USE OF CONE PENETRATION TEST IN PILE DESIGN
Introduction
The prediction of pile capacity is complicated by the large variety of soil types and installation procedures. Many methods have been proposed to predict the axial capacity of single piles. These methods can be divided in three main groups:
- Full scale pile load test: This test exactly describes the piles behaviour with a load-settlement curve. At the moment this is the best method to predict the capacity of a single pile. The disadvantages of this method are: the costs of such a test are high, and it is rarely feasible in the stage of planning.
- Calculation methods based on results of laboratory tests: The low accuracyof these methods makes their economical use difficult.
- Calculation methods based on results of ‘in-situ’ tests: Among the numerous in-situ devices, the cone penetration test (CPT) and the piezocone (CPTu) represent the most versatile tools for geotechnical design. One of the earliest applications of this device was to predict a pile’s axial capacity. As a model pile it is pushed into the ground and measurements are made of the resistance to penetration of the cone. Using this test the pile capacity can be predicted time- and cost-efficiently even in the stage of planning. Nevertheless the accuracy of thismethod will not achieve that of full scale tests, but the reliability of predictions based on CPT is improving.
Test Site
Most of the tests were performed at the construction of highwayM3 in Hungary, but other tests were also performed at different sites in Hungary. The tested piles were CFA (continuous flight auger) piles. The length of the piles ranged from 6.20 m to 22.00 m, the diameter ranged from 0.60 m to 1.00 m. The pile load tests were performed in different soil conditions.
Interpretation of Pile Load Tests
The ultimate bearing capacity of the tested piles, determined by the method described in the Hungarian Code (MI-04-190), ranged from 830 kN to 3900 kN. The pile axial capacity consists of two components: end bearing load (base resistance) and side friction load (shaft resistance). At the pile load tests only total capacity of the piles is measured, so it was necessary to calculate the base resistance and the shaft resistance of the piles separately, based on the results of the performed tests. These values were determined based on detailed analysis of the test results and soil conditions.
Prediction of Pile Capacity Using CPT Data
The maximum bearing resistance of each single pile was predicted using the following methods:
• DIN 4014 (German Standard) method
• Bustamante and Giasenelli (1982) method (LCPC method)
• EUROCODE 7-3 method (process recommended by EC 7-3)
• ERTC3 method (De Cock, F. – Legrand C., 1997)
In all cases the pile capacity is derived from the known formula:
Fmax = Fmax, base + Fmax, shaft ,
where:
- Fmax is the maximum bearing resistance of the pile;
- Fmax, base is the maximum base resistance;
- Fmax, shaft is the maximum shaft resistance;
DIN 4014 Method
The German Standard provides different methods for cohesive and non-cohesive soils. In case of non-cohesive soils the unit base and shaft resistance of a bored pile can be calculated using the following tables (Table1, Table 2):
For the calculation of unit base resistance the average cone tip resistance is determined between the pile tip and depth of three times the pile diameter under the tip. In case of cohesive soils the unit base resistance and the unit skin friction can be determined with an indirect method, using the following tables (Table 3, Table 4):
Similarly to the non-cohesive case average c u is determined between the pile tip and depth of three times the pile diameter under the tip.
So the unit base and shaft resistance are calculated using the undrained shear strength (cu) of cohesive soils. There are numerous methods to predict c u based on CPT values, but the method used for this purpose has a great influence on the estimated pile capacity. So the final predicted pile capacity may be different using
different methods for undrained shear strength prediction.
Bustamante and Giasenelli (1982) Method
In this case the unit base and shaft resistances of a pile are calculated directly from the cone resistance (qc) using the following equation:
where:
EUROCODE-7-3 method
In this method the maximum unit base and shaft resistance can be derived form the following equations:
where :
qc,I, mean : is the mean of the qc,I values over the depth running from the pile base level to a level (critical depth) which is at least 0.7 times and at most 4 times the pile base diameter deeper. (Critical depth: where the calculated value of pmax, base becomes a minimum)
qc,II, mean : is the mean of the lowest qc,II values over the depth going upwards from the critical depth to the pile base
qc,III, mean : is themean value of the qc,III values over a depth interval running from the pile base level to a level of 8 times the pile base diameter higher. This procedure starts with the lowest qc,II value used for computation of qc,II, mean.
qc,z is qc at depth z
αp is the pile class factor
αs is a factor depending on the pile class and soil conditions (Table6).
