Instrumented Cone Penetrometer for Dense Layer Characterization

sensors

Article Instrumented Cone Penetrometer for Dense Layer Characterization

Jong-Sub Lee 1 and Yong-Hoon Byun 2,*

1 School of Civil, Environmental and Architectural Engineering, Korea University, Seoul 02841, Korea; [email protected] 2 School of Agricultural Civil & Bio-Industrial Engineering, Kyungpook National University, Daegu 41566, Korea * Correspondence: [email protected]; Tel.: +82-53-950-5732

Received: 5 September 2020; Accepted: 11 October 2020; Published: 13 October 2020

Abstract: Subsurface characterization is essential for a successful infrastructure design and construction. This paper demonstrates the use of an instrumented cone penetrometer (ICP) for a dense layer characterization at two sites. The ICP consists of a cone tip and rods equipped with an accelerometer and four strain gauges, which allow dynamic driving, in addition to quasi-static pushing of the cone. The force and velocity of the cone are measured using the ICP instrumentation and compared with the N value, dynamic cone penetration index, and static cone resistance. A strong correlation has been observed between the total cone resistance estimated from the ICP and the dynamic cone penetration index and static cone resistance. After the correction of the dynamic cone resistance eﬀect, the static component of the total cone resistance can be used as an alternative to a static cone resistance. This novel approach of soil resistance estimation using the ICP may be useful for dense layer characterization.

Keywords: cone penetrometer; dense layer; dynamic response; in situ test; subsurface characterization

1. Introduction One of the most important stages needed to achieve successful infrastructure design and construction is subsurface characterization. In situ tests are widely employed in geotechnical site characterization and soil property determination. Speciﬁcally, for cohesionless soil and highly fractured rock, in situ testing is suitable for soil resistance estimation without the requirement of any undisturbed samples utilized in laboratory tests. Accordingly, various methods for in situ testing have been established, as described in the following paragraphs. The standard penetration test (SPT) is conducted by dropping a hammer with a weight of 623 N from a height of 760 mm on top of an anvil connected to the rods. At the end of the rods, the impact of the hammer drives the split-barrel sampler into the subsoil. The N value, the so-called blow count, denotes the number of blows required for the 300 mm penetration. However, to avoid gross errors in soil resistance estimation, the N value must be corrected for various factors, such as rod length, overburden pressure, sampler type and condition, borehole diameter, and transferred energy [1,2]. Moreover, a blow count must be referred to the energy transferred to the rod head and corrected to the same energy level for which any method or correlation was previously established. The cone penetration test (CPT) is conducted using an electric cone penetrometer, cylindrical rods, rig, and data acquisition system. The electric cone penetrometer, which is mounted with transducers in and behind the cone tip, is quasi-statically pushed into the subsoil at a standard penetration rate of 20 mm/s; however, other rates may be applied, depending on the soil type so that accurate results can be obtained [3–6]. The continuous proﬁles of static cone resistance, sleeve friction, and pore water pressure

Sensors 2020, 20, 5782; doi:10.3390/s20205782 www.mdpi.com/journal/sensors Sensors 2020, 20, 5782 2 of 19 are obtained from the CPT. In recent decades, the advantages of the CPT have been acknowledged, such as its simplicity, repeatability, and accuracy [7,8]. Furthermore, reliable correlations between the CPT and other penetration tests have been observed [9,10]. Despite the usefulness of the large-sized CPT rig for penetration through the dense layers, difficulties may arise in conducting the CPT in dense layers of intermediate geomaterials, such as weathered rock, till, or soils containing considerable gravel content, due to the limited availability of large-sized CPT rigs. The dynamic cone penetrometer (DCP), which is a portable device designed for application to a shallow depth, is utilized for the evaluation of the in situ strength of granular materials ranging from fine-grained subgrades to weakly cemented layers. The standard DCP test provides the blow count per the desired depth, and in this test, a series of correlations between the DCP index and other engineering parameters of sandy soils have been observed [11]. For a more accurate characterization, an instrumented DCP was developed and used to evaluate the soil strength at a shallow depth [12–17]. However, the application of standard and instrumented DCPs is limited to fine-grained materials owing to their smaller cone diameter compared with coarse-grained materials. Accordingly, modification of the cone penetrometers is required for the characterization of deep and dense layers of intermediate geomaterials. The dynamic probing super heavy (DPSH) tests specified in EN ISO 22476-2 [18] are commonly used in geotechnical site investigations of deep and dense layers. The DPSH equipment consists of a cone with a diameter of 50.5 mm and an apex angle of 90◦, a rod with a diameter of 32 mm, and a hammer with a weight of 623 N dropped from a height of 750 mm. Similarly, the Texas cone penetrometer (TCP) has been utilized for the determination of the bearing capacities for pile design in Texas and Oklahoma [19–21]. The TCP with a diameter of 76 mm is driven by dropping a hammer with a weight of 760 N at a height of 610 mm. Similar to the SPT, the N value recorded from the TCP test exhibits a correlation with the compressive strength of intermediate geomaterials. However, the measured N values can be influenced by the energy transferred to the rod head. Moreover, the TCP test has not been widely accepted. Especially in the dense layer where the driving of cone is difficult or refused, a number of blow counts should be obtained, which is time consuming. Therefore, instrumentation of a large-sized cone penetrometer can be a promising method for a prompt and quantitative dense layer characterization. The Becker penetration test (BPT) has been employed for the characterization of gravelly soils [2,22,23]. The large-diameter Becker drill exhibiting a closed-ended tip has the advantage of applicability for large-sized particles; however, the shaft resistance mobilizes along the drill string as it is driven into the subsoil. Ghafghazi et al. [24] developed an instrumented BPT to estimate the energy transferred at the tip and then to correct the blow count. However, a large driving system and blow counts are required for the instrumented BPT to reach a certain penetration distance owing to its large diameter. This study presents the development of a new cone penetrometer incorporated with a drilling rig exhibiting high availability for dense layer characterization. It also suggests the use of a new index of soil resistance obtained from the dynamic response of the cone tip. The paper first provides a detailed description of the instrumented cone penetrometer (ICP), including the design, measurement system, mechanical force calibration, and test procedure. Then, the test results obtained at two sites are presented, and several correlations of the soil resistance index newly suggested by using the ICP are then discussed.

2. Instrumented Cone Penetrometer

2.1. Design The ICP consists of a cone tip, driving rod, and rod head. Figure1 shows a schematic drawing of the cone tip and rod head equipped with electrical sensors. The cone tip and rod head have the same outside diameter of 44.5 mm as an AW-type rod. The cone tip has an apex angle of 60◦, Sensors 2020, 20, 5782 3 of 19

which is smaller than the cone apex angle (90◦) of dynamic probing specified in EN ISO 22476-2 [18]. An accelerometer and four strain gauges were installed on the cone tip (Figure1a). In the SPT, it was observed that the magnitude of acceleration above the sampler matched and even exceeded 10,000 g [25]. The accelerometer in the ICP can measure acceleration as large as 100,000 g in the frequencySensors range2020, 20, of x FOR 1–10,000 PEER REVIEW Hz; it is rigidly mounted behind the cone tip. The acceleration3 measured of 18 at the cone tip can be applied for the calculation of velocity and displacement via integration or doublemounted integration. behind For the force cone measurement,tip. The acceleration four strainmeasured gauges at the (350 coneΩ) tip were can attached be applied at thefor conethe tip. calculation of velocity and displacement via integration or double integration. For force The strain gauges were configured as a full bridge circuit and diametrically attached opposite to the measurement, four strain gauges (350 Ω) were attached at the cone tip. The strain gauges were surface of the inner rod in order to compensate for the effects of temperature change and eccentric configured as a full bridge circuit and diametrically attached opposite to the surface of the inner rod loadingin order [26–28 to]. compensate When the for force theand effects velocity of temperat signalsure duringchange aand blow eccentric are known, loading the [26–28]. transferred When the energy can beforce calculated. and velocity The signals rod head during also a blow consists are known, of an accelerometer the transferred energy and four can strain be calculated. gauges, The which rod also allowhead for thealso evaluation consists of an of accelerometer the energy transferred and four strain from gauges, the hammer which also to allow the rods for the (Figure evaluation1b). For of the verificationthe energy of the transferred instruments, from the the Pile hammer Driving to Analyzerthe rods (PDA)(Figure sensors,1b). For twothe strainverification gauges of andthe two accelerometers,instruments, were the Pile also Driving mounted Analyzer on the (PDA) rod head. sensors, two strain gauges and two accelerometers, were also mounted on the rod head.

Strain gauges (PDA)

Accelerometers (PDA)

Strain gauges Strain gauges (ICP) (ICP)

Accelerometer Accelerometer (ICP) (ICP)

44.5 mm 44.5 mm (a) (b)

FigureFigure 1. Schematic 1. Schematic drawing drawing of of the the instrumented instrumented cone penetrometer penetrometer (ICP): (ICP): (a) (conea) cone tip; tip;(b) rod (b) head. rod head. PDA—PilePDA—Pile Driving Driving Analyzer. Analyzer.

