Infrastructure systems include many embankments for highways and railroads. In seismically sensitive areas west of the Rocky Mountains and over broad areas of the eastern and central US - estimated to cover as much as 40% of the continental US, earthquake engineering for these facilities is very important. A key design issue for such facilities is whether or not liquefaction - or the loss of shear strength of saturated sands - will occur during an earthquake. If such a possibility exists, then one must either relocate the embankment or improve the potentially liquefiable soil to the point where the improved soil will not liquefy under the expected earthquake. Loose sands are the soils most susceptible to liquefaction. Two questions arise during design: (i) will liquefaction occur under a given earthquake loading, and (ii) what are the consequences of liquefaction? The most egregious effect is a flow failure of the embankment through the liquefied soil. Because highway and railroad embankments traverse large areas, the costs of mitigating the effects of the liquefiable soils are large, as are costs related to realignment, if this option is followed in design. To improve the ground over large areas, densification of loose sands by controlled blasting is an economical approach. Blast densification consists of placing charges within the loose sand layer requiring treatment. The charges are detonated with multiple delays to generate cyclic loads, similar to an earthquake. Large amounts of gas also are released in the ground with each explosion. The "design" usually relies on historic or previous contractor experience, as there is no rigorous theory that accounts for the parameters that influence the densification process. Case studies have shown that loose sands compress almost immediately after blasting, but when common penetration tests are conducted to verify the increase in density, these results provide erratic and, at times, rather counterintuitive results. If taken soon after the blast, the penetration resistance may decrease, and at times never increases to levels above the pre-blast level. At the same time, the ground surface settles almost immediately after blasting, implying that loose sands in the subsurface have increased density. However, the lack of increase in penetration resistance suggests that the strength and stiffness of the soil apparently does not. This leads to questions about future performance. Have the loose sands really been improved to the point where liquefaction is not a possibility? Furthermore, how the released gas affects the soil behavior is ignored in design and verification, e.g., what form does the gas take in the ground, how does it dissipate and how does it affect the behavior of the sands in response to subsequent static and cyclic stresses.
The objectives of this GOALI research are to develop (i) a methodology to quantify the amount of densification required to make the soil resistant to liquefaction and flow in the presence of shear stresses with explicit evaluation of the effects of gasses released during blasting, and (ii) a means of reliable in situ verification of the ground improvement. To achieve the first objective, a laboratory experimental program will be conducted to define the constitutive response of reconstituted "gassy" sand specimens. The laboratory program will account for in situ stresses and gas concentrations and include both monotonic and cyclic tests. To achieve the second objective, a field verification program will monitor the soil and pore fluid responses before, during, and after densification at a production blast test section at a municipal waste disposal facility near Charleston, SC. Blasting has been used at the site for more than a decade to densify loose sands, and Geosyntec Consultants is the engineer of record for this project. The field studies will include pore pressure measurements, surface settlements, in situ testing, pore fluid sampling, pore fluid pressure measurements, and in situ gas concentration and composition measurements. Combining the field and laboratory studies to evaluate the effects of the released gas during blasting provides a unique opportunity to remove the empiricism in design and the contradictory results obtained during verification testing.
In addition to application to municipal waste fills, the results of this research will have direct application to infrastructure systems including many embankments for highways and railroads. Also defining the behavior of moderately dense gassy sands will provide valuable data that currently does not exist which can have an impact in offshore applications, tailing dams and other conditions where gassy soils exist.