ERTC3 Method
This method has the same process for calculation of unit base resistance as the EC 7-3 method, but the determination of the shaft resistance is modified. It can be calculated with the same process, but αs values are different (Table 7).
A. MAHLER
Fig. 1. DIN 4014 method
Fig. 2. LCPC method
Reliability of the Prediction Methods
In the following figures the results of the pile capacity predictions are shown. In each figure the predicted pile capacity values are plotted against the results of pile load tests (measured pile capacity). The continuous lines represent the perfect prediction ‘zone’.
Fig. 1 shows the results of the DIN 4014 method. As it was mentioned earlier it is an indirect method in case of cohesive soils, so the prediction method of undrained shear strength (cu ) has a serious effect on the calculated pile capacity. So it is obvious that the reliability of this method is varying with various c u prediction
methods.
LCPC (BUSTAMANTE and GIASENELLI, 1982) method provides a direct calculation process using solely the measured cone tip resistance (q c) values. The results of this method are shown in Fig. 2. By comparing the two figures we can say that this process gives similarly accurate pile capacity values. Fig. 3 shows the results of the EUROCODE 7-3 method. Among the investigated methods this one has the most complicated process for estimation of unit base resistance of a pile. This calculation process results in very accurate results, but the prediction of unit skin friction is less accurate in some cases. As it can be seen in Fig. 3 in these particular cases the pile capacities are extremely overestimated. These piles were of large diameter, long piles in stiff clays of high plasticity and in each case the skin friction of the piles were overestimated.
The ERTC 3 method has the same process for determination of unit base resistance, but the calculation process of unit skin friction is modified. In Fig.4 the results of this prediction method are shown. We can see that this method gives reliable results even in cases in which the pile capacities were extremely overestimated by EUROCODE 7-3 method.
Fig. 4. ERTC 3 method
Summary and Conclusion
Four pile capacity methods were evaluated using CPT data for thirteen full scale axial pile load tests. The test piles were CFA piles with various geometry and were measured in various soil conditions. DIN 4014, LCPC, EUROCODE 7-3 and ERTC 3 methods were used for pile capacity predictions.
Although the number of performed tests is not enough for statistical analysis, the results of performed tests are summarized in the following table (Table8):
In case of the EUROCODE 7-3 method the standard deviation and average estimated value are higher than the respective values of othermethods. Nevertheless these differences are caused by the overestimation of pile capacities of some piles. Without these results the EUROCODE 7-3 method gives similarly reliable results for pile capacity.
Table 8. Results of performed tests
It is necessary to note that the pile capacities were generally overestimated. To decide whether it is caused by the differing soil conditions or by differences in the pile installation process requires further studies. Nevertheless it is clear that these methods represent a useful tool in geotechnical design.
References
- BUSTAMANTE,M.–GIANESELLI, L., Pile Bearing Capacity Prediction by Means of Static Penetrometer CPT. Proceedings of the 2nd European Symposium on Penetration Testing, ESOPT-II, Amsterdam, 2 (1982), pp. 493–500.
- DE COCK,F.–LEGRAND, C. ed. Design of Axially Loaded Piles, European Practice, 1997.
- DE COCK,F.–LEGRAND,C.–LEHANE,B.ed. Survey Report on Design Methods for Axially Loaded Piles, European Practice, 1999.
- EUROCODE 7: Geotechnical Design – Part 3: Design Assisted by Fieldtesting; European Committee for Standardization, 1999.
- DIN 4014 Bohrpfähle – Herstellung, Bemessung und Tragverhalten; Deutsche Institut für Normen, 1990.
Source : PERIODICA POLYTECHNICA SER. CIV. ENG. VOL. 47, NO. 2, PP. 189–197 (2003)
Introduction
The standardized cone-penetrometer test (CPT) involves pushing a 1.41-inch diameter 55o to 60o cone (Figs. 1 thru 3) through the underlying ground at a rate of 1 to 2 cm/sec. CPT soundings can be very effective in site characterization, especially sites with discrete stratigraphic horizons or discontinuous lenses. Cone penetrometer testing, or CPT (ASTM D-3441, adopted in 1974) is a valuable method of assessing subsurface stratigraphy associated with soft materials, discontinuous lenses, organic materials (peat), potentially liquefiable materials (silt, sands and granule gravel) and landslides.
