch.6, simult Case3 Case4 0~40 41~80 61~130 0~40 41~80 61~130 Create angle18 Stretch.4 Up 6, stretch.6, simult Create angle18 Stretch.4 Up 6, stretch.6, simult Inc.112 60,000 psi Mat l shape Tensile specimen Tensile specimen Tensile specimen Inc.112 70,000 psi Rectangular 5
The below fig.9 shows the change of the thickness in the case 3. An area of the end and of the cutout become thinning and sharing the stress. the thinnest area Fig. 8 Variation of thickness of material shaped tensile specimen Stretch forming Analysis for 5911409-239 Skin The size of 5911409-239 skin is wider and longer than it of 5910343-19 Skin, So it is more difficult to form 5911409-239 Skin than 5910343-19 Skin. The reason why the analysis of 5910343-19 Skin in parallel with 5911409-239 skin performed is that it is possible to verify the analysis for only 5910343-19 Skin through the experimental stage which can be progressed in our plant immediately due to the raw material on hand. As explained in the analysis of stretch forming for 5910343-19 Skin, there is no difference in the formability between two cases of using an angled jaw and a rounded jaw, but it is also necessary to confirm whether, in 5911409-239 skin, results of two cases are the same or not, because the size of 5911409-239 skin is wider than it of 5910343-19 Skin. The analysis comparing with two cases of using different jaws in stretch -forming for 5911409-239 Skin is conducted too. Fig.9 shows the contact condition and the von -mises stress in the increment 210 of the two cases of which load history is the same. The von-mises stress in the case of using the rounded jaw is 76,230, the other s is 70,630. The result of the analysis of the two cases represents that the contact condition in which a angled jaw used is not so good as that of a rounded jaw but the highest stress is lower than that of the rounded jaw. In the case of using the angled jaw, there is a margin of more stretching to make the contact condition better because the highest stress of the case using an angled jaw is still lower than that of using a rounded jaw. As the result, it is concluded that the machine modification to have a rounded jaw is not needed in stretchforming for 5911409-239 skin too. Fig. 9 Contact status at increment 210: (left) using the angled jaw, (right) using the rounded jaw 6
Fig. 10 Highest stress at increment 210: (left) using the angled jaw, (right) using the rounded jaw As you see Fig.10, The stress in the case of using a angled jaw reached at the limit of the ultimate stress and its contact condition is not so good as satisfied. Being more stretched to get a better contact condition, the workpiece would be failed before it is stretched sufficiently. Through the analysis, it become known that other method which can be disperse the stress concentrated on the extreme end of workpiece is needed to form it successfully. A workpiece shaped a tensile test specimen as a means of other methods can also be considered. The following is the analysis of the stretch forming with the workpiece shaped a tensile test specimen. Fig.11 shows the stress in the deformed condition of a workpiece shaped rectangular and that of the a workpiece shaped a tensile test specimen at the same increment. As you see in figures, the highest von-mises stress in the case of using the workpiece shaped a tensile test specimen is 79,500 psi and the other s is 71,210 psi. Accordingly, the highest von-mises stress in the case of using the workpiece shaped a tensile test specimen can be low because the highest stress on the extreme end is dispersed to the area of the cutout. Fig. 11 stress on the end of material: (left) rectangular shape, (right) tensile specimen shape In the analysis of the stretch forming with the workpiece shaped a tensile test specimen for 5911409-239 Skin, it is investigated how much the adjustment of an amount of the moving distance between jaw and die and the alternation of the order of those have influenced on formability. The process of the analysis is 7
as below: Create angle 18 (Inc.0~50) stretch 0.6 (Inc.51~110) stretch 1.4 and moving up 7 simultaneously (Inc.111~250) stretch 1.0 (Inc.251~350) The more the forming progress, the higher the stress on ends is. So, it is impossible to form further because the stress reached at a ultimate stress before it is formed sufficiently. But, when only tension without die s movement is applied from the increment 250, not only the contact condition is improved but the highest stress is not increased. Contact conditions at the increment 250 and 350 are shown on fig.12. Compared with the contact condition at the increment 350 with that at the increment 250, it shows that of increment 350 is much better improved. Fig. 12 Contact condition: (left) at increment 250, (right) at increment 350 When both tension and ascending die are applied simultaneously, the stress history plot of the node having the highest stress is shown on fig.13 (left), and it, when only a tensile force without ascending die is added from the increment 250, is shown on fig.13 (right). The result shows that, when a tensile force and moving die are applied simultaneously, the stress on extreme ends is increasing continuously in proportion to the movement of jaw and die. However, when only a tensile force without a moving die is applied after then, the increase of the stress stops. In a while, a stress of other area is increased, this phenomenon means that post-tension after moving die makes workpiece stretch-form more and reduce the amount of the springback. Fig. 13 History of the stress in progress, (left) the case of moving die and tension simultaneously, (right) the case of additional tension only after moving die and tension simultaneously 8
