Geotechnical and Geological Engineering 20: 89^121, 2002. # 2002 Kluwer Academic Publishers. Printed in the Netherlands. 89 Applicability of piezocone and dilatometer to characterize the soils of the Venice Lagoon GIUSEPPE RICCERI, PAOLO SIMONINI and SIMONETTA COLA Dipartimento di Ingegneria Idraulica, Marittima e Geotecnica, Universita' di Padova, Via Ognissanti, 39, 35129 Padova, Italy (Tel: ++39-049-8277980; Fax ++39-049-8277988; e-mail: giuseppe.ricceri@unipd.it; e-mail: paolo.simonini@unipd.it; e-mail: simonetta.cola@unipd.it) (Received 22 January 2001; revised 8 June 2001; accepted 7 August 2001) Abstract. The effectiveness of two geotechnical investigation tools0the piezocone and the dilatometer0to characterize the soils forming the shallowest deposits of the upper quaternary basin of the Venice lagoon soil is examined in this study. For this purpose, the results of a comprehensive site and laboratory investigation carried out recently over a small lagoon area0the Malamocco Test Site0are used to evaluate the applicability of the most widely used charts or correlative equations to characterize soil pro¢le and estimate the main geotechnical properties of these soils, when applied to the interpretation of CPTU and DMT results. The particular interest of this site0apart from its unquestionable historical relevance0is the presence, apparently without any regular pattern in depth and site, of a predominantly silty fraction combined with clay and/or sand, thus forming an erratic interbedding of various types of sediments. This case represents therefore the opposite condition of that which has been normally utilized in the past to calibrate the two devices, namely the presence of particularly homogeneous natural deposits or arti¢cially sedimented homogeneous layers of sand or clay. The Malamocco Test Site may therefore be considered as test benchmark for the applicability of the two devices to characterize highly heterogeneous silty deposits. Key words: dilatometer, heterogeneous soil, piezocone, silt, site investigation, Venice lagoon 1. Introduction The Venice lagoon is suffering an overall rapid deterioration, including dramatic changes in the water and sediment balance of the basin with corresponding environmental damage. Following an accelerated rate in human-induced land subsidence between 1946 and 1970, the city of Venice has observed an increase in the frequency of £ooding with a record tidal level of nearly 2 m above mean sea water level measured in November of 1966. Since then, numerous engineering solutions have been proposed, including the use of movable gates located at the three Lagoon inlets (i.e. Malamocco, Lido and Chioggia) to control water levels within the lagoon. An extensive study on the characterization of these soils has been undertaken in the design of these mobile gates (Ricceri, 1997). Within this study, a special test site had been selected at the Malamocco inlet0referred to here as the ‘‘Malamocco 90 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA Test Site’’ (MTS)0where, in a limited area, several geotechnical investigations including deep boreholes together with piezocone, dilatometer, self-boring pressuremeter and cross/down hole tests have been concentrated and carried out on contiguous verticals. In addition, in order to obtain high quality undisturbed samples, a 220 mm diameter piston sampler was used. Some results of these investigations concerned mainly with the small and large-strain response of these soils measured in the laboratory have been already published (Cola et al., 1998; Cola and Simonini, 1999). From these studies it appeared that the main feature of these soils, that strongly in£uenced their overall behaviour, is the presence of a predominant silt fraction, the latter being a consequence of mechanical degradation and breakage of the original sand particles. These silty materials present intermediate characteristics between ¢ne and coarse grained soils, namely they are poorly structured materials and highly sensitive to stress-relief due to sampling. However, specimens can be trimmed for mechanical testing in the same way as for cohesive soils. The silty fraction in the Venice soils is combined with either clay or sand or both together, without any regular pattern with depth. One of the aims of the MTS is to evaluate the effectiveness and reliability of two in-situ testing devices, namely the piezocone and dilatometer, to characterize soil pro¢le and estimate the main geotechnical properties, the latter approximately determined through the use of a series of semi-empirical correlative equations applied to the readings recorded throughout the test. When necessary, some modi¢cations are introduced to the correlations to suit these very particular silty soils. The ¢nal goal of the study is to provide a calibration of both site investigation tools for an extensive characterization of the ground at the inlets of the Venice lagoon and avoided the use of more expensive borehole and laboratory testing as well as the scale effects inherent in these highly heterogeneous soils. 2. The Piezocone and the Dilatometer 2.1. THE PIEZOCONE TEST The cone penetrometer test (CPT) has been used for many years as a standard investigation tool essentially for determining the ground pro¢le quickly and cheaply. The ¢rst attempts to measure porewater pressure throughout the process of penetration into the ground were reported separately by Torstensson (1975) and Wissa et al. (1975). They showed the advantage of incorporating a pore pressure transducer into the static penetrometer so that continuous coupled pro¢les of tip resistance and/or sleeve friction could be obtained. Since the 1980s this new instrument, referred to as a Piezocone (CPTU), has developed rapidly with new varieties becoming available, the majority of them built with the standard dimensions of the Dutch cone (60‡ apex angle and base area of 10 cm2). Unfortunately the position of the pore pressure ¢lter has not been APPLICABILITY OF PIEZOCONE AND DILATOMETER 91 standardized. Its position at the tip, along the cone or just behind it along the shaft, has a signi¢cant in£uence on the magnitude of the measurements. It becomes therefore particularly important to know the type of the piezocone utilized in the investigations. The results of the piezocone test are customarily represented in terms of the three pro¢les of cone resistance, sleeve friction and pore pressure measured throughout the penetration. If the cone is stopped in cohesive soils exhibiting positive excess pore pressure generated during penetration, the decay of the excess pore pressure can be recorded and interpreted to provide consolidation properties. Alternatively, in sandy formations the pore pressure transducer indicates the water head at that depth. In the tests carried out at the MTS, the piezocone was the standard Dutch cone provided with the pore pressure transducer located along the shaft just behind the cone with a net area ratio a ¼ 0:83. This ratio is needed in order to correct the resistance allowing for the effects of pore pressure acting on the back of the cone. The following symbols have been used in the interpretation of the piezocone tests (IRTP, 1999) qc = cone resistance; u2 = pore pressure measured at the cylindrical extension of the cone; fs = measured sleeve friction; qt ¼ qc þ ð1  aÞu2 = corrected cone resistance; qn ¼ qt  svo = net corrected cone resistance (svo ¼ total overburden stress); Du2 ¼ u2  uo = excess pore pressure, where uo is water pore pressure in hydrostatic condition; Rf ¼ 100fs =ðqt  svo Þ = normalized sleeve friction ratio; Bq ¼ Du2 =ðqt  svo Þ = pore pressure ratio. 