Creep behavior of geosynthetics using confinedaccelerated

Transcription

1 Creep behavior of geosynthetics using confinedaccelerated tests F. A. N. França 1 and B. S. Bueno 2 1 Graduate student, Laboratory of Geosynthetics, Geotechnical Engineering Department, School of Engineering of University of São Paulo at São Carlos, University of São Paulo, São Carlos, São Paulo, , Brazil, Telephone: , Telefax: , fagner@sc.usp.br 2 Professor, Laboratory of Geosynthetics, Geotechnical Engineering Department, School of Engineering of University of São Paulo at São Carlos, University of São Paulo, São Carlos, São Paulo, , Brazil, Telephone: , Telefax: , bsbueno@sc.usp.br Received 7 January 2010, revised 2 December 2010, accepted 22 June 2011 ABSTRACT: The creep behavior of geosynthetics has commonly been determined using standardized creep tests, which are time consuming and very expensive. In addition, these tests involve the use of in-isolation specimens. Thus they are not likely to consider the overall effect of soil confinement. Confined creep tests conducted at elevated temperature can be used to address these negative aspects of standardized creep tests. This paper presents a pioneering laboratory apparatus developed in order to conduct confined, accelerated and confined-accelerated creep tests. In addition, preliminary tests were performed to assess the new equipment s capability of conducting confined and accelerated creep tests. These tests were performed using a biaxial geogrid, a woven geotextile, and a nonwoven geotextile. The new equipment allowed different conditions to be reproduced. The creep behavior of the nonwoven geotextile and the geogrid was found to be very sensitive to soil confinement. On the other hand, the woven geotextile presented a creep behavior independent of soil confinement. The geogrid results did not agree with reports in the technical literature. Accordingly, these results showed the importance of characterizing the effect of soil confinement in geosynthetics creep behavior. Additionally, these preliminary results showed the potential of the new device to overcome the main negative aspects of standardized creep tests on geosynthetics. KEYWORDS: Geosynthetics, Creep behavior, Soil confinement, Elevated temperature REFERENCE: França, F. A. N. & Bueno, B. S. (2011). Creep behavior of geosynthetics using confinedaccelerated tests. Geosynthetics International, 18, No. 5, [ gein ] 1. INTRODUCTION Creep behavior plays an important role in geosyntheticreinforced soil (GRS) structures, because it reduces the design short-term tensile strength of the reinforcement. In addition, creep strains obtained in laboratory tests may also be used to evaluate the behavior of GRS structures prior to their construction. These laboratory tests have commonly been used to determine the creep behavior of geosynthetics. They are conducted at standard values of temperature and relative humidity using in-isolation specimens. Furthermore, the recommended test duration may reach up to h (e.g. ASTM D 5262, ISO/TR (ISO 2007)). Consequently, the full determination of geosynthetics creep behavior is both time consuming and very expensive. Moreover, these tests do not consider the # 2011 Thomas Telford Ltd possibly significant effect of soil confinement on the stress strain behavior of geosynthetics, as in-isolation specimens are used. Thus in-isolation creep tests may lead to conservative results, which increase construction costs. In order to overcome these drawbacks, many studies have suggested different approaches to evaluate geosynthetics creep behavior. Creep tests conducted at elevated temperature may be used to reduce the test duration. In addition, they may be performed under soil confinement, in order to consider its effect on specimens creep behavior. Both improvements in geosynthetics creep tests soil confinement and heating are quite simple to implement when they are considered by themselves. As presented in the next section, several researchers have published results obtained from improved test arrangements, which were able to consider each of these effects one at a time. 242

2 Creep behavior of geosynthetics using confined-accelerated tests 243 This paper presents a pioneering piece of equipment that allows confined-accelerated creep tests to be conducted. In addition, preliminary results obtained using three different geosynthetics and four different types of creep tests (conventional, confined, accelerated, and confined-accelerated) are shown. The new creep-testing equipment was developed in the Laboratory of Geosynthetics of the School of Engineering of the University of São Paulo at São Carlos, in Brazil. Further analyses, such as time temperature superposition plots, are not discussed in this paper. Additional tests are planned, and the analyses of their results will contribute to understanding of the creep behavior of the material used so far in this research. 2. THEORETICAL REVIEW OF GEOSYNTHETICS CREEP BEHAVIOR Creep behavior refers to time-dependent deformations that any material is subjected to under a constant load. Polymers may present significant creep deformations, owing to their molecular structure. Also, the creep behavior of geosynthetics is one of the most important properties in the design of GRS structures. Therefore their longterm performance should be determined (Koerner 2005) in order to provide reliable design parameters. Creep deformations in geosynthetics are commonly associated with both the macro- and microstructure of geosynthetics. Bueno et al. (2005) illustrated this by discussing the creep behavior of nonwoven geotextiles. These authors defined the intrinsic creep behavior, which is related to elongation of polymeric fibers, and the structural creep behavior, which is associated with the slippage between geosynthetic fibers. Creep strains may be divided into three distinct phases, with noticeable strain-rate distinctions. Figure 1 presents the result of a conventional creep test (ASTM D 5262) performed using a polypropylene woven geotextile. The specimen was subjected to 70% of its short-term ultimate tensile strength (UTS) in the cross-machine direction. This test was conducted up to specimen failure in order to illustrate