Primary and secondary settlements in normally consolidated organic soils

n d International Conference on New Developments in Soil Mechanics and Geotechnical Engineering, 8-30 May 009, Near East University, Nicosia, North Cyprus Primary and secondary settlements in normally consolidated organic soils Halil Murat ALGIN Department of Civil Engineering, Harran University, Turkey; hmalgin@harran.edu.tr KEYWORDS: Organic soils; Ultimate settlement; Consolidation; Organic content; Embankments. ABSTRACT: The technique used in practice to predict the primary and secondary settlement of normally consolidated organic soils on the basis of experimental model including organic content and vertical stress is simplified and analytically evaluated. The results based on the variations of stress and organic content are presented in this paper for the primary and secondary settlement of organic soils on the basis of the vertical strain equations considered. Paper demonstrates that the vertical strains which constitute the compressibility parameters can directly be determined without taking the numerical integration of the primary and secondary strain equations presented by Gunaratne (006). The presented analytical solutions employ the compressibility parameters previously developed from the excessive experimental data that were analyzed using the rheological model to develop expressions for primary and secondary compressibilities in terms of organic content and consolidation pressure. The analytical solution presented in this paper can readily be applied into the practice to accurately predict the expected ultimate settlement of an organic soil subjected to an average stress increase. 1 INTRODUCTION Organic soils are defined as soil deposits that contain a mixture of soil particles and organic matter. They may be identified by observation of peat-type materials, a dark color, and/or a woody odor. In the Unified Soil Classification System the O designation defines the organic soil which can be more specifically defined by the ASTM D 487 method. In this method, the distinction between organic and inorganic soils can be made by performing two liquid limit tests on the same material. If the liquid limit of the oven-dried sample is less than about 0.75 times that for the air-dried sample, the soils may be classed as organic. Oven-drying also lowers the plastic limits of organic soils, but the drop in plastic limit is less than that for the liquid limit. Organic soils that plot below the A line have visible organic material. It is necessary to use both the qualifier organic silt or organic clay and the symbol OL or OH in the classification. However, organic soils are further classified and defined such as the organic-rich soils generally referred to as muck, and when they contain more than 75% organic matters, as peat. Muck is a thin watery mixture of soil and organic material. Peat is fibrous material with a sponge-like structure, composed almost entirely of dead organic matter, which can form to extensive thickness. Organic silts and clays form in lakes and estuarine environments, where they can attain a thickness of 5 m or more. Since the measurement of thickness of organic deposits is difficult, the piezocone penetration test (PCPT) is used to identify the organic soil with the soil behaviour type classification system (Robertson et al. 1986). Organic soils create serious problems for road engineers in many temperate regions. Organic soils that are highly unsuitable for use as a subgrade are classified as A-8. When the organic materials are encountered in roadway alignments, they are often regarded as the undesirable foundation material for highways because of their high compressibility and poor strength. As stated by Mitchell and Soga (005), the compacted density and strength decrease significantly with the content of organic matter. The increased organic content also causes an increase in the optimum water content for compaction. The large increase of compressibility is due to high organic content in clay. The effect of organic matter on the strength and stiffness of soils depends largely on whether the organics are decomposed or consist of fibers. The fibers can act as reinforcements, thereby increasing the strength (Mitchell 14

