Acta Polytechnica CTU Proceedings doi:10.14311/APP.2017.10.0056 Acta Polytechnica CTU Proceedings 10:56–60, 2017 © Czech Technical University in Prague, 2017 available online at http://ojs.cvut.cz/ojs/index.php/app DESIGN OF SUBSOIL IMPROVEMENT BELOW HALL FLOORS Peter Turček∗, Monika Súľovská STU Bratislava, Faculty of Civil Engineering, Radlinského 11, Bratislava ∗ corresponding author: peter.turcek@stuba.sk Abstract. The construction of an industrial park is now being prepared near the town of Nitra. The investor fixed very strict conditions for the bearing capacity and, above all, the settlement of halls and their floors. The geological conditions at the construction site are difficult: there are soft clay soils with high compressibility and low bearing capacity. A detailed analysis of soil improvement was made. Stone columns were prepared to be fitted into an approximately 5 m thick layer of soft clay. The paper shows the main steps used in the design of the stone columns Keywords: Engineering geological investigation, soil improvement, gravel columns, decreasing the settlement, consolidation. 1. Introduction The construction of an industrial park was prepared near the town of Nitra in 2015. Difficult geological conditions, created by a layer of soft to firm Quater- nary clays approximately 5 m in thickness resulted in unfavourable soil characteristics: low bearing capacity and high settlement. For this reason, it was decided to improve the soil properties. Following the evaluation of a preliminary engineering geological investigation, an initial design of stone columns was made during detailed investigation works. Before the final design, the efficiency of the stone columns was verified by field tests. Three test fields for three various improvement technologies were prepared. Penetration tests were applied before and 10 days after the stone columns completion. In test field No. 1, stone columns with a diameter of 500 to 600 mm were prepared using the technology of deep vibrating compaction, and filled with gravel of grain size 4/15 mm, 8/32 mm and 4/32 mm. The stone columns were arranged to form a square and triangular grid. In test field No. 2, dynamic compaction com- bined with stone columns preparation using a non- traditional technology was used for soil improvement. First, a borehole with a diameter of 600 mm was made and filled with gravel of grain size 0/63 mm or 0/90 mm containing low contents of fine particles. Soil particles smaller than 0.063 mm formed less than 5 %. The boreholes were not protected by casing pipes; they had to be filled by gravel as soon as possi- ble to avoid the collapse of boreholes. The next step was compaction of the stone column in two phases by 50 hits. The stone columns were spaced within a square and triangular grid with varied distances from each other. In field test No. 3, dynamic consolidation with a weight of 20 t falling in free fall from a height of 10 m was performed. Using this technology created craters approximately 2 m in diameter. This squishy space was filled by stones with grain size 0-400 mm. The content of fine soil was very low. The space of compacted columns formed a square and triangular grid with a spacing of more than 5 m. After the evaluation of data from field experiments [1], dynamic consolidation was rejected, because the results had proved this technology improper for ap- plication in local geological conditions. Also, the set- tlement of columns produced by dynamic compaction was higher than by the other technologies. As the depth reached by dynamic compaction was not suf- ficient for the project requirements, this technology was also rejected. Field experiments helped to prepare recommenda- tions which became input data for soil improvement. More halls with different floor loading will be con- structed in the industrial park in the near future. The paper focuses on the design of soil improvement for uniform floor load intensity of 60 kPa. 2. Design of stone columns Various connections of columns to other structural layers were analysed before the design of the stone columns. After removing the