CORRELATION BETWEEN ACOUSTIC EMISSIONS AND GLUING DEFECTS IN CONICAL JOINTS D.Croccolo, R.Cuppini , V.Dal Re DIEM, Dipartimento di Ingegneria delle Costruzioni Meccaniche Nucleari Aeronautiche e di Metallurgia, Facoltà di Ingegneria, Università degli Studi di Bologna, viale Risorgimento, 2, 40136, Bologna, ITALY, dario.croccolo@mail.ingunibo.it , rossano.cuppini@mail.ing.unibo.it , vincenzo.dalre@mail.ing.unibo.it Introduction. Object of this work is to find out the presence and the entity of the gluing defects in conical steel couplings joint with anaerobic adhesive (Loctite 638) and interference by applying the acoustic emission (AE) technique. The adhesive defect, as shown in [1], influences a lot the releasing moment but it is very difficult to be discovered with the traditional non-destructive testing techniques once the parts are assembled. Because of these reasons, it was decided to correlate the AE signals recorded during the releasing tests with the defect entities and, if it’s possible, to foresee the final releasing moment of the couplings. Fundamental goal, therefore, is to verify if there is a correlation between the characteristic parameters of AE, the defect entities and the releasing load of the joint. Materials and Methods The conical couplings have been designed by the authors [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13] and realized at the DIEM laboratories in order to perform some releasing tests in presence of three interference levels and three adhesive’s defect sizes [14]. The testing goal is the evaluation of the releasing moment which is demonstrated mainly influenced by the gluing defect [1]. The defect presence is, unfortunately, very difficult to be discovered with traditional non- destructive testing techniques [15, 16] once the parts are assembled. Some recent works have pointed out a good correlations between the very low or very high adhesive’s defect size and the cumulative counts of the AE [17, 18, 19, 20, 21, 22, 23], therefore we decided to investigate if a correlation may also exist between the middle adhesive’s defect size and the cumulative counts. Finally we investigated if it’s also possible to foresee the final releasing moment of the couplings using AE technique. The torque moment was applied with a load constant ramp of 15 Nm/s applied in a torque machine. The fundamental parameters of acoustic emissions observed and analysed with the PAC (Physical Acoustics Corporation) control unit installed on an electronic computer are the cumulative counts and the energy emitted as acoustic waves. Results The graphs of the cumulative counts related to the applied load have been plotted in figg.1 and 2; in fig.2 the graphs are cut at 100.000 counts. In these figures the blue curves represent the 0% of adhesive defect, the red ones the 50% of adhesive defect and the yellow ones the 100% of adhesive defect. The cumulative counts at 4kNm was observed. This value is equal to the 30% of the maximum releasing moment tested (equal to 13,01kNm) obtained with no defect presence and no interference [1]. If the correlation between the number of cumulative counts and the releasing moment of the joints is demonstrated it is possible to use this technique as non-destructive test during the production phase. In fact, applying a small moment (equal, for instance, to the 30% of the maximum releasing moment) and comparing the acoustic emissions to those produced by a known test specimen, it could be possible to foresee the releasing moment of the coupling and, therefore, to decide if discarding or 0 100000 200000 300000 400000 500000 -1 4 9 14 kNm cont.cum P401 P401R2 P401R3 P405 P405R2 P405R3 P400 P400R2 P455 P455R2 P455R3 P410R2 P415 P415EA P450EA P450EA1 Fig. 1. Cumulative counts related to the applied load 0 50000 100000 -1 4 9 14 kNm cont.cum P401 P401R2 P401R3 P405 P405R2 P405R3 P400 P400R2 P455 P455R2 P455R3 P410R2 P415 P415EA P450EA P450EA1 Fig. 2. Cumulative counts (below 100.000 units) related to the applied load