European Journal of Environmental and Civil Engineering ISSN: 1964-8189 (Print) 2116-7214 (Online) Journal homepage: http://www.tandfonline.com/loi/tece20 In situ tests on improvement of collapsible loess with large thickness by downhole dynamic compaction pile Yu-chuan Zhang, Yong-guo Yao, An-gang Ma & Chen-lin Liu To cite this article: Yu-chuan Zhang, Yong-guo Yao, An-gang Ma & Chen-lin Liu (2017): In situ tests on improvement of collapsible loess with large thickness by downhole dynamic compaction pile, European Journal of Environmental and Civil Engineering, DOI: 10.1080/19648189.2017.1370393 To link to this article: http://dx.doi.org/10.1080/19648189.2017.1370393 Published online: 30 Aug 2017. Submit your article to this journal Article views: 21 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=tece20 Download by: [University of Florida] Date: 25 October 2017, At: 02:14 European Journal of Environmental and Civil Engineering, 2017 https://doi.org/10.1080/19648189.2017.1370393 In situ tests on improvement of collapsible loess with large thickness by down-hole dynamic compaction pile Yu-chuan Zhanga,b , Yong-guo Yaoa,b, An-gang Mac and Chen-lin Liua,b School of Civil Engineering and Mechanics, Lanzhou University, Lanzhou, China; bKey Laboratory of Mechanics on Disaster and Environment in Western China, Ministry of Education, Lanzhou University, Lanzhou, China; cGansu Zhonglian Construction Engineering Technology Co., LTD, Lanzhou, China Downloaded by [University of Florida] at 02:14 25 October 2017 a ABSTRACT This paper presents three in situ tests about down-hole dynamic compaction pile of 35 m, down-hole dynamic compaction pile of 50 m and pipe sinking compaction pile. To develop a method on improvement of collapsible loess foundation with large thickness and investigate improvement effect and quality control of the method. The tests results showed that (1) down-hole dynamic compaction pile can improve collapsible loess at a thickness of 50 m under reasonable design and required construction quality, improvement effect can be guaranteed; (2) for three tests, collapsible coefficients of soil between piles are all less than .015 in range of improvement depth, the collapsibility elimination effect has a downtrend along depth, whereas the compacting range and extent have no obvious difference, the depth is not a major factor influencing compacting effect; (3) the hammer intersection angle can influence improvement effect of downhole dynamic compaction pile (4) a calculation formula of tamping number is given, which can simplify pre-tamping work because of acquiring tamping number only through field measurement to meet quality requirement. The tests results confirm a useful method for improvement project of collapsible loess foundation with large thickness, and to make those areas become improved and usable engineering construction site. ARTICLE HISTORY Received 26 February 2017 Accepted 16 August 2017 KEYWORD Large-thickness collapsible loess; downhole dynamic compaction pile; improvement depth; tamping number Introduction Loess is a clastic, predominantly silt-sized sediment that is formed by the accumulation of wind-blown dust (Dijkstra, 2001; Lecturer, 1996; Pye, 1995). When soaked by water, some of loess will occur prominent and rapid collapse under the self-weight pressure of overburden soil or self-weight pressure and additional pressure, which is called collapsible loess (Derbyshire, 2001).The loess is spread widely over the world, with an area of about 1.3 × 107 km2 (Wang, Zhu, & Huang, 2014). In China, the loess covers 6.4 × 105 km2, the collapsible loess accounts for 4.3 × 105 km2, mainly distributing in central and eastern Gansu, southern Ningxia, north-west and central Shaanxi or other areas (Figure 1). Among these areas, in eastern and western Gansu, collapsible loess thickness is between 15 and 40 m (local than 50 m) and self-weight collapsibility is outstanding, belongs to the large-thickness self-weight collapsible loess (Zhang, Yang, & Zhang, 2010). Loess collapse and subsidence can lead to buildings or structures deformation in various degrees, such as uneven settlement and incline, affecting safety and daily use seriously. Particularly, in self-weight CONTACT Yu-chuan Zhang email@example.com © 2017 Informa UK Limited, trading as Taylor & Francis Group Downloaded by [University of Florida] at 02:14 25 October 2017 2 Y.