the creep phases. First, an immediate elastic strain (å i ) was noticed in the geosynthetic specimen, due to tensile load application. As the tensile load was Axial strain, ε (%) 30 ε 3 ε 2 20 ε 1 10 e i Primary creep Secondary creep Tertiary creep Creep strain rate Figure 1. Creep phases of geosynthetics Strain rate, ε (%/s) maintained, creep strains (å cr ) occurred. These strains are permanent, and are divided into three different phases: primary (å 1 ), secondary (å 2 ) and tertiary (å 3 ) creep. The main differences between these creep stages concerns their rate of strain. As shown in Figure 1, the creep strain rate decreased during primary creep, and remained approximately constant throughout secondary creep. Finally, tertiary creep was characterized by a significant increase in creep strain rate, which led to specimen rupture. Creep data are usually represented by means of the logarithmic function å ¼ a þ bln ðþ t (1) where å is the specimen elongation (in %), a and b are logarithmic regression parameters, and t is the elapsed test time (in hours). In this function, parameter a corresponds to the creep strain at unitary time, and parameter b indicates the rate of strain. Equation (1) can also be used to calculate the initial strains of the specimen, commonly considered 1 min after the beginning of tensile load application. Therefore the comparison between parameters a and b obtained from creep tests in different conditions can be used to evaluate the creep behavior of geosynthetics. Concerning creep data visualization, both primary and secondary creep strains are approximately linear in a semi-logarithm plot. The creep behavior of geosynthetics has been considered in the design of GRS structures either as a reduction of the UTS of the material, or as a constraint of the maximum strain of the geosynthetic. In the first case, reduction factors are used to decrease the allowable longterm strength of the geosynthetic. These factors are related to creep, to some degradation processes (e.g. installation damage, and both chemical and biological degradation), and to uncertainties regarding their determination. Of these, the reduction factor due to creep is commonly the most significant. As a result, the design tensile strength becomes considerably lower than the UTS obtained from short-term tensile tests (Kongkitkul et al. 2010). In addition to this approach, the creep data can also be used to determine the geosynthetic s maximum allowable strain. The corresponding load is taken from short-term tensile test results, and is taken as the design tensile strength of the field GRS structure. Standard creep tests (e.g. ASTM D 5262, ISO/TR (ISO 2007)) have been used to determine the creep behavior of geosynthetics. Referred to in this paper as conventional creep tests, they consist in the application of a constant tensile load on an in-isolation geosynthetic specimen, while specimen elongation measurements are performed and strains are computed over time. Consequently, it is possible to either plot a curve similar to that in Figure 1 or express the data in a semi-logarithmic plot. Despite their widespread use, conventional creep tests present two main negative aspects. First, owing to their very long duration, they may be both time consuming and expensive. Second, they do not consider the effect of soil confinement, since in-isolation specimens are used in these tests. These drawbacks may be addressed by means

3 244 França and Bueno of two main approaches: tests performed at elevated temperature and under soil confinement, respectively. Conventional creep tests can be performed at elevated temperature in order to expedite the characterization of creep behavior (Bueno et al. 2005). Thus several conventional creep tests may be conducted at the same load level, but at different temperatures. Subsequently, timetemperature superposition (TTS) techniques are used to interpret their results: that is, creep strains recorded at elevated temperature can be interpreted as creep strains at reference temperature on a different timescale. This improves the prediction of geosynthetics long-term creep behavior at ambient temperature. Several successful researches have been conducted concerning the acceleration of geosynthetics creep response by means of elevated temperature tests (Jeon et al. 2002; Zornberg et al. 2004; Bueno et al. 2005; Jones and Clarke 2007; Tong et al. 2008; Yeo and Hsuan 2008). This issue is already well established in the technical literature, and its discussion is not within the scope of this paper. In addition, ASTM D 6992 presents the stepped isothermal method (SIM), initially developed by Thornton et al. (1998). In this approach, the creep behavior of geosynthetics is evaluated by tests conducted with temperature increments in one single specimen. Thus the material variability does not interfere in the analysis of the results. In some cases, geosynthetic stress strain behavior is strongly dependent on soil confinement. FHWA (1998) suggest three mechanisms of soil geosynthetic interaction that may cause this. First, the internal friction between fibers and yarns is restrained under soil confinement. Second, the basic components of a geosynthetic specimen may present themselves as tortuous, and soil confinement may constrain alignment of these curved elements. Finally, soil penetration into openings or apertures in a geosynthetic may reduce the reorientation of fibers and yarns. Regarding these assumptions, the stress strain behavior of nonwoven geotextiles is considered to be the most affected by soil confinement, followed by that of woven geotextiles and geogrids. Even though Boyle et al. (1996) found woven geotextiles stress strain behavior to be completely independent of soil confinement, FHWA (1998) showed some results where woven geotextile stress strain behavior was affected by soil confinement. Based on such discrepancies, FHWA (1998) suggest that each geosynthetic should be characterized regarding its confined stress strain behavior. Similarly to short-term stress strain behavior, the creep behavior of geosynthetics may also be affected by soil confinement. Thus, in order to overcome the second negative aspect of conventional creep tests, they