n d International Conference on New Developments in Soil Mechanics and Geotechnical Engineering, 8-30 May 009, Near East University, Nicosia, North Cyprus and Soga, 005). The undrained strength and stiffness, or modulus, are usually reduced as a result of the higher water content contributed by the organic matter. Although they are not economical, concrete piles may be considered for organic soils, but these soils contain acids which can damage the concrete. Because of their high acidity, organic materials are usually highly corrosive to steel and concrete. The highly organic deposits do not respond well or at all to the deep dynamic compaction (Bergado 1996). Stone columns are not applicable to thick deposits of peat or highly organic silts or clays. The properties of very soft clays, silts and organic soils can be improved with deep mixing, a soil improvement technique in which stabilizing agents, such as lime and/or cement are mixed into the soil in situ by using auger-type mixing tools. Deep mixed columns are nowadays extensively used to reduce settlements and to improve the overall stability of road and railway embankments on soft soils (EuroSoilStab, 00). Preloading is another technique on normal to lightly overconsolidated organic deposits to preconsolidate and stabilize such materials. The complete excavation or demucking of such material is another option. Depending on the application, it may also be necessary to remove undecomposed organic components prior to hauling or spreading. In some situation, these techniques may be impractical and therefore, such costly construction practices can be avoided if the resulting ultimate settlement can be estimated accurately and determined to be tolerable for the intended structure. Thus, practical and accurate settlement estimation are invaluable for highway construction operations on natural organic soil deposits. Gunaratne et al. (1998) presented a solution technique to predict the ultimate, one-dimensional (1D) settlement of natural organic soils with reasonable accuracy, using their basic index properties. Many researches have demonstrated that the most significant index property of a organic soil is its organic content. Consistency limits such as the plasticity index and liquid limit are not practically meaningful for fibrous organic soils. Even for amorphous organic soils the liquid limit can be experimentally determined, the plasticity index however is more difficult to obtain. Gunaratne et al. (1998) conducted a number of consolidation tests on normally consolidated Florida organic soil samples. Their results demonstrated the definitive trends of ultimate settlement variations with organic content and consolidation pressure. Their ultimate laboratory settlements also agreed with the predictions of the Gibson and Lo (1961) long-term solution for Merchant and Taylor s (1940) consolidation theory. The main attractiveness of the Merchant and Taylor theory lies in its generalization of Terzaghi s 1-D primary consolidation theory to include creep effects. Hence, using this theory, analytical relations were generated for primary and secondary compressibilities in terms of organic content and consolidation pressure (Gunaratne et al. 1998). The developed expressions are generalized by Gunaratne (006) to determine simply the ultimate 1-D compression of an organic soil, avoiding the need to differentiate the primary and the secondary settlements on a time scale. In this technique Florida organic soil samples are used, which are often completely saturated, the main reason of settlement due to a stress increase is expulsion of pore fluids and consequently rearrangement of the organic soil structure. Since the secondary compression is usually the predominating settlement component, the total settlement may only be consisting of the primary and secondary compressions which have been postulated to occur concurrently. The coefficient of secondary consolidation ( C α ) is a commonly adopted parameter to express the secondary settlement behaviour of organic soils and it is defined as the slope of the linear portion of a vertical strain versus logarithm-of-time plot. Although it is convenient for settlement rate predictions, such a linear plot does not adequately explain any termination in secondary compression. Berry and Poskitt (197) provided an analytical base to the above approach and validate the C α concept for both amorphous and fibrous organic soils. Their theoretical formulation is based on bonding the primary and secondary compression to discrete void ratio changes in the soil skeleton, this concept was used in the study conducted by Gunaratne et al. 1998 as well. Fox et al. (199) showed that C α / Cc concept does not apply to peat. This empirical approach also suffers from its inability to predict the ultimate settlement. Whereas the technique used in practice presented by Gunaratne (006) predicts the ultimate settlement of normally consolidated organic soils. This technique is simplified and the strain expressions are analytically evaluated in this paper so that it 15