Fig.14 shows that, at inc.250, the stress value in the area where a spring back predicted is 35,000 psi, but, at inc.350, the stress value is 50,000 psi. Only a tensile force without moving die is applied from inc.250 to inc.350. Compared to results, it is concluded that an amount of springback can be reduced by only an additional tension at the position after which shows a contact condition good. Fig. 14 Stress in which springback predicted Experimental result of actual stretch forming for 5910343-19 Skin The actual stretch forming test for 5910343-19 Skin based on the analysis of stretch forming was performed. Following Figures are photographs of workpiece being processed in the experimental stretchforming. The workpiece formed 18 and stretched by the jaw is shown on fig.15. Fig.16 shows the appearance of workpiece stretched to the point of the ultimate strength. Having been stretched more, the workpiece have failed. A crack was started from the edge clamped by the jaw and led to a failure. A separated piece of workpiece in the jaw and the other on the die is shown fig.17. When the workpiece was formed at its maximum, total amount of going die up is 2.8 and distance of stretching is 0.9. Though springback occurred in the predicted area, the formed condition was so good that the formed workpiece could be touched with die by a finger force. There has been no deviation between the result in experimental forming and in the simulation using MSC.Marc program. The workpiece clamped by the jaw is shown on fig.15. Fig. 15 Appearance of the material before bended 18 9
The workpiece bended 18 and stretched at its maximum is shown on fig.16. Fig. 16 Aspect of the formed workpiece stretched at its maximum The workpiece failed by stretching beyond the limit of ultimate stress is shown on fig.17. It shows that there is a piece of the workpiece clamped inside of the jaw, the other is placed on the die. The crack is developed from an extreme end to a center of the workpiece. Fig. 17 Appearance of the failed workpiece 10
4 Discussions There have been several failures in stretch-forming test for 5910343-19 Skin. At that time, it has been not known from where the crack on workpiece originated. So, proper action which makes workpiece form at its maximum could not be found out. Through the finite element analysis using MSC.Marc program, it is possible to make out the magnitude of stress in whole area of workpiece and the area where a crack come into being. The failure were caused by concentration of stress of the end due to clamping and not having a optimized process that combines the amount of tension with moving die for the best formability. Accordingly, the proper technique under such circumstance could be adopted and the test could be succeeded. 5 Conclusions Based upon the present investigation, the following conclusions can be drawn. First of all, it is revealed that Skins can be formed in our plant without modification of machine by adopting longitudinal stretchforming direction. The second, the method which can make the material form at its maximum without failure is that only tensile force applies at the position where both end sides of workpiece is brought tangentially into contact with the tool. The third, if a method reducing the concentration of stress in edges is adopted, the formability would be better than now. These conclusions are verified and the optimized process is confirmed by experiment of 5910343-19 Skin. Accordingly, the analysis of 5911409-239 Skin is a help to a success in actual stretch forming. In sheet metal forming analysis, MSC.Marc program provides not only accurate analysis results but also multiple functions which make various forming techniques apply. Acknowledgement The author wish to express thanks to our members who have performed experimental stretch forming test based on the FE analysis and Dr. Sung-hoon Kim, in MSC.Software Corporation, who has given technical assistance in the sheet metal forming analysis related to solving our various practical problem. Reference 1. Kurt Lange, professor, University of Stuttgart, HandBook of Metal Forming McGraw-Hill Book Company. (1975) 2. Charles Wick. CmfgE, John T. Benedict, Raymond F. Veilleux. Tool and Manufacturing Engineers Handbook, Vol.II. Forming, Society of Manufacturing Engineers.(1983) 3. MSC.Marc User s Manual, Vol.C Program Input, MSC.Software Corp.,U.S.(2000) 4. MSC.Marc User s Manual, Vol.B Element Library, MSC.Software Corp.,U.S.(2000) 5. MSC.Marc User s Manual, Vol.A User Information, MSC.Software Corp.,U.S.(2000) 11