Displacement-based Seismic Design for Graving Docks

Graving docks are used to dry dock large ships for maintenance. Seismic vulnerability concerns for graving docks have arisen in recent years in seismically active regions because seismic damage can interrupt operations and pose a significant safety risk. In modern seismic engineering, the displacement-based design approach is recognized as the standard of care for unique and valuable infrastructure. Displacement-based design is more cost-effective than force-based design and provides a better prediction of anticipated seismic performance. However, code and industry standards for applying displacement-based design principles to graving docks do not exist. This paper provides a summary of a displacement-based design method developed during the design of a graving dock for the US Navy.

The challenges of applying traditional displacement-based design to graving docks will be explored. These challenges are primarily due to the dynamic soil-structure interaction (SSI) between the walls and the retained soil. Traditional displacement-based design for elevated structures, such as piers, wharves, and bridges, represent the structure as an equivalent single-degree-of-freedom (SDOF) oscillator using a secant stiffness from the structure’s capacity curve. For these structures, performance at peak displacement response is estimated using the response spectrum as a demand curve. Because of the complexity of the SSI, a SDOF assumption is not applicable to the design of graving dock walls and the development of a demand curve directly from the response spectrum is not possible.

Progressively more complex approaches to account for the SSI effects when determining an appropriate demand curve are discussed. These begin with soil-pressure-displacement relationships for the walls and conclude with time-history analysis of graving dock cross sections using 2D SSI software to simulate soil response to dynamic forces. In each case, the result is a demand curve that can be paired with the structural capacity curve in a traditional displacement-based design approach. It may be counterintuitive to use time-history analysis to generate demands for a less-sophisticated nonlinear static capacity analysis, and this is discussed. Primary reasons include flexibility to changes during the design process and the ability to easily perform analysis iterations when optimizing the design.

A procedure to extend the 2D results to a 3D analysis will be summarized. Reasons for conducting this 3D analysis in traditional structural analysis software (as opposed to 3D SSI software) will be discussed. These reasons include cost, efficiency, flexibility, and the need to validate results. Modeling observations are provided. These include a discussion of the importance of developing accurate, idealized seismic pressure surfaces for graving dock walls. The benefits of using distributed-plasticity elements will also be discussed. The paper will conclude with a discussion of key takeaways and recommendations for other design professionals involved in the seismic design of new graving docks or the evaluation of the seismic performance of existing dry docks.