Primary and Secondary Settlements in Normally Consolidated Organic Soils Algin, H.M. can be used in practice more readily. The accuracy of this technique was shown by Gunaratne et al. (1998). ORGANIC CONTENT The most common laboratory method which determines the amount of organic matter usually in peats, clays, silts and mucks is the ignition test (ASTM-D 979; Jarrett, 1983, ASTM D-974, DIN 1818, AASHTO T67-86, AASHTO T194-87; Schmidt 1970). It is a high-temperature combustion technique in which the percentage organic matter in a soil is determined by subtracting the weight of ash from the initial dry weight of a soil sample. Organic material is made of carbon compounds and if it is heated to high temperatures in the ignition process, it is converted to carbon dioxide and water. The organic matter in the soil is given off as gases. This results in a change in weight which allows for calculation of the organic content (OC) of the sample. The organic content of a soil is defined as Wo OC = (100)% (1) Ws Where, W o is the weight of organic matter in the soil sample and W s is the total weight of the solids in the soil sample. The OC ratios are usually used by many international standards to classify the organic soils. Some of these standards are summarized in Figure 1. Von Post (194) also classified the organic soils in a scale with respect to the decomposition of the organic matter. Figure 1. Some international standards used to classify the organic soils with respect to OC ratio. 3 COMPUTATION OF SETTLEMENT IN ORGANIC SOILS The secondary compression in an organic soil is usually the predominating settlement component and the total settlement refers to the primary and secondary compressions are assumedly take place concurrently. This is particularly the case when the organic content of the soil deposit is significant. The structures (such as highway embankment or foundation) constructed in organic soils exhibit long-term settlement due to secondary compression, which is relatively larger in magnitude than the primary consolidation. Therefore, the engineers who do not recommend the removal of organic soils 16

n d International Conference on New Developments in Soil Mechanics and Geotechnical Engineering, 8-30 May 009, Near East University, Nicosia, North Cyprus from potential construction sites, must alternatively employ specific analytical techniques to estimate the expected secondary compression component that predominates the total settlement of the structure. The analytical treatise based on Gunaratne et al. (1998) presented by Gunaratne (006) addresses this requirements. The samples used in the study of Gunaratne et al. (1998) have not indicated any appreciable overconsolidation during conventional consolidation testing performed on them. Thus, all the samples were deemed to be normally consolidated and to be attributed to the shallow depths where they were retrieved. This statement is also valid for the compressibility relations given by Gunaratne (006). Since there was no traceable stress history, such as reported previous loadings, there was no need to include that factor, except to state that the presented compression parameters by Gunaratne (006) are valid for normally consolidated organic soils. Retrieval of significantly overconsolidated samples and modifying the relations with the overconsolidation ratio (OCR) parameter can be an extension to the research presented. Gunaratne et al. (1998) presented the following relationships for Florida organic soils. e = 0.46 + 1. 