2.2. THE DILATOMETER TEST The £at dilatometer test (DMT) was introduced by Marchetti in 1975 (Marchetti, 1975, 1980) and its use as a simple in-situ test has largely increased, especially in North America. The use of the DMT in Italy is reserved for particular relevant site investigations. The standard dilatometer utilized at the MTS is a £at blade, 14 mm thick, 95 mm wide, and 220 mm long, with a £exible stainless steel membrane 60 mm in diameter located on one face of the blade. The membrane is in£ated with high-pressure nitrogen gas and two values of gas pressure are recorded, one to lift the membrane off the sensing device located just beneath the membrane and one to cause 1 mm de£ection. These two readings are corrected for membrane stiffness and noted as p0 and p1 respectively. The test was subjected to standardization procedures (ASTM, 1986; Eurocode 7, 1999). 92 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA From the two corrected pressure readings Marchetti derived the following index parameters, namely:    the material index: ID ¼ p1  p0 =p1  u0 the horizontal-stress index: KD ¼ p0  u0 =s0vo the dilatometric modulus: ED ¼ 34:7ðp1  p0 Þ where s0vo = in-situ vertical effective stress. The standard DMT output in Italy is proposed in terms of corrected pressure readings, of the three index parameters ID , KD and ED and of a series of empirically derived properties through correlations based on the above index parameters, such as soil type, soil unit weight, total overburden stress, effective overburden stress, vertical constrained modulus M, undrained shear strength cu , coef¢cient of pressure at rest K0 and overconsolidation ratio OCR (e.g. Totani et al., 1999). Several studies have been carried out in the past on both the investigation devices and from the basic readings, attempts have been made to determine the largest number of soil properties, the latter evidently being estimated in a fairly reliable way (Campanella and Robertson, 1983). Many semi-empirical correlations have been proposed by researchers comparing the CPTU and DMT data with laboratory or ¢eld test values at a speci¢c site. It was found that adjustments in the correlations were needed to adapt the latter to local geological soil conditions. For example, Mayne and Martin (1998) collected over 230 correlations for the interpretation of dilatometer test readings. The advantage of these correlations is their straightforward application for a simple even fairly approximate description of soil properties. On the contrary, their main limitation is that they often provide soil properties which are not directly measured or derived on the basis of the principles of mechanics. Consequently they are only strictly applicable to the geotechnical description of the site where they have been calibrated. The determination of a series of relevant geotechnical properties of the Venice lagoon soils will be considered, namely the soil classi¢cation, the overconsolidation ratio OCR, the coef¢cient of earth pressure at rest Ko , the undrained shear strength cu , the friction angle f0 , the horizontal consolidation coef¢cient ch , the constrained modulus M and the small-strain shear stiffness Gmax . The values of the above properties carefully measured in the laboratory will be compared with those estimated through the use of charts or correlations in order to evaluate the degree of applicability of the two testing devices. 3. Brief Geological History of the Venice Lagoon and Mineralogy of the Sediments The Quaternary basin reaches depths of approximately 800 m over the whole lagoon area. During the Quaternary period the lagoon area underwent alternating periods APPLICABILITY OF PIEZOCONE AND DILATOMETER 93 of marine transgression and regression, so both marine and continental sediments coexist. Particularly, the deposits forming the upper 50^60 m below mean sea level, the depth of interest here, are characterized by a complex system of interbedded sands, silts and silty clays, deposited during the last glacial period of Pleistocene (Wˇrm) when the rivers transported £uvial material from the Alpine ice ¢elds. The Holocene is only responsible for the shallowest lagoon deposits, about 10^15 m thick. The top layer of Wˇrmian deposits is composed of a crust of highly overconsolidated clay, commonly referred to as caranto, where many historical Venetian buildings are founded. It was subject to a process of overconsolidation as a result of oxidation during the 10,000 year emergence of the last Pleistocenic glaciation. Figure 1 shows a view of the Venice lagoon in the present con¢guration with the location of the MTS at the Porto di Malamocco inlet. The depositional patterns of the Venetian sediments are rather complex due to the combined effects of geological history and human action, which signi¢cantly modi¢ed the morphology of the lagoon, inlets and channels over the centuries. Subsurface pro¢les are characterized by irregular alternation of three soil types, sand, silt and very silty clay, with a few thin layers of compacted peat. Therefore, for classi¢cation purposes the soil types have been reduced to three: medium to ¢ne sand (SP-SM), silt (ML) and very silty clay (CL) according to the Uni¢ed Soil Classi¢cation System. Despite the highly heterogeneous grain-size conditions, the basic material properties vary over a relatively narrow range due to their common mineralogical origin and depositional environment. Typical grain size distribution and mineralogical composition of the soils at the MTS are compared in Figure 2. Sand composition is mostly calcareous and siliceous, with a predominance of calcite and dolomite crystals, especially at higher depths. When the carbonate and quartz-feldspar fractions decrease the clay minerals increase, although they never exceed 20% in any sample. Silt and clay particles were generally formed by mechanical crushing in a continental environment and are aggregated in an irregular assemblage, characteristic of a predominant £occulated structure. Clay minerals are mainly composed of illite, prevalently 2M muscovite, with chlorite, kaolinite and smectite as secondary materials. 4. Soil Pro¢le at the Malamocco Test Site Figure 3 shows the soil composition with depth at the MTS (the ground level at the MTS is at 10.5 m below m.s.l.) determined from several sieve and sedimentation analyses. As discussed above, the classes of Venice soils were reduced to three, namely sand, silt and clay. The predominance of silt and sandy fractions can be clearly appreciated at the MTS. From borehole log and particle size analyses, the three classes of soils apparently occur in the proportion of 35% sand, 20% silts, 40% very silty clay 94 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA Figure 1. View of the Venice lagoon with the location of Malarocco Test Site. and only 5% medium plasticity clays and peat. Percentages of silt exceeding 50% are present in 65% of the samples analysed. These values are only slightly different from those reported by Rowe (1973) in a previous study and relative to a site lying close to the historic city, about 12 km distance from the MTS. There the soil nature down to 62 m was predominantly due to sandy fraction (61%) with less silts and very silty clays (34%) present. The second column reports the values of D60 and D10 whereas the Atterberg limits of cohesive materials are reported in column three. The average value of liquid limit LL and of plasticity index PI is respectively equal to 36  9% and 14  7%. Saturated bulk density gsat and void ratio e, both determined in the laboratory on undisturbed samples, are also reported in Figure 3. Typical values of e range from APPLICABILITY OF PIEZOCONE AND DILATOMETER Figure 2. 