Engineering characteristics and classification based on JSCE in portal zones of Tunnel A, Yogyakarta - Bawen toll road, Magelang District, Central Java Province

The Yogyakarta - Bawen Twin Tunnels emerge as an environmentally conscious solution for Toll Road Phase 5 (STA 6+300 to STA 27+640), addressing the ecological impact of open-cut excavation, which demands a staggering 1,209,056 m3. Open-cut excavation not only incurs substantial operational costs, encompassing disposal and treatment, but also poses threats like groundwater loss and erosion-induced landslides. The twin tunnels mitigate these concerns by limiting excavation volume to a mere 1/8, averting substantial costs and ecological damage. Challenges persist, particularly in the portal zones of Tunnel A, characterized by a shallow overburden and exposure to weathering. The Japan Society of Civil Engineers (JSCE) Method proves instrumental in assessing ground quality in these zones, considering factors such as competence, unit weight, unconfined compressive strength, depth of cover, elastic wave velocity, boring core condition, and geological characteristics. The analysis reveals a ground quality classification of E for the portal zone in Tunnel A, emphasizing the necessity for a meticulous design approach that incorporates analytical methodologies alongside empirical considerations. Notably, the unconfined compressive strength ranges from 32.89 to 472.88 kPa, and the elastic wave velocity spans 402 to 2000 meters/sec, reinforcing the imperative for a comprehensive and analytically informed design for excavation and support systems.


Introduction
Yogyakarta -Bawen Twin Tunnels are located in Magelang District, Central Java Province.These tunnels are part of Yogyakarta -Bawen Toll Road Phase 5 (STA 6+300 to STA 27+640) which aims to improve connectivity between Central Java and the Special Region of Yogyakarta.These tunnels are the solution to environmental preservation because open-cut excavation of 1,209,056 m 3 could be prevented and only 1/8 of the volume could be carried due to tunnels.These tunnels also minimize the loss of groundwater infiltration and prevent the potential of erosion in open-cut excavations, which can cause landslides.Since the road tunnels will be constructed underground, engineering characterization and classification are essential.
Yogyakarta -Bawen Twin Tunnels consist of Tunnel A (Direction to Yogyakarta) dan Tunnel B (Direction to Bawen).Based on the regional geological map, tunnels are located in the quarternary 2 volcanic zone.Volcanic breccia, lava flows, tuff, and weathered volcanic rock are probably present.Moreover, the portal zone of Tunnel A is located in front of the river at the inlet and has the smallest overburden at the outlet.Portal zone, where the overburden generally is small, is frequently located in a generally weathered zone and unconsolidated deposits.Because of weathered zones and unconsolidated deposits, the base of portal zones often suffers from settlement and deformation because of the insufficient bearing capacity of the ground [1].Because of the uncertainty of underground conditions in portal zones of Tunnel A, this research is necessary to be carried out.This research aims to assess comprehensively the quality of the ground condition of portal zones of Tunnel A based on megascopic evaluation, microscopic evaluation, laboratory test, and geophysics data so that a better understanding of the ground will be performed.
Seven boreholes have been evaluated directly.Megascopic evaluation is conducted at the core box.Thin section analyses, by using a petrographic microscope, were conducted with two kinds of approach: soil micrograph analysis and volcanic rock analysis.Based on the study soil micrograph of 6 (six) samples, underground materials have been altered into clay minerals in about 95% of the tunnel area.Six sample thin sections still show crystals like opaque, iron oxide, quartz, and plagioclases at the microphotograph.However, thin section analysis at TN 14A 74-75 M (Borehole ID: TN 14A; Depth from 74 to 75 meters) shows differences.Minerals like hornblende, clinopyroxenes, and plagioclases appear.The laboratory test result was produced from 26 samples underground (Figure 2).Data showed the underground materials consisted of fat clay (CH), elastic silt (MH), and silty sand (SM) by using the USCS (Unified Soil Classification System) [2].

