Modified Time of Setting Test for Fly Ash Paste and Fly Ash Soil Mixtures

Transcription

1 Technical Note Modified Time of Setting Test for Fly Ash Paste and Fly Ash Soil Mixtures Xin Kang 1 ; Gi-Chun Kang 2 ; and Louis Ge, M.ASCE 3 Downloaded from ascelibrary.org by NTU on 3/14/13. Copyright ASCE. For personal use only; all rights reserved. Abstract: A survey of current literature has revealed that the testing protocol for determining the time of setting of fly ash soil mixtures is not available. Although the current ASTM standard recommends using the Vicat needle method with a water-to-fly ash weight ratio of.35, it was originally proposed for testing the time of setting of cement. The current study showed that the Vicat needle method yielded poor test results for some common fly ashes. Moreover, the water-to-fly ash weight ratio of.35 may not be appropriate for some fly ashes because the chemical components are varied in different fly ashes. Therefore, a British fall cone was adopted to determine the time of setting of fly ash paste and fly ash soil mixtures. It suggests that the best starting water-to-fly ash weight ratio should be at its liquid limit. As the water content increases, the initial and final times of setting of fly ash paste increase accordingly. For the fly ash soil mixture, the initial and final times of setting decrease as the ash-to-soil weight ratio increases. Finally, a linear correlation was established based upon the water content, ash-to-soil weight ratio, and the ratio of calcium oxide to silicon dioxide present in the fly ash. DOI: 1.161/(ASCE) MT American Society of Civil Engineers. CE Database subject headings: Fly ash; Soils; Mixtures; Time factors. Author keywords: Time of setting; Fly ash-soil mixture; British fall cone method. Introduction Fly ashes are classified as Class C or Class F by ASTM. Class C fly ash contains a large portion of calcium oxide and exhibits some self-cementing ability when mixed with water (Parsa et al. 1996). Class F fly ash usually does not have cementing and is normally used with lime or other additives for soil improvement and stabilization. Because of its excellent cementitious properties, large amounts of Class C fly ashes have been used for stabilizing soft ground soils and embankment fills (Zia and Fox 2; Edil et al. 22, 26). Hydration and time of setting play important roles in strength gain. The calcium oxide content can form cementitious products when hydrated; therefore, the final strength might be largely governed by the time of setting. ASTM D (24) uses a series of test methods to determine the characteristics of fly ash. For the time of setting property, ASTM recommends a water-to-ash weight ratio of.35 and a test by Vicat needle in accordance with the test method ASTM C191 (28). Based on preliminary test results, however, the ratio of.35 did not work for several tested class C fly ashes because the fly ash paste was too soft and slurry-like, and was never formed into the Vicat needle container. The Vicat 1 Graduate student, Dept. of Civil, Architectural, and Environmental Engineering, Missouri Univ. of Science and Technology, Rolla, MO xkb4c@mail.mst.edu 2 Postdoctoral Research Fellow, Dept. of Civil, Architectural, and Environmental Engineering, Missouri Univ. of Science and Technology, Rolla, MO kangg@mst.edu 3 Associate Professor, Dept. of Civil Engineering, National Taiwan Univ., Taipei, Taiwan, 1617 (corresponding author). louisge@ ntu.edu.tw Note. This manuscript was submitted on December 1, 211; approved on June 5, 212; published online on August 27, 212. Discussion period open until July 1, 213; separate discussions must be submitted for individual papers. This technical note is part of the Journal of Materials in Civil Engineering, Vol. 25, No. 2, February 1, 213. ASCE, ISSN /213/ /$25.. needle is 5 mm long and 1 mm in diameter. The needle easily penetrated through the fly ash paste and made it difficult to get any readings in the beginning. The time of setting of fly ash paste is determined by linear interpolation of the penetration time plot [ASTM C191 (28)]; however, its curve is largely nonlinear, thus, linear interpolation might result in interpretation errors. Most of the research work in the literature has skipped the step of determining the time of setting of fly ash and fly ash soil mixtures, and instead has focused only on the strength gain after stabilization. Some researchers adopted different approaches; for example, White et al. (25) used a ratio of.275 and a pocket