2.2. Data2.2. Data Acquisition Acquisition System System The dataThe data acquisition acquisition system system consists consists of of a a bridgebridge box and and a afour-channel four-channel data data logger, logger, which which were were connected with the strain gauges and accelerometer, respectively. The data logger can be used to connected with the strain gauges and accelerometer, respectively. The data logger can be used to obtain obtain data with a sampling rate of 96,000 Hz per channel simultaneously. During the static data withpenetration, a sampling the sampling rate of rate 96,000 of the Hz strain per channel gauges simultaneously.used in force measurement During theis set static to 10 penetration, Hz to the samplingmeasure ratethe static of the cone strain resistance gauges with used high in forceresolution. measurement Conversely, is setfor tothe 10 dynamic Hz to measure response, the the static conesampling resistance rates with of highsignals resolution. detected by Conversely, strain gauges for and the accelerometers dynamic response, are set theto 96,000 sampling Hz. The rates of signalsdynamic detected response by strain can gaugesbe recorded and during accelerometers 310 ms, give aren setthat to the 96,000 number Hz. of The data dynamic per blow response is set to can be recorded30,000. In during the data 310 logger, ms, given the low-pass that the filter number can be of controlled data per blowup to a is frequency set to 30,000. of 20,000 In the Hz. data logger, the low-pass ﬁlter can be controlled up to a frequency of 20,000 Hz. 2.3. Calibration 2.3. CalibrationThe strain gauges installed in the cone tip and rod head of the ICP are utilized for the mechanical Theforce strain measurement. gaugesinstalled Using an inaxial the loading cone tip frame and with rod a head maximum of the compression ICP are utilized capacity for of the 1600 mechanical kN, calibrations of the strain gauges for the tip and the head were separately performed. The calibration force measurement. Using an axial loading frame with a maximum compression capacity of 1600 kN, results are presented in Figure 2. The slopes of the two force-voltage lines are the calibration factors calibrations of the strain gauges for the tip and the head were separately performed. The calibration for data acquisition.

Sensors 2020, 20, x FOR PEER REVIEW 4 of 18

Sensors 2020, 20, 5782 4 of 19 results are presented in Figure2. The slopes of the two force-voltage lines are the calibration factors for data acquisition. Sensors 2020, 20, x FOR PEER REVIEW 4 of 18

Figure 2. Mechanical force calibration of strain gauges. Input voltage is 5 V.

2.4. Test Procedure A static–dynamic cone penetration test was conducted using the ICP, as shown in Figure 3. After drilling down the borehole to the desired depth, the ICP, which was connected to a typical SPT drilling rig with a weight of approximately 40 kN, was seated at the bottom of the borehole. Prior to static penetration, the ICP was allowed to hang freely to ensure a zero load and to account for temperature effectsFigure Figure[29,30]. 2. Mechanical 2. MechanicalBy using force theforce calibrationdrilling calibration ri g, of thestra strainin ICP gauges. gauges. was Input then Input voltage pushed voltage is 5 up isV. 5 to V. 600 mm into the subsoil to minimize the effect of borehole-drilling operations. Treen et al. [31] suggested that the 2.4.penetration Test2.4. ProcedureTest distanceProcedure required for the effect of borehole drilling to become negligible varies from 5 to 10 timesA static–dynamic theA static–dynamic cone diameter. cone cone In penetration penetrationthis study, testthe test waspe wasnetration conduc conductedted distance using using the of ICP, 445 the as mm ICP, shown corresponds as in shown Figure in3. to After Figure 10 times 3. Afterthe conedrilling drilling diameter. down down the theTherefore, borehole borehole to it tothe is the desiredexpected desired depth, depth,that the the theICP, cone ICP, which whichresistance was was connected measured connected to a tointypical athis typical SPTmanner SPT drillingapproximatesdrilling rig with rig the with a cone weight a weight resistance of of approximately approximately that would 40 be kN, kN, measured was was seated seated during at the at the abottom regular bottom of theCPT. of borehole. the By borehole.using Prior the to Priorsame static penetration, the ICP was allowed to hang freely to ensure a zero load and to account for todrilling static rig penetration, with a hammer the ICP having was alloweda weight toof hang623 N freelydropped to ensurefrom a aheight zero loadof 760 and mm, to the account dynamic for temperature effects [29,30]. By using the drilling rig, the ICP was then pushed up to 600 mm into the temperaturecone penetration effects was [29 performed,30]. By using in sequence the drilling followin rig, theg the ICP completion was then pushedof the static up to penetration. 600 mm into In fact, thesubsoil available to minimize distance the foreffect dynamic of borehole-drilling penetration operations.depends on Treen the typeet al. and[31] depthsuggested of subsoils. that the The the subsoilpenetration to minimize distance therequired effect for of the borehole-drilling effect of borehole operations. drilling to become Treen negligible et al. [31] varies suggested from 5 that to the results of dynamic cone penetration tests using the ICP were obtained mostly within the first 90 mm penetration10 times distance the cone required diameter. for In this the estudy,ffect ofthe borehole penetration drilling distance to becomeof 445 mm negligible corresponds varies to 10 from times 5 to 10 penetration distance. Considering the similarity between the CPT and pile load test, the cone resistance timesthe the cone cone diameter. diameter. Therefore, In this study, it is theexpected penetration that the distance cone resistance of 445 mm measured corresponds in this to 10manner times the conecould diameter.approximates be influenced Therefore, the only cone by it resistance issoil expected no more that that wouldthan the twice be cone measured the resistance cone during diameter measured a regular below in CPT. thisthe cone mannerBy using tip [32]. approximatesthe sameIt should thebe noted conedrilling resistance that rig the with 90 that amm hammer would distance having be measured correspond a weight during ofs to623 twice aN regular dropped the cone CPT. from diameters By a height using theof of 760 samethe mm, ICP. drilling the dynamic rig with a hammerTestingcone having penetration was a conducted weight was performed of 623 at two Ndropped in sites sequence located from followin aon height theg the southwest of completion 760 mm, coast theof the dynamic(site static A) penetration. and cone the penetration southern In fact, the available distance for dynamic penetration depends on the type and depth of subsoils. The wasinland performed (site B) inof sequencethe Korean following peninsula, the completionas shown in of Figure the static 4. penetration.Two types of In in fact, situ the tests available were results of dynamic cone penetration tests using the ICP were obtained mostly within the first 90 mm distanceconducted: for SPTs dynamic and penetrationstatic–dynamic depends cone onpenetration the type andtests depth using of the subsoils. ICP at each The resultssite. For of site dynamic A, the penetration distance. Considering the similarity between the CPT and pile load test, the cone resistance coneindex penetration tests were tests conducted using the on ICP the were disturbed obtained samples mostly withinobtained the firstusing 90 the mm split-spoon penetration distance.sampler. Contrarily,could be the influenced subsoils only of site by soil B included no more than weather twiceed the rock cone anddiameter fill materials below the thatcone containtip [32]. Itconsiderable should Consideringbe noted the that similarity the 90 mm between distance thecorrespond CPT ands to pile twice load the test, cone the diameters cone resistance of the ICP. could be influenced gravel content. Adequate samples were not obtained using the sampler due to the large-sized only by soilTesting no more was thanconducted twice at the two cone sites diameter located belowon the thesouthwest cone tip coast [32 ].(site It shouldA) and bethe noted southern that the particles. The groundwater level for site A was 5 m, whereas that for site B was not detected up to 90 mminland distance (site correspondsB) of the Korean to twice peninsula, the cone as diametersshown in Figure of the ICP.4. Two types of in situ tests were the finalconducted: depth ofSPTs the and tested static–dynamic boreholes. cone penetration tests using the ICP at each site. For site A, the index tests were conducted on the disturbed samples obtained using the split-spoon sampler. Contrarily, the subsoils of site B included weathered rock and fill materials that contain considerable gravel content. Adequate samples were not obtained using the sampler due to the large-sized particles. The groundwater level for site A was 5 m, whereas that for site B was not detected up to the final depth of the tested boreholes.

FigureFigure 3.3. Procedure for aa static–dynamicstatic–dynamic conecone penetrationpenetration teststests usingusing ICPICP atat aa depth.depth.

Testing was conducted at two sites located on the southwest coast (site A) and the southern inland

(site B) of the Korean peninsula, as shown in Figure4. Two types of in situ tests were conducted: SPTs and static–dynamicFigure cone 3. Procedure penetration for a static–dynamic tests using the cone ICP penetration at each tests site. using For ICP site at A, a depth. the index tests were

Sensors 2020, 20, 5782 5 of 19 conducted on the disturbed samples obtained using the split-spoon sampler. Contrarily, the subsoils of site B included weathered rock and ﬁll materials that contain considerable gravel content. Adequate samplesSensors were 2020, not20, x obtainedFOR PEER REVIEW using the sampler due to the large-sized particles. The groundwater5 of 18 level forSensors site 2020 A was, 20, x 5 FOR m, whereasPEER REVIEW that for site B was not detected up to the ﬁnal depth of the tested boreholes.5 of 18

Figure 4. Location of sites and layout of the soundings in South Korea. BH—borehole.

3. Results Figure 4. Location of sites and layout of the soundingssoundings in South Korea. BH—borehole. 3. Results3.1. Standard Penetration Tests 3. ResultsSPTs were conducted to obtain the blow count profiles for the two sites. Figure 5 presents the N 3.1. Standardvalue profiles Penetration for depths Tests from 2.3 to 20.8 m at site A and for depths from 4 to 10 m at site B. In this 3.1. Standardstudy, the Penetration N values were Tests corrected as follows: SPTs were conducted to obtain the blow count profiles for the two sites. Figure5 presents the N SPTs were conducted to obtain the blow count profilesER for the two sites. Figure 5 presents the N value profiles for depths from 2.3 to 20.8 mN at=N∙ site A and∙𝐶 for depths from 4 to 10 m at site B.(1) In this 60% study,value profiles the N values for depths were correctedfrom 2.3 to as 20.8 follows: m at site A and for depths from 4 to 10 m at site B. In this study,where the NER values denotes were the energycorrected ratio as defined follows: as the ratio of the transferred energy at the rod head to the theoretical potential energy of the hammer, andER CR denotes the factor of rod length. The N = N ERC (1) transferred energy at the rod head was evaluatedN60 =N∙ us·60%ing the· ∙𝐶force–velocityR (FV) method. The corrected (1) 60% blow count (N60) from the surface to the depth of 10.3 m at site A remains approximately unchanged wherewhereunder ERER denotes denotesthe 10 blows, thethe energy energyincreasing ratio ratio with defined defined depth as fromas the the ratiothat ratio point of of the on.the transferred At transferred a depth energyof 20.8energy m, at theat thesoil rod rodis head very head to the to theoretical potential energy of the hammer, and C denotes the factor of rod length. The transferred the theoreticaldense, with apotential corrected energyN60 equal of to the112. hammer,ConsideringR and site C B,R thedenotes N60 starts the from factor 11 blowsof rod at alength. depth The energytransferredof 5 at m the and energy rod rapidly head at the increases was rod evaluated head with was depths using evaluated up the to force–velocity133 us blows.ing the force–velocity (FV) method. (FV) The method. corrected The blow corrected count (N ) fromThe the disturbed surface soil to the specimens depth of sampled 10.3 m during at site Athe remains SPTs at approximatelysite A were used unchanged to determine under the the blow60 count (N60) from the surface to the depth of 10.3 m at site A remains approximately unchanged index properties of the soils, which are summarized as a function of depth in Table 1. The surface 10under blows, the increasing 10 blows, withincreasing depth fromwith depth that point from on. that At point a depth on. of At 20.8 a depth m, the of soil 20.8 is verym, the dense, soil is with very a correctedlayer of N siteequal A was to made 112. up Considering of fill materials site down B, the to Na depthstarts of 5 from m. The 11 soils blows obtained at a depth from a ofdepth 5 m and dense, with 60a corrected N60 equal to 112. Considering 60site B, the N60 starts from 11 blows at a depth rapidlyof 6.3 increases to 10.3 m with were depths comprised up to of 133 silt blows. and clay (ML or CL), whereas those obtained from greater of 5 mdepths and rapidlywere classified increases as silty with sand depths (SM). up to 133 blows. The disturbed soil specimens sampled during the SPTs at site A were used to determine the index properties of the soils, which are summarized as a function of depth in Table 1. The surface layer of site A was made up of fill materials down to a depth of 5 m. The soils obtained from a depth of 6.3 to 10.3 m were comprised of silt and clay (ML or CL), whereas those obtained from greater depths were classified as silty sand (SM).

