Prediction of rock mass along tunnels by geostatistics

Kaewkongkaew, K.; Phienwej, N.; Kham-ai, D.

doi:10.1007/s12205-014-0505-3

Cited by 16 publications

(5 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As an advanced machine learning algorithm, NN represents an alternative to the well-established pool of response surface methods (see Li et al, 2016), such is the Kriging model, used because of its recognized ability to provide high quality predictions. However, some studies (Kaewkongkaew et al, 2015) showed that ordinary Kriging does not work well in estimating rock mass quality along tunnel alignments in complex geological settings. Further, Shi et al (2019) state that the NN surrogate model can accurately estimate the geological conditions prior to excavation when compared with the other soft computing methods, while Santos et al (2014) conclude that model errors obtained with the different estimation methods (linear regression, geostatistical Kriging and NN algorithms) are very similar.…”

Section: Development Of Netrheo Neural Networkmentioning

confidence: 99%

Assessment of long-term deformation of a tunnel in soft rock by utilizing particle swarm optimized neural network

Kovačević

Bačić

Gavin

et al. 2021

Tunnelling and Underground Space Technology

View full text Add to dashboard Cite

The continuous monitoring of long-term performance of tunnels constructed in soft rock masses shows that the rock mass deformations continue after construction, albeit at a rate that reduces with time. This is in contrast with NATM postulates which assume deformation stabilizes shortly after tunnel construction. This paper proposes the prediction of long-term vertical settlement performance of a tunnel in soft rock mass, through the inclusion of a Burger's creep viscous-plastic constitutive law to model post-construction deformations. To overcome issues related to the complex characterization of this constitutive model, a neural network NetRHEO is developed and trained on a numerically obtained dataset. A particle swarm algorithm is then employed to estimate the most probable rheological parameter set, by utilizing the long-term in-situ monitoring data from several observation points on a real tunnel. The paper demonstrates the potential of the proposed methodology, using displacement measurements of two adjacent tunnels in karstic rock mass in Croatia. The complex interaction of a railway tunnel Brajdica and a road tunnel Pećine, conditioned by the character of the surrounding rock mass as well by the chronology of their construction, was evaluated to predict the future behavior of these tunnels.

show abstract

Section: Development Of Netrheo Neural Networkmentioning

confidence: 99%

Assessment of long-term deformation of a tunnel in soft rock by utilizing particle swarm optimized neural network

Kovačević

Bačić

Gavin

et al. 2021

Tunnelling and Underground Space Technology

View full text Add to dashboard Cite

show abstract

“…As such, it has been used in the geotechnical domain (Brito et al 1997) and has been incorporated in many software offering custom-built surrogate models for fundamental aspects of uncertainty quantification, such as the open-access platform UQLab (2020). Despite being popular and providing relatively high accuracy, some geotechnical studies showed that ordinary Kriging did not work well in estimating rock mass quality along tunnel alignments in complex geological settings, with a large difference between the estimated and actual values (Kaewkongkaew et al 2015).…”

Section: An Architecture Of the Nettunn Neural Networkmentioning

confidence: 99%

Application of neural networks for the reliability design of a tunnel in karst rock mass

2021

View full text Add to dashboard Cite

This paper offers a solution to overcome time-consuming numerical analysis for the evaluation of the impact of tunnel construction in a complex karst environment by implementing Monte-Carlo simulation (MCS) using a neural network (NN) tool. The rock mass is described using three parameters, Geological Strength Index, the uniaxial compression strength of the intact rock and the Hoek-Brown parameter for the intact rock mi. By using their probabilistic distribution as an input, a developed neural network NetTUNN produces probabilistic distributions of tunnel crown displacement, rock bolt axial load and shotcrete uniaxial compression stress. A full MCS is then applied on these NetTUNN outputs to determine the reliability index and probability of failure for the relevant limit states. To demonstrate the potential of NN in tunnel design, a case study tunnel Pećine in Croatia is used, where the NetTUNN-assisted MCS assessment served as a benchmark to evaluate approximate reliability assessment techniques. It was shown that the developed NN can be used as an accurate surrogate model for determination of probabilistic distributions of tunnel design parameters. Further, it was shown that approximate reliability assessment techniques generally overestimate the reliability index and underestimate the probability of failure when compared to the NetTUNN assisted MCS.