penetrometer to test the strength gain versus the time of setting. Additionally, the fly ash soil mixture has a time of setting that is slightly different from that of fly ash itself. The time of setting of fly ash soil mixtures directly relates to the compaction delay (Misra 1998; Mackiewicz and Ferguson 25; Senol et al. 26). The American Coal Ash Association (ACAA 1999) suggests that all mixing, compaction, and final shaping should be completed in one or two hours of initial mixing. However, there is no clear study on the time of setting characteristics of fly ash soil mixtures. In this paper, a simple method is proposed to determine the time of setting of fly ash and fly ash soil mixtures. A fall cone device was utilized to determine the time of setting characteristics. The water-to-ash weight ratio was modified. Initial and final times of setting were determined from the two tangent method, which is similar to the determination of the pre-consolidation pressure in a one-dimensional consolidation test. Test Materials Five fly ashes used in this study were shipped from LaCygne, Nearman, Labadie, Rush Island, and Meramec power plants in Missouri. Their physical compositional properties from X-ray diffraction (XRD) and scanning electron microscopy energy dispersive system (SEM EDS) are summarized in Table 1. All fly 296 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING ASCE / FEBRUARY 213 J. Mater. Civ. Eng :

2 Downloaded from ascelibrary.org by NTU on 3/14/13. Copyright ASCE. For personal use only; all rights reserved. Table 1. Percentage of Chemical Compositions in Fly Ash Parameters Rush Island (A) Chemical compositions (%) LaCygne (B) Nearman (C) Meramec (D) Labadie (E) SiO Al 2 O Fe 2 O CaO MgO SO K 2 O P 2 O TiO Na 2 O Loss on ignition Specific gravity CaO=SiO Classification C C C C C ashes were derived from the combustion of coal and were collected using electrostatic precipitators. All fly ashes had high contents of calcium oxide (CaO) and silicon dioxide (SiO 2 ). The CaO-to-SiO 2 ratios of the fly ashes ranged from.66 to.93. The loss on ignition of all fly ashes was fairly small, except for the Meramec fly ash. Grain size distribution curves of all fly ashes are presented in Fig. 1 and the coefficients of uniformity and curvature are tabulated in Table 2. Nearman, Rush Island, and Labadie fly ashes are gap graded and their grain size distribution curves were located to the right, which indicated that these three fly ashes are coarser than the other two fly ashes. Approximately 5% of the finer fly ash particles were smaller than.75 mm; however, for the coarser fly ashes, only 15% passed the #2 sieve (.75 mm). The natural soil used in this study was obtained from a road shoulder in Atchison, Missouri. The grain size distribution curve of the soil is plotted in Fig. 1. The plasticity index of the soil is 13 and the soil is classified as low plasticity clay (CL). The liquid limit (LL) of the soil and fly ashes are tabulated in Table 2. Set time Percentage Finer (%) Soil Rush Island LaCygne Meramec Labadie Nearman Grain Size (mm) Fig. 1. Grain size distribution of the soil and fly ashes Table 2. Uniformity and LL of Soil and Fly Ashes Parameters Soil Rush Island LaCygne Meramec Labadie Nearman C u C c LL Table 3. Summary of Tests on Different Fly Ash Soil Mixtures Source 5% 1% 15% 2% 25% 3% 4% 5% 1% A: Rush Island N T N T N T N N T B: LaCyne N T T T T T T N T C: Nearman T T N T N T T T T D: Meramec N T N T N T N N T E: Labadie N T N T N T N N T Note: T = tested; N = untested. tests were conducted on fly ashes and fly ash soil mixtures with different water contents and different ash-to-soil mix ratios. The testing program is summarized in Table 3. Modified Testing Method Fly ash percentage The strength of the fly ash paste is highly variable and dependent on the water-to-ash ratio. For different fly ashes (varied chemical contents) with the same water-to-ash weight ratio, the initial and final time of setting would vary significantly (CaO hydration procedure). The Class C fly ash paste normally goes through liquid, plastic, semi-solid, and solid states, respectively. Because the LL state is the beginning state at which a fly ash paste starts to gain strength, it is reasonable to assume the LL as a reference point for determining the time of setting. Most research work has skipped the step of determining the time of setting of fly ash and fly ash soil mixtures, and instead has focused only on the strength gain after stabilization. Some researchers did