Research of Characteristics of Deep Cement Mixing Columns in Treatment of Soft Soil

Research of Characteristics of Deep Cement Mixing Columns in Treatment of Soft Soil

Tuan Anh Nguyen, Lecturer, PhD.

Ho Chi Minh City University of Transport 2 Vo Oanh Street, Binh Thanh District, Ho Chi Minh City, Viet Nam.

Dat Thanh Nguyen

Ho Chi Minh City University of Transport 2 Vo Oanh Street, Binh Thanh District, Ho Chi Minh City, Viet Nam.

Abstract – The article presents the content assessment of the application of deep cement mixing technology. From laboratory and field experiments with geological conditions in Duyen Hai – Tra Vinh areas, Vietnam, we will find out the factors that affect the quality and durability of the deep cement mixing column. At the same time, we can determine the optimal ratio about content of cement and water for soil samples after being reinforced to meet the economic – technical requirements. Finally, test data is analysed to serve calculations and simulation based on linear regression models of Microsoft Excel.

Keywords- Deep cement mixing columns, soft soil, optimal ratio of water and cement, linear regression models.

1. INTRODUCTION

   Deep cement mixing columns (DCMCs) are the soil at the construction site and the cement is grouted to the ground by the injection grouting pump. The drill bit is drilled down for loosening the soil until it reaches the depth of the soil layer, that needs to be reinforced, then it comes back and moves up. In the process of moving up, cement is grouted into the ground. This is a new technology applied in flooded areas where other types of columns do not meet the requirements.

   DCMCs are widely applied in the treatment of foundation and soft ground for public construction works such as bridges, ports, embankments, repairing the leaking for the sewer sides and sewer bottoms, retaining walls, reinforcing the soil around the tunnel, preventing landslide of the slope, strengthening the roadbed, bridge abutments, etc. Especially in the Mekong Delta, DCMCs are frequently used because this area is often flooded due to high tides and climate change.

   With specific advantages in soft ground treatment, the technology of deep cement mixing column is widely used to reinforce the ground, control the subsidence when the work is put into use. The calculation for design of the ground reinforced by the method of DCMCs is based on many different points of view and assumptions.

   However, based on the specific conditions of soft soil, topography, geological conditions, construction methods, working conditions of DCMCs and experience, the most appropriate calculation method is chosen.

   In addition, the theory of calculating DCMCs is quite limited, when parameters of durability of the pile are used in calculation, they need to be experimented with each specific soil and construction.

2. RESEARCH METHODS

   Empirical method: Performing experiments in the laboratory and on the site to determine the specific characteristics and factors affecting the durability and quality of DCMCs.

   Theoretical method: Summarizing and selecting experimental data based on linear regression model.

3. TESTING METHODS FOR DETERMINING CHARACTERISTICS OF DCMCs

1. Parameters of soil and materials used for the experiment

   Samples of soil ground used in the experiment were taken in the field of Duyen Hai Thermal Power Plant, Duyen Hai District, Tra Vinh Province, Vietnam which all have the same characteristics of soft grey brown clay silt. Samples are taken in the field and carefully stored to ensure the natural criteria for the experiment. Parameters of ground soil and materials used for the experiment are synthesized in Tables I-IV.

   TABLE I. PHYSICAL-MECHANICAL PARAMETERS OF SOIL

   Parameter	Unit	The average value
   Moisture W	%	41.6
   Unit weight	kN/m3	17.36
   Void ratio e	–	1.123
   Cohesive force C	kN/m2	7.4
   Angle of internal friction	Degree	1.74
   Plastic limit WP	%	22.92
   Liquid limit WL	%	39.99
   Plasticity index IP	%	17.07
   Liquidity index IL	–	1.1

   Parameter	Unit	Standard (TCVN)	Limit	Sample test results
   Specific weight	kN/m3	4030-2003	–	29.6
   Volumetric mass	kN/m3	1772-87	–	1.21
   Standard consistency	%	6017-1995	–	26
   Time to start setting	minute	6017-1995	> 45	115
   Time to end setting	minute	6017-1995	< 420	320
   Volumetric stability	mm	6017-1995	< 10	2.33
   Fineness of grinding of the remainder on the 0.09mm sieve	%	4030-2003	< 10	1.5
   Bending strength at 28 days	kN/m2	6017-1995	–	9.75
   Compressive strength at 28 days	kN/m2	6017-1995	40	42.9