95 Comparison between typical grain size distribution and mineralogical composition of soils. about 0.6 to 1.1. Higher values, around 2^3, are due to the laminations of organic soils, not depicted in column 1, included in the cohesive formations. It is interesting to note that the void ratio between approximately 19 and 36 m below mean water (sea) level (MWL) lies in the range between 0.7 and 1.0; at greater depths the majority of e values are somewhat lower and fall in the range 0.6^0.75. Speci¢c weight Gs was determined on several samples with its average value turning out to be equal to 2:77  0:03. 5. Stress History The estimate of soil stress history is generally a dif¢cult task and in the case of Venetian soils is even more so, due to the coupled effect of a complex geological history and continuous modi¢cation of the natural morphology by human action. Overconsolidation ratio OCR was determined on the basis of the estimate of preconsolidation stress calculated from oedometric tests using the traditional Casagrande construction and comparing it to the overburden stress. Many dif¢culties were encountered in applying the method because a low-pronounced yielding 96 Figure 3. GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA Soil pro¢le and basic geotechnical properties at the Malarocco Test Site. curvature, dividing overconsolidated from normally consolidated states, was observed in oedometric curves. This is probably due to the in£uence of sampling disturbance, which is more important in non-structured and low-plasticity silty soils such as the Venetian ones. With reference to the in£uence of sampling disturbance on the compression response of these silty materials, no speci¢c measurement was carried out. An attempt to evaluate the effect of sample disturbance was performed by applying the method suggested by Andersen and Kolstad (1979), who observed that the vertical strain in oedometic compression at the overburden stress increases with the level of disturbance. In our case, this reference strain is signi¢cantly higher than suggested values (e.g. Holtz et al., 1986), thus suggesting the presence of a high degree of disturbance. It should be noted however that this method was successfully applied only to materials characterized by high plasticity, for which the yielding in one-dimensional is clearly detectable, that is not the case for typical Venice lagoon soils. APPLICABILITY OF PIEZOCONE AND DILATOMETER 97 Additional OCR data needed to provide a more signi¢cant pro¢le with depth were indirectly obtained from the undrained shear strength cu , which is a function of the current vertical overburden stress and OCR. For this purpose one of the most accepted empirical relationships is that proposed by Ladd and Foott (1974) and Ladd et al. (1977), which was validated on the basis of the results of consolidated undrained triaxial compression/extension tests and simple shear tests: cu ¼ S OCRm s0vo ð1Þ where S ¼ cu =s0vo is the undrained strength ratio for NC clays, dependent on the conditions within which the soil is being sheared, and m ¼ 0:8 is an experimental exponent. According to Mesri (1975), Koutsoftas and Ladd (1985) on the basis of the analysis of the behaviour of a marine plastic clay suggested S ¼ 0:22 as average reasonable value for undrained stability analysis of soft clay deposits, irrespective of their degree of plasticity. Although based on ¢eld vane data, we however preferred, in this context, to evaluate S with the use of the well-known relationship proposed by Skempton (1957) for normally consolidated soils, such as: S ¼ 0:11 þ 0:0037PI ð2Þ which allows the clay plasticity to be accounted for in the evaluation of the undrained strength ratio. For plasticity index PI ¼ 14, the typical value for the Venetian silty formations, the relationship given by Equation (2) provides a ratio S equal to 0.16, somewhat lower than that suggested above. Resolving Equation (1) for OCR, undrained strength data from unconsolidated undrained triaxial compression tests carried out on samples taken at Malamocco were used for a very rough estimate of OCR, whose values are reported with the oedometric data in the last column of Figure 3. Note that, except the upper layer of caranto, the silty formations are slightly over consolidated with a few layers with OCR exceeding 3^4. It should be again emphasized that, in both methods used to estimate OCR, we observed a large in£uence of sample disturbance. More particularly, we noted large variations of cu even from specimens trimmed at the same level of a sample, that cannot be justi¢ed only by accounting for soil heterogeneity. 6. Piezocone and Dilatometer Tests at MTS Piezocone and dilatometer tests have been carried out very close and at increasing distances from the borehole verticals. The approximate distances from borehole verticals of those piezocone and dilatometer tests, the results of which are discussed 98 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA here, and of the selfboring pressuremeter (SBPMT) and cross-hole tests (CHT) are reported in Table 1. Figure 4 shows the results of three typical piezocone tests SCPTU19, CPTU8 and CPTU9. The corrected cone resistance qt together with sleeve friction fs and pore pressure pro¢les u2 are reported in Figure 4. The pro¢les con¢rm the high non-homogeneity of the Venetian soils to the depths investigated. Sharp variations of cone resistance, sleeve friction and pore pressure pro¢le are observed with the latter being below the hydrostatic level, not just within the Table 1. Distance of various tests from the borehole vertical Test Distance from borehole (m) SCPTU19 CPTU8 CPTU9 DMT2 DMT3 DMT4 CHT (three verticals) SBPMT 8.1 136.6 66.3 8.4 32.1 96.6 12:3  13:3  17:5 5.6 Figure 4. Pro¢le of qt, fs and u from three typical CPTU tests. APPLICABILITY OF PIEZOCONE AND DILATOMETER 99 caranto, but also in some cases in the sandy layers. This latter effect has been observed in special circumstances of dense cohesionless soil that exhibit dilation when sheared (Wroth, 1984). However, in our case this particular effect has been recorded especially in the upper few decimetres of any granular layer when the piezocone tip is driven from a cohesive layer into the granular one. For that reason, it is presumed here that the negative excess pore pressures measured are due not to the dilatative soil behaviour but to some delay in establishing the hydraulic contact between sand and the porous ¢lter. Note that Bq rarely exceeds 0.4 and is much lower than the higher values, over 0.8^1.0, reported by the literature for normally consolidated or slightly overconsolidated clays. The results of three dilatometer tests DMT2, DMT3 and DMT4 are reported in Figure 5 in terms of corrected lift off pressure p0 and expansion pressure p1 . The distance among the tests is also reported in Figure 5. Test DMT2 was driven close to the seismic piezocone test SCPTU19. The main feature to note is the continuous and coupled oscillations of the two pressures, thus con¢rming a high degree of heterogeneity already shown by the piezocone tests. The lift off pressure varies in a narrower range in all the three tests, whereas the expansion pressure shows a larger variation, especially at depths Figure 5. Pro¢le of p0 and p1 from three DMT tests. 