Cone rigs can usually penetrate normally consolidated soils and colluvium, but have also been employed to characterize d weathered Quaternary and Tertiary-age strata. Cemented or unweathered horizons, such as sandstone, conglomerate or massive volcanic rock can impede advancement of the probe, but the author has always been able to advance CPT cones in materials of Tertiary-age sedimentary rocks. The cone is able to delineate even the smallest (0.64 mm/1/4-inch thick) low strength horizons, easily missed in conventional (small-diameter) sampling programs. Some examples of CPT electronic logs are attached, along with hand-drawn lithologic interpretations.
Most of the commercially-available CPT rigs operate electronic friction cone and piezocone penetrometers, whose testing procedures are outlined in ASTM D-5778, adopted in 1995. These devices produce a computerized log of tip and sleeve resistance, the ratio between the two, induced pore pressure just behind the cone tip, pore pressure ratio (change in pore pressure divided by measured pressure) and lithologic interpretation of each 2 cm interval are continuously logged and printed out.
Tip Resistance
The tip resistance is measured by load cells located just behind the tapered cone (Figure 4). The tip resistance is theoretically related to undrained shear strength of a saturated cohesive material, while the sleeve friction is theoretically related to the friction of the horizon being penetrated (Robinson and Campanella, 1986, Guidelines for Use and Interpretation of the Electric Cone Penetration Test, 3rd Ed.: Hogentogler & Co., Gaithersburg, MD, 196 p.). The tapered cone head forces failure of the soil about 15 inches ahead of the tip and the resistance is measured with an embedded load cell in tons/ft2 (tsf).
Local Friction
The local friction is measured by tension load cells embedded in the sleeve for a distance of 4 inches behind the tip (Figure 4). They measure the average skin friction as the probes advanced through the soil. If cohesive soils are partially saturated, they may exert appreciable skin friction, negating the interpretive program.
Figure 1 (left) – Cone tip exposed beneath truck, just before being advanced into the ground at a rate between 1 and 2 cm/sec. The cone has an area of either 10 or 15 cm and is typically advanced in 1 m increments because the rods are of that length.
Figure 2 (right) – Hogentogler CPT rig operated by Ertec in Long Beach, CA. Cone rigs weigh about 18 tons, so are capable of exerting considerable normal force on the advancing rod. The cone tip includes a plumb meter to warn of out-of-vertical penetration.
Figure 3 – Manufacturing and operating tolerances of cones, taken from ASTM D5778.
Figure 4 – Schematic section through an electric friction-cone penetrometer tip, taken from ASTM D3441.
Friction ratio
The friction ratio is given in percent. It is the ratio of skin friction divided by the tip resistance (both in tsf). It is used to classify the soil, by its behavior, or reaction to the cone being forced through the soil. High ratios generally indicate clayey materials (high c, low Ø) while lower ratios are typical of sandy materials (or dry desiccated clays). Typical skin friction to tip friction ratios are 1 % to 10%. The ratio seldom, if ever, exceeds 15%. Sands are generally identified by exhibiting a ratio < 1%.
Piezocones also measure insitu pore pressure (in psi), in either dynamic (while advancing the cone) or static (holding the cone stationary) modes. Piezocones employ a porous plastic insert just behind the tapered head that is made of hydrophilic polypropylene, with a nominal particle size of 120 microns (Figure 5). The piezocell must be saturated with glycerin prior to its employment. The filter permeability is about 0.01 cm/sec (1 x 10-2 cm/sec).
When using the cone to penetrate dense layers, such as cemented siltstone, sandstone or conglomerate, the piezo filter element can become compressed, thereby inducing high positive pore pressures. But, the plastic filters do not exhibit this tendency, though they do become brittle with time and may need to be replaced periodically. In stiff over-consolidated clays the pore pressure gradient around the cone may be quite
high. This pore pressure gradient often results in dissipations recorded behind the CPT tip that initially increase before decreasing to the equilibrium value.
Figure 5- Schematic section through a piezocone head, showing the piezo-element and friction sleeve.Taken from ASTM D5778.
