Instrumentation can provide valuable data during the construction process and confirmation of design performance in the st-construction period. It is the principal means by which we can check design theory against actual behaviour in the ground and should form a vital part of any engineering trial.
An informed choice of instruments, proper installation and a suitable monitoring regime are required to gain the maximum benefit from the financial commitment.
Geotechnical research has traditionally provided the impetus for innovations in measuring ground behaviour and has spawned most of the 'standard' instrumentation systems commercially available today.
The December 1997 issue of Ground Engineering described two new trials at the Bothkennar soft soil research site in Scotland. This described a 'light' raft foundation built by BRE's Centre for Ground Engineering & Remediation as part of a DETR Construction Directorate research programme to investigate the design and performance of low-rise building foundations on soft ground.
BRE and Bauer Foundations had at this time also begun a trial as part of a DETR Partners in Innovation project to study the use of vibro stone columns in soft cohesive soils. This was to include the installation of stone columns using the dry bottom-feed technique to support a series of individual pad foundations on treated and untreated ground and a raft foundation similar to the one already in place on untreated ground.
Bauer installed a total of 44 columns (including two individual trial columns). It subsequently constructed eight individual strip and pad foundations and the 8.1m square semi-reinforced raft and provided kentledge and loading facilities (Figure 1).
Before ground treatment and the earlier construction of the untreated raft, comprehensive instrumentation systems were installed by BRE and a programme of monitoring put in place.
The techniques employed on the Bothkennar trials reflect a mixture of standard methods and more specialised techniques adapted for the particular purpose.
The overall requirement was that the instruments had to survive and give accurate data during the installation of the vibro stone columns and subsequently to measure ground behaviour and foundation performance during construction and loading of the raft and pad foundations.
Figure 2 shows, for example, how a range of monitoring systems, as well as precise levelling, were employed to monitor the vibro ground treatment and subsequent performance of the treated raft foundation.
Vertical ground movement and foundation settlement
Three borehole magnet extensometer gauges were installed before construction of the untreated raft to measure distribution of settlement at depths up to 16m inside and outside the foundation area. A similar gauge was installed between stone column locations for the treated raft.
Before foundation construction, the settlement gauges on both rafts were monitored using a BRE portable ground frame which ensures measurements within the 1mm accuracy normally quoted for such systems.
For greater precision, the basic tape/reed switch probe was combined with an adapted micrometer head which can be located on stainless steel tubes cast into the concrete. Figure 3 shows the head in position on the treated raft. In the ground treated by vibro stone columns the settlement gauge was installed to 4m below the maximum treatment depth and was used to measure the distribution of vertical ground displacement due to the treatment.
All foundation and ground surface settlement monitoring was carried out by precise levelling using an automatic quickset level with optical vernier. Levels were related to a deep datum, founded in gravel at approximately 22m below ground level.
Temporary shallow datums closer to the foundations were found to rise and fall seasonally with the crust layer by several millimetres.
Sleeved ground settlement points were installed at 0.4m below ground level radiating away from the raft and stainless steel studs with domed nuts were cast into the foundations. Survey closure errors were routinely significantly below 1mm while measured foundation movements were relatively much larger. The technique is proving to be very effective in mapping the distortion of the raft foundations.
Lateral ground movement
To measure lateral ground displacement during stone column construction and subsequent foundation loading, a special shallow inclinometer system was devised comprising individual electro-level measur-ing units mounted in rigid, articulated aluminium box sections (Figure 4).
This was grouted into a small diameter borehole and measured ground displacement up to 2.5m below ground level due to a column installation and subsequent loading by a pad foundation. Displacement was related to the top of the gauge which was monitored by optical means. The degree of resolution of the electrolevels is greater than the torpedo systems, although the range is generally much smaller. The system is very flexible and has been used in a wide variety of applications for ground and structural monitoring.
A conventional servo-accelerometer torpedo type inclinometer gauge was installed to measure lateral ground displacement immediately out- side the area of the untreated raft foundation. The technique was of limited value as lateral ground movements have been very
Measurement of insitu earth pressures and pore water pressures
A vital part of the trials was to measure changes in lateral and vertical earth pressure and pore water pressures at depth below the pads and raft foundations. It was also desirable to monitor changes that occurred during the installation of the vibro stone columns.
A total of 24 BRE miniature earth pressure cells were installed in the soft clay beneath untreated and treated foundations from 150mm and 200mm diameter boreholes. A special placing device was used to push cells horizontally up to four borehole diameters, orientated to measure horizontal or vertical total pressure around stone columns.
Cells were also installed to measure vertical pressure beneath the base of stone columns constructed to different depths to support loaded pads and similarly under a typical column supporting the raft. Figure 5 shows the placing device with a cell loaded to measure vertical earth pressure.
All the cells were installed prior to ground treatment and stress changes were continuously monitored during poker penetration and column construction (Figure 6). Pneumatic piezometers were installed at similar locations to the total pressure cells around and between columns to measure pore pressure changes during and subsequent to column installation and foundation loading.
Measurement of stress distribution below the loaded foundations
Flatjack type 300mm diameter pneumatic pressure cells were installed beneath the treated pad and raft foundations to measure stress distribution between columns and the intervening soil. Figure 7 shows a cell being installed in the firm clay crust immediately beneath formation level for a pad foundation. The top of a typical stone column can be seen uncovered.
Cells were also installed in the top section of columns below most of the pad foundations and selected columns under the raft. The top of the column was excav- ated by hand and cells placed within pockets of graded granular material prior to replacing and re-compacting the stone (Figure 2). Embedment cells operating on the vibrating wire principle were installed to measure contact pressure on the underside of the untreated raft.
All the instruments installed prior to vibro treatment survived the installation process and are providing very valuable and, in some cases, unique data associated with stone column and foundation performance in soft soil. In particular, the borehole extensometer, miniature pressure cells and special inclinometer measuring lateral soil displace- ment are likely to provide valuable data in relation to vibro stone column construction in soft soil.
Both raft foundations were loaded in two stages along the reinforced edge downstand to simulate a maximum structural line load of just over 50kN/m. The strip/pad foundations were also loaded in two stages to a similar line load, equivalent to a bearing pressure of approx.70kN/m2.
Additional load was applied in two further stages to two of the treated and an untreated pad to investigate the performance of the interactive stone column/soil system as failure was approached and the contribution made by the shallow crust.
It is anticipated that the trials overall will provide a valuable insight into shallow foundation performance and improve design for applications of stone columns in soft soil.
Ken Watts, Centre for Ground Engineering & Remediation, BRE