100 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA between 18 and 28 m and again between 40 and 50 m. This con¢rms the nonhomogeneity of Venetian ground, which is characterized by continuous variation of soil layering, not only in the vertical but also in the horizontal direction. For the sake of comparison between the two tests, Figure 6 reports the pro¢les with depth of signi¢cant quantities qt , fs , u2 , p0 , p1 and Vs , where the latter is the shear wave velocity measured at various depths by the piezocone. The test SCPTU19 was in fact carried out using a special seismic piezocone (SCPTU), in order to perform down-hole wave measurements during penetration. To this purpose two geophones were installed along the shaft (spaced 1.0 m) near the tip (e.g. Tanaka et al., 1994). Wave velocity records, reported on the last column of Figure 6, were provided at depth intervals equal to 1 m. A general increase of velocity with depth is observed with very small variations among the different soil formations recorded in the cross hole test (CHT). Some large variations of velocity, measured with SCPTU19 especially at depths between approximately 29^34 m and 45^52 m below MWL, may be due to the presence of some thin peaty layers, which could in£uence the propagation of shear waves across the horizontal soil layering. Figure 6. Comparison among the pro¢le of qt, fs, u, p0, p1 and Vs at the centre of Malarocco Test Site. APPLICABILITY OF PIEZOCONE AND DILATOMETER 101 On the basis of the pro¢les depicted in Figure 6, due to the tests SCPTU19 and DMT2, located at distances of about 8 m from the borehole verticals, the estimate of the main geotechnical properties are considered and discussed in the following sections. It should be kept in mind the dif¢culty encountered in comparing the relevant properties from laboratory tests with those indirectly estimated from in situ tests, due to the in£uence of heterogeneity. That is, no clear horizontal layering provided the absolute certainty to carry out the comparison of properties within exactly the same type of soil. Nevertheless, the variation of the basic material characteristics over a relatively narrow range due to mineralogical origin and depositional environment, together with the a relevant presence of the silty fraction in the great majority of samples, suggested that a straightforward comparison could be attempted. 7. Test Interpretation 7.1. SOIL CLASSIFICATION 7.1.1. Classi¢cation with Piezocone Soil classi¢cation with the piezocone can be obtained by using tip resistance coupled with the sleeve friction (Olsen and Farr, 1986; Robertson, 1990) and/or with the pore pressure measured during penetration (Senneset et al., 1982; Wroth 1984; Robertson, 1990). All the classi¢cation methods are based on empirical charts, and the most widely accepted are those due to Robertson (1990) in which the soils are grouped into 9 classes, accounting also for OCR. Figures 7(a) and (b) report the piezocone data superimposed onto Robertson’s charts, dividing the soils in the four classes, SM-SP, ML and CL and organic soils. From Figure 7(a) it can be noted that the proposed subdivision is suitable for the classi¢cation of the Venetian soils, which mostly belong to the groups 4, 5 and 6. Data due to the organic formations lie in the right part of the chart, in the zone of the clays, where the caranto falls in the groups 8 and 9. More particularly, the contribution of the sleeve friction, with respect to the tip resistance seems very important for classi¢cation purposes. The classi¢cation is not improved by combining tip resistance with the pore pressure ratio Bq , as shown in Figure 7(b). However, one of the advantages of the piezocone with respect to the standard cone is the availability of a continuous pore pressure pro¢le, which is, more than the qt pro¢le, an excellent indicator of the presence of interstrati¢cations or laminations, particularly relevant in heterogeneous deposits such as those at the MTS. Therefore the pore pressure parameter Bq was linked to the friction ratio Rf , as shown in Figure 8. 102 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA Figure 7. Soil classi¢cation at Malarocco Test Site with the Robertson’s charts. Figure 8. Speci¢c chart to classify the Venice Lagoon soils. APPLICABILITY OF PIEZOCONE AND DILATOMETER 103 The SM-SP materials with negative or zero Bq and Rf generally not exceeding unity are located in the lower part of the chart. ML soils show similar pore-pressure response but the sleeve friction is somewhat higher, approximately between 1 and 2. In the case of CL soils, Bq lies in the range between 0:1 and 0.4, as a function of OCR. A qualitative scale of OCR, determined on the basis of the results of laboratory tests on homogeneous groups of data has been superimposed on the chart and is valid only for CL materials. As regards the Venice Lagoon soils the use of this chart (Fig. 8) instead of the above two provides a better estimate of the soil type as well as a qualitative judgement of the stress history. 7.1.2. Classi¢cation with Dilatometer The material index ID was originally used as a simple and approximate parameter for identifying the soil type, i.e. ID W 0:6 for clay, 0:6 < ID < 1:8 for silt and ID X 1:8 for sand. However, it was later recommended to combine ID with the dilatometric modulus ED for a better soil classi¢cation. Figure 9 summarizes the data obtained from the MTS and plotted on the improved chart proposed by Marchetti and Crapps (1981). The two extreme classes of soil, namely the silty clay (CL) and silty sand (SM-SP) appear to be better de¢ned than those of silts (ML). In the latter case, a much larger Figure 9. Classi¢cation by means of Marchetti and Crapps chart. 104 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA variation of ID can be noted and in some instances these soils are not properly classi¢ed as silts but as silty sands or as silty clays. Marchetti and Crapps’ soil classi¢cation chart provides indications about the unit weight of soils. When comparing the predicted values of g with those measured in the laboratory, we observe that the range of maximum scatter around the average values does not exceed 15%. Nevertheless, the chart tends to underestimate the bulk density of silty clays while overestimating that of sands. 7.2. OVERCONSOLIDATION RATIO 7.2.1. Overconsolidation Ratio from Piezocone The level of pore pressure generated during cone penetration is a function of the type of soil and of its overconsolidation ratio. OCR values were therefore plotted in Figure 10 versus the pore pressure ratio Bq calculated within silty clay layers. A general decrease of OCR along with an increase of Bq can be observed, but no reliable correlation between these two quantities can be established, the level of pore pressure being generated being much more dependent on the type of soil than on OCR. Another attempt to link OCR with piezocone results could be performed as suggested by Mayne (1991), who proposed a relationship between OCR and Figure 10. Overconsolidation ratio versus pore pressure ratio. APPLICABILITY OF PIEZOCONE AND DILATOMETER 105 (qt  u2 Þ=s0vo on the basis of the coupled use of cavity expansion theory and of the Modi¢ed Cam Clay model:
 31:33 qt  u2 7 6 s0vo 7
OCR ¼ 26 0  5 4 3 sin f 1:95 þ 1 6  sin f0 2 ð3Þ where 3 sin f0 =ð6  sin f0 Þ is M, the slope of critical state line in the (mean effective stress)- (deviatoric stress) plane. The experimental data reported in Figure 11 con¢rm the increase of OCR with normalized cone resistance: however the curve given by Mayne’s theoretical approach for a friction angle of 30‡, characteristic of the Venetian clayey silts, represents approximately an upper limit of OCRs determined indirectly from the undrained strength by means of Equations (1) and (2), as explained in Section 5. Figure 11 also reports the result of a power regression analysis only on the data from oedometric tests:
qt  u2 OCR ¼ 0:36 s0vo 0:91 ðr2 ¼ 0:85Þ ð4Þ Equation (4) could be used to predict reasonable tentative values of OCR in the slightly overconsolidated range. Figure 11. Overconsolidation ratio versus (qt  uÞ=s0vo . 106 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA 7.2.1. Overconsolidation Ratio from Dilatometer The estimation of OCR for clays was proposed by Marchetti relating KD to OCR from oedometer tests with the following correlation: OCR ¼ ð0:5KD Þ1:56 ð5Þ The use of correlation (5) is restricted to materials with ID < 1:2, free of cementation which have experienced simple one-dimensional stress histories. Among improved relationships available in literature, that proposed by Lacasse and Lunne (1988): OCR ¼ 0:225KD1:35 1:67 ð6Þ takes into account a large range of soil plasticity in the exponent, that varies from 1.35 for plastic clays and up to 1.67 for low plasticity materials. The applicability of correlations (5) and (6) to our data was veri¢ed in Figure 12. The Marchetti correlation seems to provide a rather good estimation of OCR at the lowest values of KD , when OCR is estimated from the oedometric curves. Large differences compared to laboratory results occur at higher horizontal stress indexes (KD > 10), that is, for the highly overconsolidated caranto lying at shallow depths. No improvement of data relating to any of the range of KD was obtained by adopting the Lacasse and Lunne relationship. If we apply the power regression to all our data independently of ID and stress history we have the following relationship: OCR ¼ 0:66KD1:05 Figure 12. ðr2 ¼ 0:85Þ Overconsolidation ratio versus horizontal stress index. ð7Þ APPLICABILITY OF PIEZOCONE AND DILATOMETER 107 which could provide, similarly to the piezocone, an OCR prediction in the range 1:7 < KD < 70. 7.3. COEFFICIENT OF EARTH PRESSURE AT REST FROM DILATOMETER The dilatometer test appears to be particularly suitable for measuring the coef¢cient of lateral pressure at rest. The original correlation (Marchetti, 1980) between at-rest coef¢cient Ko and KD is: Ko ¼ ðKD =1:5Þ0:47  0:6 ð8Þ where the reference Ko values were empirically obtained from OCRs and plasticity indices on ten Italian soils. Lacasse and Lunne (1988), on the basis of direct Ko measurements, argued that the Marchetti relationship tends to overestimate Ko . To determine Ko in clays, they suggested the following correlation: Ko ¼ 0:34KD0:44 0:64 ð9Þ where the lower exponent value is associated with highly plastic clays. To estimate Ko from DMT in granular materials Schmertmann (1983) suggested the use of a chart based on CPT experience, that relates friction angle f0 , in-situ vertical effective stress s0vo and static cone penetration resistance qt . In Figure 13 Ko values measured in CKoU/D triaxial tests are plotted, distinguishing both the different types of soils and the way they have been determined Figure 13. Coe⁄cient of earth pressure at rest versus horizontal stress index. 108 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA (Ko -oedometer cell and the self-boring pressuremeter). Correlations (8) and (9) together with the Schmertmann curves for granular materials are also sketched in Figure 13. No signi¢cant differences in Ko values are noted among the data from different tests, the three classes of soils nor according to the material index ID . The points are located below both the Marchetti, and Lacasse and Lunne curves and seem to be more in accordance with Schmertmann’s suggestions. This is probably due to the prevalently silty nature of Venetian cohesive soil, characterized by low chemical bonds between soil particles, together with a high sensitivity to the stress relief inherent in the sampling disturbance. A correlation between Ko and KD does not seem possible in this case using the data collected so far. 7.4. UNDRAINED SHEAR STRENGTH 7.4.1. Undrained Strength from Piezocone From the tip resistance in cohesive soil the undrained shear strength pro¢le cu with depth can be determined by applying the bearing capacity equation to the cone resistance. Hence, we have: cu ¼ qt  svo Nk ð10Þ where Nk is the cone bearing capacity factor. Several theoretical approaches are available for Nk (Yu and Mitchell, 1998), these being dependent on the type of analytical solution assumed to describe the undrained penetration process (the limit equilibrium method, the cavity expansion theory, the strain path method). The bearing factor Nk is theoretically characterized by a large range of variation, between 11 and 19 for normally consolidated clays, approaching 25 for overconsolidated ones. On the basis of undrained shear strength measured in the laboratory from unconsolidated undrained triaxial compression tests on undisturbed samples, values of Nk were calculated for silty clay formations. The result is shown in Figure 14, where the two lines corresponding to 11 and 25, the upper and lower limit of our range, are plotted. The scatter around the average value of 18.5, most likely due to soil heterogeneity together with sample disturbance, is relatively large; consequently therefore Equation (10) can be used only for a preliminary and approximate estimate of undrained strength. An alternative derivation of undrained strength can be obtained by dividing the excess pore pressure by the cone pore pressure factor NDu as suggested by Robertson et al. (1986): cu ¼ Du2 NDu ð11Þ APPLICABILITY OF PIEZOCONE AND DILATOMETER Figure 14. Undrained shear strength from triaxial compression tests versus cone resistance. Figure 15. NDu from triaxial compression tests versus pore pressure ratio. 109 with the cone pore pressure factor related to the pore pressure ratio Bq : NDu ¼ 18:6Bq þ 0:13 ðr2 ¼ 0:85Þ ð12Þ Note the better ¢tting obtained in this case was by using the relationships (11) and (12), with respect to the previous one (10). 110 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA 7.4.2. Undrained Strength from Dilatometer Considering the dependence of cu =s0vo on OCR, Marchetti proposed a correlation between cu and KD : cu ¼ 0:22ð0:5KD Þ1:25 s0vo ð13Þ Correlation (13) was recommended, similarly for correlation (5), only for material with ID < 1:2 and simple loading history. In our case, cu from unconsolidated undrained triaxial tests is plotted against KD in Figure 16: a relatively large scatter among all data can be observed. Nevertheless, a general increase of undrained strength as a function of KD is appreciated: the power ¢tting proposed here, similar to that of Marchetti, showed a regression coef¢cient of r2 ¼ 0:79. It may therefore be suggested that the original correlation could be utilized for a preliminary estimate of the trend of cu with depth also for all the range of KD . 