topsoil, the separa- tion geotextile was put on the ground level. Then, a 500 mm thick gravel layer was laid as a base below the embankment. The 500 mm layer was made up of two layers of crushed stone with grain size 0/125 mm compacted to min. ID = 0.85. From this working plane, the stone columns were assembled. After they had been completed, the ground surface was levelled and compacted, as needed, to min. ID = 0.85 again. The embankment was built on this base layer. Its first three layers levelled the non-horizontal terrain. The main part of the embankment was designed from these layers: 200 mm of the bottom layer of crushed stone 0/125 mm, 300 mm of pit-run gravel with small volumes of fine soil (grain size max. 100 mm), then 56 http://dx.doi.org/10.14311/APP.2017.10.0056 http://ojs.cvut.cz/ojs/index.php/app vol. 10/2017 Design of Subsoil Improvement below Hall Floors again 200 mm of crushed stone 0/125 mm and 300 mm of pit-run gravel with small volumes of fine soil (grain size max. 100 mm). The last 800 mm of the embank- ment were prepared from a 400 mm layer made up of crushed stone 0/63 mm and two 200 mm layers of crushed stone of grain size 4/32 mm. The main reason for using different types of material was the possibility of their delivery to the site. The length of the construction site was 1380 m. A schematic sectional view of improved subsoil is in Fig. 1, while Fig. 2 presents an overall view of making the stone columns. One part of the construction site will contain halls whose floors will be loaded by 60 kPa. The investor’s requirement for the settlement of the floors was 10 mm. The thickness of the embankment situated on im- proved subsoil was 1.5 m. Based on a parametric study, a raster of stone columns (SC) 1.8 · 1.8 m in a square grid was selected. The average length of the stone columns was 5.5 m. This length resulted from the requirement that the bottom of the columns would extend into the quarternary gravel layer. It was necessary to verify the bearing capacity, settle- ment and consolidation of the stone columns for the prepared project. The bearing capacity validation manifested great reserve and it will be not be dealt with this paper. A traditional analytical calculation method was applied for the calculation of settlement and consolidation, partly complemented by Priebe’s theory for vertical gravel stone columns. 2.1. Traditional calculation of settlement and consolidation After the withdrawal of a humus layer, a 0.5 m thick gravel layer would be prepared on the ground. The stone columns would be built from this level. Cal- culations were made for a simulated model area of 20 · 20 m, as there must a reserve left for transport corridors and, if necessary, a free space for technology in the halls. When a stone column with a diameter of 600 mm is made, the cross-sectional area equal to one stone column is A1 = π r2 = π 0.32 = 0.283 m2 (1) Using the distribution of the stone columns in a square grid with a distance between the columns of 1.9 · 1.9 m, 27.67 columns will be placed in an area of 10 · 10 m (5.26 columns fit in a line up to 10 m in length, the distance between the rows is 1.9 m and the 10 m section holds 5.26 columns). Therefore, the total area of the columns will be Ac = A1 · 27.67 = 0.283 · 27.67 = 7.83 m2 (2) The original soil will occupy the area As = 100− 7.83 = 92.17 m2 (3) The following bedrock model was compiled for the area where a hall with a floor loaded by 100 kPa will be placed: • 3.3 m layer of clay of high plasticity with mainly soft consistency; where Edef = 3 MPa; Eoed = Edef/β = 3/0.37 = 8.1 MPa (4) • 1.2 m thick layer of clayey sand with mainly firm consistency; where Edef = 5.95 MPa; Eoed = Edef/β = 5.95/0.62 = 9.6 MPa (5) • 1.0 m thick layer of silty gravel with medium relative density; where Edef = 35 MPa; Eoed = Edef/β = 35/0.74 = 47.3 MPa (6) • Tertiary clay with high plasticity and firm consis- tency was found below silty gravel; where Edef = 4 MPa; Eoed = Edef/β = 4/0.37 = 10.8 MPa (7) The average oedometric modulus value in a 3.3 m thick clay layer was determined using the equation: Eoed = E1 ·A1 + E2 ·A2 