may be performed with in-soil specimens. Tests conducted with this approach are more likely to consider the overall effect of soil confinement in the creep response of geosynthetics. Two main types of equipment are described in the technical literature to perform confined creep tests. In the first, as used in a pioneering study conducted by McGown et al. (1982), the geosynthetic is directly subjected to the tensile load by means of clamps connected to it. The second type of equipment comprises a different load system, where the confining soil is subjected to a vertical stress and is allowed to deform laterally. As a result, the soil itself transmits the tensile load to the specimen. Figure 2 schematically presents both types of confined creep test apparatus. The confining stress is applied throughout the test in both configurations. As mentioned above, the specimen is directly loaded in the first type of confined creep equipment. Thus an external loading system is used, which usually consists of deadweights. One of the main disadvantages of this confined creep test setup consists in the reduction of tensile load due to soil geosynthetic friction. This may be overcome either by imposing a low-friction boundary condition on the confining soil geosynthetic contact, or by allowing the confining box to move during the test (FHWA 1998). In the second type of confined creep equipment, the confining stress causes an increase in the soil vertical stresses, which leads to an increase in the horizontal stresses. Thus movable faces on both sides of the test box will be subjected to a horizontal force. Since the specimen is attached to these faces, the geosynthetic will be subjected to a tensile load. Despite its reliable reproduction of field load application, this type of confined creep test usually leads to very small tensile load values. Therefore it would require very high vertical stress values. Additionally, the loading process generally results in very small loading rates (e.g. 0.05%/min), compared with those in in-isolation creep tests. ASTM D 5262 prescribes that the load be applied smoothly at a strain rate of 10% 3%/min. Thus results from the two tests cannot be compared, owing to the difference in strain rate during load application, and the effect of the soil is not clearly defined (FHWA 1998). This issue was also discussed by Confining stress Confining stress Tensile load Soil Geosynthetic Soil Geosynthetic Hardened portion of the specimen Movable faces Figure 2. Types of confined creep test devices

4 Creep behavior of geosynthetics using confined-accelerated tests 245 Walters et al. (2002) in a study of reinforcement stiffness in different conditions (in-isolation and in-soil specimens). FHWA (1998) presented several types of equipment used in different studies to measure the confined stress strain properties of geosynthetics. These authors also described some studies where pullout tests were used to characterize the in-soil stress strain behavior of geosynthetics. Despite the convenience of pullout tests, FHWA (1998) suggested that alternative tests be considered to conduct confined creep studies, such as those presented in Figure 2. Additionally, Boyle and Holtz (1996) emphasized the importance of direct measurement of reinforcement tension in in-soil tests in order to consider some reduction in the load while the geosynthetic continues to strain. This may occur because of a soil tendency to set up. Some recent studies on confined creep tests are presented by Costa (2004), Mendes et al. (2007), Ding et al. (2008) and Kamiji et al. (2008). Despite the number of successful attempts, there is not a standard procedure to conduct in-soil creep tests. The approaches to address the two main negative points of conventional creep tests have been comprehensively published in the technical literature. However, the combination of the two techniques in one creep test would both consider the effects of the confining soil surrounding the geosynthetic and expedite the tests. Thus reliable results would be produced, with a reduction of test duration and costs. 3. EXPERIMENTAL 3.1. Creep-testing equipment Confined, accelerated and confined-accelerated creep tests were performed with the new equipment developed during this research, which is illustrated in Figure 3. The central metal box (400 mm mm) may be filled with soil in order to conduct confined and confined-accelerated creep tests. Each side of the equipment has both clamping and loading systems. Consequently, the specimens (200 mm wide) were symmetrically loaded from both sides with the same load level by using deadweights. Additionally, a pulley system was used in order to multiply the applied deadweight load by a factor approximately equal to Since some friction may occur between different parts of the new equipment, a pair of calibrated load cells were used to determine precisely the tensile load on both sides of the specimen. The entire loading system was calibrated. In this procedure, several different deadweight levels were used to establish their respective readings in each load cell. Then their relationship was used to compute the necessary deadweight to reach the desired load to be applied to the specimen. Readings from two telltales installed on the specimen allowed the calculation of geosynthetic strains during each test. The initial length (L i ) between the two telltale fixation points needed to be measured before the test. Each telltale was linked to a calibrated linear variable differential transducer (LVDT) by means of inextensible steel wires. Readings from each LVDT were taken con- Pressurized air Temperature controller 11 1 Geosynthetic specimen 6 Roller grip Cushion (b) 2 Reinforced-lubricated portion of the specimen 3 Upper part of testing chamber (confining medium and specimen) 4 Lower portion of testing chamber (heating system) 5 Pressurized air bag (a) 7 Load cell 8 Telltale weight 9 LVDT 10 Pulley system 11 Dead weight 12 Hydraulic jack 6 (c) 7 Figure 3. Overall view of the new creep-testing equipment: (a) schematic side view; (b) general view; (c) close-up view of roller grip and load cell position