7.5. FRICTION ANGLE 7.5.1. Friction Angle from Piezocone For a given granular soil, the cone resistance is dependent on the friction angle and on the current effective stress level acting in the soil. Therefore a correct test interpretation should include the above three quantities. It was also observed that the dependence of cone resistance is much higher on horizontal stress than on the vertical overburden stress (Houlsby, 1998). However due to the dif¢culty in estimating the former, it is commonly preferred to express the dependence on stress level by considering the latter. Several relationships have been proposed in the past Figure 16. Normalized undrained strength versus horizontal stress index. APPLICABILITY OF PIEZOCONE AND DILATOMETER 111 but the simplest and most widely accepted is that due to Durgunoglu and Mitchell (1975), modi¢ed subsequently by Robertson and Campanella (1983). This is based on the direct correlation between friction angle, cone resistance and overburden stress with the following analytical expression:
qt f ¼ arctan 0:10 þ 0:38 log 0 svo 0  ð14Þ A similar relationship was used in our case and the regression analysis
qt f ¼ arctan 0:38 þ 0:27 log 0 svo 0  ðr2 ¼ 0:50Þ ð15Þ is proposed in Figure 17, where the dashed area de¢nes the range of possible excursion between the upper and lower values of friction angle (about 1.5‡). 7.5.2. Friction Angle from Dilatometer Observing a linear correlation between KD and CPT penetration resistance, Marchetti adapted the relationship (14) of Durgunoglu and Mitchell, thus presenting a new chart that provides the peak triaxial friction angle f0 of sand from KD (if Ko is known or determined with correlation (8)). Recently Marchetti (from Totani et al., 1999) suggested two direct empirical correlations, which could be used to give lower Figure 17. Friction angle and cone resistance in sands and silty sands. 112 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA and upper bounds of the range of possible friction angles: f0max ¼ 31 þ KD =ð0:236 þ 0:066KD Þ ð16Þ f0min ¼ 28 þ 14:6 log KD  2:1 ðlog KD Þ2 ð17Þ A comparison between laboratory values of peak and critical friction from triaxial tests on SP-SM and ML soils and the angle predicted by Equations (16) and (17) is plotted in Figure 18. In this case, the ¢tting of the two above correlations is rather limited showing a poor dependence of friction angle on KD . 7.6. COEFFICIENT OF CONSOLIDATION A proper advantage of the piezocone is the possibility to determine in-situ pro¢les of consolidation coef¢cients, affected only by the disturbance caused by steady penetration of the cone tip. To this purpose, the piezocone dissipation data are interpreted to calculate the coef¢cient of consolidation. The estimation of the consolidation parameter requires the use of an appropriate theoretical method of analysis for the evaluation of the consolidation process induced in the soil surrounding the device tip. Several methods, differing in the theoretical and/or numerical approach adopted, have been proposed in the past to predict the initial excess pore pressure distribution generated by driving the piezocone and its subsequent dissipation (Torstensson, 1977; Baligh and Levadoux, 1980; Gupta and Davidson, 1986; Teh and Houlsby, 1991). In our case we used Figure 18. Friction angle versus horizontal stress index in sands and silty sands. APPLICABILITY OF PIEZOCONE AND DILATOMETER 113 the method proposed by Gupta and Davidson to calculate the consolidation coef¢cient. The position of the ¢lter, as used in these tests, induced prevalently horizontal £ow, providing consolidation coef¢cient ch related to prevalently horizontal drainage. Figure 19 shows the trend of ch with depth determined by the dissipation tests. Its pro¢le shows relatively large oscillations with depth. For the sake of comparison, the same ¢gure also reports the values of vertical coef¢cient of consolidation cv determined from laboratory oedometric consolidation tests using the Casagrande construction. As expected, higher £ow rates were measured in-situ with the piezocone, but no particular comment is possible with respect to the in£uence of the soil heterogeneity and/or inherent anisotropy. Figure 19. Consolidation properties from pore pressure dissipation in CPTU tests and comparison with vertical consolidation properties. 114 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA 7.7. CONSTRAINED MODULUS To estimate one-dimensional constrained modulus M, the correlation with RD is used in the form: M ¼ RM ED ð18Þ where RM is a function of KD and ID as a result of some relationships not reported here for the sake of brevity. Comparisons between M from DMT with laboratory oedometric modulus for various types of soil proposed by Lacasse and Lunne (1988), showed that the ratio MDMT =Moed could vary from 0.5 for high-plasticity clays to 2.0 for sandy or silty soils. In Figure 20 the constrained modulus estimated with relationship (18) is plotted against that determined by oedometric tests, carried out on all the types of Venetian soils, including sands. In our case, the ratio MDMT =Moed is always greater then the unity and is not in£uenced by the type of soil or by its ID value. Even if RM appears to decrease with both KD and ID , it does not seem possible to ¢nd a correlation between RM and KD and ID . 7.8. MAXIMUM SHEAR STIFFNESS 7.8.1. Maximum Stiffness from Piezocone Several correlations have been proposed in the past between the CPT resistance and Gmax or Go for a large variety of soils either granular (Baldi et al., 1989) or cohesive (Mayne and Rix, 1993) or both soils (Hegazy and Mayne, 1995). The major Figure 20. Constrained modulus from oedometric tests compared with that calculated from DMT. APPLICABILITY OF PIEZOCONE AND DILATOMETER 115 criticism of all these correlations is that Gmax is a parameter determined at very small shear strain levels whereas qt is a quantity measured at large deformations involving yielding and failure of the soil surrounding the cone. However, as pointed out by Mayne and Rix (1993), both quantities, Gmax and qt , show, for a given soil, similar dependence on the same parameters, namely the mean effective stress level and void ratio. The reconstruction of the void ratio pro¢le is a dif¢cult task. As observed in the previous sections, the advantage of using the piezocone is given by the continuous pro¢le of the pore pressure ratio Bq which, in these soils, is mainly a function of pore size distribution rather than OCR. Therefore, the Bq pro¢le may act as a simple substitute of e pro¢le, in order to take into account soil type and its structural condition. To analyse the dependence of Gmax on qt and Bq we used the data from the SCPTU, data from CHT and from laboratory measurements performed with the bender element system and resonant column equipment (Simonini and Cola, 2000). The following multiple regression function applied on data from all types of Venice lagoon soils was selected: Gmax ¼ 21:5qt0:79 ð1 þ Bq Þ4:59 ðMPaÞ ðr2 ¼ 0:63Þ ð19Þ Performing regression analysis separately on four groups of data determined from the various types of seismic tests, the highest coef¢cient of correlation was determined for CHT data as follows: Gmax ¼ 21:5qt0:82 ð1 þ Bq Þ6:85 ðMPaÞ ðr2 ¼ 0:73Þ ð20Þ The data in Figure 21 show that an association between Gmax , qt and Bq is quite possible in the Venetian soils: the relationship (19) could therefore ¢nd a moderate applicability for a preliminary estimate of Gmax pro¢le with depth in the Venetian ground in the absence of direct measurements of shear wave velocity. 