test the time of setting of fly ash; however, their methods were not consistent with the ASTM C191 (e.g., White et al. 25). Unlike cement, the strength of fly ash paste is highly variable. The Vicat needle is too heavy to test the time of setting if the fly ash paste is not at consistent condition; however, a conetype needle can gradually increase its resistance as it penetrates. Therefore, in this study, the British fall cone was found appropriate to use and a simple set time testing method was proposed. The fall cone used in this study weighs 32.5 g with an angle of 3. The LLs of fly ash and fly ash soil mixtures were predetermined (LL is defined in the range of a depth of penetration between 15 and 25 mm in terms of the British fall cone device). The fly ash/ fly ash soil mixture paste in the LL state was prepared and placed in the Vicat cup with its surface leveled horizontally. The cup was then carefully transferred underneath the British fall cone and the cone tip was lowered until it touched the paste surface. The cone was immediately released and freely fell from the frame. The penetration value from the British fall cone meter was noted after approximately 1 s (until the cone stabilized in the paste). The unit of the penetration values was shown in degrees (1 ¼.1 mm). The time intervals used were initial, 3 s, 1, 2, 4, 8, 15, and 3 min, and 1, 2, 4, and 8 h, respectively. In terms of data interpretation, the recorded penetration value and its corresponding time in logarithmic scale are plotted. In ASTM C191, the initial time of setting is defined as the time corresponding to a 25-mm penetration. This value is obtained by linear interpolation of the penetration versus JOURNAL OF MATERIALS IN CIVIL ENGINEERING ASCE / FEBRUARY 213 / 297 J. Mater. Civ. Eng :

3 time plot; however, the time of setting curve is largely nonlinear. A semi-logarithmic scale is more appropriate to show the trend of strength gain as the fly ash sets. The details of this data reduction method are presented later in this paper. Results Downloaded from ascelibrary.org by NTU on 3/14/13. Copyright ASCE. For personal use only; all rights reserved. Time of Setting of Fly Ash Paste The time of setting curves of the fly ashes are plotted in Fig. 2. The legends A to E in the figure represent different fly ashes, as described in Table 1. All fly ash paste specimens were tested at their LL. Within approximately 2 min, Nearman, Rush Island, and Meramec fly ashes achieved initial setting, and at approximately 1 min, all reached final setting. Because of the fast setting property, however, Rush Island fly ash did not show an initial setting in the plot. The times of setting of all fly ash pastes are listed in Table 4. Penetration values versus log time curves showed that there were two turning points on each curve (except A). In the beginning, the curve extended flatly, however, after initial setting (first turning point), the ash paste quickly gained strength and the slope of the curve sharply increased. When the paste evolved into the final setting, the curve flattened again and the second turning point occurred. Water content also plays an important role in governing the time of setting of fly ash paste. Different water contents were adopted in this study, and the corresponding time of setting curves are presented in Fig. 3. As the water content increased, the initial and final times of setting increased (Fig. 3). Compared to the final setting, the initial setting appears to be more sensitive to water content. It is known that tricalcium aluminate gel dominates the initial hardening of cement. Similarly, it can be also hypothesized that tricalcium aluminate controls the set of fly ash. Janz and Johansson (22) concluded that the compression strength is related to the waterto-cement weight ratio. When the cement sets, the hydration product cement gel is found to be porous and contains chemically combined water. If the water content changes, the amount of water Penetration Reading Initial Set Point Final Set Point A B C D E Fig. 3. Time of setting of Fly Ash A at different water contents absorbed in the porous cement gel will change; thus, the strength of the cement gel will also change. As water-to-cement ratio increases, the compressive strength gradually decreases. Similar behavior was also observed in the fly ash paste; as the water-to-ash weight ratio increased, the curves shifted toward the upper right, which indicated that the strength of the paste decreased. To test the strength of the