Geological Condition
Based on the regional map, the underground condition of Tunnel A's portal zone probably was composed of Gilipetung Volcanics Lithology (Qg) and Kaligetas Formation (Qpkg).Gilipetung Volcanics Lithology was formed from vesicular lava flows.Kaligetas Formation was formed with more variety.This formation consists of volcanic breccia, lava flows, tuff, tuffaceous sandstone, and claystone.Dashes on the map (Figure 1. (a)) show lineament was produced from air photos and remote sensing images.Moreover, there was no found of lineament in the research area from outcrop.Based on megascopic evaluation at the surface, the outcrop in the research area is mostly a highly weathered category [3].Merapi -Merbabu Fault [4] is found on the right side of the tunnel.However, it was outside the 10-km radius of the tunnel.

Material and Methods
The Japan Society of Civil Engineers published Standard Specification for Mountain Tunnel in 2007.This standard is well known as The JSCE [4]1.The classification from The JSCE works well for mixed face (either soil face or soil-rock face).Ground competence factor, unit weight of the ground, unconfined compressive strength, depth of cover, elastic wave velocity, boring core condition, and geological conditions are parameters in determining the ground's quality based on The JSCE (2007).The ground competence factor is established as follows: Where qu = unconfined compressive strength of the ground (kN/m 2 ); γ = unit weight of the ground (kN/m 3 ); H = depth of cover (m).For ground in which the presence of fractures can be ignored, the unconfined compressive strength of the specimen can be adopted as the unconfined compressive strength of the ground, but for ground in which the effect of fractures is large, the quasi-strength of rock mass qu' (kN/m 2 ) shall be used and calculated as follows: Where vp = elastic wave velocity of the ground (P wave, km/s); up = ultrasonic wave velocity of the specimen (P wave, km/s); and qu = unconfined compressive strength of the ground (kN/m 2 ).The value of qu can be determined from an unconfined compressive strength (UCS) test in the laboratory.In finegrained soil, qu is closely related to the value of undrained shear strength (Su) which has the equation as follows: However, both qu and Su can be determined also from the N-SPT value.Many studies show a correlation between qu and N-SPT values.The first research to determine the correlation between qu and N-SPT value was done by Terzaghi & Peck (1967).Hettiarachchi & Brown (2009) [5] showed an empirical relation between Su and N-SPT values which has the equation as follows: For weak rocks, Clayton (1995) [6] provides the relationship between unconfined compressive strength (UCS) and the corrected N-SPT value (N60) as follows: N60 is the number of blows that require penetration with 60% of the energy to the falling hammer.The N60 [2] is given as follows: N60 =   .  .  .  . 0.60 (6) Where Em is the hammer efficiency; CB is the borehole diameter correction; CS is the sample barrel correction; CR is the rod length correction; and N is the raw N-SPT value recorded in the field.
On the other hand, the samples at TN 03A 69-70M, TN 04A 65-66M, TN 14A 66-67M, and TN 14A 74-75M showed differently.Discolored rock was present as corestones at TN 03A 69-70M, TN 04A 65-66M, TN 14A 66-67M, and TN 14A 74-75M.Thin section analysis was analyzed for corestones to define the name of rock by using classification and nomenclature of volcanic rock.The thin section analysis was conducted at TN 14A 74-75M (Figure 3) which resulted from andesite [8] with an abundance of 50% of clay minerals.This condition was aligned with the sieve analysis at TN 03A 69-70M and TN 04A 65-66M which resulted total fine-grained material was about 45%.This sample belongs to extremely weak rock because could be indented by a thumbnail [9].Based on the weathering degree [3], this material belongs to highly weathered rock because more than half of the rock is decomposed and disintegrated into soil.When using a classification based on USCS, TN 14A 74-75M belongs to silty sand strata with the percentage of course-grained material is about 55%.Based on zonal characteristics, TN 14A 74-75M is categorized as soil with corestones: less than 30% rock.Figure 4, Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, and Figure 10 showed photomicrographic appearance taken with plane-polarized light (PPL) and with coressed polars (XPL) by petrographic microscope.Figure 4 until Figure 9 were analyzed by