This study presents a new design approach which can optimize position and size of a rockfall protection wall with the help of a numerical simulation. In the proposed framework, sufficient numbers of rockfall paths with information of kinematic energy are accumulated by rockfall simulation. A protection wall is then virtually overlapped on the simulated results, and values of a safety function are checked under the different combinations of design parameters. A response surface of the safety function is then approximately constructed, and a cost function is defined to express the penalty in unreasonable situation. Once the two functions are obtained, an optimization problem can be formulated. In order to demonstrate the capability of the proposed optimal design approach, a numerical example is presented in this study. The obtained result indicates that the proposed framework has the possibility of developing optimal design of rockfall protection walls.

1 Introduction

Once rockfalls reach to roads or housing area, houses and infrastructures are easily damaged or destroyed. The construction of a protection wall is one of the typical measures, which can withstand rockfall impacts with few meters in diameter. In order to construct safety and reasonable protection walls, we have to estimate the path and the kinetic energy of rockfall. It is, however, still hard to predict the movement of rockfall with high accuracy, because rockfall movement strongly depends on a local situation. In addition, strong variability is generally involved in rockfall movement and its path. Under such uncertainties, there are a lot of difficulties in the design of rockfall protection walls. Thus, the information of past rockfall records and engineering knowledge has been widely utilized in the design procedure. The problems are that the design of protection walls strongly depends on the designers' ability and experience, and that quantitative design procedures have not been established completely.

This content is only available via PDF.
You can access this article if you purchase or spend a download.