Research outline

Study of River Channels Plans Using Hydraulic Models

How far will riverbed scouring around bridge piers progress? What should be done to protect the bridge from scouring? (Left: actual site; right: model)

Water flow causes waves (sand waves) in the riverbed.
The rise and fall of dunes, which are typical sand waves, has a major influence on water flow and is an important point when considering the reproducibility of phenomena.
(Left: actual site; right: model)

Overhead view of a river hydraulic model

“Why do you conduct tests using river models?” Very simply: to confirm whether or not projects that are planned onsite will be appropriate. Desktop studies, such as calculations, and onsite measurements are important steps in the study process, and today a lot can be understood through these activities alone. However, there are limits to their effectiveness, and thus they cannot be relied upon alone. For example, it is difficult to calculate how water flow will scour around bridge piers or how the riverbed will undulate. And that is why tests are conducted using river hydraulic models. The following presents thinking regarding river model scale and points to bear in mind when studying river channel plans.

When conducting tests using river hydraulic models, it is important not only to reproduce phenomena that occur at the project site but also to predict phenomena that may occur in the future. However, because a model is merely a model, simply running water through a reduced (or enlarged) model of a river cannot faithfully reproduce actual phenomena, and thus limitations naturally occur here. One major reason is that water has viscosity, and if a model is too small, the factors needed to produce the effects of this viscosity (e.g., water flow, water volume, and movement of sand on the riverbed) will not be identical to those at the actual site. For this reason, it is important to conduct studies by choosing model scale and type based on what is being examined and the degree of precision that is required.

As is demonstrated above, when using river hydraulic models, it is vitally important to approximate actual phenomena as faithfully as possible. Therefore, the quality of onsite flood data that serves as verifying data greatly influences the accuracy of the test.

As part of river channel planning, studies are conducted that consider very rare large-scale flooding (i.e., flooding that occurs once in 100 years). If the verifying onsite flood size is not very large, efforts are made to obtain more precise study results by adding local observations and studies and engaging in mutual compensation through hydraulic accounting.

(Contact: River and Dam Hydraulic Engineering Research Team)

Shaking Table Tests for Mountain Tunnels: “Elucidating the Mechanisms of Seismic Damage”

Damage in the Wanazu Tunnel

Fallen concrete in the Wanazu Tunnel, National Rte. 17, Kawaguchi Town Courtesy of the Road Department, Hokuriku Regional Development Bureau, MLIT (Photo taken October 24, 2004)

Codes indicating positions within the tunnel

The upper section of the vertical axis refers to tensile strain; the lower section refers to compressive strain
The horizontal axis shows position in the tunnel.

Thus far, mountain tunnels that are mainly constructed in bedrock have not suffered great damage during earthquakes (excluding tunnels in sections with extremely poor ground, such as fault fracture zones, and tunnel entrances). Consequently, it was thought experience proved that tunnels are highly earthquake-resistant structures. However, during the Mid-Niigata Prefecture Earthquake in 2004, relative large damage - albeit in a limited number of locations - occurred in areas that were thought to be resistant to seismic damage. Such damage included collapses of tunnel linings. Because Japan is a country with numerous tunnels, it is important to elucidate the mechanisms by which seismic damage occurs and to establish rational earthquake-proofing measures in order to minimize this kind of damage. Thus, this research project is studying such steps through numerical analyses and model shaking table tests.

In the case of the Mid-Niigata Prefecture Earthquake in 2004, significant damage occurred in tunnels built in areas with soft ground and relatively large earth coverings - places in which major seismic damage had never been seen before. In one such tunnel, the Wanazu Tunnel, compression failure occurred at the crown that caused a piece of lining measuring several meters to fall. Although a number of research efforts are underway toward preventing this kind of damage, the mechanism that caused it remains unknown. Thus, this project conducted a shaking table test using a model to study the seismic performance of tunnels in soft-ground areas.

In this test, a tunnel lining model with a diameter of 15cm and made of 1mm thick aluminum plate was buried in a rectangular block of dry sand. The block was then vibrated from the bottom. When the amplitude was low and when acceleration was low, the tunnel experienced distinct expansion and contraction in a diagonal direction, and concentrated distortion occurred at the shoulder. However, when both amplitude and acceleration were increased, expansion and contraction occurred in horizontal and vertical directions in addition to the diagonal direction. Moreover, as is shown in the graph, severe compressive strain occurred at the crown. These results suggest the possibility that strong compression force occurs in the tunnel crown when strong seismic movement acts on a tunnel in soft ground, and that the crown may suffer compression-caused failure depending on the stress state of the tunnel lining before the earthquake occurs.

There are still many areas in which the mechanism that cause seismic damage in mountain tunnels are unknown. Thus, plans call for this project to move forward with studies that are better matched to actual phenomena as it continues to improve testing and analytical methods.

(Contact: Tunnel Research Team)