soil-micromorphology-analysis since the megascopic evaluation showed completely weathered rock material that had been decomposed and disintegrated into soil.The present of nodules at TN 04A 19-20M (Figure 6) and TN 15A 39-39.5M(Figure 8) indicated the nodule was inherited from the parent material which could be formed in situ when weathering [7].Lithic appeared also at TN 03A 9-10M (Figure 5), TN 14A 39-40M (Figure 7), and TN 16A 22-23M (Figure 9) which have been weathered.Minor minerals like opaque, quartz, plagioclases, and iron oxide appeared as well when analyzing these samples.However, TN 14A 74-75M (Figure 10) showed differently.Plagioclases (Pl), clinopyroxenes (Cpx), hornblende (Hb), and clay minerals (Cly) were identified from the corestone (Figure 3).After plotting at QAPF diagram for volcanic rocks, andesite resulted based on classification and nomenclature of volcanic rocks.Even though megascopic evaluation showed the original mass structure of the ground was still intact and the fabric was still preserved, the origin minerals of fresh rock materials are hard to identify since the majority of minerals are only clay minerals (Figure 4 -Figure 9, Table 2 -Table 3).Based on the ground material analysis, completely weathered rock material that decomposed and disintegrated into the soil could be simplified into 3 layers referring to soil classification using USCS (  3 showed laboratory test results from specific gravity, unit weight, water content, liquid limit, plastic limit, plasticity index, initial void ratio, porosity, p200 (percentage material passing through the No. 200 sieve), and unconfined compression test.The unconfined compression test was only conducted for borehole BH-01 and did belong to fat clay (USCS Symbol: CH).The ground condition thus could be modelled into 4 layers including a highly weathered andesite layer in which corestones less than 30% appeared at this layer (Figure 11 and Figure 12).
To model the layer, soil consistency [5] was described as well on 4 (four) boreholes: TN 01A, TN 18A, TN 19A, and TN 20A.Data showed the N-SPT value of TN 18A, TN 19A, and TN 20A was less than 20, 24, and 19, respectively.Forth layer at TN 01A could be identified since Pwave > 1500 m/s and N-SPT > 60.The data of qualitative evaluations and megascopic evaluation was also gathered from 7 (seven) boreholes.However, only the selected height from every borehole was evaluated.All the data, including laboratory data, was then arranged in core running for every single borehole (Figure 11 and Figure 12) so that a better understanding of the ground's layers was performed.N-SPT data analyses due to soil consistency evaluation were conducted as well at inlet Tunnel A. Even though the first and third layers have the same group symbol they have different range consistency of soil.Based on N-SPT data, the mean value of Layer 1 is 2.87 (mean N-SPT value = 2.50 TN 01A; 1.80 at TN 03A; 3.00 at TN 04A; 2.67 at TN 05A).On the other hand, the mean value of Layer 3 is 29 (mean N-SPT value = 23.18 at TN 01A; 29.40 at TN 04A; 25.55 at TN 05A).
Since the model of layers had been considered and consisted of 4 layers (Figure 11 and Figure 12), both inlet and outlet Tunnel A, the engineering characteristic for JSCE was then analyzed and compiled from the laboratory data (Table 2 and Table 3), the empirical calculation (Eq.( 3), ( 4), (5), and ( 6)) dan geophysical data.Index properties were compiled from 26 samples from 7 (seven) boreholes: TN 03A, BH-01, TN 04A, TN 14A, TN 15A, TN 16A, and TN 17A.Elastic wave velocity was concluded with two kinds of geophysical data: downhole seismic and seismic refraction testing.Furthermore, the RQD of Layer 4 was derived from the limitation value RQD of weak rock [10].Tunnel A penetrates 3 (three) layers at the portal zone at outlet Tunnel A (Figure 11.B).However, the tunnel only penetrates 1(one) layer (Figure 12. B).
Layer 1, Layer 2, Layer 3 and Layer 4 have values of unit weight below 17 kN/m 3 .Layer 2 has a unit weight from (γ)from 13.34 to 16.57 kN/m 3 , with an average of 15.2 kN/m 3 .Parameter qu for Layer 2 had already been derived from a lab test, with an average of 74.20 kPa and a slightly different empirical calculation.Because the calculation is dependent on the depth of cover, the evaluation will be given to 5 (five) selected points (see Figure 11.B and Figure 12.B).It resulted in the competence factor (Eq. 1) of 5 points being less than one and the ground category both inlet and outlet zone did belong to the E category (Table 5).

