Problems in dealing with difficult ground in Earth Pressure Balance (EPB) machines and jacking pressures in pipe-jacking have led to industrial sponsorship contributing to research projects at Oxford and Cambridge Universities. The evening’s papers showed overviews of work in progress on long-term research projects being conducted under the supervision of Professor Guy Houlsby at Oxford and Professor Robert Mair at Cambridge.
Soil conditioning treatments for EPB machines are often selected on a trial and error basis, which sometimes result in inappropriate use of conditioning products. In pipe-jacking, the effects of the lubricant additives on the jacking forces need to be further investigated to help select the optimal lubricant for different ground conditions. Systematic comparison of the properties of various conditioning agents, their performance in laboratory tests and in the field is the information that engineers need to select soil conditioning treatments. The Engineering and Physical Sciences Research Council, the Pipe Jacking Association together with a group of other industrial sponsors have contributed to the research funds for projects which will gain more precise knowledge about these topics.
Conditioning of clays – field Investigations
The research carried out in Cambridge concentrates on the conditioning of clays. Xavier Borghi, of Cambridge University, opened the proceedings with a report on his field investigations of soil conditioning for EPBs. He began by describing the operational principles of an EPB machine. He then explained the importance of the screw conveyor in controlling the pressure in the excavation chamber, and described the factors affecting the pressure dissipation along the screw conveyor. He pointed out the difficulties that are encountered with the operation of the machine in different ground conditions and stressed the importance of soil conditioning to modify the ground properties to suit the machine requirements. Appropriate use of soil conditioning results in a range of benefits, in particular:
Usual types of conditioning agents are foams and polymers. The aim is to produce a paste of undrained shear strength ranging from 5kPa–25kPa. The selection of the type of conditioners, their concentration, the volumes injected and location of the injection points, represent many variables affecting the performance of the EPB machines. The following phases were defined:
Data from instrumented EPB machines are being collected on Contract 220 and 250 from the UK’s CTRL project. The initial field measurements looked at the earth pressures at various points in the excavation chamber and along the screw conveyor. Further work will relate in detail the conditioning parameters and the performance of the machines. Conditioned soil samples are collected at the discharge of the screw conveyor and are characterised to support the interpretation of the monitoring data.
Comparisons will be done between data gathered in the field, the properties of conditioned clay samples as measured in the laboratory tests, and findings from model screw-conveyor test.
Conditioning of clays – laboratory tests
Andrew Merritt, of Cambridge University, continued the presentation with a description of his laboratory tests measuring the undrained shear strength of clays conditioned with different foam and polymer treatments, using a large-scale fall cone apparatus and a shear vane. He then presented results from tests with a model EPB screw conveyor that has been developed at Cambridge University. Instrumentation measures the mechanics of the conveyor operation and the pressure gradient along the screw during tests with different conditioned soil samples and conveyor operating conditions. The measured pressure gradients were linear, and depend on the strength of the soil sample, the applied ‘head chamber’ pressure, the screw speed, the discharge outlet restriction, and the screw geometry (T&TI, April p29).
Lubrication in pipe-jacking
The final contribution from the Cambridge team was a presentation by Xavier Borghi on his investigations into lubrication for pipe-jacking. A small-scale model has been commissioned to simulate the installation of a pipe-jack. The apparatus, involves an instrumented pipe that is installed in a clay model of controlled stress history. Lubricants are injected into an overcut excavated around the model pipe, and the stresses and pore pressures acting on the pipe are measured during the installation and the subsequent equilibration stage. Through these tests, the effects of different swelling inhibiting additives can be quantified in terms of contact stresses on the model pipe, allowing conclusions on their effectiveness to reduce jacking forces.
Conditioning of sands
Miguel Peña, of Oxford University, talked about his research with Sotirios Psomas on the conditioning of sand. Sands used for the research were:
The research also used a variety of foams and polymers supplied by Nishimatsu Construction Co. The foam properties were investigated in a similar way to Merritt’s project (Figures 1 and 2).
Sand/foam and sand/foam/polymer mixes were subject to consolidation tests to investigate the compressibility of the mixes (Figure 3).
The shear properties of foamed sand were also investigated in small and large shear box tests. The small shear box results here showed the friction reducing properties of the foam conditioner, as in Figure 4.
The tests in the large shear box gave scope for more careful measurement of pore pressures and thus calculation of effective stress.
Lastly slump tests were carried out to investigate its use as a simple index test for soil conditioned materials. Results were obtained for both foam mixtures and polymer mixtures. The results for foam mixtures are shown Figure 5.
In conclusion foam can be used to increase the void ratio of these soils even at fairly high stresses. At a high void ratio the strength is much lower than at conventional void ratios. It is thought this might be due to the pore pressures developed in the foams. Sand/conditioning agent mixtures have a much lower permeability then sand itself. The slump test provides a simple index of conditioned soil behaviour.
At this point the Chairman opened the presentations to further discussion from the floor (see box).
Related Files
Foam drainage time comparison table
Compressibility results
Comparison of shear behaviour
Comparison of foam mixture slumps
Foam expansion comparison table