Differential Pore Pressure
The Differential Pore Pressure Ratio is used to aid in soil classification according to the Unified Soil Classification System (USCS). When the cone penetrates dense materials like sand, the sand dilates and the pore pressure drops. In clayey materials high pore pressures may be induced by the driving of the cone head. If transient pore pressures are being recorded that seem non-hydrostatic, most experienced operators will ask that the penetration be halted and allowed at least 5 minutes to equilibrate, so a quasi-static pore pressure reading can be recorded. Sometimes equilibration can take 10 to 30 minutes, depending on the soil. In practice experienced operators try to stop the advance and take pore pressure measurements in recognized aquifers and just above or adjacent to indicated aquacludes.
Temperature sensor
One great advantage of the electric cone is the temperature sensor. This has been found to be very useful in assessing the precise position of the zone, or zones, of saturation, which is of great import in slope stability and consolidation studies. A temperature shift of about 6o F is common at the groundwater interface, even perched horizons within landslides.
Corrected Logs
Most CPT rigs are equipped with one or several automated interpretation programs, which classify 1 cm horizons according to the Unified Soil Classification System. The most widely employed routine has been that originally developed by Robinson and Campanella, available from Hogentogler & Co., of Gaithersburg, MD or from the Natural Sciences and Engineering Research Council of Canada. An alternative interpretation program was developed by Dr. Richard Olsen (available through www.liquefaction.com).
The interpretation programs evaluates all of the measured properties and classifies the horizon according to its behavior (in lieu of petrology). For instance, when classifying a clayey material the interpretive programs consider undrained shear strength, tip resistance and differential pore pressure. A high differential pore pressure is assumed diagnostic of more clayey materials.
Importance of “ground-truthing”
Like geophysical techniques, CPT soundings are most meaningful when “ground-truthed” with established lithologic horizons. The easiest method to “ground truthing” CPT data is to advance a sounding next to a bucket auger or conventional boring, from which subsurface samples are collected. In this way the electronic “signature” of the sounding can be compared with the various lithologies already identified in the substory. This comparison can prove especially valuable in identifying potentially liquefiable materials and old landslide slip surfaces. Once the CPT sounding is “ground-truthed”, the rig can traverse the job site, commonly advancing 10 or 12 soundings in a single day. This allows for an expanded data set, which allows superior three-dimensional characterization of the site under evaluation, and allows construction of reliable geologic cross sections through the area.
Notes of Caution
Some notes of caution are advised when applying the CPT method to evaluating discrete low-strength horizons or partings, such as landslide slip surfaces. The 60o tip of the cone forces a passive failure of the ground in front of the advancing tip. The instrumented tip senses soil resistance about 21cm (8.4 in) ahead of the advancing tip.
This means that the tip resistance reported as “undrained shear strength” is actually an average value, taken over the zone within 21 cm of the cone tip. If the tip penetrates low strength horizons less than 21 cm thick, such as a landslide slip surface, the tip resistance reported on the CPT log may be much higher than actually exists on the discrete plane of slippage, which maybe only a fraction of an inch thick.
Another problem with the CPT method is that cone soundings advanced through desiccated clay will often be interpreted as sand or silt mixtures (by the computerized lithologic interpretation routine) because of recorded sleeve friction. The opposite problem occurs when reporting Standard Penetration Test (SPT) blow counts after advancing drive samples through clayey horizons! The SPT test is only intended for granular materials, and blow counts in such materials must be regarded with some degree of skepticism as they may shift dramatically upon later absorption of moisture.
Sample CPT logs
The attached logs are representative of the features common to electronic friction cones. They include raw data sensed by the cone as it is pushed through the ground. This data includes: Friction Ratio, Local Friction, Tip Resistance, Pore Pressure, Differential Pore Pressure Ratio and an interpreted lithologic profile (often printed out on a separate sheet, depending on which interpretation program is being utilized).
Cone penetration tests (CPT)
An
Electric Cone Penetration Test (CPT) is a geomechanical probing technique for shallow subsurface exploration. Probing through weak ground to locate firmer strata at depth has been practised since 1917, but CPT developed into its final form in the Netherlands during the 1930’s (Lunne et al. 1997). CPT combines rapid and cheap insight in the mechanical composition of the subsurface in the upper tens of meters. The widest application is currently found in geomechanical applications, i.e. surveys for road and railway constructions and the foundation of buildings and houses in areas with weak subsurface. The principles of
CPT are published in Lunne et al. (1997) and Coerts (1996).
