Abstract This paper describes the model testing of auger displacement piles in clay, a recent development that provides significant environmental advantages over conventional piling. Constructed model piles were loaded to failure in compression and tension and were physically examined after construction and loading. The work presented investigates factors which affect the design of such piles such as the adhesion factor, shape and integrity of the concrete flanges and the location of shear displacement surfaces which determine the effective diameter. Simple laboratory model tests provided a fast, inexpensive and effective technique for studying the physical processes governing pile performance and have complimented full-scale tests.
Improved auger design led to the construction of piles with full flange integrity and an increased load capacity.
Introduction Auger displacement piles were developed in Belgium in the 1960s as a new method of pile construction combining the best aspects of bored and displacement piles. The Atlas and Omega piles are probably the best known types of auger displacement piles and have gained widespread popularity in mainland Europe.However, there is relatively little practical experience in UK soil conditions reported in the literature. Similarly, there is much that needs to be understood in how this novel pile type modifies the soil stress state during installation and how the pile flanges behave during loading. This paper describes a oneyear laboratory research project that was carried out as part of an MEng degree to investigate the behaviour of auger displacement piles. The continuous helical displacement (CHD) pile recently developed by Roger Bullivant, which has similar attributes to the Atlas pile (Figure 1), was modelled.
A steel hollow stem boring head with helical flights is screwed into the ground while displacing the soil laterally. When the founding depth is reached, the direction of rotation is reversed and the auger is withdrawn. On withdrawal, concrete is pumped at high pressure through the hollow stem and tip, forming a pile with flanges that follow the path created by the helical flight. This method of pile construction involves no ground vibration and no soil arisings that require off-site transportation and disposal. Soil is displaced laterally and not transported up the auger flights thus increasing the strength of the soil and consequently enhancing pile capacity, while less concrete and manpower is required.
The objective of the project was to model the field construction and pile behaviour. Simple laboratory model tests provided a fast, low cost and effective technique for studying the physical process governing pile capacity and performing parametric studies under controlled conditions. Valuable insight was gained in terms of soil deformation during construction, shear displacement surfaces during loading, integrity and uniformity of the concrete flanges and short-term loading capacity and behaviour. The controlled parameters of the tests will help understand , explain and improve full-scale construction and design.
Theory Typically, auger displacement piles in clay attribute about 90% of their total bearing capacity to shaft capacity. To calculate this, a reliable estimation of skin friction, qsu, and shaft area, A s, is needed. Auger displacement piles have so far been designed using field test data (Hollingsworth, 1992; Van Impe, 1984, 1987, 1988; Bustamante & Gianeselli, 1992) and previous research concentrated on deriving semi-empirical correlations based on field (CPT, SPT, PMT) and model tests. Shaft capacity in clay is traditionally calculated in terms of the undrained shear strength, cu, by: Qu=S(A s. q su )=SA s(a c cu )Where a is the adhesion factor. To calculate the shaft area the effective pile diameters recommended were functions of the core or flange diameters (Figure 2) depending on flange thickness. One of the aims of the project was to clarify which diameter to use for designing the CHD pile and to provide guidance for the a values to use.
Apparatus and experimental techniques Three model augers were machined out of aluminium alloy (Figure 3).These represent 1:10 scale models of the CHD auger heads and are summarised in Table 1, where D is the diameter. Auger 3 was made by modifying auger 1.
The model construction process was set up to resemble the full-scale rig (Figure 4).A variable speed motor rotated the hollow shaft which could be moved up and down, while a peristaltic pump sent pressurised grout to the auger head. The applied torque, pushing and pulling forces and grout pressure had to be carefully co-ordinated to produce piles with a good shape and to minimise soil disturbance.
Well graded sand, rapid hardening Portland cement, ground granulated blastfurnace slag and plasticiser were needed to produce a grout with similar proportions to the concrete used on site. High quality E-grade Kaolin clay (PL=30%, LL=51%, I p=21%) was used to produce compacted and consolidated samples, with c uwas measured by a hand vane.
Piles were loaded by maintained load increments and were considered to fail when the settlement rate was not decreasing significantly. Compression and tension tests were carried out.A total of 25 piles were constructed, and most of them were loaded. Discussion of some of the findings follow.
Results and discussion Shear surfaces, construction technique and flange integrity The principal displacement shear surface in clay was found to occur over a very thin region surrounding the pile. Figure 5, which is a section through a sample containing horizontal layers of dye, shows that this occurs at the outer flight diameter and this was the case for all piles.
Incomplete flanges broke off and led to the development of a second shear surface during loading for soils with a stiffer base layer. Figure 5 shows this effect schematically: the forces on the cylinder between the core and flange diameters cannot be balanced if the flanges fail, so the second shear surface develops before the base resistance of the outer cylinder is fully mobilised.
Diagonal marks on some extruded unloaded samples indicated a circumferential shear surface at the outer flange diameter during construction (Figure 6).The main flange of augers 1 and 3 extend for more than two revolutions around the core, causing clay to be trapped between the flanges and shear circumferentially during construction.