-C. ZHANG ET AL. Figure 1. Loess distribution in China (Liu, 1988). collapsible loess field with large-thickness, serious collapsibility and large thickness bring higher harm to construction project. So it is necessary to eliminate the collapsibility harm for construction works on collapsible loess site (GB 50025-2004, 2004; Ma et al., 2014; Qian & Wang, 1985). Changing the physical and mechanical properties of collapsible soil layer by foundation improvement is the most fundamental method to eliminate or weaken collapsibility harm, improve the foundation bearing capacity and so on. Traditionally, loess foundation is improved by cushion, dynamic compaction and compaction (Dai, 2014). Cushion (Zhang, 2011) is suitable for shallow improvement, the improvement depth is about 3 m; the effective reinforcement depth of dynamic compaction (He & Fan, 2007; Mayne, Jones, & Dumas, 1984; Zhao & Xie, 2014) is not more than 10 m; pipe sinking and compacting (Mi & Yang, 2012) is a way to improve collapsible loess foundation economically and effectively, it also has some advantages to adjust the foundation non-uniformity and improve impermeability, but the superiority of pipe sinking and compacting is only reflected within 12 m, when improvement depth greater than 12 m, there will be difficult to hole and extubation. On the other hand, with the national economic development and the implementation of western development, a large number of engineering projects need to be carried out in large-thickness collapsible loess site. Commonly used improvement methods of loess foundation have been not competent for the demanded improvement depth. It is in extremely urgent to study effective improvement method of collapsible loess foundation with large thickness. Down-hole dynamic compaction (DDC) pile (Si & Tang, 1999) is a pile that is created by filling soil material and tamping strongly in pre-drilled hole, to compact and consolidate filling material in vertical and soil between piles in lateral at the same time. DDC pile can reach a larger improvement depth and compaction effect is not engendered in progress of drilling but in progress of compacting filling material (Feng, Shi, Shen, & Li, 2015). So the compaction effect is limited and quality is hard to control, it is deemed to ‘Have contribution in improving foundation bearing capacity, but little effect in eliminating collapsibility’. The role in eliminating collapsibility is not underachieving, which brings difficulties to application in improvement of large-thickness collapsible loess foundation. Different to above viewpoint, some researchers think as long as the diameter of DDC pile can be ensured, the ability of eliminating collapsibility will be sure in theory. Moreover, DDC pile has advantage Downloaded by [University of Florida] at 02:14 25 October 2017 EUROPEAN JOURNAL OF ENVIRONMENTAL AND CIVIL ENGINEERING 3 in improvement depth. There have been researches on applying DDC pile in improvement of large-thickness collapsible loess foundation. Based on foundation improvement cases about Ⅲ self-weight collapsible loess site in north-west Shanxi, Chen, Guan, and Liu (2007) mainly focused on working mechanism, design, construction technology and application of DDC pile. Zhang, Huang, Zhu, and Zeng-Hong (2011) had done useful works for different length of DDC pile affecting to foundation improvement effect by soaking load tests. All of literature has proved that the method has some advantages and feasibility in improving large-thickness collapsible loess foundation. However, there is little information available in literature about application in larger depth more than 20 m, which makes the method restricted in the thicker collapsible loess site. Therefore, this paper reports some new tests on improving fully larger-thickness collapsible loess foundation by DDC pile. To study the feasibility, improvement effect and quality control about this method, we conducted contrast tests between DDC pile with pipe sinking compaction (PSC) pile (pressing a steel pipe with tip into soil, the soil can be compacted when forming hole, then filling soil material in the hole to form a pile body). Improvement depth reaches 35 and 50 m, respectively, which is rare in China and around the world. The results have promoted the further scientific study, test methods and engineering applications in improvement of large-thickness collapsible loess foundation. Test site The in situ tests were conducted in Heping Town of Lanzhou, Gansu Province, China (Figure 2). Test site is located in transition zone from loess slope terraces to valleys. Strata (from top to bottom) are loess-like silt, pebble, clay rock or the sand conglomerate. Surface of site is covered continuously by Late Pleistocene of Quaternary Q3 loess. The loess has light yellow or grey-yellow, macro pores, loose structure, uniform texture, no vertical joints developed and thickness is approximately 30–70 m. Geotechnically, the loess foundation is evaluated by collapsibility level of loess and collapsibility grade of site. Evaluation parameter of collapsibility level (collapsibility coefficient) is defined in formula 1. Figure 2. Sketch map of test site (from geotechnical engineering report of ‘Hengda City’ project in Lanzhou). 4 Y.