(a) (b)

FigureFigure 5. N 5. value N value proﬁles profiles measured measured fromfrom standardstandard penetration penetration test test (SPT): (SPT): (a) (sitea) siteA; (b A;) site (b) B. site B.

(a) (b)

Figure 5. N value profiles measured from standard penetration test (SPT): (a) site A; (b) site B.

Sensors 2020, 20, 5782 6 of 19

The disturbed soil specimens sampled during the SPTs at site A were used to determine the index properties of the soils, which are summarized as a function of depth in Table1. The surface layer of site A was made up of ﬁll materials down to a depth of 5 m. The soils obtained from a depth of 6.3 to 10.3 m were comprised of silt and clay (ML or CL), whereas those obtained from greater depths were classiﬁed as silty sand (SM).

Table 1. Properties of soil sampled at site A.

Liquid Plastic Sampling Specific Median Gradation Uniformity Water Plasticity US Depth Limit Limit N Depth Gravity Diameter Coefficient Coefficient Content Index CS (m) 60 LL PL (m) G D (mm) C C w (%) PI (%) * s 50 c u c (%) (%) 0–3.3 6 2.3 2.88 1.64 0.31 26.76 14 - - - SP 3.3–5.3 6 4.3 2.72 4.12 0.94 38.3 16 - - - SP 5.3–7.3 2 6.3 2.68 0.13 0.71 2.54 37 31 27 4 ML 7.3–9.3 7 8.3 2.72 0.49 0.77 3.43 26 45 23 22 CL 9.3–11.3 6 10.3 2.71 0.34 0.59 3.12 32 39 28 11 CL 11.3–13.3 13 12.3 2.72 0.33 0.7 2.61 32 34 29 5 SM 13.3–15.3 25 14.3 2.66 0.42 0.84 3.08 24 33 27 6 SM 15.3–17.1 52 16.3 2.68 0.49 0.8 3.52 28 33 27 6 SM 17.1–18.6 58 17.8 2.65 0.58 0.9 3.54 18 32 26 6 SM 18.6–20.1 55 19.3 2.59 0.5 0.86 3.09 24 32 28 4 SM 20.1–21.6 112 20.8 2.66 0.56 0.85 3.07 14 26 23 3 SM * USCS denotes unified soil classification system.

3.2. Static Cone Penetration Tests The static CPTs were conducted using the ICP at sites A and B. At site A, two boreholes, BH1 and BH2, located at a 3.6 m spacing near the borehole for SPTs were utilized for the ICP application. The profiles of static cone resistance (qs) measured at each depth in BH1 are plotted in Figure6. Figure6 shows that the qs at depths of 2 and 4 m rapidly increases over 20 MPa. It should be noted that at site A, the poorly graded sands with relatively large particles were laid down to a depth of 5 m. Below the depth of 6 m, the qs hardly increases at depths ranging from 8 to 12 m below 10 MPa; however, the qs starts to fluctuate at depths exceeding 14.5 m. Eventually, the qs significantly increases as the depth of 21 m is approached; however, it was impossible for the generated reaction force to push the cone all the way to a penetration distance of 445 mm at depths of 4, 20, and 21 m. The qs profiles at site B are plotted in Figure7a. At site B, the static penetration tests were performed only at depths of 5 to 7 m, making it possible to push the cone only less than 445 mm penetration distances. For the depths of 6 and 7 m, where the qs reached approximately 30 MPa, the cone could not penetrate into the subsoil due to the lack of reaction force. Considering the weight of the drilling rig employed in this study, the maximum reaction force corresponds to approximately 35 MPa. The static cone resistances at each depth were averaged from the values measured for the last 45 mm penetration. The boring operation performed before the static cone penetration test at each depth may lead to stress relief of the borehole and reduce the qs compared with the cone tip resistance (qc) obtained at the same depth in CPT. To minimize the effect of borehole-drilling operations on the measured qs value, the values averaged for the last 45 mm penetration were applied. Note that the 45 mm penetration corresponds to 1 diameter of the cone above the cone tip at the final position. Considering that the equivalent average cone resistance is determined above and below 1 diameter of the pile tip [33], the qs averaged for the last 45 mm penetration can be correlated with the results of dynamic cone penetration tests obtained below the final position of the cone tip after the completion of the static cone penetration test. The profile of the average qs along the entire depth is plotted in Figure8. At site A, the trends of average q s obtained from BH1 and BH2 were similar, except for the depth of 4 m. The variance in qs can be expected, as the sample at the 4 m depth includes particles larger than 19 mm. The qs at site B was excluded due to the lack of reaction force. Note that at depths deeper than 20 m at site A, the required penetration distance of 445 mm was also not achieved to estimate the average qs. Sensors 2020, 20, 5782 7 of 19 SensorsSensors 2020 2020, 20, 20, x, xFOR FOR PEER PEER REVIEW REVIEW 7 7of of 18 18

(a(a) ) ( b(b) ) (c()c )

FigureFigure 6. 6. ProfilesProﬁles Profiles of of static static cone cone resistances resistances and and cumulati cumulative cumulativeve blow blow counts counts obtained obtained from from ICP ICP in in BH BH 1 1 at at sitesite A: A: ( a()a ) 2–6 2–6 m; m; (( b(b)) ) 8–178–17 8–17 m;m; m; (( cc())c 20–21)20–21 20–21 m.m. m. SP SP and and DP DP denote denote the the staticstat staticic and and dynamic dynamic cone cone penetration penetration tests,tests, respectively. respectively. NoteNote Note thatthat that thethe the data data points points of of dynamicdy dynamicnamic penetration penetration at at depths depths of of 20 20 and and 21 21 m m indicate indicate penetrationpenetration depths depths at at every every three three blows.blows. blows.

(a(a) ) (b(b) )

FigureFigure 7. 7. ProfilesProﬁles Profiles of of static static cone cone resistances resistances and and cumulati cumulative cumulativeve blow blow counts counts obtained obtained from fromfrom ICP ICP at at site site B: B: (a()a ) 5–75–7 5–7 mm; m;;( (b b()b) 8–11)8–11 8–11 m.m. m. SP SP and and DP DP denote denote the the static static and and dy dynamic dynamicnamic cone cone penetrationpenetration tests, tests,tests, respectively. respectively.respectively. NoteNote that that the the data data points points at at the the depths depths of of 8–118–11 8–11 mm m indicateindicate indicate penetrationpenetrat penetrationion depthsdepths depths atat at everyevery every threethree three blows.blows. blows.