fly ash paste at different set time stages, a compressive strength test was conducted by using a pocket penetrometer at adequate time intervals on Fly Ash Paste A. The strength versus time is plotted in Fig. 4. The solid line is a best fit line for these test data. As shown in the figure, the ash paste very quickly gained strength in the first few minutes. Approximately 15 min later, the fly ash paste was very stiff; its strength increased to 5 tsf (24 kpa ), and even the cone tip could not penetrate the paste. Time of Setting of Fly Ash Soil Mixture Although all fly ashes set very quickly, once mixed with natural soil, the initial and final setting were greatly changed. Different ash-to-soil weight ratios were used in this study to examine the effect on the set time characteristics of fly ash soil. The fly ash soil mixture was prepared by first mixing the dry soil with fly ash, followed by the addition of water. After that, the mixture was thoroughly mixed before testing. Fig. 5 shows the time of setting curve of Fly Ash Soil Mixture A; the mix ratios were 1, 2, and 3%. The water content of the soil was fixed at 4% (water-to-dry soil weight ratio) which was at its LL state. As displayed in the chart, when the ash-to-soil mix ratio increased, the initial and final times of setting decreased. Compared with Fly Ash Paste A, the fly ash soil mixture set relatively slowly. The final time of setting of Fig. 2. Time of setting curves of different fly ashes Strength (kpa) Table 4. Initial and Final Set Times of Fly Ash Time of setting A B C D E Initial set (min) Final set (min) y = ln(x) R² = Fig. 4. Strength gain versus time of setting of Fly Ash A 298 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING ASCE / FEBRUARY 213 J. Mater. Civ. Eng :

4 Downloaded from ascelibrary.org by NTU on 3/14/13. Copyright ASCE. For personal use only; all rights reserved. Penetration Reading Final Set Range 1% 2% 3% Fig. 5. Time of setting of Fly Ash Soil Mixture A Fly Ash Paste A was approximately 8 min; however, for the mixture, it was more than 1 min. Fig. 6 shows the time of setting of Fly Ash Soil Mixture B; the ash-to-soil ratio ranged from 1 to 4%, the water content of the soil was kept constant at 4%, and different amounts of dry fly ash were added. As shown in the figure, the initial time of setting was decreased as the ash-to-soil weight ratio increased. Although the final time of setting decreased slightly, the changes were slight. The initial and final times of setting of the mixtures are tabulated in Table 5. The initial time of setting was very large for lower ash-to-soil weight ratios; however, as more fly ash was added, the initial time of setting decreased. As shown in the graph, all curves exhibited similar trends: all of the time of setting curves show similar slopes and are parallel to each other in the final set stage. Because the mixtures were mixed with different amounts of fly ash, their starting strengths were not the same and the time of setting curves began at different locations on the vertical axis. As the ash-to-soil ratio increased, the final set strength of the mixture also increased. A pocket penetrometer was used to determine the final strength (Fig. 7). The best fit line was drawn through the Penetratino Reading Initial Set Range Final Set Range 1% Fig. 6. Time of setting of Fly Ash Soil Mixture B 15% 2% 25% 3% 4% Table 5. Initial and Final Set Times of Fly Ash Soil Mixture B Mix ratio 1% 15% 2% 25% 3% 4% Initial set time (min) Final set time (min) Final Set Strength (kpa) y = x x R² = Fly Ash to Soil Raio Fig. 7. Final set strength versus ash to soil weight ratio of Fly Ash Soil Mixture B Water Content Change y =.4484x x R 2 = Fly Ash to Soil Ratio Fig. 8. Water content change due to fly ash hydration of Fly Ash Soil Mixture B center of the dots, as shown in the graph. Below an ash-to-soil ratio of 2%, the final set strength was relatively small; however, beyond that, the final set strength increased significantly. This trend shows that Fly Ash B could work effectively as an additive in stabilizing some weak soils because the strength could be greatly increased by the bonding between fly ash and soil particles. Fig. 8 exhibits the effectiveness of fly ash working as a drying agent in natural soil. As the ash-to-soil weight ratio increases, water content can be greatly reduced. Therefore, fly ash could be used as a dryer and stabilizer in weak and wet foundation soils, especially those in wet land areas under emergency circumstances. Fig. 9 shows