5 246 França and Bueno tinuously during every test with a precision of 0.01 mm. Therefore the displacement of each fixation point was constantly measured and registered throughout the test. Finally, specimen strains were calculated by means of the equation å ¼ d A þ d B L i (2) where å is the specimen elongation (in %); d A and d B are the readings from each telltale (in mm); and L i is the initial length between the two fixation points of the telltales (in mm). Since the new creep test device was designed to conduct creep tests at elevated temperature, it is also equipped with a heating system. This comprises two thermocouples, an expanded polystyrene cover, and a heating chamber filled with loose sand, in which a set of electric resistances is installed. Thermocouple 1 (TC-1) is located near the electric resistances, and is used to control the programmed temperature set point inside the heating chamber. TC-1 can also be used to perform any heating path. The second thermocouple (TC-2) is placed 20 mm above the specimen. Therefore the specimen temperature was considered to be equal that read by TC-2. The whole heating system was calibrated in order to determine the relationship between the readings from each thermocouple. As a result, the temperature near the specimen was accurately determined in each test. Regarding confined creep tests, the confining stresses were reproduced by means of a pressurized air bag placed on top of the surrounding soil. The pressure inside the bag was kept constant during the tests. Two pieces of highdensity polyethylene geomembrane were used to reduce soil geosynthetics friction on both sides. In addition, the portions of each specimen in contact with the geomembrane were reinforced, covered with a polyester Mylar1 sheet (0.075 mm thick, manufactured by DuPont1), and lubricated. This procedure resulted in a geosynthetic free length of 100 mm. Thus the specimen s width/length ratio was equal to 2: Geosynthetics Three different categories of geosynthetics were used in this research. They were selected in order to provide a wide variety of geosynthetic types. Geosynthetic A was a biaxial polyester geogrid with 28 mm wide apertures. Geosynthetic B was a polypropylene woven geotextile with mass per unit area (ASTM D 5261) equal to 276 g/m 2 (COV ¼ 3.03%) and thickness (ASTM D 5199) equal to 0.94 mm (COV ¼ 4.21%). Both geosynthetics A and B were tested in the machine direction. Geosynthetic C was a polyester nonwoven geotextile manufactured with needled-punched short fibers. The thickness (ASTM D 5199) and mass per unit area (ASTM D 5261) of geosynthetic C were 2.59 mm (COV ¼ 6.80%) and 313 g/m 2 (COV ¼ 4.64%), respectively. Table 1 summarizes the geosynthetics information Confining media: soils Confined creep tests were performed with two different soils used as confining medium. Their grain size distributions curves (ASTM D 422; ASTM D 6637) and some geotechnical characteristics (ASTM D 698; ASTM D 854; ASTM D 4253; ASTM D 4254) are shown in Figure 4. Soil A is pure sand (SP; ASTM D 2487), and was used at 45% of relative density in dry condition. Soil B is clayed sand (SC; ASTM D 2487), and was compacted in the test chamber to 90% of the maximum dry density, and at standard Proctor optimum moisture content Tests The first step in the creep testing was to perform shortterm, wide-width tensile tests (ASTM D 4595) using each geosynthetic in order to determine their actual UTS and strain at break. Thus the load level in every creep test is referred to in this paper as a percentage of the respective UTS. The creep behavior of the geosynthetics was characterized in this research by means of four different types of test: conventional, confined, accelerated, and confinedaccelerated creep tests. Conventional creep tests were those standardized by ASTM D 5262, where an inisolation specimen was submitted to a constant tensile load while axial strains were registered during the test. In addition, conventional creep rupture tests were conducted with geosynthetic A at different load levels. Three specimens were used at each load level, and the average time to break of each load level was computed in this set of tests. Conventional creep tests were performed using standard creep devices in order to allow a comparison of their results with current practice. Three types of creep test were performed with the new equipment. Accelerated creep tests were conducted at elevated temperature with in-isolation specimens. The heating system was programmed to reach the desired temperature set point, and the specimen was loaded after Table 1. Geosynthetics subjected to creep tests in this research Geosynthetic Description Mass per unit area (ASTM D 5261) (g/m 2 ) Nominal thickness (ASTM D 5199) (mm) Tested direction A Biaxial polyester geogrid, 28 mm wide apertures n/a n/a Machine direction B Polypropylene woven geotextile Machine direction C Polyester nonwoven geotextile, needle-punched Cross-machine direction n/a: Not applicable.

6 Creep behavior of geosynthetics using confined-accelerated tests Property Soil A Soil B Particles smaller than size shown, n (%) γ s (kn/m³) γ d,max (kn/m³) γ d,min (kn/m³) w opt (%) Soil A Soil B Particle diameter, d (mm) Figure 4. Grain size distribution curves and geotechnical properties of the soils used as confining media temperature equilibrium was reached. Confined creep tests refer to those where in-soil specimens were used in the new creep test equipment. These tests were conducted at room conditions (temperature and relative humidity), which were kept within a strict range. Finally, the confined-accelerated creep tests were conducted at elevated temperature and under soil confinement, simultaneously. A specific set of creep tests was performed using each geosynthetic (Table 2). None of the tests reported in this paper was conducted up to h, as prescribed by ASTM D 5262 and ISO/TR (ISO 2007). However, these tests were clearly appropriate to compare the creep response of the geosynthetics studied so far in this research. Wide-width specimens (geotextiles 200 mm wide and seven-rib geogrid specimen) were used in every creep test, except for one confined creep rupture test performed with geosynthetic A at 90% of UTS. In order to reach such a high load with the new equipment, this test was conducted with a