Experimental study on the physical modeling of loess tunnel-erosion rate

Abstract

Loess tunnel erosion is a distinctive soil erosion pattern that commonly occurs in collapsible loess areas. The resulting soil erosion is a serious and complex problem. At present, most of the research methods on loess tunnel erosion are based on field investigation, and the knowledge obtained is, therefore, mostly qualitative. Erosion characteristics and erosion rates of loess tunnels are rarely revealed quantitatively by laboratory physical modeling or field investigation. In this paper, we report the results of a small model test based on three types of archetypal geological structures that are recognized by field investigation, i.e., vertical fractures, horizontal fractures, and round holes, These results reveal the effect of different initial flow cross sections on the erosion characteristics of loess tunnels. We found that, where water crosses an area of loess, waterfalls dominate initial erosion. We also studied the mud erosion rate to quantitatively reveal tunnel-erosion rates and their variation with time and found that tunnel erosion occurs in three stages: a rapid-erosion stage, an erosion slow-down stage, and a slow uniform-erosion stage. Using a model test to undertake quantitative studies into the relationship between the erosion rate of a loess tunnel and both initial dry density and initial water content, we demonstrated a significant linear negative correlation between these parameters. Through physical simulations combined with field investigation and analysis of survey results, we found that the erosion of loess tunnels occurs with a certain periodicity and that the tunnel-erosion base level plays a decisive role in the erosional features and final morphology of loess tunnels. By increasing our understanding of the internal mechanisms of loess tunnel erosion, our work will contribute to the prevention of cave erosion and reduction in the risks of associated disasters in the Loess Plateau of the Yellow River in China and in other loess-containing areas around the world. Our studies also provide a theoretical basis for carrying out soil and water conservation work in similar areas.