A laboratory method to simulate seismic waves induced by underground explosions

The seismic waves induced by underground explosions generate geological hazards affecting deep buried tunnels such as rockbursts and engineering-induced earthquakes. This issue is difficult to study through full-scale testing due to the expense and unpredictable danger. To solve this problem, the authors developed experimental apparatus and presented a laboratory method to simulate seismic waves induced by underground explosions. In this apparatus, a combined structure of a diffusive-shaped water capsule and a special-shaped oil capsule was designed. This structure can provide an applied confining stress and freely transmit the stress wave generated by external impact. Therefore, the coupled loading of in situ stress and seismic waves induced by underground explosions in the deep rock mass was simulated. The positive pressure time and peak value of the stress wave could be adjusted by changing the pulse-shaper and the initial impact energy. The obtained stress waves in the experiments correspond to that generated by 0.15-120 kt of TNT equivalent explosion at a scaled distance of 89.9-207.44 m/kt.

Hydromechanical characterization of gas transport amidst uncertainty for underground nuclear explosion detection

Given the challenge of definitively discriminating between chemical and nuclear explosions using seismic methods alone,surface detection of signature noble gas radioisotopes is considered a positive identification of underground nuclear explosions(UNEs).However,the migration of signature radionuclide gases between the nuclear cavity and surface is not well understood because complex processes are involved,including the generation of complex fracture networks,reactivation of natural fractures and faults,and thermo-hydro-mechanical-chemical(THMC)coupling of radionuclide gas transport in the subsurface.In this study,we provide an experimental investigation of hydro-mechanical(HM)coupling among gas flow,stress states,rock deformation,and rock damage using a unique multi-physics triaxial direct shear rock testing system.The testing system also features redundant gas pressure and flow rate measurements,well suited for parameter uncertainty quantification.Using porous tuff and tight granite samples that are relevant to historic UNE tests,we measured the Biot effective stress coefficient,rock matrix gas permeability,and fracture gas permeability at a range of pore pressure and stress conditions.The Biot effective stress coefficient varies from 0.69 to 1 for the tuff,whose porosity averages 35.3%±0.7%,while this coefficient varies from 0.51 to 0.78 for the tight granite(porosity<1%,perhaps an underestimate).Matrix gas permeability is strongly correlated to effective stress for the granite,but not for the porous tuff.Our experiments reveal the following key engineering implications on transport of radionuclide gases post a UNE event:(1)The porous tuff shows apparent fracture dilation or compression upon stress changes,which does not necessarily change the gas permeability;(2)The granite fracture permeability shows strong stress sensitivity and is positively related to shear displacement;and(3)Hydromechanical coupling among stress states,rock damage,and gas flow appears to be stronger in tight granite than in porous tuff.

Simulations of explosion-induced damage to underground rock chambers

A numerical approach is presented to study the explosion-induced pressure load on an underground rock chamber wall and its resultant damage to the rock chamber.Numerical simulations are carried out by using a modified version of the commercial software AUTODYN.Three different criteria,i.e.a peak particle velocity (PPV) criterion,an effective strain (ES) criterion,and a damage criterion,are employed to examine the explosion-induced damaged zones of the underground rock chamber.The results show that the charge chamber geometry,coupling condition and charge configuration affect significantly the dynamic pressure exerted on the rock chamber wall.Thus the chamber is damaged.An inaccurate approximation of pressure boundary ignoring the influences of these factors would result in an erroneous prediction of damaged area and damage intensity of the charge chamber.The PPV criterion yields the largest damaged zone while the ES criterion gives the smallest one.The presented numerical simulation method is superior in consideration of the chamber geometry,loading density,coupling condition and rock quality.The predicted damage intensity of rock mass can be categorized quantitatively by an isotropic damage scalar.Safe separation distance of adjacent chambers for a specific charge weight is also estimated.

Effects of underground explosions on soil and structures

Much effort has been dedicated to the study of underground explosions because they pose a major threat to people and structures below or above the ground.In this regard,it is especially important to model the propagation of blast waves in soil and their effects on structures.The main phenomena caused by underground explosive detonation that must be addressed are crater or camouflet formation,shock wave and elastic–plastic wave propagation in soil,and soil-structure interaction.These phenomena can be numerically simulated using hydrocodes,but much care must be taken to obtain reliable results.The objective of this study is to analyze the ability of a hydrocode and simple soil models that do not require much calibration to approximately reproduce experimental and empirical results related to different buried blast events and to provide general guidelines for the simulation of this type of phenomena.In this regard,crater formation,soil ejecta,blast wave propagation in soil,and their effects on structures below and above the ground are numerically simulated using different soil models and parameters;the results are analyzed.The properties of soil have a significant effect on structures,the ejecta,and the propagation of shock waves in soil.Thus,the model of the soil to study these phenomena must be carefully selected.However their effect on the diameter of a crater is insignificant.

Models of wave duration and event frequency of explosion aftershocks

The contained underground explosion (CUE) usually generates huge number of aftershocks. This kind of after-shocks induced by three CUEs was investigated in the paper. The conclusions show that the duration of aftershock waveforms are rather short, 70 percent of them range from 2 to 7; the occurrences of the aftershocks conform to negative power function, which has the power of -1.6. The aftershock sequence attenuates a little bit faster, with power of -1.0, within two weeks of post-explosions. During the early stage of post-explosions the aftershocks show up in a cluster, however, they usually show up individually during the late stage of post-explosions. The number of aftershocks generated by the compatible explosions differs by several times because of different me-dium and geological structure; within one month after an explosion with Richater magnitude of 5.5, the number of aftershocks attenuates to the background. Hereafter there are still tiny numbers of aftershocks.

Study on automatic recognition of the first motion in a seismic event

In this paper, we have studied the waveforms of background noise in a seismograph and set up an AR model to characterize them. We then complete the modeling and the automatic recognition program. Finally, we provide the results from automatic recognition and the manual recognition of the first motion for 25 underground explosions.

Applicability of P/S amplitude ratios for the discrimination of low magnitude seismic events

Applicability of regional P/S amplitude ratios for the discrimination of low-magnitude seismic events was tested and proved using earthquakes and explosions in Central Asia. Results obtained show that regional P/S amplitude ratios which may discriminate medium or large magnitude events well, are also applicable to low magnitude events Their performances for low magnitude events are almost as good as that for medium or large events. Statistical comparisons based on 25 P/S discriminate from the four seismic stations WMQ, BLK, MUL and MAK showed that the average misclassification rate for low-magnitude seismic events averagely was only 2 percent higher than that for medium and large magnitude seismic events.

My Research Connections with Russian Scientists over the Past Half Century

Over the past half century, I have maintained research connections with Russian scientists during investigations in seismology and mineral physics. These studies have focused on detection and discrimination of underground nuclear explosions and measurements of the physical properties of minerals at high pressures and temperatures. During this period, I have also visited many research laboratories in Russia, including Moscow, Chernogolovka, Novosibirsk and St. Petersburg. The objective of this paper is to relate this history.