A1 +A2 = = 166 · 7.83 + 8.1 · 92.17 100 = 20.46 MPa (8) Similarly, for the average value of clayey sand the oe- dometric modulus determined was Eoed = 21.85 MPa, for silty gravel Eoed = 56.59 MPa and for Tertiary clay Eoed = 22.95 MPa. The calculation of settlement was made (as de- scribed above) for an area of 20 · 20 m. The embank- ment layer thickness of 1.5 m acts on the surface of the original ground by a stress with an intensity of 1.5 · 20 = 30 kPa. It was anticipated that a soft clay layer 3.3 m thick with clayey sand and Tertiary clay would be compressed after improvement by 5.5 long stone columns in a raster of 1.9·1.9 m (fixing the stone columns 1.0 m into the gravel layer was considered). The calculation is processed in Tab. 1. In the same way, the settlement due to a uni- form surcharge of the floor by 60 kPa was calcu- lated. The expected settlement under this load is s2 = 9.012 mm. The total estimated settlement (embankment + surcharge of the floor) will reach stot = 4.886+9.012 = 13.898 mm. But the magnitude of this settlement needs to be corrected taking into account consolidation, because part of the settlement will occur during the construction time. The calculation of consolidation in a drainage sys- tem consists of two parts: consolidation into vertical elements (stone columns) and consolidation into the horizontal permeable subbase. Then, using Terza- ghi’s theory the average degree of consolidation can be expressed as 57 Peter Turček, Monika Súľovská Acta Polytechnica CTU Proceedings Figure 1. Schematic sectional view of improved subsoil. Figure 2. Overall view on making stone columns. Point no. h z z/B I2 σ2 σor m2 · σor Eoed s (m) (m) (kPa) (kPa) (kPa) (MPa) (m) 1 1.6 0.80 0.040 0.98 29.4 16 1.6 20.46 0.002174 2 1.7 2.45 0.122 0.85 25.5 49 4.9 20.46 0.001712 3 1.2 3.90 0.195 0.71 21.3 78 7.8 21.85 0.000741 4 1.0 5.00 0.250 0.62 18.6 100 10.0 56.59 0.000152 5 2.8 6.90 0.345 0.52 15.6 138 13.8 47.30 0.000107 6 2.0 9.30 0.465 0.43 12.9 186 18.6 10.80 - 7 2.0 11.30 0.565 0.38 11.4 226 22.6 10.80 -∑ s1 = 0.004886 Table 1. Settlement of the embankment (σ = 30 kPa) after improvement by SC in a raster of 1.9 · 1.9 m. Explanatory notes: h - thickness of the soil layer; z - distance between the footing bottom and the point at which vertical stress is calculated; B - width of the foundation; I2 - coefficient of stress spreading influence; σ2 - vertical stress from the foundation at a depth z; σor - original vertical stress at depth z; m2 - coefficient of structural strength; Eoed - oedometric modulus of soil; s - settlement of the soil layer with a thickness h. 58 vol. 10/2017 Design of Subsoil Improvement below Hall Floors U = 1− (1− Uh) · (1− Uv) (9) where: Uh - degree of consolidation for horizontal dewater- ing, Uv - degree of consolidation for vertical dewatering. Firstly, the average degree of subgrade consolidation was determined: Eoed = ∑ Ei · hi∑ hh = 8.1 · 3.3 + 9.6 · 1.2 3.3 · 1.2 = 8.5 MPa (10) In the calculation of consolidation in the vertical direction, the time factor Tv was determined: Tv = tvkvEoed γwh2 = 1.0368 · 107 · 1 · 10−9 · 8.5 0.01 · 4.52 = 0.435 (11) where: tv - duration of vertical consolidation - 4 months (4 · 30 · 86400 = 1.0368 · 107 s); kv - filtration coefficient in the vertical direction (1 · 10−9 m/s); Eoed - oedometric modulus of soil (8.5 MPa); γw - specific gravity of water (0, 01 MNm−3); h - thickness of the drainage layer of soil (4.5 m). From Terzaghi’s consolidation graph, we obtained the degree of consolidation Uv = 83 %. These boundary conditions have been taken into account for the calculation of consolidation in the horizontal direction: de = 1.13 · d = 1.13 · 1.9 = 2.147 (where d is the axial distance of drains). The time factor Th for determining the degree of consolidation Uh is: Th = thkhEoed γwd2 e = 1.0368 · 107 · 1 · 10−9 · 8.5 0.01 · 2.1472 = 1.91 (12) where: th - duration of horizontal consolidation - 4 months (4 · 30 · 86400 = 1.0368 · 107 s); kv - filtration coefficient in the horizontal direction (1 · 10−9 m/s); For n = de d1 = 2.147 0.60 = 3.58 ⇒ in this case, auxiliary graphs yielded Uh = 