7.8.2. Maximum Stiffness from Dilatometer A very small-strain modulus can also be estimated from dilatometric modulus. The ratio: RG ¼ Gmax =ED ð21Þ is, in some cases, expressed as a function of KD or relative density (Jamiolkowski et al., 1988), but these relationships cannot be applied straightforward to all soils. Hryciw (1990) expressed Gmax as a more general function of Ko , gDMT and s0vo , in which Ko is given by correlation (8) and gDMT is the total unit weight determined from DMT data by means of Marchetti and Crapps’ chart. As presented above, the errors in estimating gDMT and Ko for Venice soil are not negligible. Consequently, we preferred to propose a direct relationship between RG and dilatometric indices: the best ¢tting was found in the form of the logarithmic 116 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA Figure 21. Relationship between maximum sti¡ness and cone resistance. Figure 22. Relationship between RG and material index. relationship between RG and ID , as shown in Figure 22. Using Gmax data measured both in-situ and in the laboratory, the relationship can be written as: RG ¼ 6:71  14:2 log ID ðr2 ¼ 0:68Þ ð22Þ Similarly to the results obtained using the piezocone, if we consider a regression APPLICABILITY OF PIEZOCONE AND DILATOMETER 117 analysis of data only from CHT, we get: RG ¼ 8:11  18:1 log ID ðr2 ¼ 0:85Þ ð23Þ characterized by a signi¢cant improvement of ¢tting with respect to the regression on all data. 8. Conclusions An analysis of the applicability of two geotechnical investigation tools, the piezocone (CPTU) and the dilatometer (DMT), to characterize the soils forming the shallowest deposits of upper quaternary basin of the Venice lagoon has been presented. For this purpose, the results of a comprehensive site and laboratory investigation carried out recently in a small area have been used to evaluate the reliability of the most widely used charts or correlative equations. The aim was to characterize the soil pro¢le and to estimate the main geotechnical properties of these soils, mostly composed of slightly overconsolidated silts with ¢ne sand or low plasticity clay lying in an erratic interbedding of various types of sediments. Emphasis should be given to the dif¢culties encountered in comparing the relevant properties from laboratory tests with those indirectly estimated from in situ tests due to the in£uence of heterogeneity. More particularly, the absence of any clear horizontal layering provided no absolute certainty in carrying out the comparison of material properties within exactly the same type of soil. In addition, the effect of sample disturbance, relevant in these sandy/silty soils but dif¢cult to be quanti¢ed, have in£uenced the estimate of the mechanical properties in the laboratory. Nevertheless, the variation of the basic material characteristics over a relatively narrow range due to common mineralogical origin and depositional environment, the continuous and relevant presence of the silty fraction together with the small degree of overconsolidation in the great majority of typical samples, suggested that a straightforward comparison could be attempted. From the experimental investigation carried out so far and from the comparison between in situ and laboratory results, some interesting conclusions can be in fact drawn: 1. The piezocone represents an indispensable investigation tool to characterize the highly interbedding of Venice ground, thus allowing for a very accurate description of soil pro¢le. More particularly, the combined use of sleeve friction ratio and pore pressure parameter provides a simple and quick soil classi¢cation. Reliable classi¢cation can be also performed with the dilatometer, but with some uncertainties in de¢ning the intermediate class of Venetian soils, namely the silts. No reliable value of unit soil weight seems to be determinable with the DMT. 118 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA 2. Both CPTU and DMT can be used to reasonably reconstruct the stress history, especially for the slightly overconsolidated layers. For these, approximate values of the overconsolidation ratio can be determined to sketch tentative pro¢les of OCR with depth. It should be noted that the estimate of stress history of Venetian soils with traditional oedometer-based methods was also particularly di⁄cult, due to the signi¢cant in£uence of sampling disturbance, more important in low-plasticity silty soils, even in specimens trimmed from high-quality large diameter undisturbed samples. 3. For similar reasons, the method of determination of K0 with DMT cannot be de¢nitely validated on the basis of the comparison with pressuremeter test data and laboratory results. The latter was particularly in£uenced by the typical stress-relief due to the non-structured nature of Venetian sediments, characterized by low chemical interaction. Nevertheless, the estimate of K0 with the DMT, keeping in mind the relatively small level of overconsolidation of most of these soils, appears to be reasonable for standard in-situ and laboratory testing techniques. 4. The applicability of both CPTU and DMT to calculate the undrained shear strength is rather limited, with a preference for CPTU. With CPTU, a better prediction can be obtained using a correlation based on a cone pore pressure factor, which seems to be more sensitive than the traditional bearing factor in the interpretation of CPTU strength data. 5. With reference to the e¡ective friction angle of sands and silty sands, the well accepted Durgunoglu and Mitchell relationship was calibrated for CPTU to suit these soils: the friction angle could be therefore reasonably calculated with a scatter of 1.5‡. Improved empirical relationships for DMT recently proposed by Marchetti to determine the maximum and minimum friction angles seem unfortunately not particularly applicable to these materials. 6. CPTU dissipation tests are useful to calculate the coe⁄cient of consolidation, which is, using a piezocone with the ¢lter just behind the cone, mostly related to horizontal drainage. Comparison with laboratory data obtained from vertical consolidation tests showed higher in situ values of ch typical of silty clay, but no particular comment is possible with respect to the in£uence of the inherent anisotropy. 7. Constrained modulus from DMT was rather di¡erent from that measured in the laboratory on small oedometric samples. However, as discussed in the previous sections, the noticeable dispersion of data may probably be attributed to the sensitivity of this parameter even to small variations of grain size composition. 