HOME > Publications > Ph.D Dissertaitons

저자명 | 하지웅 |
---|---|

년도 | 2013 |

This paper newly proposes dynamic failure criteria for laser welds to describe strain-rate dependent failure behavior of laser welds in crash analysis of an auto-body. The failure criteria had been deduced from experimental results with nine different loading conditions including the normal loading and the pure-shear loading condition. Quasi-static and dynamic failure tests for 25 mm stitch-type laser welds of SPRC340 1.2t and SPCC 1.0t, 20 mm stitch-type laser welds of CR340 1.2t and *f* 8 mm O-type laser welds of SPRC340 1.2t were examined in this paper. A universal testing machine of INSTRON 5583 was used in the quasi-static failure tests and a high speed material testing machine was utilized in the dynamic failure tests at the intermediate strain rates of 1/sec, 10/sec and 100/sec, as the strain rate most relevant to an automotive crash event lies in the intermediate strain rate range. In order to obtain the failure load under the combined normal and shear loading condition including a pure-shear failure loading condition, the testing systems were newly designed and prepared. The prepared apparatus guarantees prevention of unfavorable rotation of the bead using guide plates during failure tests under the combined loading condition. A new testing fixture and a specimen were designed to evaluate the pure-shear strength of laser welds fabricated using the same welding condition with a two-layered lap joint. With the aid of the proposed testing fixtures and specimens, the normal and the shear failure loads of laser welds could be acquired accurately by simple trigonometric functions of the applied failure load and the initial loading angle.

Failure loads and failure behavior of laser welds were evaluated using the experimental results. In the failure tests of the stitch-type laser welds and the O-type laser welds, two types of failure modes were observed: base metal failure took place from 0° to about 60°; and interfacial failure took place at the mid plane of a bead from 80° to 90°. There is a transition region between the angle of 75° and 80° where the failure mode of laser welds changes from base metal failure to interfacial failure. From the failure loads measured in the dynamic failure tests, the strain-rate dependent failure contour of laser welds were constructed by decomposing the failure loads at various loading angles into the two components along the normal and shear directions and decomposed ones were plotted in the force domain. It was found that the failure contour of the laser welds expands with increasing strain rates.

Based on the experimental data, a new failure criterion was proposed for prediction of failure behavior of laser welds. In the case of the base metal failure, the shear failure load increases as the normal failure load decrease with increase of the loading angle. The base metal failure contour shows a near-elliptical contour in terms of the normal and shear loads. Therefore, the failure of laser welds can be predicted by a similar failure criterion to that for spot welds. The transition point is observed around a loading angle of 75° due to the change in the failure mode from base metal failure to interfacial failure. In the case of the interfacial failure, the shear failure load decreases as the normal failure load decrease with increase of the loading angle. Consequently, the new criterion introduces to describe the interfacial failure other than the failure criterion for the base metal failure. In this paper, the dynamic failure criterion is divided into two regions of a base metal failure and an interfacial failure. The dynamic failure criterion for the base metal failure region is expressed as a function of the normal and the virtual shear load, which is the so-called *b*-norm function, while the dynamic failure criterion for the interfacial failure region is explained in a form of a modified power law function. The dynamic failure criterion for the base metal failure region is verified based on the assumption of a stress field around a bead with the aid of finite element analysis which is induced by the normal and the shear loads. The strain-rate sensitivity values of the normal and the virtual shear failure load are assigned as an exponential function of the logarithm of the strain rate. The dynamic failure criterion for the interfacial failure criterion is proposed to describe the failure contour in the interfacial failure region which is represented empirically. The strain-rate sensitivity value of the corresponding shear failure load for the normal load acting on laser welds is assigned as an exponential function of the logarithm of the strain rate and the change of shear failure load hardening due to the change of strain rate is considered. The maximum value of the failure criteria is obtained after calculating the base metal failure criterion and the interfacial failure criterion with the normal load and the shear load acting on laser welds. The failure mode and the failure of laser welds are determined when the maximum value reaches to 1. The failure criterion newly proposed provides a fairly accurate prediction of the failure load obtained from the experiments under combined normal and shear loading conditions.

*b*-norm function toward the normal load axis direction and the shear load axis direction. There are five coefficients which are obtained from interpolation result for the failure contours of laser welds. The strain-rate sensitivity values of each coefficient are assigned as an exponential function of the logarithm of the strain rate. The modified dynamic failure criterion describes the failure contour of stitch-type laser welds simply and accurately. There are very few differences between the newly proposed dynamic failure criterion and the modified dynamic failure criterion to describe the failure contour of stitch-type laser welds. However, there are large differences between the proposed failure criterion and the modified failure criterion to describe the failure contour of O-type laser welds at the transition point. The modified failure criterion can describe the experimental results accurately like stitch-type laser welds when the failure loads change gradually at the transition point of a failure mode. However, the modified failure criterion cannot describe the failure contour of O-type laser welds when the failure loads change sharply around the transition point of a failure mode. Therefore, the newly proposed dynamic failure criterion is better than the modified failure criterion to describe the failure contour of laser welds.