-C. ZHANG ET AL. / � s = (hp − hp ) h0 (1) s – Collapsibility coefficient, evaluation parameter of collapsibility, when s < .015, the effect of collapse settlement to buildings is small and can be neglected, it’s stipulated that loess with s < .015 is non-collapsible loess in GB 50025-2004 (2004); the collapsible level can be divided into three types according to the value of s, (1) .015 ≤ s ≤ .03, slight collapsibility (2) .03 < s ≤ .07, medium collapsibility (3) s > .07, serious collapsibility; the collapsibility coefficient under saturated self-weight pressure is z s (the self-weight collapsibility coefficient). Downloaded by [University of Florida] at 02:14 25 October 2017 hp Height of specimen under a certain pressure (mm) ′ hp Height of specimen after compressing and soaking stably (mm) h0 Original height of specimen (mm) The average collapsibility coefficient of loess in site is .039–.110, the average self-weight collapsibility coefficient is .007–.059 (Figure 6). The loess belongs to medium or serious collapsible loess. The collapsibility grade of site is evaluated by entire self-weight collapsibility and total collapsibility, which are defined in formulas 2 and 3, respectively. Δzs = 0 n ∑ (2) zsi hi i=1 Δzs zsi hi 0 n Entire self-weight collapsibility, evidence to dividing collapsibility grade of site (mm) Self-weight collapsibility coefficient of layer in number i Thickness of layer in number i (mm) Correction coefficient varying with region, getting values according to test data Number of collapsible loess layers in calculating range, layer with zsi < .015is not accumulation Δs = n ∑ (3) si hi i=1 Δs Total collapsibility, evidence to dividing collapsibility grade of site (mm) si Collapsibility coefficient of layer in number i Correction coefficient considering factors, such as lateral extrusion and soaking rate, getting values according to test data Based on Δzs and Δs, the test site evaluation is IV (find in Table 1) self-weight collapsible loess, the collapsibility of site is very serious. So the site needs to be improved as foundation of superstructure. The stable groundwater level is lower than 70.0 m below the surface, and the influence of which can’t be considered. The test site was divided into two sections: A and B, respectively. The 35 m DDC pile tests were carried out in section A, mainly researched improvement effect, design parameters and Table 1. The collapsibility grade of collapsible loess foundation. Non-self-weight collapsible site Total collapsibility (mm) Δs ≤ 300 300 < Δs ≤ 700 Δs > 700 Δzs ≤ 70 mm Ⅰ (slight) Ⅱ (medium) – Self-weight collapsible site 70 mm < Δzs ≤ 350 mm Ⅱ (medium) Ⅱ or Ⅲ Ⅲ (serious) Notes: when Δs > 600 and Δzs ≥ 300, can be judged as Ⅲ, others isⅡ. GB 50025-2004 (2004). Δzs > 350 mm – Ⅲ (serious) Ⅳ (very serious) EUROPEAN JOURNAL OF ENVIRONMENTAL AND CIVIL ENGINEERING 5 Table 2. The specific tests design and related parameters. Partition Area (m2) Thickness of loess (m) Maximum collapsible soil layer thickness (m) Water content Void ratio s zs Test type Downloaded by [University of Florida] at 02:14 25 October 2017 Abbreviation A 35 × 32 38–42 ≥35 B 20 × 20 ≥50 ≥50 12.2–18.5% (After humidification) .880–1.014 .011–.044 .005–.053 12 m pipe sinking compaction pile 35 m down-hole dynamic compaction pile PSC-12 DDC-35 6.6–11.6% .774–1.175 – .025–.076 50 m down-hole dynamic compaction pile DDC-50 construction technology and so on, meanwhile carried out contrast tests of PSC pile in section A; the 50 m DDC pile tests were carried out in section B to mainly study the feasibility of construction technology. There have been humidification before tests in section A (in principle, it should be humidified to construct conveniently for sections A and B, but based on following considerations to section B: (1) higher improvement depth will lead to larger difficult, longer time and higher cost to humidify; (2) the water content in original conditions is higher than section A; (3) tests were carried out to only research the feasibility of construction technology, so there have not been humidification in section B). See Table 2 and Figure 3 for the specific tests design and related parameters. Foundation improvement design and construction The design of DDC pile DDC pile is formed by tamping filling material in pre-drilled hole to compact soil between piles. Essentially, the compaction process of DDC pile is a process about cylindrical hole expansion (Gong, 1999). Soil between piles is compacted with filling and tamping material, the key to ensure compaction quality (the range of effective compaction area) is to ensure the tamped expanded body diameter. For same foundation soil, the diameter of PSC pile determines the effective compaction range, further to determines pile spacing; unlike PSC pile, the diameter of pre-drilled hole and tamped expanded body determine the effective compaction range of DDC pile, and then determine pile spacing. DDC pile is generally distributed as triangle. As shown in Figure 4, reasonable pre-drilled hole diameter is 400–700 mm. To make the compaction coefficient of soil between piles meet design value, the relationship (formula 4) between tamped expanded body diameter and pile spacing can be derived from the principle that quality of soil in unit compaction area with unit thickness doesn’t change before and after compaction. √ s2 c d max − s2 d + .9d 2 d (4) D= .9c d max s d D d c ρdmax Pile spacing (m) The diameter of pre-drilled hole (m) The diameter of pile body after tamping and expanding (m) Average dry density of soil between piles before compaction (g/cm3) Average compaction coefficient of soil between piles Maximum dry density of soil between piles determined by compaction test (g/cm3) Generally in engineering, pile spacing need to be ascertained by natural foundation condition and engineering requirement at first. When the pile spacing is s = 2.0–3.0 d (CECS197:2006, 2006), the DDC pile can eliminate the collapsibility economically and reasonably. The theoretical tamped expanded 6 Y.