The observations helped in understanding how the soil and concrete behaved, and gave clear evidence as to how to modify the auger head. Experiments indicated that the large bottom flange of auger 1 was blocking pressurised grout from reaching the flanges on time, which was essential to prevent them from collapsing during construction. Removing the base flange of auger 1 to make auger 3 had a remarkable effect on flange integrity (Figure 7) and produced thick solid flanges.
Auger 2, with the auger flight extending for just over one revolution and a refined bottom flange produced solid flanges and did not shear the soil circumferentially during construction. The net result was a clearly defined effective diameter at the outer flange diameter and a 20% increase in a values (see Investigation into design parameters, below).
Changes in moisture content and clay strength For a clay within the same consolidated sample, a lower moisture content generally implies a higher strength. Figure 8 shows the moisture content around a pile after loading. All points are within just +/-4% of an average of 37.8%.The points that are furthest away from the pile are within 0.3% of this average and this indicated no significant change, so the pore water in the vicinity of the pile is redistributed locally after piling and later during loading. The first point was taken from between the flanges of the extruded piles. It has a higher moisture content than average, as water may have diffused out of the grout. At r = R pile , where the soil was sheared, the water content is high compared to the rest of the points. Moisture is lowest directly next to the pile and then increases away from the pile, to reach a plateau at r = 3R pile .Figure 8 implies stiffening of the soil next to the pile, which softens after failure.
Investigation into design parameters Ultimate shaft capacity was assessed either as the difference between the measured compressive load and the calculated base capacity, or as the tensile failure load. The actual measured shear surface areas were used, with appropriate treatment of tapered parts, to derive values for a=q su /c u. The effect of auger design can be seen in Figure 9, which shows the variation of a values against the shear strength of the soil. The following can be noted: There is a clear trend for a decrease in a as c uincreases. This is not unexpected as softer clay displaces laterally easier than firm clay, hence causing less disturbance. Stiffer clays are generally more brittle.
The top ellipse shows the results from piles constructed with auger 2, which does not cause circumferential shear during construction as discussed in the section on shear surfaces, construction technique and flange integrity. a is higher than other auger tension tests by about 20%.
For augers 1 and 3, the compression tests gave lower a values than the corresponding tension tests. The pointed shape of the pile tip reduces the base capacity, which was calculated using A b.N c. c u, where N c= 9 for piles. Alternative base treatments would converge the compression test results to the tension tests. This difference did not appear in piles constructed with auger 2 because of the more rounder pile tip, so tension and compression results are grouped together.
This variation shows that a is not a simple number but depends on several soil and construction parameters which may not be easily identified through field tests.
Conclusions Laboratory-scale model CHD piles were used to investigate pile construction and design parameters in consolidated clay. The following conclusions were reached: The principal displacement shear surface was found to occur at the outer flange diameter and this determines the effective diameter. Incomplete flanges led to the development of a second shear surface at the pile core.Two of the augers caused circumferential shear of the soil during construction, thus reducing the adhesion factor.
Moisture content distribution around a pile indicates soil stiffening due to pile construction, which improves the shaft load capacity.
a values decrease as clay strength increases and depend on auger design.
An improved auger design resulted to the construction of piles with full flange integrity and reduced soil disturbance during construction, leading to a 20% increase in a values and hence in skin friction.
This work has showed how simple tests can be used to investigate a construction method and pile loading behaviour in different soil conditions. Laboratory model tests provided a fast, inexpensive and effective technique for studying the physical processes governing pile performance and have complimented full-scale tests. Further research can be aimed at examining the effect of these parameters closely in order to provide improved design methods.
Acknowledgements I would like to express my gratitude to my project supervisor Dr Abir Al-Tabbaa of Cambridge University and to Dr Fiona Chow of Geotechnical Consulting Group for their guidance, support and constructive comments. The financial support provided by Roger Bullivant is greatly appreciated.
Hollingsworth JR and Imbo-Burg RM (1992).The Atlas screw pile - construction, design and performance. Piling: European practice and world-wide trends, pp170-176.Thomas Telford, London.
Bustamante M and Gianeselli L (1993). The bearing capacity and settlement of displacement piles. Deep foundations on bored and auger piles, Van Impe (ed) pp21-34.Balkema, Rotterdam.
Van Impe WF (1984). Pile loading test results on an Atlas auger pile. In proceedings of the 6th conference on soil mechanics and foundation engineering, Budapest, (G Petrasovits ed), pp419-426.
Van Impe WF (1987).Discussion of some loading test results on cast insitu auger piles. In proceedings of the 8th Asian reg. conference on soil mech. and foundation engineering, Kyoto, Japan, pp423-426.
Van Impe WF (1988).Consideration of the auger pile design. Deep foundations on bored and auger piles, Van Impe (ed), pp193-218.Balkema, Rotterdam.