show abstract

“…The RMR defines the geomechanical quality of a rock mass as the sum of five rates referred to the following rock and rock mass parameters: the uniaxial compression strength of rock matrix, the Rock Quality Designation (RQD), the discontinuity spacing, the condition of discontinuities and the water presence The use of the RMR index as a unique regionalized variable, as usually done [24][25][26][27][28][29][30][31], can constitute a conceptual mistake, because the RMR considers parameters with different origin, assigning them different weights, and so each parameter is not considered in an independent way. It is worth noting that, considering only the final RMR value and not the individual parameters, geostatistical analysis becomes easier and faster; this approach could be reasonable to assess the rock mass quality in a wide area and especially to individuate the critical sites without understanding why low RMR values occur, i.e.…”

Section: Local Rock Mass Propertiesmentioning

confidence: 99%

“…Furthermore the indicator kriging needs an indicator transformation, which always implies a loss of information: the extra information about significant high or low values which fall within the same class is lost, actually whether a value is only a little bigger or very bigger than the chosen threshold does not play a role. The ordinary kriging, which has been already used two times in the RMR estimation [26,31], has been chosen to take in account the entire data set. The ordinary kriging is the technique that provides the Best Linear Unbiased Estimator of unknown fields [56,58], furthermore this method is a local estimator that provides the interpolation and extrapolation of the originally sparsely sampled data in the whole domain, assuming that the values are reasonably characterized by the Intrinsic Statistical Model.…”

Section: Predictionmentioning

confidence: 99%

“…Several authors have applied the geostatistical approach to analyze rock mass fracture-distribution [7][8][9][10][11][12][13][14][15] or rock mass specific properties [16][17][18][19][20][21][22][23]. The RMR index has been estimated using geostatistical analysis since 2004 [24][25][26][27][28][29][30][31], especially for tunnel projects; the kriging method has usually been applied to borehole data, sometimes integrated by geophysical surveys, with a secondary and only qualitative role. RMR values have always been considered as a single regionalized variable, and not as the sum of more variables.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Rock Mass Rating spatial estimation by geostatistical analysis

Ferrari

Apuani

Giani

2014

International Journal of Rock Mechanics and Mining Sciences

View full text Add to dashboard Cite

This work aims to estimate the Rock Mass Rating of 200 km 2 area of the Italian Central Alps, along San Giacomo Valley (province of Sondrio). The regional geological setting is related to the Pennidic Nappe arrangement, which is characterized by the emplacement of sub-horizontal gneissic bodies, separated by meta-sedimentary cover units. The resulting RMR map can be a useful tool to forecast the quality of outcropping rock masses as well as to derive their geomechanical behaviour. Almost 100 geomechanical field surveys have been carried out in the research area, in order to characterize the outcropping rock masses; afterwards rock mass quality indexes have been evaluated in each surveyed site. In order to estimate the Rock Mass Rating values in un-sampled locations, different geostatistical techniques (kriging and simulations) have been applied, using both bi-dimensional and almost three-dimensional approaches. The validation process shows that kriging tends to produce smoothened distributions, while conditional simulations allow respecting local extreme values. Although geostatistical analysis reveals that geomechanical properties show spatial correlations, it is to remind that rock mass quality is strongly related to its geological and structural history.

show abstract

Prediction of rock mass along tunnels by geostatistics

Cited by 16 publications

References 3 publications

Assessment of long-term deformation of a tunnel in soft rock by utilizing particle swarm optimized neural network

Assessment of long-term deformation of a tunnel in soft rock by utilizing particle swarm optimized neural network

Application of neural networks for the reliability design of a tunnel in karst rock mass

Rock Mass Rating spatial estimation by geostatistical analysis

Contact Info

Product

Resources

About