-C. ZHANG ET AL. Downloaded by [University of Florida] at 02:14 25 October 2017 Figure 3. The layout of section A with piles and exploratory wells position. Figure 4. The ketch map of dynamic compaction and expansion. body diameter can be deduced from the formula 4, of course, the final tamped expanded body diameter should be ascertained by in situ tests. Construction of DDC pile According to the design, drilling machines include power-driven Luoyang pipe and long twist drill of CFG-40. As shown in Figure 5, firstly, drill with long twist drill (hole diameter is 400 mm, hole depth is up to 20 m), then expand pre-drilled hole (expand to diameter of 700 mm) until presupposed depth with Luoyang pipe. Pre-drilling is to improve the speed of creating hole. Mix filling material and fill holes through loader, tamp and expand filling material with electric self-propelled rammer of JK-B. Hammer is cone cylinder shape with intersection angle of 40°, weighs 2.78 t, 500 mm diameter and entire lengths is 2.55 m including hammer tip. Take 700 mm to pre-drilled diameter and 1.4 m (2.0 d) to pile spacing, then calculate tamped expanded body diameter by formula 4. Consideration that tamped expanded body diameter directly affects compaction effect and it is difficult to control in construction, the value of 1.26 g/cm3 (minimum dry density) is assigned to d in calculation, .97 for c . After these, the tamped expanded body diameter was calculated as .957 m, taking 960 mm. In order to measure the tamped expanded body diameter conveniently after opening cut, use 1:9 lime soil as filling material. To ensure the tamped expanded body diameter, the field pre-tamping was carried out before construction to determine construction parameters, such as falling distance of rammer, height of filling material and tamping number. When pre-tamping, the sampling tests of dry density can be converted to measurement of filling material height, so as to get the tamping number required for certain tamping Downloaded by [University of Florida] at 02:14 25 October 2017 EUROPEAN JOURNAL OF ENVIRONMENTAL AND CIVIL ENGINEERING 7 Figure 5. The construction of DDC pile. quality, which simplifies pre-tamping work and improves efficiency. As shown in Figure 4, once the filling distance of hammer and the height of filling material are determined, the required diameter of tamped expanded body and compactness will be met by tamping filling material from h to H, H can be calculated by formula 5. H= 0 d 2 h .97D2 d max (5) ρ0 Dry density of filling material before dynamic compaction (g/cm3) ρdmax Maximum dry density of filling material (g/cm3) h Height of filling material for each time (m) d The diameter of pre-drilled hole (m) D Required diameter of tamped expanded body (m) In the tests, taking falling distance of rammer as 5 m, 2.6 m for h, .98 g/cm3 for ρ0, 1.69 g/cm3 for ρdmax, H can be obtained from the formula 5 that is .82 m. Through pre-tamping, the tamping number of tamping filling material from 2.6 m (corresponding d) to .82 m (corresponding D) is 12. Which is simple and easy to apply, we can also use it in construction process to recheck the quality of tamping. Evaluation of foundation improvement Evaluation methodology Foundation improvement was evaluated by sampling with exploration wells, indoor and outdoor tests. Excavating exploration wells (layout is shown in Figure 3) and sampling along the improvement depth at 1 m intervals (having removed the loose layer of surface at a thickness of 1 m). Indoor and outdoor tests mainly include minimum compaction coefficient cmin (defined in formula 6), self-weight collapsibility coefficient zs, compaction coefficient of pile body c (defined in formula 7) and tamped expanded body diameter. Sampling of two former tests were conducted at the centroid of triangle consisted with three neighbouring piles, where compaction effect is weakest. Sampling of latter two were done at distance of 2/3 hole radius from pile centre (GB 50025-2004, 2004; the tests data at distance of 2/3 hole radius from pile centre is approximately their average in radius direction). Based on above four parameters, 8 Y.