   Parameter	Unit	Standard (TCVN)	Limit	Sample test results
   Specific weight	kN/m3	4030-2003	–	29.6
   Volumetric mass	kN/m3	1772-87	–	1.21
   Standard consistency	%	6017-1995	–	26
   Time to start setting	minute	6017-1995	> 45	115
   Time to end setting	minute	6017-1995	< 420	320
   Volumetric stability	mm	6017-1995	< 10	2.33
   Fineness of grinding of the remainder on the 0.09mm sieve	%	4030-2003	< 10	1.5
   Bending strength at 28 days	kN/m2	6017-1995	–	9.75
   Compressive strength at 28 days	kN/m2	6017-1995	40	42.9

   TABLE II. PHYSICAL-MECHANICAL PARAMETERS OF CEMENT

   TABLE III. PHYSICAL-MECHANICAL PARAMETERS OF DOMESTIC WATER USED FOR SAMPLE PREPARATION

   Determination parameter	Unit	Water used to prepare the samples
   		Result	TCVN 302- 2004
   Color	Level	Colorless	Colorless
   Greasy scum	Level	No scum	No scum
   pH	Degree	7.4	4÷12.5
   Total amount of dissolved salt	mg/L	9.2	200
   Content of (SO4)2-	mg/L	23.5	600
   Content of Cl-	mg/L	52.6	350
   Total amount of suspended solid (SS)	mg/L	25	200

   TABLE IV. PHYSICAL-MECHANICAL PARAMETERS OF WATER AT THE SAMPLING LOCATION USED FOR SAMPLE PREPARATION

   Parameter	Unit	Water at the sampling location
   		Amount	According to TCXD 3994- 1985
   Temperature	0C	28.125
   pH	–	7.885
   Clearness	Sensing	Turbid
   Odor	Sensing	Non
   Free CO2	mg/L	12.194	No corrosion
   Corrosive CO2	mg/L	0
   Na+ + K+	mg/L	9109.644
   Ca2+	mg/L	285.82
   Mg2+	mg/L	863.208	No corrosion
   Cl-	mg/L	15575.86
   (SO4)2-	mg/L	1772.608	Strong corrosion
   (HCO3)-	mg/L	318.066
   (CO3)2-	mg/L	0.75
   Total hardness	mg/L	85.25
   Total mineralization	mg/L	27927.95	Medium corrosion

2. Uniaxial unconfined compression test

   1. Experimental results for mixing samples using domestic water with a ratio of water / cement respectively as 1.4; 1.0; 0.8

      1. Experiment 1: Water/Cement =1.4

         – Based on Fig. 1, we can see that the value of qu (kPa) from

         7 to 14 days of age increases rapidly by about 70 times compared to the original soil without reinforcement, then it tends to increase slowly until 28 days of age.

         • The intensity of deep cement mixing sample is proportional to the amount of cement in preparation. But when the amount of cement quickly grows from 14% to 20%, the intensity of cement tends to increase slowly. (Table V)

           TABLE V. UNIAXIAL COMPRESSIVE STRENGTH QU FOR MIXING SAMPLE USING DOMESTIC WATER WITH THE RATIO OF W/C =1.4

           Day	Uniaxial compressive strength qu (kPa)
           	M100 (6%)	M150 (9%)	M200 (12%)	M250 (14%)	M300 (17%)	M350 (20%)
           0	15	15	15	15	15	15
           7	218.3	533.3	687.3	790.2	914.3	1206.7
           14	293.4	702.6	1068.1	1226.8	1376.3	1609.1
           28	322.2	943.9	1339.9	1687.7	1779.5	1960.4

           Fig. 1. Uniaxial compressive strength of experiment 1

      2. Experiment 2: Water/Cement =1.0

         Uniaxial compressive strength qu for mixing sample using domestic water with the ratio of W/C=1 is synthesized in Table VI.