8. Maximum shear sti¡ness can be tentatively estimated in the absence of direct measurement of shear wave velocity from both CPTU and DMT. In the ¢rst case the better result was obtained using an association between Gmax , qt and pore pressure factor Bq, the latter representing a substitute for void ratio. Compared to CPTU, the DMT leads to a more reliable prediction of maximum sti¡ness pro¢le with depth, especially if it is compared with the CHT in-situ wave velocity measurements. APPLICABILITY OF PIEZOCONE AND DILATOMETER 119 References Andersen, A. and Kolstad, P. (1979) The NGI 54 mm samplers for undisturbed sampling of clays and representative sampling of coarser materials. Proc. of the Int. Symp. On Soil Sampling, Singapore, pp. 13^21. ASTM Subcommittee D 18.02.10 (1986) Suggested method for performing the £at dilatometer test. ASTM Geot. Testing. J., 9(2), 93^101. Baldi, G., Bellotti, R., Ghionna, V.N., Jamiolkowski, M. and Lo Presti, D.C.F. (1989) Modulus of Sands from CPT and DMT. Proc. XII Int. Conf. on Soil Mech. Found. Eng., Rio de Janeiro, Vol. 1, Balkema, Rotterdam, pp. 165^170. Baligh, M.M. and Levadoux, J.N. (1980) Pore pressure dissipation after cone penetration. Research Report R.80-11, MIT, Cambridge, Mass. Campanella, R.G. and Robertson, P.K. (1983) Flat plate dilatometer testing: research and development. Soil Mech. Report 68, University of British Columbia. Cola, S., Ricceri, G., and Simonini, P. (1998) Small strain stiffness of Venetian soils from ¢eld and laboratory tests, XI Danube-European Conf. on Soil Mech. and Geot. Eng., Porec, Croatia, Balkema, Rotterdam, pp. 679^686. Cola, S. and Simonini, P. (1999) Some remarks on the behavior of Venetian silts, Int. Sym. on Pre-failure behaviour of geomaterials, IS Torino 99, Balkema, Rotterdam, pp. 167^174. Durgunoglu, H.T. and Mitchell, J.K. (1975) Static penetration resistance of soils: I. Analysis. Proc. ASCE, Specialty Conference on In Situ Measurements of Soil Properties, Raleigh, NC, Vol.1, pp. 151-172. Eurocode 7 ^ CEN, European Committee for standardization (1999) Geotechnical design, Part 3: Design assisted by ¢eld tests, Section 9: Flat Dilatometer Test (DMT). Final edition, ENV 1999-3. Gupta, R.C. and Davidson, J.L. (1986) Piezoprobe determined coef¢cient of consolidation. Soils and Foundations, 36(3), 12^22. Hegazy, Y.A. and Mayne, P.W. (1995) Statistical correlation between Vs and cone penetration data for different soil types. Int. Symposium on Penetration Testing, CPT ‘95, Linkoeping, Vol. 2, Swedish Geotechnical Society, pp. 173^178. Holtz, R.D., Jamiolkowski, M. and Lancellotta, R. (1986) Lessons from oedometer tests on high quality samples. J. of Geot. Eng. Div., ASCE, 112(8), 768^776. Houlsby, G.T. (1998) Advanced interpretation of ¢eld tests. First Int. Conf. on Geotechnical Site Characterization, ISC ’98, Vol. 1, Balkema, Rotterdam, pp. 99^112. Hryciw, R.D. (1990) Small-strain-shear modulus of soil by dilatometer. Journal of Geotechnical Engineering, ASCE, 116(11), 1700^1716. IRTP, International Reference Test Procedure (1999) Report of the ISSMGE Technical Committee 16 on Ground Property Characterization from In-situ Testing. Proc. of the 12th Eur. Conf. on Soil Mech. and Geot. Eng., Amsterdam, the Netherlands, Balkema, Rotterdam, pp. 2195^2222. Jamiolkowski, M., Ghionna, V.N., Lancellotta, R. and Pasqualini, E. (1988) New correlations of penetration tests for design practice, Penetration Testing 88, ISOPT-1, Balkema, Rotterdam, pp. 263^296. Koutsoftas, D. and Ladd, C.C. (1985) Design strengths for an offshore clay. J. of Geot. Eng. Div., ASCE, 111(3), 337^355. Lacasse, S. and Lunne, T. (1988) Calibration of dilatometer correlations, Penetration Testing 88, ISOPT-1, Balkema, Rotterdam, pp. 539^548. Ladd, C.C. and Foot, R. (1974) New design procedure for stability of soft clays. J. of Geot. Eng. Div., ASCE, GT7, pp. 763^786. 120 GIUSEPPE RICCERI, PAOLO SIMONINI AND SIMONETTA COLA Ladd, C.C., Foot, R., Ishihara, K., Schlosser, F. and Poulos, H.G. (1977) Stress-deformation and strength characteristics. State of Art Report, Proc. 9th ICSMFE, Tokyo, Vol. 2, pp. 412^494. Marchetti, S. (1975) A new in-situ test for the measurement of horizontal soil deformability. Proc. Am. Soc. Civ. Engrs. Spec. Conf. In-situ measurements of soil properties, 2, 255^259. Marchetti, S. (1980) In situ tests by £at dilatometer, J. of Geot. Eng., ASCE, 106(GT3), 299^321. Marchetti, S. and Crapps, D.K. (1981) Flat Dilatomer Manual, Schmertmann and Crapps Inc. Gainsville. Mayne, P.W. (1991) Determination of OCR in clays by piezocone tests using cavity expansion and critical state concepts, Soils and Foundations, 31(2), 65^76. Mayne, P.W. and Rix, G.J. (1993) Gmax -qc relationships for clays. Geotechnical Testing Journal, 16(1), 54^60. Mayne, P.W. and Martin, G.K. (1998) Commentary on Marchetti Flat Dilatometer Correlation in Soils, Geotechnical Testing Journal, GTJODJ, 21(3), 222^239. Mesri, G. (1975), Discussion of ‘‘New design procedure for stability of soft clays’’ by C.C. Ladd and R. Foot, J. of Geot. Eng. Div., ASCE, 101(GT4), 409^412. Olsen, R.S. and Farr, J.V. (1986) Site characterization using the Cone Penetrometer Test, Proc. Int. Symp. IN-SITU ’86, Blacksburg, USA, pp. 854^868. Ricceri, G. (1997) Geotechnical Problems for the Barriers. Symposium on ‘‘Venice and Florence: a complex dialogue with water’’, May 24th, Florence, ITCOLD, Bologna. Robertson, P.K. (1990) Soil classi¢cation using the cone penetration test, Can. Geotechnical Journal, 27, 151^158. Robertson, P.K. and Campanella, R.G. (1983) Interpretation of cone penetration tests. Part I: Sands, Can. Geotechnical Journal, 20(4), 718^734. Robertson, P.K., Campanella, R.G., Gillespie, D. and Greig, J. (1986) Use of piezometer cone data. Proc. In-situ ’86, ASCE Spec. Publication, Blacksburg, Virginia, pp. 1263^1280. Rowe, P.W. (1973) Soil mechanic aspect of the cores of the deep borehole VE 1 in Venice. A critical analysis and recommended future investigations. Technical Report N. 57. National Research Council, Venice. Schmertmann, J.H. (1983) Revised procedure for calculating Ko and OCR from DMTs. DMT Workshop. Gainsville. Senneset, K., Janbu, N. and Svanoe, G. (1982) Strength and deformation parameters from cone penetration tests. Proc. 2nd Eur. Symp. on Penetration Testing, Amsterdam, Vol. 2, Balkema, Rotterdam, pp. 863^870. Simonini, P. and Cola, S. (2000) Use of piezocone to predict maximum stiffness of Venetian soils. Journal of Geotechnical and Geoenvironmental Engineering, 126(4), 378^382. Skempton, A.W. (1957) Discussion on the planning of the new Hong Kong Airport. Proc. Inst. of Civ. Eng., 7, 306. Tanaka, H., Tanaka, M. Iguchi, H. and Nishida, K. (1994) Shear modulus of soft clay measured by various kinds of tests. Proc. Int. Symp. on Pre-failure deformation of Geomaterials, Sapporo, Vol. 1, Balkema, Rotterdam, pp. 235^240. Teh, C.I. and Houlsby, G.T. (1991) An analytical study of the cone penetration test in clay. Geotechnique 41(1), 17^34. Torstensson, B.A. (1975) Pore pressure sounding instrument. Proc. Am. Soc. Civ. Engrs. Spec. Conf. In-situ measurements of soil properties, 2, 48^54. Torstensson, B.A. (1977) The pore pressure probe. Geotechnical Day Meeting, Oslo, Paper No. 34. APPLICABILITY OF PIEZOCONE AND DILATOMETER 121 Totani, G, Marchetti, S, Monaco, P. and Calabrese, M. (1999) Impiego della prova dilatometrica (DMT) nella progettazione geotecnica. XX National Geotechnical Congress, Parma, Italy, Vol. 1, pp. 301^308. Yu, H. S. and Mitchell, J. K. (1998) Analysis of cone resistance: review of methods. J. of Geotechnical and Geoenvironmental Engineering, 124(2), 140^149. Wissa, A.E.Z., Martin, R.T. and Garlanger, J.E. (1975) The piezometer probe. Proc. Am. Soc. Civ. Engrs. Spec. Conf. In situ measurement of soil properties, Vol. 1, pp. 536^545. Wroth, C.P. (1984) The interpretation of in situ soil tests. Geotechnique. 34(4), 449^489.