-C. ZHANG ET AL. Table 3. Evaluation methodology of improvement foundation. Objects Parameters Sampling point zs Vertical Horizontal Compaction effect of soil between piles cmin 1 m intervals Centroid of triangle consisted with three neighbouring piles c Quality of pile body Tamped expanded body diameter Distance of 2/3 hole radius from pile centre we can get compactness and collapsibility of soil between piles and pile body quality (reflected in compaction coefficient and tamped expanded body diameter). See Table 3 for evaluation methodology. Downloaded by [University of Florida] at 02:14 25 October 2017 c min = d0 d max (6) cmin – Minimum compaction coefficient of soil between piles after dynamic compaction representing compactness of soil, it’s stipulated in GB 50025-2004 (2004) that cmin ≥ .88 represents principle of meeting required compactness and eliminating collapsibility is come true. d0 Dry density of soil between piles in range of compaction depth (g/cm3) dmax Maximum dry density of soil between piles (g/cm3) c = �d � d max (7) c Compaction coefficient of pile body, representing compaction quality of pile body, compactness can be met when c ≥ .90 ′d Dry density of pile body (g/cm3) ′ dmax Maximum dry density of pile body (g/cm3) Tests results Collapsibility of soil between piles Figure 6 shows the distribution to self-weight collapsibility coefficients of soil between piles before and after improved with DDC-35, DDC-50 and PSC-12. Figure 6 indicates that the original soil has a serious collapsibility which characterised by collapsibility coefficients as high as between .045 and .075. By three improvement tests, the collapsibility coefficients of soil between piles become less than .015, thereby achieving the purpose of eliminating collapsibility. It also can be known from distribution of scatter that collapsibility coefficients of three tests are all in a law of smaller top and greater bottom, the elimination effect of collapsibility is the best within 20 m, the collapsibility coefficients are generally less than .01 but blow 40 m are about .005–.015. Compaction effect of soil between piles Figure 7 shows the distribution to minimum compaction coefficients of soil between piles (the average value of all detection holes) after improved with DDC-35, DDC-50 and PSC-12. Figure 7 indicates that there is a decreasing trend to minimum compaction coefficients of soil between piles as it changes with depth (explained in part of ‘evaluation and discussion of improvement method’). The compaction effect is the best in range of 5–15 m. There is no obvious difference on values and distribution of EUROPEAN JOURNAL OF ENVIRONMENTAL AND CIVIL ENGINEERING 9 Self-weight collapsible coefficients 0.000 0 0.005 0.010 0.015 0.020 0.04 0.05 0.06 0.07 0.08 DDC-35 DDC-50 PSC-12 Before improvement 10 20 Depth/m Downloaded by [University of Florida] at 02:14 25 October 2017 0.03 30 40 50 Figure 6. Distribution chart of self-weight collapsible coefficients of soil between piles. Minimum compaction coefficients of soil between piles 0 0.86 0.88 0.90 0.92 0.94 0.96 10 20 30 40 DDC-35 DDC-50 PSC-12 50 Figure 7. Distribution chart of the minimum compaction coefficients of soil between piles. minimum compaction coefficients of soil between piles about three tests, that is, there is no significant difference on compaction range and extent between three tests. Both DDC-35 and DDC-50 appear low compaction coefficients of soil between piles at bottom of pile, the phenomenon may be caused by thick loose soil at bottom of pile, which affects compactness of pile body afterwards affects compaction effect of soil between piles. 10 Y.-C. ZHANG ET AL. Compaction effect of pile body Figure 8 shows distribution to the compaction coefficients of pile body (the average value of all detection holes) about DDC-35, DDC-50 and PSC-12. Figure 8 indicates that the compaction coefficients of pile body have little change with depth, which are generally between .91 and .96. Where the pile body compaction coefficients of DDC pile is small corresponds to where the minimum compaction coefficients of soil between piles is also small, but this phenomenon (explained in part of ‘evaluation and discussion of improvement method’) doesn’t take place for PSC pile. So we can conclude that the compaction effect of PSC pile depends on design parameters including pile spacing and so on instead of the backfill quality of pile body, however, the DDC pile’s mainly depends on backfill quality of pile body. Downloaded by [University of Florida] at 02:14 25 October 2017 Tamped expanded body diameter The tamped expanded body diameters measured by excavating exploration wells are shown in Figure 9. For 38 cross sections, the maximum is 1190 mm, the minimum is 700 mm and the average is 972 mm, meeting the required tamped expanded body diameter of 960 mm basically. According to the distribution of tamped expanded body diameters along improvement depth, which varies with depth slightly, and variation range is between +3 and –2% from required 960 mm. Tamped expanded body diameters varies with a larger amplitude in the depth about 30–40 m, but returns to normal below that depth. The change of tamped expanded body diameters may be caused by change of strata or construction quality, the improvement depth is not main factor affecting tamped expanded body diameters. Evaluation and discussion of improvement method The improvement effect of DDC pile in collapsible loess foundation with large thickness is evaluated by improvement effect of soil between piles and pile body quality, specific evaluation parameters are showed