Principles of cone penetration tests
Cone resistance, sleeve friction and friction ratio
CPT surveying involves the penetration of a metal electrical cone with a surface of 10 cm2 into the subsurface (Fig. 4.9.1). From beneath a heavy truck, the cone is penetrated at a constant rate of 1 cm/s. During penetration, a number of variables are recorded at the cone head or along the sleeve. At the cone head the cone resistance (qc) is recorded (in MPa), which expresses the resistance of the sediments to penetration. Along the cone the sleeve friction (fs) is recorded (also in MPa); indicative for the adhesive strength of the material.
Terminology for cone penetrometers
(from Lunne et al. 1997).
From the cone resistance and the sleeve friction the friction ratio (Rf) can be calculated according:
Rf = [(fs /qc)*100]
Numerous analyses of data have lead to an empirical relationship between Rf and inferred lithology (Table Below). The friction ratio is, in combination with cone resistance, broadly used in geomechanical applications.
Empirical relation between the dimensionless friction ratio and inferred lithology in CPT.
Pore water pressure
Another useful parameter that can be recorded in CPTU surveying (the so-called piezocone test) is
the pore pressure u (Fig. 4.9.1). In the saturated or vadose zone increasing values occur with increasing depth, expressed in MPa. Also perched ground water tables can be detected using this technique.
Data interpretation
CPT interpretation mainly involves pattern analysis of the cone resistance and friction ratio
curves. In common practice it is possible to define CPT „facies“ for certain sedimentary deposits. In buried valley environments for example, the friction ratio curve characteristics of “pot clay“ (or Lauenburger Ton) are well known. Similar typical CPT facies units can be defined for cover sands, boulder clay (till), several fluvial deposits and so forth. Figure 4.9.2 demonstrates an example of a CPT plot in which typical “pot clay“
patterns can be recognized between 22 and 29 m depth. Less distinct are the clayey deposits between 0 and 22 m depth.
Fig 2
Fig.2 - CPT log showing sedimentary units associated to buried valley infill, including “pot clay“ deposits (PENI). PE=Peelo Formation; PENI = Nieuwolda Member of the Peelo Formation, BX = Boxtel Formation (Weichselian deposits).
Ideally, site-specific CPT „facies“ have to be verified with borehole data or observations from exposures to ascertain relationships with actual lithofacies units.
Application of CPT in the study of buried valleys
CPT is a useful, fast and cost-effective technique that can be used for the following applications related to the characterisation of buried valleys;
1. the establishment of the occurrence, extenand upper boundary of “pot clay“ bodies
2. the establishment of protecting impermeable beds above buried valley aquifer systems (such as “pot clay“, boulder clay, etc.)
3. characterisation of the upper sedimentary records outside buried valleys.
Figure 5.6.8 presents a
CPT transect, combined with lithological columns of boreholes, of the
Groningen Burval project area. Particularly between 2 and 4 km along the profile, a large clayey body is identified. The presence and rough outline of this unit was also demonstrated by Helicopter Electromagnetics . Below this unit other sediments associated with the Peelo Formation occur, while near the surface Weichselian deposits are present (Boxtel Formation). Figure 4.9.3 shows an enlarged part of Figure 5.6.8 to demonstrate the correlation more clearly.
The CPT characteristics of the “pot clay“ (PENI) are clearly defined. The deeper undifferentiated deposits of the Peelo Formations are characterised by a strong lateral and vertical heterogeneity.
Fig. 4.9.3: Enlarged part of Figure 5.6.8. CPT transect with lithological columns of boreholes across unit of low-resistivity in HEM data. PE=Peelo Formation; PENI=Nieuwolda Member of the Peelo Formation, consisting of “pot clay“. BX =Boxtel Formation (Weichselian deposits).
Some remarks on the application of CPT
CPT can be broadly used in unconsolidated sediments; however there are certain limitations that have to be kept in mind:
- The empirical relationship presented in Table 4.9.1 is based on observations below the ground water table. Above the ground water table a clayey bed (for example) can be partly dried out, leading to higher cone resistance and lower sleeve friction, hence lowering the friction ratio number. However in a climate with excess rainfall like that in northwestern Europe, hanging water is likely in the unsaturated zone. Hence it should be possible to discriminate clayey beds from their sand or gravely counterparts in the unsaturated zone.