55OC () OC = 0.136 wo +.031 (3) Where, e and wo are the ultimate void ratio and the initial water content, respectively. Since OC determination in laboratory requires special procedures, the OC ratio for organic soils can be obtained from water content values using Eq 3. The linear relationships as demonstrated above are also determined by Andersland et al. (1980). The ultimate 1D compressibility of organic soils constitutes a primary compressibility component a and a secondary compressibility component b as expressed below by Gunaratne (006) e ult = Δσ [ a + b] (4) Parameters a and b can be expressed in terms of the primary and secondary void ratio components ( e p and es respectively) of the initial void ratio ( e o ), thus, 1 e p a = (5) (1 + e) σ 1 e b = s (6) (1 + e) σ Gunaratne et al. (1998) also determined that 1 d d a = I p ( σ ) + oc M p ( σ ) (7) (1 + e) dσ dσ 1 d d b = I s ( σ ) + oc M s ( σ ) (8) (1 + e) dσ dσ Where, I p (σ ), M p (σ ) and I s (σ ), ) M s (σ are stress-dependent functions associated with primary and secondary compressibilities, respectively. The derived primary and secondary compressibility parameters by Gunaratne et al. (1998) can simplified as follows based on the analysed data for Florida organic soils..7(95.3 + σ )(17.75 + σ )(.5(144.56 + σ ) + OC(1189 + σ (5745.18 + 37.31σ ))) a = (9) (77 + σ ) (144.56 + σ ) ( 0.14(17.75 + σ )(176. + σ ) 0.13OC(95.3 + σ )(895.48 + σ )) 0.1(95.3 + σ )(17.75 + σ )(97.39(78.1+ σ ) + OC140.49(193.64 + σ ) ) b = (10) (78.1+ σ ) (193.64 + σ ) ( 0.14(17.75 + σ )(176. + σ ) 0.13OC(95.3 + σ )(895.48 + σ )) Parameters a and b are dependent on the vertical stress and OC ratio. The vertical strain in a layer of thickness ( Δ z ) can be expressed in terms of its total (primary and secondary) 1D settlement Δ ) as, ( s p + s Δs p+ s ε ult = (11) Δz Therefore, Eq 4 can be written as follows, 17

Primary and Secondary Settlements in Normally Consolidated Organic Soils Algin, H.M. ε ult σ ' +Δσ av = [ a( σ ) + b( σ )]( dσ ) (1) σ ' Where, σ ' is the effective stress and soil layer. The above integration is successfully solved introducing Eqs 9 and 10 into Eq 1 and due to its complex nature the results are presented in a graphical form shown in Figure. Δ σ av is the average vertical stress increase in the organic Figure. ε p0, ε p1, ε s0 andε s1 values dependent on the OC ratio and vertical stress. As shown in Figure, the strains of ( ε p 0 andε s 0 ) and ( ε p 1 andε s 1) represent the resulting strain values for primary and secondary compressions and are corresponding ( σ ' ) and ( σ ' + Δ σ av ) stress values respectively (Fig. ). Therefore, by using the sign convention assumed in Figure, Eq 1 can be rewritten as, ε = ε ε ) + ( ε ε ) (13) ult ( p0 p1 s0 s1 Eq 13 shows that if the effective vertical stress, OC ratio (not in percentage) and the average stress increase values are known, by using Figure the primiar and secondary compression of normally consolidated organic soils can swiftly be obtained. Therefore, if the thickness of organic soil layer is denoted as H, then, the total (primary and secondary) 1D settlement ( Δ ) in normally consolidated organic soil can be determined by using Eqs 11 and 13, as Δ s + = H ε ε ) + ( ε ε ) (14) p s [ ] ( p0 p1 s0 s1 s p + s 4 AVERAGE VERTICAL STRESS INCREASE FOR EMBANKMENT LOADING This paper now presents the hitherto not available closed-form analytical solutions for the average vertical stress increase under an embankment loading. The average stress concept is usually utilized in the computation of primary consolidation settlement of foundations. The current practice of estimating the average stress is usually undertaken with sub-layer method or mid-depth stress approach. For accurate calculation of average stress the current point-stress formulas are generally run simultaneously for each sub-layer by using a special software and accurate modeling. Since the compressibility coefficient of organic soil is considered to be constant in the Gibson and Lo (1961) long-term solution which is based on the Merchant and Taylor s (1940) consolidation theory (Gunaratne et al. 1998), there is no need for the sub-layer method to estimate the average stress within an organic soil layer. Instead, the average stress increase can directly be calculated by the formula presented below. The vertical Boussinesq point-stress solution under a semi-embankment is provided by U.S. Army Corps of Engineers (1990) and given in Figure 3(a). This point-stress formula for can be rewritten with a dimensionless multiplier generally denoted as influence factor (I). Thus, Δσ = q I. This influence factor is generally a function of the loaded area dimensions and 18