in Table 4. Combined with the variation of evaluation parameters in part of ‘tests results’, a similarity in terms of minimum compaction coefficients between DDC pile and PSC pile can be found, compacting well; collapsibility elimination is ideal except little points at depths of DDC-50; compaction Compaction coefficients of pile body 0 0.88 0.90 0.92 0.94 0.96 10 20 30 40 50 Figure 8. Distribution chart of compaction coefficients of pile body. DDC-35 DDC-50 PSC-12 0.98 1.00 EUROPEAN JOURNAL OF ENVIRONMENTAL AND CIVIL ENGINEERING -1500 -1000 -500 0 500 1000 11 1500 Downloaded by [University of Florida] at 02:14 25 October 2017 Depth/m 10 20 30 40 50 Figure 9. Tamped expanded body diameters. Table 4. The tests data statistics of improvement effect. Test results statistics zs cmin c Tamped expanded body diameter (mm) Sample number Maximum Minimum Average Percent (zs < .015) Sample number Maximum Minimum Average Percent (cmin ≥ .88) Sample number Maximum Minimum Average Percent (c ≥ .90) Sample number Maximum Minimum Average Percent (D ≥ 960) PSC-12 78 .004 0 .0005 100% 78 .99 .83 .91 80.77% 78 .99 .80 .92 65.38% – DDC-35 244 .013 0 .002 100% 244 .99 .81 .89 77.46% 244 .99 .79 .93 80.74% 28 1100 850 986 64.3% DDC-50 300 .019 0 .005 97.30% 300 .99 .82 .89 72.70% 300 .99 .81 .92 80% 38 1190 700 970 63.15% coefficients of pile body for DDC pile meet the requirement basically, DDC pile is significantly higher than PSC pile in percentage of c ≥ .90, however compaction coefficients of pile body for DDC pile are still small when compared with the limit .93 in JGJ 79-2002, 2002, compaction quality is not good enough; the tamped expanded body diameters reach design requirements. So, for DDC pile, we can conclude that foundation improvement achieves desired results, the method have feasibility and advantages in improvement of collapsible loess foundation with large thickness. Meanwhile, we found three key problems about DDC pile in tests, doing following discussions at now. 12 Y.-C. ZHANG ET AL. Downloaded by [University of Florida] at 02:14 25 October 2017 Discussion one Analysing collapsibility coefficients, minimum compaction coefficients, compaction coefficients of pile body and tamped expanded body diameters at same depth, we find an obvious corresponding relationship exist in them. When the tamped expanded body diameters meets design requirement, the corresponding compaction coefficients of pile body, collapsibility coefficients and minimum compaction coefficients of soil between piles can also meet requirements. At the improvement depth about 33 m of DDC-50, self-weight collapsibility coefficient of soil between piles is greater than .015,where the minimum compaction coefficient is less than .88, the compaction coefficient of pile body is less than .90, and the tamped expanded body diameter is less than 900 mm. That is, the quality of pile body (compaction coefficient and diameter) and the compaction effect of soil between piles (compaction coefficient and collapsibility coefficient) are all not satisfied at this point. Essentially, because of the poor backfill quality of pile body at some positions, the compaction coefficients of pile body are low, further resulting the tamped expanded body diameters are not up to design requirements, then both compaction range and extent of soil between piles reduce, without doubt the compaction coefficients will be small and collapsibility is not eliminated completely. So, we can get conclusions: (1) The backfill quality of tamped expanded body is the key to ensure eliminate collapsibility for loess foundation improved by DDC pile. It is not accurate to use a single parameter to evaluate improvement effect, because the difference of foundation soil and the heterogeneity of backfill soil can all lead to some test parameters too large or too small. Only when the quality of pile body (compaction coefficient and diameter) and the compaction effect of soil between piles (compaction coefficient and collapsibility coefficient) have a corresponding relationship can we judge construction quality or foundation improvement effect. (2) When the designed pile spacing (or tamped expanded body diameter) is reasonable and the quality of pile body meets requirement, it is feasible to improve loess foundation at a thickness of 50 m with DDC pile, improvement effect can be ensured. Discussion two We put forward that the compaction coefficients of pile body are not ideal in previous section. To reason, we had done tests and analyses, turned out it is mainly related to hammer intersection angle. Analysing the mechanics principle of foundation soil, the tamped expanded compaction effect is formed by hammer tamping and extruding backfill soil in pre-drilled hole, further to compact soil between piles. Stress process of backfill soil has two kinds. First is that hammer compacts the filling soil under its bottom, making these soils move downward as a soil plug. Because the pressure of soil under hammer exceeds soil strength, soil structure is destroyed, softened and lateral pressure coefficient increases. Soil is