JEC COMPOSITES MAGAZINE - Issue #70 - January/February 2012 - 58

welding
LSS was to maximize it at 40 MPa, within a minimum value of 37 MPa, while for melting time the target was to minimize it at 400 ms, within a maximum of 800 ms. Maximum power was minimized at 30%, within a maximum of 60%. The coefficients of variation (COV) of the responses were included in the optimization, as they were sometimes too high, and seemed to have some connection with the parameters. The results from the Response Optimizer including these extra factors did not provide any useful information. Hence, they were not included in the discussion. Under these settings, the optimum conditions predicted by the Response Optimizer, and the experimental results, were those shown in Table 4. Melting time optimization The predicted responses for this optimization, 61 ms and 43 MPa, were quite far from the experimental results. However, these results, 310 ms and 37.24 MPa, complied with what was expected from the optimization, although the prediction did not correspond with these values. These optimized welding conditions were compared to Run 12 (see Table 3), with a lower travel (30% instead of 80%). As shown before from the analysis of the tendencies, decreasing the travel results in a decreased melting time, as less time is needed to reach a lower travel under the same welding conditions. The melting time was indeed decreased (247.8 ms), but the LSS was also lower (33.6 MPa). This LSS decrease could have been caused by the excessive amount of resin in the interface, due to a lower squeezing out. Optimization of the maximum power The differences between the predicted responses and the experimental results were not very pronounced in this case. LSS resulted to be within the admitted limits, although the resulting COV was too high (7.43%), which is believed to be a sign of an unstable welding process. However, in the analysis of the tendencies it was concluded that maximum power decreased when decreasing amplitude and trigger. Run 4 from the experiment (see Table 3) was welded under these conditions showing a better response for maximum power than the predicted one (38% against 57.6%), resulting in the same LSS, with lower COV (36.71 MPa, COV 4.67%), within the same melting time (887.8 ms). ters having the most impact in the welding process. At low amplitude, intermolecular friction is lower and heat generation less effective. The LSS seemed to be hardly affected by the parameters of the welding process. Two sets of welding conditions were found for the optimization of the melting time and maximum power involved in the melting process of ultrasonic welding. The set of conditions for minimizing the melting time was: amplitude 86.2 µm, travel 80%, trigger 1500 N, build-up 1000 N/s, solidification force 300 N and solidification time 1200 ms, for which the responses were maximum power 87.4%, melting time 310 ms and LSS 37.24 MPa. The set of conditions for minimizing the maximum power was: amplitude 86.2 µm, travel 30%, trigger 300 N, build-up 0 N/s, solidification force 3000 N and solidification time 1200 ms, for which the responses were maximum power 57.6%, melting time 746 ms and LSS 36.82 MPa. However, low amplitude was expected in the optimization of maximum power. Samples welded this way gave lower power than the predicted response, within the admitted LSS values and with less scattered responses. n
More information: lr.tudelft.nl/en/ m.barrosoromero@tudelft.nl References Ali Yousefpour, Mehdi Hojjati, Jean-Pierre Immarigeon: Fusion Bonding/Welding of Thermoplastic Composites, J. of Thermoplastic Composite Materials, 17 (2004) 303. 2. I. Fernández-Villegas, D. Stavrov, H.E.N. Bersee: Ultrasonic Welding of Advanced Thermoplastic Composites: An Investigation on Energy Directing Surfaces, J. Advances in Polymer Technology, Thermoplastic Composite Materials, 29, (2010) 112. 3. Shih-Jung Liu, I-Ta Chang, Shiu-Wan Hung: Factors affecting the Joint Strength of Ultrasonically Welded Polypropylene Composites, Polymer Composites, 22, (2001) 140. 4. Handbook of Plastics Joining A Practical Guide, Plastics Design Library is a trademark of William Andrew, 1997, Chapt. 3, p. 19. 5. Mark J. Anderson, Patrick J. Whitcomb: DOE Simplified, Productivity Press, New York, 2000, Chapt. 1, p. 12; Chapt. 5, p. 105; Chapt. 2, p. 32. 6. Shih-Jung Liu, I-Ta Chang: Optimizing the Weld Strength of Ultrasonic Welded Nylon Composites, J. of Composite Materials, 3, No 05/2002 (2002) 621. 7. L. Moser, I. Fernandez-Villegas, P. Mitschang, H.E.N. Bersee, A. Yousefpour: Comparison of three welding processes for joining advanced thermoplastic composites, Proceedings of the SAMPE Seattle 2010 Conference (2010). 1.

Conclusions
Design of experiments was a suitable method to obtain a significant amount of valuable information on ultrasonic welding of CF/PEI composites. It helped to describe and quantify the influence of parameters and their interactions in the welding process. Amplitude and trigger were found to be the parame-

58 jec composites magazine / No70 January - February 2012

http://lr.tudelft.nl/en/

JEC COMPOSITES MAGAZINE - Issue #70 - January/February 2012

Table of Contents for the Digital Edition of JEC COMPOSITES MAGAZINE - Issue #70 - January/February 2012