three-rib specimen. For this reason, a specific set of in-isolation creep tests was performed with both threerib and seven-rib specimens in order to compare their creep behavior. Creep data are represented by means of Equation 1 in this paper. Therefore creep data were evaluated by comparing the specimens initial strain and parameters a and b. Additionally, readings were taken at very short intervals ( s), but only a few points are shown in the plots presented in this paper, to illustrate the geosynthetics overall creep behavior. Nonetheless, every regression line is plotted based on the whole set of readings taken during the test. 4. RESULTS AND DISCUSSION 4.1. Geosynthetic A Wide-width tensile tests Wide-width tensile tests (ASTM D 4595) conducted with geosynthetic A in the machine direction resulted in a UTS of kn/m (COV ¼ 1.34%) and a strain at break equal to 10.87% (COV ¼ 3.44%). The following sections pre- Table 2. Creep tests conducted with each geosynthetic, and their main characteristics Geosynthetic Conventional creep test Accelerated creep test Confined creep test Confined-accelerated creep test A B C 10h % of UTS 10h 20 80% of UTS 100 h 20% of UTS 10 h 50% of UTS 43.08C 10 h 30% of UTS 38.18C None 10 h 50 & 90% of UTS Soils A and B 50 kpa 10 h 30% of UTS Soil A 50 kpa 100 h 20% of UTS 30 kpa 10 h 50% of UTS Soil A: 50 kpa 44.18C None 100 h 20% of UTS Soil A: 40 kpa 49.48C

7 248 França and Bueno sent the results of creep tests performed using this material In-isolation creep tests First, geosynthetic A was submitted to conventional creep tests conducted up to 10 h. Table 3 presents the results obtained from these tests at several different load levels (20 80% of UTS). As expected, parameters a and b are dependent on the load level. Figure 5 presents the results obtained in conventional creep rupture tests with geosynthetic A. It appears that the variability of the time to break values is higher at higher load levels. This behavior is expected due to material variability. Any variation in the specimen UTS related to sample mean value causes a large time to rupture variation at higher load levels. The creep rupture tests allowed the plot of the initial portion of geosynthetic A creep rupture curve (Figure 5). Both the literature (Jewell and Greenwood 1988; Jeon et al. 2002; Koerner 2005) and the standards (ASTM D 5262; ISO 2007) recommend that extrapolation of the creep rupture curve should not exceed one order of magnitude; however, this plot was used in this paper to compare the creep rupture response of geosynthetic A under in-isolation and in-soil conditions, as presented in the following section. One single accelerated in-isolation creep test was performed with geosynthetic A at 50% of UTS. This test helped to evaluate the new creep-testing equipment concerning this type of test. It was conducted at 43.08C 1.58C over 10 h, and resulted in values of a ¼ and b ¼ 0.141, respectively (Figure 6). Unsurprisingly, the comparison between conventional and accelerated creep tests shows that parameters a and b are both greater at higher temperatures. Additional tests at elevated temperature between room temperature and 438C would allow the plot of the creep master curve of geosynthetic A, which is not in the scope of this paper. However, these preliminary results show the capability of the new creep-testing equipment to perform accelerated creep studies In-soil creep tests The results from both confined and confined-accelerated creep tests performed with geosynthetic A at 50% of UTS, over 10 h, subjected to vertical stress equal to 50 kpa are presented in Figure 7. Although the initial strain values are quite different for soils A and B, soil confinement was effective in reducing creep strains. Since parameter b indicates the creep strain rate, it is easily noticed that the strain rate in confined specimens is a small fraction of that in conventional creep tests (about 3%). This emphasizes that long-term strength taken from conventional tests, Table 3. Logarithmic regression parameters of conventional creep tests using geosynthetic A Load level (% of UTS) Parameter a Parameter b Coefficient of determination Initial strain (%) T 1.828ln t T f R Percentage of UTS, T / T f (%) Average time to break Actual time to break values In-soil creep rupture test Regression line h h Figure 5. Creep rupture tests on geosynthetic A

8 Creep behavior of geosynthetics using confined-accelerated tests ε ln t R Axial strain, e (%) e ln t R Conventional creep test Accelerated creep test Figure 6. In-isolation creep response of geosynthetic A loaded to 50% of UTS ε ln t R Axial strain, ε (%) ε ln t R ε ln t R Confined creep test in soil A Confined creep test in soil B Confined-accelerated creep test in soil A Figure 7. In-soil creep response of geosynthetic A loaded to 50% of UTS (50 kpa of overburden stress) where in-isolation specimens are used, may not reproduce the field behavior of this geogrid (geosynthetic A). Initial strains were approximately identical in in-soil creep tests conducted using soil A as confining medium, but the values of parameter b were noticeably different. As can be seen Figure 7, the increase in temperature in the confined-accelerated creep test (temperature equal to C) markedly increased the creep strain rate. Table 4 presents parameters a and b obtained from these tests, and also compares them with the results from conventional creep testing. One confined (50 kpa) creep rupture test at 90% of UTS was performed in order to compare the geosynthetic creep rupture time with that obtained in in-isolation creep rupture tests. The mean time to break in in-isolation creep rupture tests at 90% of UTS was equal to h (3.3 min), as presented in Figure 5. The time to rupture in Table 4. Logarithmic regression parameters obtained in different creep tests performed using geosynthetic A at 50% of UTS Description of test Parameter a Parameter b Initial strain (%) Conventional Confined: soil A Confined: soil B Confined-accelerated: soil A, temperature ¼ 44.18C confined creep rupture tests under 50 kpa was h (about 83.0 min). Similarly to other confined creep tests, this outcome may indicate that soil confinement markedly reduced the creep strains of geosynthetic A.