Sensors 2020, 20, 5782 8 of 19 Sensors 2020, 20, x FOR PEER REVIEW 8 of 18

Static cone resistance (MPa) 0 102030405060 0

8 Site A–BH1 10 Site A–BH2 Depth(m) 12

FigureFigure 8. 8. ProfilesProfiles of of average average static static cone cone resistances resistances obtained obtained from from ICP. 3.3. Dynamic Cone Penetration Tests 3.3. Dynamic Cone Penetration Tests The dynamic cone penetration tests were conducted after the completion of the static cone The dynamic cone penetration tests were conducted after the completion of the static cone penetration test at each depth. During the driving of the ICP, the blow counts and penetration depth penetration test at each depth. During the driving of the ICP, the blow counts and penetration depth were recorded. Figure6 presents the profiles of the cumulative number of blows obtained at site A. were recorded. Figure 6 presents the profiles of the cumulative number of blows obtained at site A. At the depths of 2 and 4 m, the cumulative blow count numbers corresponding to the first 100 mm At the depths of 2 and 4 m, the cumulative blow count numbers corresponding to the first 100 mm penetration (BC100) are 5 and 14, respectively. At a 6 m depth, the dynamic cone penetration test was penetration (BC100) are 5 and 14, respectively. At a 6 m depth, the dynamic cone penetration test was not applicable as the penetration distance per blow was greater than 450 mm. Note that the qs is not applicable as the penetration distance per blow was greater than 450 mm. Note that the qs is smaller than 1 MPa within a typical range of qs in clay. From the depths of 8 to 12 m, the BC100 is smaller than 1 MPa within a typical range of qs in clay. From the depths of 8 to 12 m, the BC100 is approximately 4 blows, whereas at depths of 17 and 18 m, it is approximately 17 blows. At the depths approximately 4 blows, whereas at depths of 17 and 18 m, it is approximately 17 blows. At the depths of 20 and 21 m, the BC100 was 15 blows for both. Overall, the variation in the BC100 along the depth is of 20 and 21 m, the BC100 was 15 blows for both. Overall, the variation in the BC100 along the depth is similar to that in the qs, but not proportional to the qs. Note that the blow count is influenced by both similar to that in the qs, but not proportional to the qs. Note that the blow count is influenced by both the tip resistance and sleeve friction, whereas the qs is influenced only by the tip resistance. The blow the tip resistance and sleeve friction, whereas the qs is influenced only by the tip resistance. The blow counts obtained at site B are plotted in Figure7. For the entire depths, the profiles of site B exhibit counts obtained at site B are plotted in Figure 7. For the entire depths, the profiles of site B exhibit larger blow counts compared with site A. The BC100 at a depth of 6 m is 11 blows and reaches 26 blows larger blow counts compared with site A. The BC100 at a depth of 6 m is 11 blows and reaches 26 blows at a depth of 8 m. At the depths of 9–11 m, a slope change in the profiles of the cumulative blow at a depth of 8 m. At the depths of 9–11 m, a slope change in the profiles of the cumulative blow counts occurs. The blow counts for the 100 mm penetration were 35 and 109 at the depths of 9 and counts occurs. The blow counts for the 100 mm penetration were 35 and 109 at the depths of 9 and 11 11 m, respectively. m, respectively. To obtain a blow count that would correspond to the same penetration length (300 mm) as the To obtain a blow count that would correspond to the same penetration length (300 mm) as the N60, the blow counts following the first 150 mm penetration distance are utilized. Note that the N60, the blow counts following the first 150 mm penetration distance are utilized. Note that the blow blow counts for the first 150 mm penetration are discarded in SPT. In the case that the dynamic cone counts for the first 150 mm penetration are discarded in SPT. In the case that the dynamic cone penetration tests were performed below the 450 mm penetration, the blow count can be multiplied by penetration tests were performed below the 450 mm penetration, the blow count can be multiplied a conversion factor inversely proportional to the penetration length, i.e., a conversion factor of 3 for the by a conversion factor inversely proportional to the penetration length, i.e., a conversion factor of 3 100 mm penetration distance after the first 150 mm. The blow counts are then corrected for the energy for the 100 mm penetration distance after the first 150 mm. The blow counts are then corrected for transferred to the rod head, so that the number corresponds to a 60% energy transfer ratio. Figure9 the energy transferred to the rod head, so that the number corresponds to a 60% energy transfer ratio. shows the corrected blow counts at each depth and demonstrates that the BH1 and BH2 profiles at site Figure 9 shows the corrected blow counts at each depth and demonstrates that the BH1 and BH2 A are similar. Compared with site A, the equivalent N60 obtained at site B rapidly increases with an profiles at site A are similar. Compared with site A, the equivalent N60 obtained at site B rapidly increase in depth. At the depth of 9 m, the equivalent N60 reaches 111 blows. Evidently, the equivalent increases with an increase in depth. At the depth of 9 m, the equivalent N60 reaches 111 blows. N60 at the depths of 10 and 11 m is greater than 150 blows, which are regarded as refusal and eventually Evidently, the equivalent N60 at the depths of 10 and 11 m is greater than 150 blows, which are removed from the profile. Due to the difference between an SPT split-spoon sampler and the cone regarded as refusal and eventually removed from the profile. Due to the difference between an SPT used in this study, the numbers should not be expected to be perfect equivalents; however, as Figure9 split-spoon sampler and the cone used in this study, the numbers should not be expected to be perfect demonstrates, the profiles of the equivalent N60 track the N60 obtained from the SPTs reasonably well. equivalents; however, as Figure 9 demonstrates, the profiles of the equivalent N60 track the N60 obtained from the SPTs reasonably well.

Sensors 2020, 20, 5782 9 of 19 Sensors 2020, 20, x FOR PEER REVIEW 9 of 18 Sensors 2020, 20, x FOR PEER REVIEW 9 of 18

The relationship relationship between between N N6060 andand the the equivalent equivalent N N60 60obtainedobtained from from sites sites A Aand and B is B plotted is plotted in The relationship between N60 and the equivalent N60 obtained from sites A and B is plotted in Figurein Figure 10 10after after matching matching the the measured measured depths depths fo forr the the SPT SPT and and ICP ICP test. test. The The slope slope of of the the linear Figure 10 after matching the measured depths2 for the SPT and ICP test. The slope of the linear relationshiprelationship andand the the determination determination coe fficoefficientcient (R ) (R corresponded2) corresponded to 0.84 to and 0.84 0.79, and respectively. 0.79, respectively. Similarly relationship and the determination coefficient (R2) corresponded to 0.84 and 0.79, respectively. Similarlyto the profile to the of theprofil equivalente of the Nequivalent60, the data N at60, thethe depthsdata at of the 10 depths and 11 mof at10 site and B, 11 which m at were site B, regarded which Similarly to the profile of the equivalent N60, the data at the depths of 10 and 11 m at site B, which wereas refusal, regarded were as excluded refusal, were from excluded the relationship. from the In relationship. fact, the dynamic In fact, cone the penetrationdynamic cone tests penetration using the were regarded as refusal, were excluded from the relationship. In fact, the dynamic cone penetration testsICP wereusing often the ICP conducted were often within conducted a 450 mm within penetration a 450 mm distance, penetration and the distance, subsoil within and the a 450subsoil mm tests using the ICP were often conducted within a 450 mm penetration distance, and the subsoil withinpenetration a 450 distancemm penetration might bedistance weaker might than be that weak at theer than penetration that at the depth penetration of 450 mm.depth Accordingly, of 450 mm. within a 450 mm penetration distance might be weaker than that at the penetration depth of 450 mm. Accordingly,the equivalent the N 60equivalentwas smaller N60 thanwas thesmaller N60. than the N60. Accordingly, the equivalent N60 was smaller than the N60. Equivalent N60 0 30Equivalent 60 90N60 120 150 0 0 30 60 90 120 150 0 2 Site A–BH1 2 Site A–BH1 Site A–BH2 4 Site A–BH2 4 Site B 6 Site B 6 8 8 10 10 Depth (m) (m) Depth 12 Depth (m) (m) Depth 12 14 14 16 16 18 18 20 20 22 22 Figure 9. Profiles of the equivalent N60 obtained from ICP. Figure 9. ProfilesProfiles of the equivalent N60 obtainedobtained from from ICP. ICP.

150 150 Site A Site A 120 Site B 120 Site B 60

60 90 90

Equivalent N 60 Equivalent N 30 30

0 0 0306090120150 0306090120150 SPT N60 SPT N60 Figure 10. Equivalent N60 versus SPT N60. Figure 10. Equivalent N60 versus SPT N60. Figure 10. Equivalent N60 versus SPT N60. 4. Dynamic Response Analysis 4. Dynamic Response Analysis 4. DynamicIn the dynamic Response cone Analysis penetration tests, the dynamic responses of the ICP were monitored using In the dynamic cone penetration tests, the dynamic responses of the ICP were monitored using the accelerometersIn the dynamic and cone strain penetration gauges tests, mounted the dynamic on the cone responses tip and of rod the head ICP were (Figure monitored 11). Figure using 11 the accelerometers and strain gauges mounted on the cone tip and rod head (Figure 11). Figure 11 theshows accelerometers the typical forceand strain and velocity gauges waveformsmounted on detected the cone at tip both and positions. rod head As(Figure the compressional11). Figure 11 shows the typical force and velocity waveforms detected at both positions. As the compressional waveshows generated the typical by force the applied and velocity energy waveforms travels down detected to the cone at both tip, thepositions. force and As velocitythe compressional waveforms wave generated by the applied energy travels down to the cone tip, the force and velocity waveforms werewave detectedgenerated earlier by the at applied the rod energy head thantravels at down the cone to the tip. cone The tip, lag the between force and the velocity two is, as waveforms expected, were detected earlier at the rod head than at the cone tip. The lag between the two is, as expected, wereL/c, where detected L denotes earlier at the the total rod lengthhead than of the at the rods cone at each tip. The depth, lag andbetween c denotes the two the is, wave as expected, velocity L/c, where L denotes the total length of the rods at each depth, and c denotes the wave velocity L/c,through where steel. L denotes the total length of the rods at each depth, and c denotes the wave velocity through steel. through steel.

Sensors 2020, 20, 5782 10 of 19 Sensors 2020, 20, x FOR PEER REVIEW 10 of 18

Sensors 2020, 20, x FOR PEER REVIEW 10 of 18

Figure 11. Schematic diagram of the force and velocity waveforms measured from the ICP. L—the total Figure 11. Schematic diagram of the force and velocity waveforms measured from the ICP. L—the length of the rods at each depth, c—the wave velocity through steel. total lengthFigure of the 11. rodsSchematic at each diagram depth, of c—the the force wave and velocity velocity through waveforms steel. measured from the ICP. L—the 4.1. Instrumentationtotal length Verification of the rods at each depth, c—the wave velocity through steel. 4.1. Instrumentation Verification The4.1. measurements Instrumentation obtained Verification from the instrumentation at the rod head were compared with those fromThe the measurements PDA sensors. Figureobtained 12a,b from shows the theinstrumentation typical force–velocity at the rod signals head obtained were compared from the with PDA The measurements obtained from the instrumentation at the rod head were compared with those from the PDA sensors. Figure 12a,b shows the typical force–velocity signals obtained from the and ICP.those Due from to the the di PDAfference sensors. in sampling Figure 12a,b frequency, shows the the typical first force–velocity peak values signals of the forceobtained and from velocity the PDA and ICP. Due to the difference in sampling frequency, the first peak values of the force and obtainedPDA from and the ICP. ICP Due at ato sampling the difference rate in of sampling 96 kHz are frequency, greater the than first those peak obtained values of fromthe force PDA and at a velocitysampling obtainedvelocity rate of obtained 1from kHz. the However,from ICP the at ICP a thesampling at trendsa sampling rate of the rateof force96 of kHz 96 and kHz are velocityare greater greater obtainedthan than those those from obtained obtained both the fromfrom ICP PDA PDA and atPDA a sampling areat similar. a sampling rate Furthermore,of 1rate kHz. of 1However, kHz. the However, estimated the trends the maximum trends of the of theforce transferred force and and velocity velocity energy obtained obtained at the rod fromfrom head bothboth of the the the ICP ICP ICP andis essentially PDAand are PDA similar. the sameare similar. Furthermore, for the Furthermore, ICP sensors the estimated the and estimated PDA maximum (Figure maximum 12 transferredc,d). transferred Thus, energy an energy equally at at the the good rod performance headhead of of the the ICPcan beis essentially expectedICP is essentially because the same thethe accelerometerforsame the for ICP the sensorsICP and sensors strain and andgauges PDA PDA (Figure installed (Figure 12c,d). on12c,d). the Thus, coneThus, tipanan areequally identical goodgood to performancethose installedperformance can on be the expectedcan rod be head. expected because because the acceleromete the accelerometer andr and strain strain gauges gauges installed installed on thethe conecone tip tip are identicalare identical to those to installed those installed on the on rod the head. rod head.