the time of setting curves of Fly Ash Soil Mixture C; the ash to soil weight ratio ranged from 5 to 5%. In this set of tests, the water content of the soil and the water content of the fly ash were kept constant. This was achieved by placing soil at its LL, adding different amounts of fly ash paste at their LLs, and thoroughly mixing. As shown in the plot (Fig. 9), a 3% (ash-to-soil weight ratio) curve seems to be a watershed; below this, the initial and final times of setting were relatively high, and beyond that, the initial and final times of setting greatly decreased. When the ashto-soil weight ratio reached 4 5%, the mixture set very quickly, almost the same as the pure fly ash paste, whose final time of setting was approximately 2 min. The data show that an increase in the ash-to-soil weight ratio can cause a decrease in the final time of setting. The initial time of setting also decreased due to the increase of the ash-to-soil weight ratio; however, the trend was not very clear JOURNAL OF MATERIALS IN CIVIL ENGINEERING ASCE / FEBRUARY 213 / 299 J. Mater. Civ. Eng :

5 Downloaded from ascelibrary.org by NTU on 3/14/13. Copyright ASCE. For personal use only; all rights reserved. Penetration Reading compared with the final set, especially when the ash-to-soil ratio was smaller than 3%. Fig. 1 shows the final strength of Fly Ash Soil Mixture C, and a similar trend was observed for Fly Ash B: the final strength increased as the ash-to-soil weight ratio increased. Fig. 11 shows the change of water content in fly ash soil mixture due to the process of hydration, and similar to Fly Ash B, the increase of fly ash content greatly reduced the water content. Final Set Strength (kpa) y = x x R² = Fly Ash to Soil Ratio Fig. 1. Final set strength versus ash to soil ratio of Fly Ash Soil Mixture C Water Content Change Initial Set Range Final Set Range Fig. 9. Time of setting of Fly Ash Soil Mixture C y = -.895x x R 2 = Fly Ash to Soil Ratio Fig. 11. Water content change due to fly ash hydration of Fly Ash Soil Mixture C 5% 1% 2% 3% 4% 5% Discussion The proposed modified set time test can effectively capture the initial and final set characteristics, as shown in the squares in Figs. 3, 5, 6, and 9. Based on the observed test data, common Missouri class C fly ash is very sensitive. The LL acts as a threshold beyond which ash paste behaves like slurry, but under which it is more plastic and easy to shape. For a given fly ash, the amount of water influences the characteristics of the time of setting. The initial and final times of setting increase as the water content increases. For a fly ash soil mixture, the initial and final sets are largely delayed due to the addition of natural soils. When the weight ratio of ash to soil increases, the initial and final times of setting decrease. Within 2% of the fly ash-to-soil weight ratio, the set properties of the mixture were almost the same; however, for the high volume fly ash soil mixtures, their initial and final sets were completed very quickly and the final set strengths were relatively high. Furthermore, fly ash not only increases the strength of the soil, but also reduces the water content, which is very helpful in stabilizing weak and soft ground soils. Bergeson and Lapke (1993) pointed out that the time of setting of fly ash is related to the CaO content. Janz and Johansson (22) concluded that the ratio of CaO=SiO 2 strongly influences the hydraulic reactions of binding agents. These findings revealed that the cementing potential relates to the percentage of CaO and SiO 2 in fly ash. As discussed earlier, the water content and ash-to-soil ratio also influence the time of setting of fly ash soil mixtures. Therefore, it is possible to define, by regression, a simple prediction formula that defines the final time of setting as a function of those parameters: TðminÞ ¼U þ Aω þ Bðash=soilÞþCðCaO=SiO 2 Þ where T = final time of setting (min); ω = water content of the mixture (%); ash=soil = ash-to-soil weight ratio (%); CaO=SiO 2 = CaO-to-SiO 2 weight ratio (%); and A and U = regression parameters ( ). The final times of setting of all fly ash soil mixtures are tabulated in Table 6. By using the final time of setting, water content, chemical composition percentage, and ash-to-soil weight ratios, statistical analysis of the Missouri fly ash was conducted and a regression equation was achieved [Eq. (2)]. Predicted times of setting are compared to the observed values shown in Fig. 12. An R-squared value of.73 was achieved for the regression. The regression coefficients A U are likely dependent on the water content, fly ash type, soil type, and the percentage of the materials in the mixture. However, further research will be necessary on this point: TðminÞ ¼114.8 þ 15.6ω 5.6ðash=soilÞ ðCaO=SiO 2 Þ ð2þ Table 6. Final Set Times of Fly Ash Soil Mixtures at Various Ash-to-Soil Weight Ratios Fly ash soil mixture final set time Ash-to-soil ratio A B C D E 5% 39 1% % 49 2% % 35 3% % 4% % 5% 4 ð1þ 3 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING ASCE / FEBRUARY 213 J. Mater. Civ. Eng :

