TPI April 2011

The increase in feeding rate has no influence on the outer diameter of the tube but the inner diameter decreases from 19.3mm to 18.5mm. Consequently the amplitude of the wave increases from 0.8mm to 1.2mm. Further, the simulation shows that the wave period also increases with increasing feeding rate. Figure 11 shows the simulated true strain distribution on the corrugated tube. For both investigated cases the true strain lies between 20% on the wave peak and 80% in the wave trough. The case with lower feeding rate shows a homogeneous strain profile over the tube while the strain for the version produced with higher feeding rate is very inhomogeneous. This leads to the conclusion that the feeding rate of 7m/min is near the upper limit for the investigated parameter set. A higher feeding rate would lead to an unstable process, ie wrinkling will occur. The resulting wall thickness on the corrugated tube (Figure 12) lies between 0.17mm on the wave peak and 0.2mm on the wave trough. The wall thickness at the wave peak experiences a small decrease compared to the initial wall thickness, while the wall thickness in the wave trough increases for 0.02mm. The main difference between the two cases with the different feeding rates is, analogous to the strain, that the wall thickness is more homogeneous for the case with 6.4m/min feeding rate. Conclusion The corrugation process of tubes presented in thiswork is a very complex manufacturing process. In many cases the variation of different process parameters directly on the machine is a very fast way indeed but often the interaction and impact of Computed strain and wall thickness of the corrugated tubes

different process parameters on the corrugation result can not be found by experimental work. Hence, a simulation model of the corrugation process was developed. With this model it is possible to perform a systematic parameter variation to investigate the influence of different process parameters without blocking any production devices. The first trials to simulate this process are presented here. It is shown that it is possible to model this process with finite elements. By varying the infeed rate the sensitivity of the model on changed process parameters is demonstrated. However, a lot of specific assumptions and a complex model are necessary to obtain accurate results in suitable computation time. In future work this model will be used to look on this process in more detail and to obtain better knowledge about the interactions of the most important parameters which influence the performance of the corrugation process. Literature [1] R. Wahl, D. Eckhardt: Herstellung von Metallschläuchen, Metallbälgen und Kompensatoren, Hochschule Pforzheim, 2008 [2] T. Estienne et al.: US Patent No. 7266886 B2 (2007) [3] S. Weiss: Proceedings of the XXVIII Conference on Metal Forming, Planneralm, A, (2009), 75-89. [4] N.N.: Abaqus 6.9 Documentation, Simulia, Providence, RI (2008) [5] X. Feaugas: Acta Materialica 47-13 (1999), 3617-3632 [6] T. Hatzenbichler, B. Buchmayr, S. Weiß; 13 th International Conference on Metal Forming; Toyohashi, Japan; 2010 [8] DIN EN 14585-1: Metallschlauchleitungen für Druckanwendungen, 2006 [9] DIN EN ISO 10380: Gewellte Metallschläuche und Metallschlauchleitungen, 2003

Figure 11 : Strain distribution on the corrugated tube produced with a feeding rate of (a) 6.4m/min and (b) 7m/min

Rosendahl Maschinen GmbH Schachen 57, PO Box 45, A-8212 Pischelsdorf, Austria

University of Leoben Franz-Josef-Straße 18, A-8700 Leoben, Austria umformtechnik@unileoben.ac.at

Figure 12 : Computed wall thickness on the corrugated tube produced with a feeding rate of (a) 6.4m/min and (b) 7m/min

66

Tube Products International April 2011

www.read-tpi.com