In the vibration isolation testing of large non-constrained metal bellows expansion joints (hereinafter referred to as expansion joints), due to their easy instability under filling pressure, test devices that balance the huge blind force generated by the internal pressure cannot be completely fixed at both ends of the expansion joints, making testing difficult and leading to the lack of a standard test method. For this reason, a device is designed to test the vibration isolation effect of expansion joints.

Double-metal expansion joints are supported against each other with both ends fixed by the pedestal support, achieving axial stabilization. Two different limiting methods（i.e., bearing cable and airbag) are designed to prevent radial instability. The axial and radial excitation of the expansion joints under pressure is carried out using a shaker, the acceleration response in the frequency range of 10 Hz to 315 Hz under the two limiting methods is collected, and the difference in vibration level between the input end and output end is calculated.

The total vibration level drops measured by the bearing cable and airbag devices respectively decrease with the increase in the internal pressure of the expansion joints, and the total vibration level drop of the airbag device under internal pressure of 0P, 0.8P and P increases by 8.59 dB, 5.70 dB and 4.80 dB respectively in the axial excitation direction of the bearing cable device. In the radial excitation direction, it increases by 4.32 dB, 3.57 dB and 4.95 dB respectively.

The designed test device can measure the vibration isolation performance of metal bellows expansion joints. As the airbag limiting method has less influence on the vibration level drop than the bearing cable limiting method, the device for testing the vibration level drop of expansion joints should adopt airbag radial limiting to obtain better measurement results.

To study the mechanical behavior of aluminum alloy sheets for ships under low-speed impact load, horizontal low-speed impact dynamic response tests are conducted on 5059-H116 aluminum alloy plates.

Based on the impact tests, the damage and dynamic response of test plates under different impact speeds and masses are compared. Based on a mixed hardening plastic model, a low-speed impact numerical model is established to numerically simulate the failure process under different impact velocities. The finite element method is used to analyze the influence of specimen size, impact position and impact head shape on critical failure energy, and a modified empirical formula for critical failure energy is proposed.

The results indicate that with the increase in impact velocity, the critical failure energy of the test plate increases correspondingly, but the increase is very small. Under the same impact energy, different impact masses have no effect on the critical failure energy of aluminum alloy sheets. The sensitivity of critical failure energy on the aspect ratio of the test plate is very small. The critical failure energy of blunt impact test plates with the same cross-sectional area can be considered equivalent.

The results of this study can provide references for research on the low-speed impact mechanical behavior and load-bearing capacity of aluminum alloy sheets.