- In buried valley environments glaciotectonized sediments can occur. The same is true for overconsolidated sediments due to glacial loading. In both situations geomechanical properties of the sedimentary record is potentially modified: sediments could respond different to CPT than expected in in situ sedimentary sequences following the stratigraphical rule (Bakker 2004). In situations with overconsolidated clay adhesive strength is relatively high, but possibly reduced due to mechanical expel of pore water. Cone resistance is enhanced (higher compaction), leading to a reduced Rf. Hence, Rf values can differ from normal sedimentaryconditions, with normal setting and compaction due to overburden.
The examples demonstrated above, attest that CPT is a powerful technique for the identification and mapping of large sedimentary units to a maximum depth of about 60 m below the surface. Combined with the low-costs CPT is a technique that is highly recommendable in environments with unconsolidated sediments such as buried valley systems.
References :
- Bakker MAJ (2004): The internal structure of Pleistocene push moraines. A multidisciplinary approach with emphasis on ground- penetrating radar. – PhD thesis, Queen Mary, University of London, 177 pp.
- Coerts A (1996): Analysis of static cone penetration test data for subsurface modelling. A methodology. – PhD thesis, Netherlands Geographical Studies 210, 263 pp.
- Lunne T, Robertson PK, Powell JJM (1997): Cone Penetration Testing in Geotechnical Practice. – Blackie Academic & Professional, London, 312 pp.
Source : Burval
Description
Cone Penetrometer Technology (CPT) is a method of providing real-time data for use in characterizing the subsurface, as opposed to older methods of analyzing subsurface conditions in the laboratory. It consists of a steel cone that is hydraulically pushed into the ground at up to 40,000 pounds of pressure. Sensors on the tip of the cone collect data. Standard cone penetrometers collect information to classify soil type by using sensors that measure cone-tip pressure and friction. CPT is often used in conjunction with Hydropunch tests, which use the CPT holes to extract groundwater for laboratory analysis. An innovation of the CPT (i.e., the wireline CPT) allows multiple CPT tools to be interchanged during a single penetration, without withdrawing the CPT rod string from the ground.
Initially developed to collect information about soil characteristics, as sensor technology was developed CPT also became a platform for collecting information about a variety of contaminants. Recent advances in sensor technology have expanded cone penetrometer capabilities to detect the presence of petroleum hydrocarbons. Sensors are being tested or demonstrated for the detection of other organics, compounds, metals, radioactivity, explosives, and soil moisture.
Cone penetration testing (CPT) is the most versatile device for in situ soil testing. Without disturbing the ground, it provides information about soil type, geotechnical parameters like shear strength, density, elastic modulus, rates of consolidation and environmental properties. Further, as it can be seen as a small scale test pile, it is the best and most cost-effective device to design piled foundations and sheet piles.
Limitations and Concerns
CPT cannot be used at some sites due to high soil density. Most sensors are now used as screening tools that provide initial site characterization data. The data is confirmed by collecting samples that are analyzed in the laboratory. This is due to limitations in sensor technology, and it will likely diminish in importance as the technology improves.
CPT is useful on sites that contain unconsolidated sediments (e.g., soil and clay that are not cemented together). On the other hand, sites with large boulders, rock or cemented layers are difficult to penetrate.
CPT sensors, such as lasers, that require a lens may be hampered by fouling of the lens due to a reaction to dust. Decontamination may be necessary if the CPT comes into contact with contaminated material.
Cone Penetration Tests are conducted to obtain the cone resistance, the side friction and, if there is a piezocone, the pore pressure. The soil type can be determined by analysing these results, the values can also be used in the design of shallow foundations through the estimation of stiffness and shear strength of cohesive soils.
A 60o cone with face area 10cm2 and 150cm2 'friction sleeve' is hydraulically pushed into the ground at a constant speed (ranging form 1.5 to 2.5 cm/s). The force required to maintain this penetration rate, and the shear force acting on the friction sleeve are recorded. The friction ratio (cone resistance / side friction) gives an indication of the soil type.
Cone Resistance qc = Fc / Ac
Side Friction fs = Fs / As
Friction Ratio Rf = fs / qc
Where Fc = pushing force, Ac = cone plan area, Fs = shear force on friction sleeve, As = area of friction sleeve.
source : civcal.media.hku.hk, www.conepenetration.com, www.cpeo.org