(a) (b)

Figure 12.FigureComparison 12. Comparison of the of dynamic the dynamic measurements: measurements: ((aa)) forceforce and and velocity velocity at atthe the ICP ICP rod head; rod head; (b) (b) forceforce and velocityand velocity(c) at the at PDAthe PDA sensors; sensors; (c) ( energyc) energy at at the the ICP ICP rod rod head; head;( (d(dd)) energy at at the the PDA PDA sensors. sensors.

Figure 12. Comparison of the dynamic measurements: (a) force and velocity at the ICP rod head; (b) force and velocity at the PDA sensors; (c) energy at the ICP rod head; (d) energy at the PDA sensors.

Sensors 2020, 20, 5782 11 of 19

4.2.Sensors Force 2020 and, 20 Velocity, x FOR PEER REVIEW 11 of 18 4.2.Based Force onand the Velocity one-dimensional wave mechanics, the force and velocity measured at the end are inﬂuencedBased by on the the boundary one-dimensional condition wave of the mechanics, end, where the theforce superposition and velocity measured of incident at and the reﬂectiveend are wavesinfluenced occurs. by Figure the boundary 13 shows condition the typical of the dynamic end, where response the superposition detected at the of coneincident tip atand three reflective depths: 8, 14.5,waves and occurs. 20 m. Figure The force13 shows and the velocity typical signals dynamic measured response at detected the cone at tipthe arecone plotted tip at three in Figure depths: 13 a. At8, a depth14.5, and of 820 m, m. the The magnitude force and velocity of the peak signals force measured is small, at whereasthe cone thetip are peak plotted velocity in Figure is large 13a. and nonzeroAt a depth velocity of 8 is m, sustained the magnitude over a relativelyof the peak long force time. is small, Conversely, whereas at the a depth peak ofvelocity 20 m, theis large peak and force is large,nonzero and velocity the nonzero is sustained force isover sustained a relatively over lo ang relatively time. Conversely, long time, at whereasa depth of the 20 amplitude m, the peak and durationforce is of large, the velocityand the nonzero signal are force reduced. is sustained The magnitudes over a relatively and durationslong time, ofwhereas the force the andamplitude velocity signalsand duration in the dynamic of the conevelocity penetration signal are test reduced. are primarily The magnitudes a function and of thedurations resistance of the experienced force and by thevelocity cone tip signals [14]. The in initialthe dynamic rising timecone detectedpenetration at the test rod are head primaril wasy set a tofunction zero, and of thatthe resistance detected at theexperienced cone tip was by L the/c. Ascone the tip rod [14]. length The initial increases rising with time increasing detected depth,at the rod the head time was corresponding set to zero, and to L /c alsothat increases. detected at the cone tip was L/c. As the rod length increases with increasing depth, the time corresponding to L/c also increases.

(a)

(b)

(c)

FigureFigure 13. 13.Typical Typical dynamic dynamic responses responses measuredmeasured at the penetration depth depth of of 100 100 mm: mm: (a ()a 8) m; 8 m; (b ()b 14.5) 14.5 m; m; (c) 20(c) m.20 m. Zero Zero on on the the x-axis x-axis corresponds corresponds toto thethe initialinitial rising rising time time of of the the force force and and velocity velocity at the at the rod rod head. head.

Sensors 2020, 20, x FOR PEER REVIEW 12 of 18

4.3.Sensors Total2020 Cone, 20, 5782 Resistance 12 of 19 The soil resistance was estimated from the dynamic penetration tests [34–37]. Schnaid et al. [34] suggested4.3. Total Cone an SPT Resistance dynamic force, considering a sampler energy and penetration distance; according to their approach, calibration of several coefficients is required to estimate the dynamic force. Kianirad [37] used Thethe soilmaximum-likelihood resistance was estimated estimation from themethod dynamic to model penetration the soil tests resistance [34–37]. Schnaidby the etdynamic al. [34] responsesuggested of an the SPT cone dynamic tip force. force, However, considering the arisi samplerng and energy falling and times penetration of an optimized distance; square according pulse to shouldtheir approach, be defined calibration by the ofoperator several in coe thefficients force–time is required history to estimateobtained the from dynamic the strain force. gauge. Kianirad In this [37] study,used the the maximum-likelihood measured dynamic estimationresponse of method the cone to tip model force the was soil used resistance as a soil by resistance the dynamic index. response of the cone tip force. However, the rising and falling times of an optimized square pulse should Total cone resistance (qt), including both the static and dynamic components, can be estimated bybe dividing defined bythe the cone operator tip force in integrated the force–time within history a certain obtained time range from by the the strain elapsed gauge. time In (∆ thist) and study, the cross-sectionalthe measured dynamic area (A) responseof the cone of tip: the cone tip force was used as a soil resistance index. Total cone resistance (qt), including both the static and dynamic components, can be estimated / by dividing the cone tip force integrated within1 a/ certainF timet dt range by the elapsed time (∆t) and the q = (2) cross-sectional area (A) of the cone tip: A ∆𝑡 where Ftip denotes the force detected at the coneR 2L tip/c using the strain gauges; ∆t is the time elapsed F (t)dt from L/c to 2L/c, which is equal to L/c. L/c and1 2L/cL/c indicatetip the times elapsed from the initial rising q = (2) time detected at the rod head (Figure 13).t TheA lower integration∆t limit (L/c) corresponds to the time at which the penetration process starts at the cone tip. The upper integration limit (2L/c) was chosen where Ftip denotes the force detected at the cone tip using the strain gauges; ∆t is the time elapsed from because,L/c to 2L /withinc, which the is time equal range to L/c. (L/c L/c to and 2L/c), 2L/c a indicate substantial the timespenetration elapsed exceeding from the initialthe conventional rising time valuesdetected of at “soil the rod qu headake” (Figurewould 13have). The been lower achieved; integration this limit allows (L/c) correspondsthe consideration to the timeof whatever at which additionalthe penetration action process occurs startsafter L/c at thewithout cone effects tip. The of upper a second-cycle integration compressional limit (2L/c) waswave, chosen which because, would withincause additional the time range penetration (L/c to [38]. 2L/c), Note a substantial that the total penetration cone resistance exceeding represents the conventional an average values stress of “soilwithin quake” a specific would time have range been obtained achieved; at the this cone allows tip. the consideration of whatever additional action The profiles of the total cone resistance estimated at each depth at site B are plotted in Figure 14. occurs after L/c without effects of a second-cycle compressional wave, which would cause additional At the depths of 5–7 m, the qt was obtained from the dynamic cone penetration tests following the penetration [38]. Note that the total cone resistance represents an average stress within a specific time static cone penetration tests, whereas the dynamic cone penetration tests were conducted only at the range obtained at the cone tip. depths of 8–11 m. For the depths of 5–8 m, the qt ranges from 4 to 40.7 MPa, and the resistance remains The profiles of the total cone resistance estimated at each depth at site B are plotted in Figure 14. almost constant along the penetration depth. Conversely, the qt at the depths of 9–11 m increases with At the depths of 5–7 m, the qt was obtained from the dynamic cone penetration tests following the increasing penetration depth. At a depth of 9 m, the value of qt reaches 42.7 MPa and then slightly static cone penetration tests, whereas the dynamic cone penetration tests were conducted only at the decreases. The qt at the depths of 10 and 11 m reaches 67.6 and 63.1 MPa, respectively. depths of 8–11 m. For the depths of 5–8 m, the qt ranges from 4 to 40.7 MPa, and the resistance remains The total cone resistance averaged for the first 90 mm penetration obtained at sites A and B is almost constant along the penetration depth. Conversely, the qt at the depths of 9–11 m increases with plotted in Figure 15. The trends of qt along the depth are almost identical between BH1 and BH2, increasing penetration depth. At a depth of 9 m, the value of qt reaches 42.7 MPa and then slightly except for the upper layer at site A, which consists of large particles. The qt at site B rapidly increases decreases. The qt at the depths of 10 and 11 m reaches 67.6 and 63.1 MPa, respectively. with the depth compared with site A.

Total cone resistance, qt (MPa) Total cone resistance, qt (MPa) 0 10203040506070 0 10203040506070 0 0

100 100

200 200 Penetration depth (mm) depth Penetration Penetration depth (mm) (mm) depth Penetration

5m 300 300 6m 9m 7m 10m 8m 11m 400 400 (a) (b)

Figure 14. ProfilesProﬁles of the total cone resistance obtained obtained at at each each depth depth at at site site B: B: ( (aa)) 5–8 5–8 m; m; ( (bb)) 9–11 9–11 m. m. Note that the data points at allall depths indicate the penetrationpenetration depths at every three blows.blows.