9 250 França and Bueno The confined creep rupture test required a specific specimen configuration in order to subject it to the desired load level. Thus a three-rib specimen was used instead of a seven-rib one. Previous comparison between three-rib and wide-width (seven-rib) specimens was performed by means of in-isolation creep tests, at 50.2% of UTS. These conventional creep tests were performed with a seven-rib specimen different from that presented in Section This measure was taken in order to ensure that the threerib and seven-rib specimens were extracted as closely as possible from each other in the geosynthetic sample. Figure 8 presents the regression lines for both tests. Although the initial strains were somewhat different, the creep strain rates in the two tests were nearly the same. The values of parameter b in these two tests were also similar to that presented in Section Therefore the two specimen configurations were in agreement for this purpose Comparison between in-isolation and in-soil creep behaviors It is suggested in the technical literature that soil confinement has either a small or no effect on the creep behavior of geogrids. However, the results obtained so far indicate the converse of this. The creep strain rates of the inisolation specimens were expressively larger than the ones obtained with in-soil specimen at the same temperature. Thus it is clear that creep strains of geosynthetic A under soil confinement are quite lower than those in in-isolation tests. This emphasizes the suggestion of FHWA (1998) that every geosynthetic should be characterized regarding its in-soil creep behavior Geosynthetic B Wide-width tensile tests The UTS of geosynthetic B obtained by means of widewidth tensile tests (ASTM D 4595) in the machine direction was kn/m (COV ¼ 4.19%), and its strain at rupture was 14.84% (COV ¼ 10.93%). The results of creep tests conducted using this material are presented in the following sections In-isolation creep tests Geosynthetic B was subjected to 10 h long conventional creep tests at different load levels (20 80% of UTS). Table 5 presents parameters a and b, and the coefficient of determination obtained in these tests, and Figure 9 presents their semi-log plot. As expected, both initial and creep strains were directly related to the loading level. From Figure 9, it can also be seen that most of the tests resulted in a non-linear plot on the semi-log scale (å against ln t). This indicates the occurrence of tertiary creep, as previously discussed. Nevertheless, only the specimen loaded to 80% of UTS failed before 10 h of conventional creep test. Since the linear portion of each curve corresponds to primary and secondary creep, the regression results presented in Table 5 correspond to this portion of each curve. One accelerated in-isolation creep test (30% of UTS) was conducted with geosynthetic B at 38.18C ( 1.08C) and was used to evaluate the new creep-testing machine concerning accelerated creep tests (Figure 9). Parameters a and b obtained in this test are also presented in Table 5. It was performed for 10 h, and its results would allow one to plot the in-isolation creep master curve for geosynthetic B, together with additional creep tests. Similarly to conventional creep data, accelerated creep results did not present good agreement with a logarithmic regression In-soil creep tests Since this paper presents preliminary tests with the new equipment, only one single confined creep test has been performed with geosynthetic B so far. Soil A was used as the confining medium in this test. As described earlier, the logarithmic regression did not lead to good agreement when the full data set was considered, owing to the Wide-width specimen Three-rib specimen Axial strain, ε (%) ε 0.095ln t R ε 0.096ln t R Figure 8. Comparison of geosynthetic A creep response regarding the width of the specimen (wide-width and three-rib specimens)

10 Creep behavior of geosynthetics using confined-accelerated tests 251 Table 5. Logarithmic regression parameters of in-isolation creep tests using geosynthetic B Load level (% of UTS) Parameter a Parameter b Coefficient of determination Initial strain (%) (temperature ¼ 38.18C) Note: These parameters refer to the linear portion of each curve (primary and secondary creep). Axial strain, e (%) % of UTS 30% of UTS 40% of UTS 50% of UTS 60% of UTS 70% of UTS 80% of UTS 30% of UTS, T 38.1 C Figure 9. In-isolation creep response of geosynthetic B occurrence of tertiary creep. Thus a power function represented by å ¼ ct d (3) was used in order to compare the in-isolation and in-soil creep test results by means of parameters c and d. Here å is the specimen elongation (in %), c and d are regression parameters, and t is the elapsed test time (in hours). Similar to parameters b and a in Equation (2), parameter d indicates how rapidly creep strains are occurring, and parameter c shows the strain at unitary time. In addition, Equation (3) may be used to calculate initial strain values. The conventional creep test at 30% of UTS resulted in c ¼ and d ¼ 0.116, whereas the confined test (under 50 kpa of vertical stress) resulted in c ¼ and d ¼ Both results are shown in Figure 10, where the vertical axis is presented on a different scale than that in Figure 9. Despite the differences in initial strain values, the creep strain rates are quite similar in the two tests. Therefore it may be concluded that soil confinement did not affect creep strains in this woven geotextile. Further in-soil creep tests are planned using geosynthetic B, including confined-accelerated tests. These will allow a confined-creep master curve to be plotted, and the longterm creep behavior of geosynthetic B under soil confinement to be predicted Geosynthetic C Geosynthetic C was submitted to the shortest set of creep tests. Therefore the results from these tests are presented in one single subsection. The UTS of geosynthetic C was equal to kn/m (COV ¼ 4.02%) in the cross-machine direction, and the strain at break was equal to 90.85% (COV ¼ 1.71%), obtained in wide-width tensile tests (ASTM D 4595). The results of creep tests conducted with geosynthetic C are presented in Figure 11. In the conventional creep test, the initial strain was 32.2%. Subsequently, creep strains occurred, which increased the total strain slightly up to 35.3% after 100 h, at a constant rate (parameter b ¼ 0.243). This small value of creep strain is justified by the low tensile load level (20% of UTS). The confined

