Publication: Evaluation system for twin tunnel interaction in kenny hill formation incorporating simplified 2d fem analysis
Loading...
Date
2022-04-01
Authors
Mohammad Zaki, Mohd Faiz
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The main risk of the Sungai Buloh – Kajang (SBK) KVMRT twin tunnels excavation project, located in the Kenny Hill Formation (KHF) and the contact zone between KHF and Kuala Lumpur Limestone is ground surface settlement; thus, an evaluation system for twin tunnel interaction is required. In the first stage, the in-situ tests (such as SPT, PMT and MASW) and laboratory tests were performed and empirical correlations were proposed. 2D and 3D subsurface models were developed based on soil properties obtained from in-situ testing. Microzonation maps were generated to clarify the location of contact and non-contact zones. Furthermore, 3D twin tunnel models were simulated as conceptual models for 2D FEM. Sensitivity analysis based on input soil parameters and model elements in 2D FEM was studied to produce an optimal tunnel conceptual model. Meanwhile, the constitutive soil model, Mohr-Coulomb (MC), Hardening Soil (HS), and Hardening Soil-Small Strain (HSS), were also considered in the analysis. The simplified 2D FEM of tunnel analysis, i.e., Lining Contraction Method (LCM), Stress Reduction Method (SRM), and Modified Grout Pressure Method (MGPM), were assessed for their performances in predicting ground settlement. Based on the results, the dominant lithology is the sandy silt soil type, with 51% of the total distribution. Meanwhile, the active MASW is efficient in assessing Vs at 5 – 15 m and the passive method is suitable for measuring Vs at depths up to 30 m. However, hybrid MASW is the most effective method due to it combining signals from both methods. For the microzonation maps, the contact zone was classified as Class C, and the non-contact zone was Class B. For correlations between NSPT to EM, Eur, and NSPT to Vs, increasing the value of NSPT is corresponding to the high value of EM, Eur and Vs. Based on the Group Method of Data Handling
(GMDH) neural network model, NSPT is the most consistent parameter that increased
with the increase of EM. Five lithology-based zones were effective in representing the
subsurface geological and geotechnical boundary models of the study area. Model 2
[exponential value of 5 in inverse distance weighting (IDW) method] accurately
produced the top-surface profile as measured on-site. For the FEM part, the efficiency
of the local mesh using ‘Very Coarse’ is less focused on tunnel elements. The higher
Rint for the soil-structure interface increases the stress distribution to the surrounding
soil. Soil heave occurs when a lower Rint = 0.01 is used. For model widths, the 100 m
and 50 m dimensions show similar ground settlement trough for parallel tunnel
configurations, while inclined and vertical tunnel configurations require a width of 100
m. The HSS model predicted the most similar ground settlement trough to the on-site
monitored data. The contraction ratio obtained is less than 1.35% using the HSS, HS,
and MC-Eur models. The HS model produces more realistic results than the HSS model
based on the stress reduction method with -values obtained ranging from 0.2 to 0.6,
which is considered relevant for heterogeneous residual soils in KHF. The minimum
face pressure for the entire tunnel section is 250 kN/m2, while the maximum face
pressure is 450 kN/m2, which is considered acceptable based on the HS model.
Nevertheless, the HS model is the most practical and reliable method for all tunnel
sections/configurations in this KHF.