99 %. The resulting average degree of consolidation is then U = 1−(1−Uh)·(1−Uv) = 1−(1−0.83)·(1−0.99) = = 0.9983 ≈ 100 % (13) Thus, it can be concluded that as long as the floor is not loaded earlier than 4 months after the completion of embankment bodies, the consolidation below the embankment will be finished. The measurable settlement of the hall structure, therefore, should only be the component induced by the operating load in the hall. The value of this part of settlement is 9.012 mm. 2.2. Calculation of settlement using partially Priebe’s theory The solution comes from the definition of the equiva- lent diameter of the cell de. Auxiliary magnitudes: equivalent diameter: de = 1.13 · 1.9 = 2.147 m cross sectional area of the stone column: Ac = πr2 = π · 0.322 = 0.283 m2 area of a cell with a diameter de: As = πr2 = π · 1.07352 = 3.62 m2 ratio of the column area to the cell area: ac = Ac/As = 0.283/3.62 = 0.078 stress in the foundation base of the floor: σ = 60 kPa The basic design parameter is the so-called concen- tration ratio of stresses n = Ec Es = 180 3 = 60 (14) where: Ec is Young’s modulus of the stone column, Es is Young’s modulus of soil around the stone column. The settlement of subsoil with reinforcement by stone columns is expressed by the parameter β This parameter represents the ratio of the reinforced soil to original soil settlement. The parameter β can be approximately determined using the equation β = 1 1 + (n− 1) · ac = = 1 1 + (60− 1) · 0.078 = 0.1785 (15) The calculation of settlement caused by a 1.5 m high embankment (σ = 30 kPa) without improvement and by the floor followed. The expected settlement of the embankment was s3 = 11.854 mm and the floor settle- ment due to loading by 60 kPa was s4 = 20.150 mm. Then, the total settlement without improvement is stot = 11.854 + 20.150 = 32.004 mm. Using Priebe’s theory the assumed settlement of the stone columns is reduced to the value s = sn · β = 32.004 · 0.1785 = 5.713 mm (16) Comparing both theoretical methods we obtained the difference: 9.012 mm against 5.713 mm. This 59 Peter Turček, Monika Súľovská Acta Polytechnica CTU Proceedings difference can be explained in this way: in classi- cal theory medium stiffness of subsoil was assumed, while in Priebe’s theory the load is carried by the stone columns to a greater extent. According to both methods, the investor’s requirement was fulfilled: the settlement was less than 10 mm. 3. Conclusion The extremely exposed construction of an industrial park near the town of Nitra required improving an approximately 5 m thick layer of soft and firm Quater- nary clays. In the first step, the suitable technology of soil improvement was selected. To this end, a field experiment in a 1:1 scale was carried out. Deep vibrat- ing compaction was evaluated as the most suitable technology. In the second step, the design of stone columns spacing was made. The paper shows the procedure of the design of soil improvement below the floors exposed to uniform loading by 60 kPa. After the accomplishment of subsoil improvements of the whole area of the industrial park, the construction of halls started. 4. Acknowledgements The paper presents part of the results from a re- search project supported by the Slovak VEGA Agency No. 1/0882/16 “The Influence of Boundary Conditions on Limit States of Geotechnical Structures”. [2] [3] References [1] K. Ströhle, A. Tomozei. Execution of soil improvement in the trial area. Construction project Jaguar Land Rower, Nitra, Slovakia. ZT-Ströhle, Wien, 2015. [2] J. Soták, et al. Project darwin. final report of detailed engineering-geological investigation (in slovak). Tech. rep., TPA, s.r.o. Bratislava, 2015. [3] P. Turčekm, M. Súľovská. Rough landscaping and design of subsoil improvement using stone columns in an industrial park near the town of Nitra (in Slovak). T-G., Bratislava, 2015. 60 Acta Polytechnica CTU Proceedings 10:57–61, 2017 1 Introduction 2 Design of stone columns 2.1 Traditional calculation of settlement and consolidation 2.2 Calculation of settlement using partially Priebe's theory 3 Conclusion 4 Acknowledgements References