11 252 França and Bueno 8 7 In-isolation specimen In-soil specimen ε t 2 R Axial strain, e (%) ε t 2 R Figure 10. Conventional and confined creep test of geosynthetic B loaded to 30% of UTS ε 0.243ln t R Conventional creep test data Axial strain, e (%) Confined-accelerated creep test data ε ln t R ε 0.040ln t Confined creep test data R Figure 11. Geosynthetic C creep test results creep test with geosynthetic C was conducted at room temperature, with vertical stress equal to 30 kpa. The initial specimen strain in this test was 15.9%, which is considerably less than that found in the conventional creep test. In addition, the creep strain rate was also reduced (parameter b ¼ 0.040). The reduction of both initial and creep strains is due to soil confinement, as expected. One single confined-accelerated creep test was conducted at elevated temperature ( C). Concerning both in-soil creep tests, the accelerated test presented the highest value of initial strain (17.37%). It also showed the highest value of creep strain (1.3%) and creep strain rate (parameter b ¼ 0.065). This behavior was expected, since temperature increases both strains and creep strain rate. Table 6 shows parameters a and b obtained from each test conducted with geosynthetic C. As expected, the conventional creep test provided the highest value of parameter b, which indicates its high creep strain rate. In addition, evaluation of parameters a and b for each test clearly indicated the influence of soil confinement on creep strains. Unsurprisingly, both parameters a and b obtained at elevated temperature were greater than those at room temperature in in-soil creep tests. Similarly to the other geosynthetics used in this research, a creep master curve would be plotted with additional creep tests in both in-isolation and in-soil conditions, conducted at different temperature values. 5. CONCLUSION This paper has presented a pioneering piece of equipment, developed to conduct confined and accelerated creep tests, either simultaneously or separately. Preliminary tests were

12 Creep behavior of geosynthetics using confined-accelerated tests 253 Table 6. Logarithmic regression parameters of creep tests using geosynthetic C Creep test type Parameter a Parameter b Coefficient of determination Initial strain (%) Conventional Confined Confined-accelerated conducted with three different types of geosynthetics. The following conclusions are drawn from the present study. The new creep-testing equipment was able to perform both in-isolation and in-soil creep tests. In addition, the two approaches were used simultaneously, introducing the confined-accelerated creep test. Soil confinement considerably reduced the creep strain of both the geogrid and the nonwoven geotextile. Concerning the nonwoven geotextile, this reduction is widely mentioned in literature; however, creep strain reduction in geogrids due to soil confinement is not usually mentioned in the technical literature. Concerning the geogrid (geosynthetic A), both soils A and B were effective in reducing creep strains under soil confinement. Also, one confined creep rupture test conducted with geosynthetic A showed an increase in time to break, owing to soil confinement. The creep behavior of geosynthetic B (woven geotextile) was not affected by soil confinement. The in-isolation and in-soil creep tests using the woven geotextile showed similar values of creep strain rate, under the same tensile load level. Soil confinement is not considered to be effective in reducing creep strains of geogrids, and is considered to be hardly capable of that in woven geotextiles. However, it has a pronounced influence on the creep behavior of nonwoven geotextiles, since it mitigates the necking process that occurs in in-isolation specimens. Necking is decidedly reduced in confined specimens, because of the restriction in filament movements. Consequently, creep strains are strongly decreased, as presented in the results for geosynthetic C. FHWA (1998) suggested that every geosynthetic should be submitted to such studies before their use. The results for geosynthetic A support this statement. In contrast with the literature, creep strains in tests performed with this geogrid were also drastically reduced by soil confinement, which may have occurred because of its structure. Apertures in geogrids may allow a different type of soil geosynthetic interaction during confined creep tests. Thus, under soil confinement, it may reduce long-term strains. On the other hand, the woven geotextile presents neither apertures nor necking behavior: therefore its creep behavior is not noticeably affected by soil confinement. Additional tests are necessary to confirm these results, to evaluate the new equipment further, and to creep master curves to be plotted. However, the tests performed so far show that the new equipment is capable of successfully addressing both the negative aspects of conventional creep tests, either simultaneously or separately. It allowed more reliable and cost-effective creep tests to be performed. ACKNOWLEDGEMENTS The authors would like to thank the State of São Paulo Research Foundation (FAPESP) and the Coordination for the Improvement of the Higher Education Personnel (CAPES) for the financial support of this research. The authors are also grateful to J. G. Zornberg for his help, discussions and advice. NOTATIONS Basic SI units are given in parentheses. a, b logarithmic regression parameters (dimensionless) c, d power function regression parameters (dimensionless) d A, d B displacements of telltale fixation points (m) L i initial length between telltale fixation points (m) R 2 coefficient of determination (dimensionless) T tensile load applied during creep tests in geosynthetics (N/m) T f tensile load at failure obtained in shortterm tensile tests according to ASTM D 4595 and ASTM D 6637 (N/m) t elapsed time (s) w opt optimum water content (%) ª d,max maximum index unit weight (N/m 3 ) ª d,min minimum index unit weight (N/m 3 ) å axial strain (dimensionless) å cr creep strain (dimensionless) å i initial strain (dimensionless) å 1 primary creep strain (dimensionless) å 2 secondary creep strain (dimensionless) tertiary creep strain (dimensionless) å 3 ABBREVIATIONS ASTM COV American Society for Testing and Materials coefficient of variation