n d International Conference on New Developments in Soil Mechanics and Geotechnical Engineering, 8-30 May 009, Near East University, Nicosia, North Cyprus the depth of interest. Therefore, the average stress increment between z=0 and z=h can be obtained as a function of the average influence factor ( I av ) over the depth of interest (H). Figure 3. (a) Vertical point-stress beneath a semi-embankment loading. (b) The average stress increase in a required soil layer. Thus, Δσ av = q I av (15) Where, I av 1 = H H 0 I dz The above integration has been analytically solved for the semi-embankment loading case using the point-stress formula given in Figure 3(a) and presented in Eq (17). 1 1 1 ( ) ( ) ( ) π t + mcsc k k1 Csc k3 + m ln m k I av = (17) π t + mbs5 ln k3 k1 + t k1 ln k3 k1 Where, ( ) ( ) m = b H, t = a H, k = m + t 1, = and k 1+ m represented in a graphical form as shown in Figure 4. (16) = (see Fig. 3a). Eq.(17) is k 3 1+ k 1 Figure 4. Influence factors for the average stress increase under the vertical edge of the semi-embankment loading. 19

Primary and Secondary Settlements in Normally Consolidated Organic Soils Algin, H.M. If the average stress is to be estimated beneath the central part of an full-embankment, the superposition method can be applied in a way that the full-embankment loading can be split by a vertical plane, parallel to the axis of the embankment, and passing through the point where the average stress is to be required. The plane cuts the full-embankment into two semi-embankments. Then, Eq.(17) can be applied for each of the semi-embankments. The average stresses are calculated for the two semi-embankments separately and added to obtain the total average stress required under the full-embankment. If the average stress within a layer is required, then the discarded area of stress needs to be subtracted as demonstrated by the expression given in Figure 3(b). 5 NUMERICAL EXAMPLE Question: Assume that we wishes to predict the ultimate 1D settlement expected in a 1-m thick organic soil layer (OC= 50%). The full-embankment dimensions are b=8 m, a=1.5 m and q=50 kpa (see Figure 3). The effective stress at the middle of the organic soil layer is given as 50 kpa. Determine the ultimate settlement under the centre of embankment. Solution: Let us first determine the average stress increase within the organic layer. The dimensions of symmetrical two semi-embankments are b=4 m and a=1.5 m. So, for Figure 4, the undimensional parameters m and t are 4 and 1.5. Thus I av = 0. 5 and σ av = 5kPa. For full-embankment the average stress increase under the centre is σ av = 5 = 50 kpa. Therefore, for Figure, σ '= 50kPa and σ ' + Δσ av = 50 + 50 = 100 kpa. By using Figure, ( ε p 0 andε s 0 ) can be obtained as 0.96 m and 0.36 m for σ ', and ( ε p 1 andε s 1) can be obtained as 0.85 m and 0.3 m for σ ' + Δ σ av respectively (see the dashed lines in Figure ). By substituting theses values into Eq 14, the required ultimate settlement can be predicted as, Δ s p+ s = H [( ε p 0 ε p 1) + ( ε s 0 ε s 1) ] = (1)[(0.96 0.85) + (0.36 0.3)] = 0.15m The above example was originally solved by Gunaratne (006) and Gunaratne et al. (1998) using the numerical integration and same result is obtained. This confirms that Figure is valid. 6 CONCLUSIONS The strain equations constitute the primary and secondary compressibility parameters are analytical solved independently and evaluated numerically to represent them in graphical forms. Paper demonstrated that the presented solutions are valid and can directly be used to predict the ultimate settlement of organic soils. It is shown that the strain of primary compressibility parameter is intensively dependent on the stress and organic content. The accuracy of predicting ultimate settlement of normally consolidated organic soils by using technique presented in this paper was validated previously by Gunaratne et al. (1998). However, the strains based on the secondary compressibility parameter provide almost identical results within the organic content ranging between 0% and 100%. Although the Gunaratne s second compressibility parameter equation is pronounced to be based on the excessive experimental studies the solutions provide that there is a little influence of organic content on secondary compressibility of organic soils in this range. Further researches on this issue should be undertaken to demonstrate whether or not the organic content is a reliable parameter for the secondary compressibility of organic soils. REFERENCES AASHTO, (1990). Standard specification for transportations materials and methods of sampling and testing, Part, ISBN 1-56051-005-6. Andersland, O.B. and Al-Khafaji, A.A.W.N., (1980). Organic material and soil compressibility. ASCE Journal of the Geotechnical Engineering Division, Vol. GT7, p. 749. 0

