Self-Piecing Rivet을 이용한 이종소재 CFRP-Steel의 접합특성에 관한 연구

Abstract

A Study on the Joining Characteristics of Dissimilar CFRP-Steel by Self-Piecing Rivet Choi Dong Won Advisor : Prof. Bang Hee Seon, Ph.D. Department of Welding and Joining Science Engineering, Graduate School of Chosun University Various light materials such as aluminum, high-strength steel, and carbon fiber are applied in the automobile industry to improve fuel efficiency due to stricter exhaust gas regulations like CAFE (Corporate Average Fuel Economy) and EURO-IV. In particular, CFRP is a composite material made of carbon fiber and resin. It is attracting a lot of attention because it weighs about 1/5th and has a tensile strength of more than five times that of general carbon steel. Laser and resistance welding has been studied for joining CFRP and metals, but obtaining sufficient strength in the joints is challenging due to the material properties of CFRP as a composite material. Therefore, mechanical joining methods such as bolts, SPR (Self-Piercing Rivets), and adhesives are commonly used. SPR involves joining by pressing rivets, eliminating the need for additional hole processing, and it is compatible with resistance point welding production facilities. Furthermore, SPR is a mechanical joining technique suitable for bonding dissimilar materials such as CFRP-Steel joints because it does not cause thermal deformation. Moreover, it is an eco-friendly method as it does not produce harmful gases like welding fumes In this study, Self-Piercing Rivet (SPR) was employed for joining CFRP/Steel materials, and the joining performance and mechanical properties were evaluated based on the force and rivet geometry factor. Furthermore, quality evaluation criteria factors were predicted using CNN(Convolutional neural network) As the rivet shank length increased, buckling occurred, and as the force increased, the interlock length increased. CFRP damage increased at the portion where the rivet was fastened as the force increased, and the thickness of CFRP decreased. In the tensile-shear test of CFRP-SPRC440 joints, the highest tensile load of 2 kN was achieved under the condition of a 4 mm shank length and a force of 27 kN. In CFRP-SPFC590 joints, the highest tensile load was 1.94 kN under the conditions of 4 mm shank length and force of 29 kN. Fractured mode observed phenomenon of rivet head pull-out and CFRP separate. It inferred to exist of relationship between CFRP damage and tensile-load. As a result of fatigue test in CFRP-SPRC440 joints, fatigue test criteria were satisfied at 50% and 60%, and the maximum load corresponding to the fatigue limit was about 1.2 kN, which is 60% of the tensile shear load, and the amplitude load was 0.54 kN. In CFRP-SPFC590 joinits, fatigue test criteria were satisfied at 50%. The fracture specimen showed the CFRP separation fracture mode, and the CFRP remaining in the rivet was confirmed. Therefore, in the case of the SPR joints of CFRP-SPRC440, the rivet shank length of 4 mm and pressing force of 27 kN showed the best cross-section and tensile strength. In the case of CFRP-SPFC590, the rivet shank length of 4 mm and pressing force of 29 kN showed the best cross-section and tensile strength. Furthermore, We performed quality prediction using two models: convolutional layers and Xception. The prediction accuracy criterion was selected as a coefficient of determination (R2) of 0.8 or more, and the learning results are as follows. The model using the Conv layer met the criteria with a maximum of training: 0.914, Test: 0.852, but did not meet the criteria when predicting the bottom thickness factor. The model using Xoption met the criteria with maximum training: 0.948 and test: 0.928, and the bottom thickness factor also met the criteria. Therefore, the following conclusions have been drawn: 1. When the thickness of the joint is 2.5 mm, the rivet shank length of 4.5 mm occurs during the SPR process, and the pressing force of 32 kN is unsuitable due to the insertion of the rivet head. 2. In the case of SPRC440, the maximum tensile shear load was about 2 kN under the condition of 4 mm rivet shank length and pressing force of 27 kN, and in the case of SPFC590, the maximum tensile shear load was about 1.94 kN under the condition of 4 mm rivet shank length and pressing force of 29 kN. In the fatigue test, SPRC440 satisfied 1 million cycles (times/Hz) up to 60% fatigue limit and SPFC590 50% fatigue limit. 3. Damage to CFRP occurring during the SPR process reduces the load transmitted to the interlock during the tensile test, resulting in a decrease in the strength of the joint. This is due to the characteristics of CFRP, a composite material, and the degree of damage to a non-flexible composite material such as CFRP is important during the SPR process. 4. Although the prediction accuracy of the bottom thickness tended to decrease slightly, the head height and interlock factors were satisfied by 0.8 or more with the coefficient of determination (R2) of training and test at a maximum of 0.928 and 0.948, and the predictability was confirmed.