         TABLE VI. UNIAXIAL COMPRESSIVE STRENGTH QU FOR MIXING SAMPLE USING DOMESTIC WATER WITH THE RATIO OF W/C=1

         Day	Uniaxial compressive strength qu (kPa)
         	M100 (6%)	M150 (9%)	M200 (12%)	M250 (14%)	M300 (17%)	M350 (20%)
         0	15	15	15	15	15	15
         7	284	569.7	982.4	1097.9	1447.3	1818.5
         14	421.5	957	1394.5	1618.2	1959.2	2360.4
         28	554.1	1365	2076.7	2247	2736.5	3389.8

         Fig. 2. Uniaxial compressive strength of experiment 2

         The result in Fig.2 indicates that the value of qu (kPa) in this experiment increases more strongly compared to that of experiment 1, specifically at 28 days of age the value of qu (kPa) increases by about 137 times in comparison with that at 28 days of age after being reinforced. The main reason is due to the change of the water amount when mixing sample. Compared with experiment 1, in this experiment, when the water amount decreases by 40% when mixing, the value of qu (kPa) of the sample will increase by 55.24% at 28 days of age.

      3. Experiment 3: Water/Cement = 0.8

      Uniaxial compressive strength qu for mixing sample using domestic water with the ratio of W/C= 0.8 is synthesized in Table VII.

      TABLE VII. UNIAXIAL COMPRESSIVE STRENGTH QU FOR MIXING SAMPLE USING DOMESTIC WATER WITH THE RATIO OF W/C=0.8

      Day	Uniaxial compressive strength qu (kPa)
      	M100 (6%)	M150 (9%)	M200 (12%)	M250 (14%)	M300 (17%)	M350 (20%)
      0	15	15	15	15	15	15
      7	394.592	777.7	1518.9	1725.4	2107.4	285.1
      14	605.706	1313.3	2381	2582.07	2974.3	4099.6
      28	838.157	1720.1	2644.3	3493.06	4448.9	25866.3

      Based on Fig. 3, it is said that the value of qu (kPa) of deep cement mixing sample in this experiment increases more significantly than those of experiments 1 and 2. The uniaxial compressive strength of natural soil increases about 211 times at 28 days of age after being reinforced. Similar to the first two experiments, when we perform the experiment of changing the ratio of water/ cement on preparing the sample, it will significantly change the value of qu (kPa) of DCMCs.

      Fig. 3. Uniaxial compressive strength of experiment 3

   2. Experimental results for mixing samples using water at the sampling location with a ratio of water/cement = 1.0

      Experiment 4: W/C=1 (water at the sampling location)

      TABLE VIII. UNIAXIAL COMPRESSIVE STRENGTH QU FOR MIXING SAMPLE USING WATER AT THE SAMPLING LOCATION WITH THE RATIO OF W/C=1

      Day	Uniaxial compressive strength qu (kPa)
      	M100 (6%)	M150 (9%)	M200 (12%)	M250 (14%)	M300 (17%)	M350 (20%)
      0	15	15	15	15	15	15
      7	159.992	259.296	505.251	562.143	903.598	1091.86
      14	216.238	400.343	871.621	1509.04	1645.84	1896.05
      28	304.228	574.236	1271.21	1613.32	2121.56	2432.66

      • Compared to experiments 1, 2, 3 (W/C = 1.4; 1; 0.8), in this experiment 4, based on Fig. 4, we find that the value of qu (kPa) at 28 days of age is quite low, it declines about 36.79% compared to experiment 1 (W/C = 1), 57.82% compared to experiment 2 (W/C = 0.8) and 1.83% lower than experiment 3 with a W/C ratio of 1.4.

      The main reason is that the water for mixing sample has too high acid content which is shown in [Table IV] then, it facilitates the process of chloride corrosion and sulfate attacking, resulting in the formation of ettringite (sulfoaluminatehydrate: 3CaO.Al2O3.3CaSO4.32H2O) and gypsum (CaSO4.2H2O) softening of cement paste, changing the microstructure to increase porosity and reduce the strength of the deep cement mixing pile.