n d International Conference on New Developments in Soil Mechanics and Geotechnical Engineering, 8-30 May 009, Near East University, Nicosia, North Cyprus Andrejko, M.J., Fiene, F. and Cohen, A.D., (1983). Comparison of ashing techniques for determination of the inorganic content of peats. In: Jarrett, P.M. (Ed.), Philadelphia, pp. 5 0. Arman, A., (193). A definition of organic soils an engineering definition. Engineering Research Bulletin, vol. 101. Division of Engineering Research, Louisiana State University, Baton Rouge. Bergado, D.T., Anderson, L.R., Miura, N. and Balasubramaniam, A.S., (1996). Soft ground improvement in Lowland and other Environments. ASCE Press, ASCE, New York, USA. CSSC, (1987). The Canadian System of Soil Classification. Agriculture Canada Expert Committee on Soil Survey, Ottawa, Canada. Davis, J.H., (1946). The peat deposits of Florida: their occurrence, development and uses. Florida Geological Survey Bulletin 30. Tallahassee, FL. DIN 1818, (1990). UvB; Bestimmung des Glühverlust, Nov. EuroSoilStab (00). Development of design and construction methods to stabilise soft organic soils. Design guide soft soil stabilisation. CT97-0351. European Commission. Industrial & Materials Technologies Programme (Brite-EU-Ram III), Brussels. Fox, P.J., Edil, T.B., and Li-Tus, L., (199). Ca/Cc concept applied to compression of peat. Journal of Geotechnical Engineering, Vol. 118, No. 8. Gibson, R.E. and Lo, K.Y., (1961). A theory of consolidation of soils exhibiting secondary compression. Acta Polytechnical Scandinavia, Ci 10 96: Scandinavian Academy of Science. Gunaratne, M., (006). The Foundation Engineering Handbook. Taylor & Francis Group, 1st Ed., New York, 113-115. Gunaratne, M., Stinnette, P., Mullins, G., Kuo, C., and Echelberger, W., (1998). Compressibility relations for Florida organic material. ASTM Journal of Testing and Evaluation, 6(1): 1 9. Jarrett, P.M., (1983). ASTM-D 979, Summary of testing of peats and organic soils. In: Jarrett, P.M. (Ed.), Testing of Peats and Organic Soils. American Society for Testing and Materials, Philadelphia, pp. 33 37. Kearns, F.L., Autin, W.J., and Gerdes, R.G., (198). Occurrence and stratigraphy of organic deposits, St. Mary Parish, Louisiana. GSA Abstracts With Programs, N.E. and S.E. Sections. Louisiana Geological Survey (LGS), GSA annual meeting, Washington, DC, USA. 30. Kivinen, E., (1963). On the spread and characteristics of bogs in Finland. In: Robertson, R.A. (Ed.), Transactions of the nd International Peat Congress (1963), Leningrad USSR. Her Majesty s Stationary Office, Edinburgh, pp. 15 6. Landva, A.O., Korpijaakko, E.O. and Pheeney, P.E., (1983). Geotechnical classification of peats and organic soils. In: Jarrett, P.M. (Ed.). ASTM, Philadelphia, pp. 37 51. Mankinen, G.W. and Gelfer, B., (198). Comprehensive use of peat in the U.S.S.R. DOE Fifth Technical Conference on Peat. U.S. Department of Energy, Bethesda, MD, pp. 44 54. Mitchell, J.K. and Soga, K. (005). Fundamentals of soil behaviour. John Wiley & Sons, Inc. 3rd Ed. Moris, N., (1989). Composition of organic materials of peat soils in Peninsular Malaysia. In: Zauyah, S., Wong, C.B., Paramanathan, S. (Eds.), Recent Developments in Soil Genesis and Classification, Kuala Lumpur, Malaysia. Malaysian Society of Soil Science, Kuala Lumpur, MY, pp. 81 87. Robertson, P.K., Campanella, R.G., Gillespie, D., and Greig, J. (1986). Use of piezometer cone data. Proceedings of the ASCE Specialty Conference on In Situ?86: Use of In Situ Tests in Geotechnical Engineering, Blacksburg, Virginia, 163-180. Schmidt, N.O., (1970). Suggested method of test for organic carbon content of soil by wet combustion. ASTM STP 479 Special procedures for testing soil and rock for eng. Purposes, 1970, 17-78. U.S. Army Corps of Engineers, (1990). Settlement analysis engineer manual. No:1110-1-1904. Von Post, L., (194). Das genetische system der organogenen bildungen schwedens. Quatrie`me commission, commission pour la nomenclature et la classification des sols, Rome, Comite International de Pe dologie, pp. 87 304. Wust R.A.J., Bustin R.M. and Lavkulich L.M., (003). New classification systems for tropical organic-rich deposits based on studies of the Tasek Bera Basin, Malaysia. Catena 53, 133 163. 1