Sensors 2020, 20, 5782 13 of 19

The total cone resistance averaged for the ﬁrst 90 mm penetration obtained at sites A and B is plotted in Figure 15. The trends of qt along the depth are almost identical between BH1 and BH2, except for the upper layer at site A, which consists of large particles. The qt at site B rapidly increases SensorsSensors 2020 2020, ,20 20, ,x x FOR FOR PEER PEER REVIEW REVIEW 1313 ofof 1818 with the depth compared with site A.

Total cone resistance, qt (MPa) Total cone resistance, qt (MPa) 0 10203040506070 0 10203040506070 0 0

2 2

4 4

6 6

8 8

10 10 Depth (m) (m) Depth Depth (m) (m) Depth 12 12

14 14 Site A−BH1 16 Site A−BH1 16 Site A−BH2 Site A−BH2 18 Site B 18 Site B

20 20

22 22 Figure 15. Profiles of the average total cone resistance. FigureFigure 15. 15. Profiles Profiles of of the the average average total total cone cone resistance. resistance. 5. Correlations with Total Cone Resistance 5.5. CorrelationsCorrelations withwith TotalTotal ConeCone ResistanceResistance 5.1. Total Cone Resistance Versus N Value 5.1.5.1. TotalTotal ConeCone ResistanceResistance VersusVersus NN ValueValue To demonstrate the correlation between the total cone resistance and the corrected blow count, To demonstrate the correlation between the total cone resistance and the corrected blow count, the NTo60 anddemonstrate average q thet at correlation the same depths between of sites the Atotal and cone B are resistance plotted in and Figure the 16corrected. The ratio blow of (qcount,t/pa)/ the N60 and average qt at the same depths of sites A and B are plotted in Figure 16. The ratio of (qt/pa)/ N60 Nthe60 Nwhich60 and resultedaverage q fromt at the two same SPTs depths and threeof sites dynamic A and B cone are plotted penetration in Figure tests 16. is The set toratio 3.5. of Robertson(qt/pa)/ N60 which resulted from two SPTs and three dynamic cone penetration tests is set to 3.5. Robertson et al. etwhich al. [ 9resulted] demonstrated from two the SPTs variation and three of (qdynamicc/pa)/N cone along penetration the median tests diameter is set to D 3.5.50 of Robertson the subsoil et al. in [9] demonstrated the variation of (qc/pa)/N along the median diameter D50 of the subsoil in SPT and SPT[9] demonstrated and CPT. Considering the variation that of the (q Dc/p50a)/Nat site along A ranges the median from 0.13diameter to 0.58, D50 except of the forsubsoil the constructed in SPT and CPT. Considering that the D50 at site A ranges from 0.13 to 0.58, except for the constructed fill layers fillCPT. layers Considering corresponding that the to D the50 at 2 site and A 4 ranges m depths, from the 0.13 value to 0.58 of, (qexceptt/pa)/ Nfor60 theobtained constructed in this fill study layers is corresponding to the 2 and 4 m depths, the value of (qt/pa)/N60 obtained in this study is slightly slightlycorresponding underestimated to the 2 comparedand 4 m depths, with the the general value trend of (q oft/p (qa)/Nc/p60a) /obtainedN reported in bythis Robertson study is etslightly al. [9]. underestimated compared with the general trend of (qc/pa)/N reported by Robertson et al. [9]. Based Basedunderestimated on an improved compared correlation with the by general Kulhawy trend and of Mayne (qc/pa [)/N39], reported the (qt/pa by)/N Robertson60 values predicted et al. [9]. forBased the on an improved correlation by Kulhawy and Mayne [39], the (qt/pa)/N60 values predicted for the fill fillon layersan improved correspond correlation to 6.2 and by 7.9,Kulhawy and those and predicted Mayne [39], for otherthe (q depthst/pa)/N60 range values from predicted 3 to 4.7, for which the arefill layers correspond to 6.2 and 7.9, and those predicted for other depths range from 3 to 4.7, which are inlayers the rangecorrespond of the datasetto 6.2 and provided 7.9, and by those Kulhawy predicted and Maynefor other [39 depths]. range from 3 to 4.7, which are inin thethe rangerange ofof thethe datasetdataset providedprovided byby KulhawyKulhawy andand MayneMayne [39].[39]. 600 600 Site A−BH1 Site A−BH1 500 Site A−BH2 500 Site A−BH2 Site B Site B 400

a 400 a / p t / p Fill layer q t 300 Fill layer q 300 (large particles) (large particles) 200 200 (qt/pa)/N60= 3.5 (qt/pa)/N60= 3.5 2 100 R 2= 0.68 100 R = 0.68

0 0 0 306090120150 0 306090120150 N60 N60 Figure 16. Comparison of the average total cone resistance obtained at each depth using ICP with N Figure 16. Comparison of the average total cone resistance obtained at each depth using ICP with N valueFigure measured 16. Comparison from SPT. of the The average pa denotes total the cone reference resistan stressce obtained of 100kPa. at each depth using ICP with N value measured from SPT. The pa denotes the reference stress of 100 kPa. value measured from SPT. The pa denotes the reference stress of 100 kPa.

5.2.5.2. TotalTotal ConeCone ResistanceResistance VersuVersuss DynamicDynamic ConeCone PenetrationPenetration IndexIndex TheThe soilsoil resistanceresistance inin thethe dynamicdynamic conecone penetratpenetrationion testtest isis commonlycommonly representedrepresented asas aa dynamicdynamic conecone penetrationpenetration indexindex (DCPI),(DCPI), whichwhich indicatesindicates ththee penetrationpenetration distancedistance perper blow.blow. ForFor comparison,comparison,

Sensors 2020, 20, 5782 14 of 19

5.2. Total Cone Resistance Versus Dynamic Cone Penetration Index Sensors 2020, 20, x FOR PEER REVIEW 14 of 18 The soil resistance in the dynamic cone penetration test is commonly represented as a dynamic thecone qt penetrationand DCPI obtained index (DCPI), at every which single indicates blow are the pl penetrationotted in Figure distance 17. The per correlation blow. For between comparison, the qthet and qt andDCPI DCPI is expressed obtained as at follows: every single blow are plotted in Figure 17. The correlation between the qt and DCPI is expressed as follows: qq==𝑎∙DCPIa DCPIb (3) t · where a and b are constants determined by regression analysis; the values of a and b are summarized where a and b are constants determined by regression analysis; the values of a and b are summarized in Table 2. The qt increases with decreasing DCPI. Because the subsoils of site A are deposited as a in Table2. The q t increases with decreasing DCPI. Because the subsoils of site A are deposited as a multilayer, the DCPI at site A exhibits a wide range, from 2 to 49 mm/blow. Most data points at site multilayer, the DCPI at site A exhibits a wide range, from 2 to 49 mm/blow. Most data points at site B B are concentrated under the DCPI of 10 mm/blow. The DCPI data at site B shows a higher coefficient are concentrated under the DCPI of 10 mm/blow. The DCPI data at site B shows a higher coefficient of of determination compared with site A (Table 2). It was found that the qt can represent soil resistance determination compared with site A (Table2). It was found that the q t can represent soil resistance with with high resolution more than the DCPI based on the strong correlation at high soil resistance. high resolution more than the DCPI based on the strong correlation at high soil resistance. Especially Especially in the dense layer, where the driving of cone is difficult or refused, the estimation of qt in the dense layer, where the driving of cone is difficult or refused, the estimation of qt under a few under a few blows can be effectively used as an alternative to the DCPI determined from a number blows can be effectively used as an alternative to the DCPI determined from a number of blows. of blows. 70

60 Site A−BH1 Site A−BH2 50 Site B

40 (MPa) t

q 30

0 0 1020304050 DCPI (mm/blow)

Figure 17. Correlation between the total cone resistance and dynamic cone penetration index. The red Figure 17. Correlation between the total cone resistance and dynamic cone penetration index. The red line indicates the trend line for the entire data set. line indicates the trend line for the entire data set. Table 2. Parameters of the correlation between the total cone resistance and dynamic cone Tablepenetration 2. Parameters index. of the correlation between the total cone resistance and dynamic cone penetration index. a b R2 2 Site A 72.33 a 0.773b R 0.60 − Site BSite A 65.23 72.33 −0.7730.641 0.60 0.82 − EntireSite B 70.87 65.23 −0.6410.757 0.82 0.73 − Entire 70.87 −0.757 0.73 5.3. Total Cone Resistance Versus Static Cone Resistance 5.3. Total Cone Resistance Versus Static Cone Resistance The static and dynamic cone penetration tests performed at each depth in a borehole produced the staticThe static cone and totaldynamic cone cone resistances. penetration At each tests depth, performed the average at each static depth cone in a resistances borehole produced obtained thefor thestatic 45 cone mm penetrationand total cone prior resistances. to dynamic At penetration each depth, were the average compared static with cone the resistances total cone resistance obtained forobtained the 45 formm the penetration 90 mm penetration prior to dynamic after the penetrat completionion ofwere the compared static penetration. with the Thetotal relation cone resistance between obtainedqt and qs atfor site the A 90 is plottedmm penetration in Figure 18aftera and the demonstrated completion asof follows:the static penetration. The relation between qt and qs at site A is plotted in Figure 18a and demonstrated as follows: q = α q R2 = 0.88 (4) s t 2 q =α∙q· (R= 0.88) (4) where α is a constant and the value of α isis 0.94.0.94. Note that the eeffectﬀect of borehole-drilling operations on q s remainsremains at at several several depths. depths. The The results results indicate indicate th thatat the the total total cone cone resistance resistance estimated from the dynamic response is generally greater than the staticstatic cone resistance duedue toto thethe dampingdamping eeffect.ﬀect. To evaluate the damping effect on the total cone resistance, a damping coefficient (CLysm) proposed by Lysmer and Richart [40] was employed as follows: 3.4R 3.4R C = 𝜌𝐺 = 𝜌𝑉 (5) 1−𝜐 1−𝜐 ,

Sensors 2020, 20, 5782 15 of 19

To evaluate the damping eﬀect on the total cone resistance, a damping coeﬃcient (CLysm) proposed by Lysmer and Richart [40] was employed as follows:

3.4R2 p 3.4R2 C = ρGmax = ρVs max (5) Lysm 1 υ 1 υ , Sensors 2020, 20, x FOR PEER REVIEW − − 15 of 18 where R denotes the radius of the cone penetrometer; υ is Poisson’s ratio of the soil; and % is the mass wheredensity R ofdenotes the soil. the As radius proposed of the by cone Baldi penetrometer; et al. [41] and υ Mayneis Poisson’s and Rixratio [42 of], the the soil; shear and wave ρ is velocitiesthe mass density of the soil. As proposed by Baldi et al. [41] and Mayne and Rix [42], the shear wave velocities Vs were estimated from the correlations between Vs and qs to evaluate the damping coefficient CLysm. VAts were each estimated depth, the from velocities the correlations measured atbetween the cone Vstip and were qs to averaged evaluate overthe damping the duration coefficient between CLysm L/c. At each depth, the velocities measured at the cone tip were averaged over the duration between L/c and 2L/c. As presented in Table3, the dynamic cone resistance (q d) was obtained by multiplying and 2L/c. As presented in Table 3, the dynamic cone resistance (qd) was obtained by multiplying the the damping coefficient CLysm and the average velocity of the cone tip. The dynamic cone resistance dampingestimated coefficient in clay is greater CLysm and than the that average estimated velocity in sand. of the cone tip. The dynamic cone resistance estimated in clay is greater than that estimated in sand. The total cone resistance was corrected to a static component of the total cone resistance qds The total cone resistance was corrected to a static component of the total cone resistance qds (=qt − qd) (=qt q ) to exclude the effect of the dynamic cone resistance. The relation between q and qs is − d ds toplotted exclude in Figurethe effect 18b. of The the constant dynamicα coincreasesne resistance. to 1.01 The with relation the determination between qds coe andfficient qs is (R plotted2) of 0.85, in Figure 18b. The constant α increases to 1.01 with the determination coefficient (R2) of 0.85, which which indicates that the qds is almost identical to the qs. For correction, data at depths deeper than indicates17 m in BH2, that wherethe qds the is almost velocity identical at the cone to the tip wasqs. For not correction, obtained, weredata at excluded. depths deeper Considering than 17 that mthe in BH2, where the velocity at the cone tip was not obtained, were excluded. Considering that the qt qt averaged along the 90 mm penetration distance is similar to the qs within the wide range, the results averaged along the 90 mm penetration distance is similar to the qs within the wide range, the results imply that the qds can be used as an alternative to qs, and the uncorrected qt, including the damping imply that the qds can be used as an alternative to qs, and the uncorrected qt, including the damping effect, should be used with caution for the estimation of qs in clay. effect, should be used with caution for the estimation of qs in clay. 50 50

Site A−BH1 Site A−BH1 40 Site A−BH2 40 Site A−BH2

30 30 (MPa) (MPa) (MPa) s s

q 20 q 20 1:1 line 1:1 line

10 10

0 0 0 1020304050 0 1020304050

qt (MPa) qds (MPa) (a) (b)

FigureFigure 18. ComparisonComparison of of the the static static cone cone resistance resistance obtained obtained using using ICP—q ICP—qss withwith ( (aa)) the the total total cone cone resistance—qresistance—qtt andand ( (bb)) static static component component of of the the total total cone cone resistance—q resistance—qdsds. .

Table 3. Estimation of the dynamic cone resistance based on Lysmer’s analog.

Average Velocity Average Dynamic Shear Wave Velocity Damping Coefficient Depth at Cone Tip Cone Resistance Vs (m/s) CLysm (kN·s/m) (m) Vtip (m/s) qd (MPa) BH1 BH2 BH1 BH2 BH1 BH2 BH1 BH2 0–3.3 120.1 117.0 0.53 0.52 3.73 5.10 1.27 1.69 3.3–5.3 192.8 162.4 0.95 0.80 2.69 3.14 1.63 1.61 7.3–8.3 516.4 233.2 2.30 1.04 3.21 - 4.76 - 8.3–9.3 292.8 280.2 1.31 1.25 - 2.93 - 2.36 10.3–11.3 199.7 241.1 0.89 1.08 3.30 2.42 1.89 1.67 12.3–13.3 206.8 200.1 0.91 0.88 2.70 2.30 1.59 1.31 13.3–14.3 211.4 215.1 0.93 0.95 - 2.18 - 1.33 14.3–15.3 217.3 233.7 0.96 1.03 2.38 - 1.47 - 15.3–16.1 233.6 248.4 1.06 1.13 - 2.11 - 1.53 16.1–17.1 247.2 262.4 1.15 1.22 1.87 1.22 1.38 0.96 17.1–18.6 263.4 260.0 1.23 1.21 1.80 - 1.42 - 18.6–20.1 280.2 258.7 1.34 1.24 1.71 - 1.47 -

6. Summary and Conclusions

Sensors 2020, 20, 5782 16 of 19

Table 3. Estimation of the dynamic cone resistance based on Lysmer’s analog.

Average Velocity Average Dynamic Shear Wave Velocity Damping Coeﬃcient Depth at Cone Tip Cone Resistance Vs (m/s) CLysm (kN s/m) (m) · Vtip (m/s) qd (MPa) BH1 BH2 BH1 BH2 BH1 BH2 BH1 BH2 0–3.3 120.1 117.0 0.53 0.52 3.73 5.10 1.27 1.69 3.3–5.3 192.8 162.4 0.95 0.80 2.69 3.14 1.63 1.61 7.3–8.3 516.4 233.2 2.30 1.04 3.21 - 4.76 - 8.3–9.3 292.8 280.2 1.31 1.25 - 2.93 - 2.36 10.3–11.3 199.7 241.1 0.89 1.08 3.30 2.42 1.89 1.67 12.3–13.3 206.8 200.1 0.91 0.88 2.70 2.30 1.59 1.31 13.3–14.3 211.4 215.1 0.93 0.95 - 2.18 - 1.33 14.3–15.3 217.3 233.7 0.96 1.03 2.38 - 1.47 - 15.3–16.1 233.6 248.4 1.06 1.13 - 2.11 - 1.53 16.1–17.1 247.2 262.4 1.15 1.22 1.87 1.22 1.38 0.96 17.1–18.6 263.4 260.0 1.23 1.21 1.80 - 1.42 - 18.6–20.1 280.2 258.7 1.34 1.24 1.71 - 1.47 -

6. Summary and Conclusions In this study, the ICP, which can detect the dynamic responses at the cone tip, was employed for dense layer characterization and soil resistance estimation. Using the dynamic response of the cone tip of the ICP, the total cone resistance was proposed as a soil resistance index. The ICP, including the cone tip and rod head, was designed to detect the dynamic responses of the subsoil. The dynamic responses obtained from the ICP and PDA instruments were compared at the rod head to verify the performance of the accelerometer and strain gauges installed on the ICP. The ICP and PDA instruments provided not only similar trends of force and velocity but also identical values of the maximum transferred energy. For the application of the ICP, the static and dynamic cone penetration tests were conducted in three boreholes at two sites. At each depth, the dynamic cone penetration tests were conducted after the completion of the static cone penetration tests. The results revealed that the average static cone resistance and equivalent N60 in two boreholes at site A were almost identical and that the increase in the average static cone resistance and equivalent N60 along the depth at site B was faster than that at site A. In the entire depth, the profiles of the average static cone resistance and equivalent N60 obtained from the ICP were similar to that of N60 obtained from the SPT. In addition, the soil properties sampled at site A demonstrated that the subsurface ground is multilayered with different soils. The analysis of dynamic responses at the cone tip started from the force and velocity detected at the cone tip. Based on the dynamic response of the cone tip force, the total cone resistance was defined as the ratio of the cone tip force integrated within a certain time range to the elapsed time and the cross-sectional area of the cone tip. In the dynamic cone penetration tests, the total cone resistances were estimated at each blow. The total cone resistance is strongly correlated with the three soil resistance indices compared with the results of the SPTs, dynamic cone penetration indices, and static cone resistances. Specifically, the total cone resistance was used effectively for dense layer characterization, where the low DCPI and high static cone resistance occurred. Therefore, the total cone resistances would be acceptable for the static cone resistance in the dense layers where the CPT cannot be performed due to the lack of reaction force. The novel approach for soil resistance estimation using the ICP may be a useful method for a prompt and quantitative characterization of the dense layers.

Author Contributions: Conceptualization, J.-S.L.; methodology, J.-S.L. and Y.-H.B.; validation, J.-S.L.; formal analysis, Y.-H.B.; investigation, Y.-H.B.; writing—original draft preparation, Y.-H.B.; writing—review and editing, J.-S.L.; visualization, Y.-H.B.; supervision, J.-S.L.; funding acquisition, J.-S.L. and Y.-H.B. All authors have read and agreed to the published version of the manuscript. Funding: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (NRF-2020R1C1C1008925). In addition, this research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (NRF-2020R1A2B5B03001470). Sensors 2020, 20, 5782 17 of 19

Acknowledgments: The authors would like to acknowledge Prof. Rodrigo Salgado of Lyles School of Civil Engineering at Purdue University for his valuable comments. Conﬂicts of Interest: The authors declare no conﬂict of interest.