compacted not only vertically but also laterally, resulting increase of reinforced zone width in horizontal direction. So the stress distribution in reinforced zone presents a horizontal and wide apple shape that is different from elliptical distribution of static load. This stress process is similar to that of the soil under dynamic compaction (Zhou, Wang, & Zhang, 2006), the hammer bottom of which is relatively flat. Second is the situation that hammer bottom is relatively sharp. At this situation, hammer tip spikes into soil, the soil suffers splitting failure then is pushed around along vertical direction of hammer bevel, compacting soil at horizontal direction to reinforce lateral soil, however the reinforcement effect to soil under the hammer is very small or not. This stress process is similar to that of PSC pile (Han & Huang, 1999). So we can conclude that the different hammer bottom shape makes stress mechanism of foundation soil different, foundation improvement effect is also different. The hammer intersection angle of previous in situ tests is 40°, which is relatively sharp, therefore it is beneficial to the compaction effect of soil around pile, but the compaction effect of pile body is limited. For supporting this analysis, the model test was also carried out. Figure 10 is the test results about compaction coefficient of pile body and soil between piles changing with the hammer intersection angle. Figure 10 indicates that it EUROPEAN JOURNAL OF ENVIRONMENTAL AND CIVIL ENGINEERING 13 is disadvantageous to compact pile body when the hammer intersection angle is less than 50 degrees, which proves well why the compaction coefficients of pile body are not ideal in in situ tests. Downloaded by [University of Florida] at 02:14 25 October 2017 Discussion three For the phenomenon that the minimum compaction coefficients of soil between piles has a decreasing trend along depth, it can be deemed that lateral pressure in deep strata is larger than shallow. But the tamping number is got from shallow test, thus, it is necessary to add tamping number appropriately for ensuring compaction effect when improvement depth is more than 20 m. In a word, the paper works verifies feasibility about large-thickness collapsible loess foundation improved by DDC pile, studies construction technology and quality control, explores and explains key issues. There is the first time to reach 50 m for improvement of collapsible loess foundation. This provides theoretical guidance and practical approach to solve the improvement challenge of large-thickness collapsible loess foundation, having significance to construction development in loess region. But there also exists following deficiencies: (1) The tamped expanded body diameter of DDC pile is difficult to control, affected by human factors and strata conditions easily, dispersing highly; the replacement rate of DDC pile is larger than that of PSC pile, which leads to high cost of foundation improvement. While the compaction effect of PSC pile depends on design, having lighter human affect, and quality is easier to ensure. So when two compaction methods can all be used, priority should be adopted to PSC pile. (2) The tamping number of construction control parameters is affected by depth, the improvement effect will decrease with depth when adopting same tamping number. There is a need to research the changing rule of tamping number with depth. Conclusions The following conclusions can be drawn based on in situ tests on the improvement of large-thickness collapsible loess foundation. (1) Down-hole dynamic compaction pile is feasible to improve 50 m collapsible loess foundation when the pile spacing of design (or tamped expanded body diameter) is reasonable and Figure 10. The curve of compaction coefficient changing with intersection angle of rammer. Downloaded by [University of Florida] at 02:14 25 October 2017 14 Y.-C. ZHANG ET AL. the quality of pile body meets requirement, improvement effect can be ensured; the focus and difficulty to construction of down-hole dynamic compaction pile are ensuring tamped expanded body diameter. According to the formula 5, we can obtain tamping number meeting the requirement of tamping quality by field measurement, which simplifies pre-tamping work and improves efficiency. (2) Three in situ tests can all make collapsibility coefficients of soil between piles less than .015, and collapsibility coefficients are both in a law of smaller top and greater bottom after improvement. Collapsibility elimination effect is the best within 20 m, collapsibility coefficients are generally less than .01. Compaction coefficients of soil between piles change as same law of collapsibility coefficients, there is no significant difference on compaction range and extent between three tests at same depth. (3) If hammer intersection angle is small, the compaction effect of pile body will be not ideal, but a larger hammer intersection angle can affect compaction range. For general collapsible loess, when hammer intersection angle is between 50 and 60 degrees, the compaction of pile body and