      Fig. 4. Uniaxial compressive strength of experiment 4

3. Comparison of laboratory experimental results

   Based on Fig. 5, we see that experiments with the same type of water [Table III] and cement but with different mixing rates, the results are quite different. Water quality has a great influence on the strength and quality of DCMCs, so it is impossible to reinforce the soft soil in this area by dry mixing technology.

   Fig. 5. Uniaxial compressive strength qu of 4 experiments in laboratory

4. Comparison with field experimental results

   Based on the results of 4 experiments in laboratory, the ratio W/C = 0.8 and the cement content of 14% will be the optimal content for mass construction. To verify the feasibility of this option, the comparison with the field results is implemented as Fig. 6.

   Fig. 6. Comparision of the value of qu (kPa) between lab and field experiments

   When the cement content is 14%, we see that the result of compressing samples in the field gives higher values than those of experiments in laboratory when the ratio of W/C = 1.4 and W/C = 1. With this content, the value of qu (kPa) of the field results is quite high but about 35.5% lower than that in the laboratory because the condition of sample preparation in the laboratory is almost ideal. Results of compression in the field also depend on: Modernity and accuracy of the mechanical system, geological conditions around the reinforced area, construction techniques, etc.

5. Direct shear test

   Using the optimal content to determine 02 basic characteristics of shear resistance as Table IX.

   TABLE IX. ANGLE OF INTERNAL FRICTION VALUE FROM DIRECT SHEAR TEST

   Day	Angle of internal friction (degree)
   	6%	9%	12%	14%	17%	20%
   0	1.74	1.74	1.74	1.74	1.74	1.74
   7	21.8	20.22	11.73	14.01	14.44	15.63
   14	20.53	17.34	14.76	12.98	12.05	10.86
   28	17.13	14.63	12.34	8.61	7.21	11.09

   Day	Angle of internal friction (degree)
   	6%	9%	12%	14%	17%	20%
   0	7.4	7.4	7.4	7.4	7.4	7.4
   7	49.06	75.145	228.31	239.41	241.51	267.05
   14	93.218	154.435	267.506	297.678	325.509	339.524
   28	116.89	208.979	316.207	448.162	450.336	457.167

   Day	Angle of internal friction (degree)
   	6%	9%	12%	14%	17%	20%
   0	7.4	7.4	7.4	7.4	7.4	7.4
   7	49.06	75.145	228.31	239.41	241.51	267.05
   14	93.218	154.435	267.506	297.678	325.509	339.524
   28	116.89	208.979	316.207	448.162	450.336	457.167

   Fig. 7. Experimental results and relation between (degree) and time TABLE X. COHESIVE FORCE C FROM DIRECT SHEAR TEST

   Fig. 8. Experimental results and relation between C (kPa) and time

   Fig. 8 indicates that: At the time of 28 days of age, the cohesive force value of the sample at the rate of 6-12% is relatively small, but for the ratio of 14-20%, the value is nearly converged and it is not much different, with the average of about 451.89 (kPa). This also confirms that the optimal content selected for mass construction is reasonable, meeting economic

   • technical requirements when investment of the work. The correlation of (degree) and C (kPa) at 28 days of age is quite close with coefficient R2 = 0.8753 and they are inversely proportional when the content of cement increases.

     Fig. 9. Relation between (degree) and C (kPa) at 28 days

6. Analyzing experimental data, select results according to linear regression model

Determination of value of qu (kPa) and summary of results of experiment 3 at 28 days of age are shown in table XI.

TABEL XI. SUMMARY OF RESULTS OF UNIAXIAL COMPRESSION TEST

The regression equation is as equation (1).

-(34.346 X1) – (3550.21 X2) + (116.92 X3) – (526.02 X4) +

(5531.98 X5) – (188.01 X6) + (325201.76 X7) + (237.94 X8) –

(181.18 X9) – (32.86 X10) + (1.14 X11) + (0.6 X12) + 58687.13 =

qui (kPa) (1)

Value of qu = 3491.90 (kPa) is found through the regression equation with the cement content of 14% (W/C=0.8).