Conclusion
The megascopic evaluation showed the original mass structure of the ground was still intact and the fabric was still preserved, the origin minerals of fresh rock materials are hard to identify from the ground at Layer 1, Layer 2, and Layer 3. Either fresh or discolored rock as a discontinuous framework or corestones wasn't found at these three layers.Thin section analysis showed the present of nodules and weathered lithic.The present of nodules at TN 04A 19-20M (Figure 6) and TN 15A 39-39.5M(Figure 8) indicated the nodule was inherited from the parent material which could be formed in situ when weathering.Sieve analysis was conducted also for these layers and the total sample was 23 samples.
Based on the sieve analysis, excluding TN 03A 69-70M and TN 04A 65-66M, total fine-grained material was from 63% to 98%, with the average was about 92%.Based on the ground material analysis, completely weathered rock material that decomposed and disintegrated into the soil could be simplified into 3 layers based on soil classification using USCS (Table 2 and Table 3).On the other hand, the samples at TN 03A 69-70M, TN 04A 65-66M, TN 14A 66-67M, and TN 14A 74-75M showed differently.Discolored rock was present as corestones at TN 03A 69-70M, TN 04A 65-66M, TN 14A 66-67M, and TN 14A 74-75M.Thin section analysis was analyzed for corestones which resulted in andesite [8].This condition was aligned with the sieve analysis at TN 03A 69-70M and TN 04A 65-66M which resulted total fine-grained material was about 45%.
To model the layer, soil consistency [5] was described as well on 4 (four) boreholes: TN 01A, TN 18A, TN 19A, and TN 20A.Data showed the N-SPT value of TN 18A, TN 19A, and TN 20A was less than 20, 24, and 19, respectively.Forth layer at TN 01A could be identified since Pwave > 1500 m/s and N-SPT > 60.The data of qualitative evaluations and megascopic evaluation was also gathered from 7 (seven) boreholes.The ground condition thus could be modelled into 4 layers including a highly weathered andesite layer in which corestones less than 30% appeared at this layer (Figure 11 and Figure 12).
The engineering characteristic for JSCE was analysed and compiled from the laboratory data, empirical calculation dan geophysical data.Index properties were compiled from 26 samples from 7 (seven) boreholes: TN 03A, BH-01, TN 04A, TN 14A, TN 15A, TN 16A, and TN 17A.Elastic wave velocity was concluded with two kinds of geophysical data: downhole seismic and seismic refraction testing.Layer 1, Layer 2, Layer 3 and Layer 4 have values of unit weight below 17 kN/m 3 .Layer 2 has a unit weight from (γ)from 13.34 to 16.57 kN/m 3 , with an average of 15.2 kN/m 3 .Unconfined compression strength (qu) = 32.89kPa, 74.20 kPa, 182.12 kPa, and 472.88 kPa, respectively.Because the calculation is dependent on the depth of cover, the evaluation will be given to 5 (five) selected points (see Figure 11.B and Figure 12.B).Analyses showed that the competence factor of 5 points was less than one so the ground's quality at portal zones of Tunnel A did belong to the E category (Table 5).For this ground category, the typical support system: is 40-cm-thickness of the arch, 40-cm-thickness of the wall, 50-cm-thickness of invert, steel support H-200 for every 1 meter or less, and 25-cm-thickness of shotcrete.The pre-support system was considered also as a face bolt, pipe for piling, backward grouting, and well-point drainage since the overburden was small and the groundwater table was shallow at the tunnel.However, based on the JSCE, for E Category support system and excavation method should be designed not only empirically but also by analytical method.This research needs to be continued since the analytical method by numerical calculation is necessary.

Figure 2 .
Figure 2. Laboratory Test for Portal Zone Evaluation at the Inlet [A] and Outlet [B] of Tunnel A.Sample ID was named based on the borehole ID followed by the depth of the sample.

Figure 3 .
Figure 3. Sample Thin Section from TN 14A 74-75M.This sample was already washed in running water flow and cleaned from soil.

Figure 11 . 9 Figure 12 .
Figure 11.Ground Evaluation and Ground Condition of Outlet Tunnel A

Table 1 .
Total Sample Test Distributed Based on Model of Layer

Table 2 .
Laboratory Test Results for Portal Zone at Outlet Tunnel A

Table 3 .
Laboratory Test Results for Portal Zone at Inlet Tunnel A

Table 4 .
Empirical Calculation for UCS Value

Table 5 .
Engineering Classification based on JSCE