soil around pile can be both taken into account. (4) It is not accurate to use a single parameter to evaluate the improvement effect, because of the difference of foundation soil and the heterogeneity of backfill soil, some test parameters will be too large or too small. Only when the quality of pile body (compaction coefficient and diameter) and the compaction effect of soil between piles (compaction coefficient and collapsibility coefficient) have a corresponding relationship can we judge construction quality or foundation improvement effect. Acknowledgements Special acknowledgements are due to my colleagues, and in particular to the Prof. Zhang Huyuan, of the Key Laboratory of Mechanics on Western Disaster and Environment (Lanzhou University, Lanzhou, China), for providing advice on writing and modifying of paper. Disclosure statement No potential conflict of interest was reported by the authors. ORCID Yu-chuan Zhang http://orcid.org/0000-0003-3454-2345 References CECS197:2006. (2006). Technical specification for down-hole dynamic compaction. Beijing: China Planning Press (in Chinese). Chen, F. M., Guan, L. J., & Liu, X. H. (2007). Research on application of DDC pile in ground improvement of thick loess collapsible under overburden pressure. Building Science, 23, 49–52. (in Chinese). Dai, P. C. (2014). Collapsible loess foundation treatment methods. Research & Application of Building Materials, 3, 40–42. (in Chinese). Derbyshire, E. (2001). Geological hazards in loess terrain, with particular reference to the loess regions of China. EarthScience Reviews, 54, 231–260. Dijkstra, T. A. (2001). Geotechnical thresholds in the Lanzhou loess of China. Quaternary International, 76–77, 21–28. Feng, S. J., Shi, Z. M., Shen, Y., & Li, L. C. (2015). Elimination of loess collapsibility with application to construction and demolition waste during dynamic compaction. Environmental Earth Sciences, 73, 1–16. GB 50025-2004. (2004). Code for building construction in collapsible loess regions. Beijing: China Architecture and Building Press (in Chinese). Gong, X. N. (1999). Soil plastic mechanics. Hangzhou: Zhenjiang University Press (in Chinese). Han, X. L., & Huang, Y. (1999). Analysis on the stress of the around the hole in hole-formation with lime soil compaction pile. Journal of Xi'an University of Architecture & Technology, 3, 256–259. (in Chinese). Downloaded by [University of Florida] at 02:14 25 October 2017 EUROPEAN JOURNAL OF ENVIRONMENTAL AND CIVIL ENGINEERING 15 He, W. M., & Fan, J. (2007). Evaluation of collapsible loess foundation treated by dynamic compaction. Journal of Rock Mechanics and Geotechnical Engineering, 26, 4095–4101. (in Chinese). JGJ 79-2002. (2002). Technical code for building foundation treatment. Beijing: China Planning Press (in Chinese). Lecturer, C. R. S. (1996). Loess: The yellow earth. Geology Today, 12, 186–193. Liu, T. S. (1988). Loess in China. Heidelberg: Springer Berlin Heidelberg. Ma, Y., Wang, J. D., Peng, S. J., Li, Y. W., Wang, J. H., & Chen, W. (2014). Immersion tests on characteristics of deformation of self-weight collapsible loess under overburden pressure. Chinese Journal of Geotechnical Engineering, 36, 537–546. (in Chinese). Mayne, P. W., Jones, J. S., & Dumas, J. C. (1984). Ground response to dynamic compaction. Journal of Geotechnical Engineering, 110, 757–774. Mi, H. Z., & Yang, P. (2012). A field experimental study of compaction piles in collapsible loess foundation. Rock and Soil Mechanics, 7, 1951–1956+1964 (in Chinese). Pye, K. (1995). The nature, origin and accumulation of loess. Quaternary Science Reviews, 14, 653–667. Qian, H. J., & Wang, J. T. (1985). Collapse loess foundation. Beijing: China Architecture and Building Press (in Chinese). Si, B. W., & Tang, Y. Q. (1999). Mechanism of technology for dynamic consolidation of soil in deep holes and relevant engineering practice. Construction Technology, 28, 48–49. (in Chinese). Wang, X. L., Zhu, Y. P., & Huang, X. F. (2014). Field tests on deformation property of self-weight collapsible loess with large thickness. International Journal of Geomechanics, 14, 613–624. Zhang, Z. (2011). Analysis on the treatment of collapsible loess foundation about multi-storey buildings with cushion method. Construction Quality, 29, 23–26. (in Chinese). Zhang, G. P., Huang, X. F., Zhu, D. Z., & Zeng-Hong, X. I. (2011). Research on soaking load tests for treatment of self-weight collapse loess with ddc piles of different depths. Journal of Water Resources & Architectural Engineering, 9, 28–33. (in Chinese). Zhang, Y. C., Yang, Q. F., & Zhang, X. Y. (2010). Foundation design of collapse loess with large thickness. Chinese Journal of Geotechnical Engineering, 32, 263–266. (in Chinese). Zhao, Z. H., & Xie, L. (2014). Application of dynamic compaction method in the collapsible loess foundation improvement. Construction & Design for Project, 1, 157–159. (in Chinese). Zhou, S. L., Wang, J., & Zhang, M. Q. (2006). The actuality and prospect of the mechanism research in dynamic consolidation. Journal of Chongqing Jiaotong University, 25, 65–70. (in Chinese).