With R2= 0.99996.

As table XII, the regression equation is as follows:

-(68.32 X1) + (0 X2) + (0 X3) + (0 X4) + (4132.37 X5) +

(81.75 X6) – (449.7 X7) + (12.39 X8) + (6.94 X9) – 57.3 = i (kPa)

(2)

The value of internal friction angle (degree) and cohesive force C (kPa) found through the regression equation with the

cement content of 14% (W/C=0.8) are = 8.80 (degree) and C= 437.10 (kPa), respectively. With R2= 0.995995.

TABLE XII. SUMMARY OF RESULTS OF DIRECT SHEAR TEST

Comment: Selecting data using a linear regression moel find that the typical values of DCMCs are 0.03% lower than the average value, at the same time, correlation coefficient R value is very high, approximately equal to 1, indicating that the relationship of the quantities and results is very close, the error during the experiment is low.

4. CONCLUSIONS

When reinforcing soft soil ground in coastal area of Vietnam, in the Southwest region in general and in Duyen Hai

• Tra Vinh area in particular by deep cement mixing technology, it should be chosen with the ratio of W/C = 0.6 – 0.8, cement content of 14 – 16% (equivalent to 250 – 270 kg cement/m3 of natural soil), this will significantly improve the bearing capacity of soil up to hundreds of times. With the optimal ratio above, the compression result in the field gives quite high results, controlling the product quality and furthermore, ensuring economic – technical requirements.

The content and quality of water when mixing samples is one of the main factors that determine the quality and durability of deep cement mixing pillar.

Using a linear regression model based on Data Analysis application of Microsoft Excel software is very useful. This tool helps us re-test the experiment process, limit errors and at the same time increase the reliability of scientific products.

REFERENCES

1. Arnold Verruijt. Soil mechanics, Delft University of Technology. 2004.

2. Bergado, D.T., Anderson L.R., Miura, N., and Balasubramaniam, A.S. Lime/cement deep mixing method. Improvement Techniques of Soft Ground in Subsiding and Lowland Environments, Rotterdam, A.A.Balkelma, 99130, 1996.

3. Boussida, M. and Porbaha. Ultimate bearing capacity of soft clays reinforced by a group of columnsapplication to a deep mixing technique, Soils and Foundations, Vol. 44, No. 3, 91101, 2004a.

4. Broms, B.B. Stabilisation of Soil with Lime Columns. Design Handbook, Third Edition, Lime Column AB, 51, 1984.

5. Braja M.Das. Principles of Foundation Engineering, Sixth Edition, 2004.

6. Holm, G. Deep mixing. Soft Ground Technology ASCE, 105122, 2001.

7. Kitazume, M. and Maruyama, K. Collapse failure of group column type deep mixing improved ground under embankment. Proc. of International Conference on Deep Mixing Best Practice and Recent Advance, 245 254, 2005.

8. Kitazume, M., Ikeda, T., Miyajima, S., and Karastanev, D. Bearing capacity of improved ground with column type DMM. Proc. of the 2nd International Conference on Ground Improvement Geosystems. Vol. 1, 503508, 1996.

9. Masaki Kitazume & Masaaki Terashi. The deep mixing method, CRC Press/Balkema, 2012.

10. Nilo Cesar Consoli, Daniel Aliati Rosa, Rodrigo Caberlon Cruz, Amandan Dalla Rosa. Water content, porosity and cement content as parameters controlling strength of artificially cemented silt soil, Engineering Geology, Volume 122, Issue 3-4, pp. 328-333, 2011.

11. Porbaha, A. State of the art in deep mixing technology. part I. basic concepts and overview. Ground Improvement, Vol. 2, No. 2, 8192, 1998.

12. Porbaha, A., Shibuya, S., and Kishida, T. State of the art in deep mixing technology: part III Geomaterial characterization. Ground Improvement, 3, 91-110, 2000.

13. https://cutram.vn.