6 Downloaded from ascelibrary.org by NTU on 3/14/13. Copyright ASCE. For personal use only; all rights reserved. Predicted Values (min) Conclusions R 2 = Observed Values (min) Fig. 12. Observed versus predicted times of setting of Missouri fly ash The time of setting characteristics of fly ash and fly ash soil mixture were studied in this paper. A simple method was proposed for testing the initial and final times of setting of fly ash and fly ash soil mixtures. Compared to the Vicat needle method, the new approach utilized a cone-type needle, which not only improved the test accuracy, but also yielded quality data. The new test configuration can ensure the same starting point and yield much more consistent penetration values. The time of setting curves in log time scale generally have two turning points that can differentiate the initial set and final set stages. The initial and final times of setting of fly ash paste increase as the water content increases. Due to the addition of natural soil, the initial and final times of setting were greatly delayed. When the ash-to-soil weight ratio increased, the initial and final times of setting of the mixture decreased. The ash-to-soil weight ratio can also affect the final set strength of fly ash soil mixtures. Meanwhile, some fly ashes were able to be used as drying agents in soil stabilization. A regression analysis was performed based on the water content, ash-to-soil weight ratio, and the ratio of CaO to SiO 2 ; however, it should be used cautiously, because the field conditions are much more variable than those in the laboratory. References American Coal Ash Association (ACAA). (1999). Soil and pavement base stabilization with self-cementing coal fly ash, Alexandria, VA. ASTM. (24). Standard test method for characterizing fly ash for use in the soil stabilization. D , West Conshohocken, PA. ASTM. (28). Standard test method for time of setting of hydraulic cement by Vicat needle. C191, West Conshohocken, PA. Bergeson, K. L., and Lapke, R. (1993). Use of fly ashes and afbc residues as synthetic aggregate material. 13th Annual Power Affiliates Rep., ERI-Ames-93-87, Engineering Research Institute, Iowa State Univ., Ames, IA. Edil, T., Tenson, C., Bin-Shafique, S., Tanyu, B., Kim, W., and Senol, A. (22). Field evaluation of construction alternatives for roadway over soft subgrade. Transportation Research Record 1786, Transportation Research Board, Washington, DC, Edil, T. B., Acosta, H. A., and Benson, C. H. (26). Stabilizing soft fine-grained soils with fly ash. J. Mater. Civ. Eng., 18(2), Janz, M., and Johansson, S. (22). The functions of different binding agents in deep stabilization. Swedish Deep Stabilization Research Center Rep. No. 9, Swedish Geotechnical Institute, Linkoping, Sweden. Mackiewicz, S. M., and Ferguson, E. G. (25). Stabilization of soil with self-cementing coal ashes. World of Coal Ash (WOCA), American Coal Ash Association, Lexington, KY, 1 7. Misra, A. (1998). Stabilization characteristics of clays using class C fly ash. Transportation Research Record 1161, Transportation Research Board, Washington, DC, Parsa, J., Munson-McGee, S. H., and Steiner, R. (1996). Stabilization/ solidification of hazardous wastes using fly ash. J. Environ. Eng., 122(1), Senol, A., Edil, T. B., Bin-Shafique, Md. S., Acosta, H. A., and Benson, C. H. (26). Soft subgrade s stabilization by using various fly ashes. Resour. Conserv. Recycl., 46(4), White, D. J., Harrington, D. S., and Thomas, Z. (25). Fly ash soil stabilization for non uniform subgrade soils, Volume I. Engineering properties and construction guidelines, Center for Transportation Research and Education, Iowa State Univ., Ames, IA. Zia, N., and Fox, P. J. (2). Engineering properties of Loess-fly ash mixtures for road base construction. Transportation Research Record 1714, Transportation Research Board, Washington, DC, JOURNAL OF MATERIALS IN CIVIL ENGINEERING ASCE / FEBRUARY 213 / 31 J. Mater. Civ. Eng :