Notations

BC100 () Cumulative blow count numbers corresponding to the first 100 mm penetration c (m/s) Wave velocity through steel C (kN s/m) Damping coefficient Lysm · CR () Factor of rod length D50 (mm) Median diameter DCPI (mm) Dynamic cone penetration index ER (%) Energy ratio Ftip (kN) Force detected at the cone tip L (m) Total length of the rods at each depth N value () Blow count measured in SPT N60 () N value corrected by the effects of the rod length and transferred energy pa (kPa) Reference stress of 100 kPa qc (MPa) Cone tip resistance measured by CPT qd (MPa) Dynamic cone resistance estimated from the dynamic cone penetration tests of ICP q (MPa) Static component of the total cone resistance (= qt q ) in dynamic cone penetration tests ds − d qs (MPa) Static cone resistance measured from the static cone penetration tests of ICP qt (MPa) Total cone resistance estimated from the dynamic cone penetration tests of ICP R (m) Radius of the cone penetrometer Vs (m/s) Shear wave velocities Vtip (m/s) Average velocity measured at the cone tip of the ICP ρ (kg/m3) Mass density of the soil υ () Poisson’s ratio of the soil

References

1. Skempton, A.W. Standard penetration test procedures and the effects in sands of overburden pressure, relative density, particle size, ageing and overconsolidation. Geotechnique 1986, 36, 425–447. [CrossRef] 2. Sy, A. Energy Measurements and Correlations of the Standard Penetration Test (SPT) and the Becker Penetration Test (BPT). Ph.D. Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, BC, Canada, 1993. 3. Campanella, R.G.; Robertson, P.K.; Gillespie, D. Cone penetration testing in deltaic soils. Can. Geotech. J. 1983, 20, 23–35. [CrossRef] 4. Randolph, M.F.; Hope, S. Effect of cone velocity on cone resistance and excess pore pressures. In Proceedings of the IS Osaka—Engineering Practice and Performance of Soft Deposits, Toyonaka, Japan, 2–4 June 2004. 5. Kim, K.; Prezzi, M.; Salgado, R.; Lee, W. Effect of penetration rate on cone penetration resistance in saturated clayey soils. J. Geotech. Geoenviron. Eng. 2008, 134, 1142–1153. [CrossRef] 6. Salgado, R. The mechanics of cone penetration: Contributions from experimental and theoretical studies. Geotech. Geophys. Site Charact. 2012, 4, 131–135. 7. Schmertmann, J.H. Measurement of in situ shear strength. In Proceedings of the Specialty Conference on In Situ Measurement of Soil Properties, Raleigh, NC, USA, 1–4 June 1975; Volume 2, pp. 57–138. 8. Schaap, L.H.J.; Zuidberg, H.M. Mechanical and electrical aspects of the electric cone penetrometer tip. In Proceedings of the 2nd European Symposium on Penetration Testing, Amsterdam, The Netherlands, 24–27 May 1982. 9. Robertson, P.K.; Campanella, R.G.; Wightman, A. SPT-CPT Correlations. J. Geotech. Eng. 1983, 109, 1449–1459. [CrossRef] 10. Kodicherla, S.P.K.; Nandyala, D.K. Use of CPT and DCP based correlations in characterization of subgrade of a highway in Southern Ethiopia Region. Int. J. Geo-Eng. 2016, 7, 11. [CrossRef] Sensors 2020, 20, 5782 18 of 19

11. Mohammadi, S.D.; Nikoudel, M.R.; Rahimi, H.; Khamehchiyan, M. Application of the dynamic cone penetrometer (DCP) for determination of the engineering parameters of sandy soils. Eng. Geol. 2008, 101, 195–203. [CrossRef] 12. Ilori, A.O. Geotechnical characterization of a highway route alignment with light weight penetrometer (LRS 10), in southeastern Nigeria. Int. J. Geo-Eng. 2015, 6, 7. [CrossRef] 13. Byun, Y.H.; Lee, J.S. Instrumented dynamic cone penetrometer corrected with transferred energy into a cone tip: A laboratory study. Geotech. Test. J. 2013, 36, 533–542. [CrossRef] 14. Byun, Y.H.; Yoon, H.K.; Kim, S.Y.; Hong, S.S.; Lee, J.S. Active Layer characterization by instrumented dynamic cone penetrometer in Ny-Alesund, Svalbard. Cold. Regt. Sci. Technol. 2014, 104, 45–53. [CrossRef] 15. Lee, J.S.; Kim, S.Y.; Hong, W.T.; Byun, Y.H. Assessing subgrade strength using an instrumented dynamic cone penetrometer. Soils Found. 2019, 59, 930–941. [CrossRef] 16. Kim, S.Y.; Hong, W.T.; Lee, J.S. Role of the coefficient of uniformity on the California bearing ratio, penetration resistance, and small strain stiffness of coarse arctic soils. Cold. Regt. Sci. Technol. 2019, 160, 230–241. [CrossRef] 17. Kim, S.Y.; Lee, J.S. Energy correction of dynamic cone penetration index for reliable evaluation of shear strength in frozen sand–silt mixtures. Acta Geotech. 2020, 15, 947–961. [CrossRef] 18. ISO. E.N. 22476-2: Geotechnical Investigation and Testing–Field Testing–Part 2: Dynamic Probing; CEN: Brussels, Belgium, 2005. 19. Nam, M.S. Improved Design for Drilled Shafts in Rock. Ph.D. Thesis, University of Houston, Houston, TX, USA, 2004. 20. Smith, J.M.; Miller, G.A. Texas cone penetrometer-pressuremeter correlations for soft rock. In Proceedings of the GeoSupport 2004: Innovation an Cooperation in the Geo-Industry, Orlando, FL, USA, 29–31 January 2004; pp. 441–449. 21. Texas Department of Transportation. Geotechnical Manual (On-Line Version); Texas Department of Transportation, Bridge Division: Austin, TX, USA, 2006. 22. Harder, L.F., Jr.; Seed, H.B. Determination of Penetration Resistance for Coarse-Grained Soils Using the Becker Hammer Drill; Report No. UCB/EERC-86/06; College of Engineering, University of California: Berkeley, CA, USA, 1986. 23. Sy, A.; Campanella, R.G. Standard penetration test energy measurements using a system based upon the personal computer. Can. Geotech. J. 1994, 30, 876–882. [CrossRef] 24. Ghafghazi, M.; Thurairajah, A.; DeJong, J.T.; Wilson, D.W.; Armstrong, R. Instrumented Becker penetration test for improved characterization of gravelly deposits. In Proceedings of the Geo-Congress 2014: Geo-characterization and Modeling for Sustainability, Atlanta, GA, USA, 23–26 February 2014; pp. 37–46. 25. Lukiantchuki, J.A.; Esquivel, E.; Bernards, G.P. Interpretation of force and acceleration signals during hammer impact in SPT tests. In Proceedings of the 14th Pan-American Conference on Soil Mechanics Geotechnical Engineering, Toronto, ON, Canada, 2–6 October 2011. 26. Yoon, H.K.; Lee, J.S. Micro cones configured with full-bridge circuits. Soil Dyn. Earthq. Eng. 2012, 41, 119–127. [CrossRef] 27. Byun, Y.H.; Hong, W.T.; Lee, J.S. Characterization of railway substructure using a hybrid cone penetrometer. Smart Struct. Syst. 2015, 15, 1085–1101. [CrossRef] 28. Hong, W.T.; Byun, Y.H.; Kim, S.Y.; Lee, J.S. Cone penetrometer incorporated with dynamic cone penetration method for investigation of track substructures. Smart Struct. Syst. 2016, 18, 197–216. 29. Lunne, T.; Edismoen, T.; Gillespie, D.; Howland, J.D. Laboratory and field evaluation of cone penetrometers. In Proceedings of the ASCE Specialty Conference In Situ ’86: Use of In Situ Tests in Geotechnical Engineering, Blacksburg, VA, USA, 23–25 June 1986; pp. 714–729. 30. Lee, C.; Kim, R.; Lee, J.S.; Lee, W. Quantitative assessment of temperature effect on cone resistance. Bull. Eng. Geol. Environ. 2013, 72, 3–13. [CrossRef] 31. Treen, C.R.; Robertson, P.K.; Woeller, D.J. Cone penetration testing in stiff glacial soils using a downhole cone penetrometer. Can. Geotech. J. 1992, 29, 448–455. [CrossRef] 32. Salgado, R. The Engineering of Foundations; McGraw-Hill: New York, NY, USA, 2008. 33. Bustamante, M.; Gianeselli, L. Pile bearing capacity prediction by means of static penetrometer CPT. In Proceedings of the 2nd European Symposium on Penetration Testing, Amsterdam, The Netherlands, 24–27 May 1982; pp. 493–500. Sensors 2020, 20, 5782 19 of 19

34. Schnaid, F.; Odebrecht, E.; Rocha, M.M. On the mechanics of dynamic penetration tests. Geomech. Geoengin. Int. J. 2007, 2, 137–146. [CrossRef] 35. Schnaid, F.; Odebrecht, E.; Rocha, M.M.; Bernardes, G.P. Prediction of soil properties from the concepts of energy transfer in dynamic penetration tests. J. Geotech. Geoenviron. Eng. 2009, 135, 1092–1100. [CrossRef] 36. Schnaid, F.; Lourenco, D.; Odebrecht, E. Interpretation of static and dynamic penetration tests in coarse-grained soils. Géotech. Lett. 2017, 7, 113–118. [CrossRef] 37. Kianirad, E. Development and Testing of a Portable In-Situ Near-Surface Soil Characterization System. Ph.D. Thesis, Northeastern University, Boston, MA, USA, 2011. 38. Schmertmann, J.H.; Palacios, A. Energy dynamics of SPT. J. Geotech. Eng. Div. 1979, 105, 909–926. 39. Kulhawy, F.H.; Mayne, P.W. Manual on Estimating Soil Properties for Foundation Design (No. EPRI-EL-6800); Electric Power Research Institute: Palo Alto, CA, USA, 1990. 40. Lysmer, J.; Richart, F.E. Dynamic response of footing to vertical loading. J. Soil Mech. Found. Div. 1966, 92, 65–91. 41. Baldi, G.; Bellotti, R.; Ghionna, V.N.; Jamiolkowski, M.; Lopresti, D.C.F. Modulus of sands from CPTs and DMTs. In Proceedings of the 12th International Conference on Soil Mechanics and Foundation Engineering, Rio de Janeiro, Brazil, 13–18 August 1989; pp. 165–170. 42. Mayne, P.W.; Rix, G.J. Correlations between shear wave velocity and cone tip resistance in clays. Soils Found. 1995, 35, 165–170. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).