13 254 França and Bueno GRS geosynthetic-reinforced soil LVDT linear variable differential transducer TC-1, TC-2 thermocouples 1 and 2 TTS time-temperature superposition UTS ultimate tensile strength REFERENCES ASTM D 422. Standard Test Method for Particle-Size Analysis of Soils, ASTM International, West Conshohocken, PA, USA. ASTM D 698. Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Standard Effort ( ft-lbf/ft 3 (600 kn-m/m 3 )), ASTM International, West Conshohocken, PA, USA. ASTM D 854. Standard Test Method for Specific Gravity of Soil Solids by Water Pycnometer, ASTM International, West Conshohocken, PA, USA. ASTM D Standard Practice for Classification of Soils for Engineering Purposes (Unified Soil Classification System), ASTM International, West Conshohocken, PA, USA. ASTM D Standard Test Methods for Maximum Index Density and Unit Weight of Soils Using a Vibratory Table, ASTM International, West Conshohocken, PA, USA. ASTM D Standard Test Methods for Minimum Index Density and Unit Weight of Soils and Calculation of Relative Density, ASTM International, West Conshohocken, PA, USA. ASTM D Standard Test Method for Tensile Properties of Geotextiles by the Wide-Width Strip Method, ASTM International, West Conshohocken, PA, USA. ASTM D Standard Test Method for Measuring the Nominal Thickness of Geosynthetics, ASTM International, West Conshohocken, PA, USA. ASTM D Standard Test Method for Measuring Mass per Unit Area of Geotextiles, ASTM International, West Conshohocken, PA, USA. ASTM D Standard Test Method for Evaluating the Unconfined Tension Creep and Creep Rupture Behavior of Geosynthetics, ASTM International, West Conshohocken, PA, USA. ASTM D Standard Test Method for Determining Tensile Properties of Geogrids by the Single or Multi-Rib Tensile Method, ASTM International, West Conshohocken, PA, USA. ASTM D Standard Test Method for Accelerated Tensile Creep and Creep-Rupture of Geosynthetic Material Based on Time-Temperature Superposition Using the Stepped Isothermal Method, ASTM International, West Conshohocken, PA, USA. Boyle, S. R., Gallagher, M. & Holtz, R. D. (1996). Influence of strain rate, specimen length and confinement in measured geotextile properties. Geosynthetics International, 3, No. 2, Boyle, S. R. & Holtz, R. D. (1996). Discussion of A performance test for assessment of long-term creep behavior of soil-geosynthetics composites by Wu, J. T. H. and Helwany, S. M. B. Geosynthetics International, 3, No. 4, Bueno, B. S., Constanzi, M. A. & Zornberg, J. G. (2005). Conventional and accelerated creep tests on nonwoven needle-punched geotextiles. Geosynthetics International, 12, No. 6, Costa, C. M. L. (2004). Time dependent deformations in geotextile reinforced soil walls. PhD thesis. Department of Geotechnical Engineering, University of São Paulo, São Carlos, Brazil. Ding, J., Tong, J. & Zhou, W. (2008). Test study on the confined creep features of HDPE geogrid in sand. Proceedings of the 4th European Geosynthetic Conference, Edinburgh, Scotland, paper no. 156, 4 pp. FHWA (Federal Highway Administration) (1998). Development of Protocols for Confined Extension/Creep Testing of Geosynthetics for Highway Applications, FHWA-RD , Elias, V., Yuan, Z., Swan Jr., R. W. & Bachus, R. C, Editors, Federal Highway Administration, McLean, VA, USA. ISO (International Organization for Standardization) (2007). ISO/TR Guidelines for the Determination of the Long-Term Strength of Geosynthetics for Soil Reinforcement. International Organization for Standardization, Geneva, Switzerland. Jeon, H. Y., Kim, S. H. & Yoo, H. K. (2002). Assessment of long-term performances of polyester geogrids by accelerated creep test. Polymer Testing, 21, No. 5, Jewell, R. A. & Greenwood, J. H. (1988). Long term strength and safety in steep soil slopes reinforced by polymer material. Geotextiles and Geomembranes, 7, No. 1 2, Jones, C. J. F. P. & Clarke, D. (2007). The residual strength of geosynthetic reinforcement subjected to accelerated creep testing and simulated seismic events. Geotextiles and Geomembranes, 25, No. 3, Kamiji, T., Bueno, B. & Costa, C. (2008). Nonwoven geotextiles: evaluation of behavior in confined creep tests. Proceedings of the 4th European Geosynthetic Conference, Edinburgh, Scotland, paper no. 288, 7 pp. Koerner, R. M. (2005). Designing with Geosynthetics, Prentice Hall, Upper Saddle River, NJ, USA. Kongkitkul, W., Tatsuoka, F., Hirakawa, D., Sugimoto, T., Kawahata, S. & Ito, M. (2010). Time histories of tensile force in geogrid arranged in two full-scale high walls. Geosynthetics International, 17, No. 1, McGown, A., Andrawes, K. Z. & Kabir, M. H. (1982). Load-extension testing of geotextiles confined in soil. Proceedings of the 2nd International Conference on Geosynthetics, Las Vegas, NV, USA, pp Mendes, M. J. A., Palmeira, E. M. & Matheus, E. (2007). Some factors affecting the in-soil load strain behavior of virgin and damaged nonwoven geotextiles, Geosynthetics International, 14, No. 1, Tong, J., Gong, B. & Liu, J. (2008). An experimental study on the creep behavior of geogrids at different temperature. Proceedings of the 4th European Geosynthetic Conference, Edinburgh, Scotland, paper no. 155, 6 pp. Thornton, J. S., Allen, S. R., Thomas, R. W. & Sandri, D. (1998). The stepped isothermal method for time-temperature superposition and its application to creep data on polyester yarn. Proceedings of the 6th International Conference on Geosynthetics, Atlanta, GA, pp Walters, D. L., Allen, T. M. & Bathurst, R. J. (2002). Conversion of geosynthetic strain to load using reinforcement stiffness. Geosynthetics International, 9, No. 5 6, Yeo, S. & Hsuan, Y. G. (2008). Evaluation of stepped isothermal method using two types of geogrids. Proceedings of the 4th European Geosynthetic Conference, Edinburgh, Scotland, paper no. 285, 9 pp. Zornberg, J. G., Byler, B. R. & Knudsen, J. W. (2004). Creep of geotextiles using time-temperature superposition methods. Journal of Geotechnical and Geoenvironmental Engineering, 130, No. 11, The Editor welcomes discussion on all papers published in Geosynthetics International. Please your contribution to discussion@geosynthetics-international.com by 15 April 2012.