![the test site will be modeled in flac3d the test site will be modeled in flac3d](http://docs.itascacg.com/pfc600/_images/embankment-geom.png)
Pillar capacities with inferred floor properties are easily fit to the Holland-Gaddy and Maleki empirical equations. An application to EGS is promised by Taron and Elsworth (Taron and Elsworth, 2009) with a poro-elastic model previously published (Taron, et.
![the test site will be modeled in flac3d the test site will be modeled in flac3d](https://acep.uaf.edu/media/273530/solartestsite_overlay2.jpg)
A representative set of properties within this region reduced capacity of the 52.4-m (172-ft) wide pillar, which has a width-to-height ratio of 17, by 46% compared to a linear extrapolation of Bieniawski's equation. EGS site, the fluid dominantly flows in the discrete EGS fracture(s) and not in the surrounding rockmass. Arching occurs in the embankment granular material, leading to load transfer onto the piles and surface settlement reduction and homogenization. The intersection of these solution sets served to bound Mohr-Coulomb properties for the mine floor. This paper proposes a three-dimensional numerical modeling of an embankment over a soft ground mass improved by vertical stiff piles, using a finite-difference continuum approach (FLAC3D). Mohr-Coulomb properties in the floor were then varied to define a boundary between heave and no-heave floor response to loading for each pillar size. Models of various pillar shapes replicated Bieniawski pillar strength equation estimates for pillar width-to-height ratios up to 8. Tunnels are excavated by using two Earth Pressure Balance Tunnel Boring Machines with a 6.5 m diameter as twin tubes with 14 m distance from center to center. Coal in mine pillars was modeled as a Hoek-Brown material with parameters tuned to match one of Bieniawski's in situ compression tests, scaled to an unconfined compressive strength of 6.2 MPa (900 psi). Geology in the study area is composed of fill, stiff clay and dense sand and hard clay. Pillars were idealized in FLAC3D models as a single square column that includes the roof, pillar, and floor. In this case, floor heave was observed near pillars approximately 52.4 m (172 ft) wide but was absent around pillars approximately 23 m (76 ft) wide, as the longwall passed. The procedure works where the presence of floor heave depends on pillar size, as has been observed in a deep western U.S. A new procedure is presented in this paper that infers Mohr-Coulomb floor properties that will synchronize floor heave behavior between mine and model. Consistency between actual in-mine and modeled ground response to mining is essential if a model is to be used as a mine design tool.