Application of the ‘CPT 2012’ model of AFNOR standard for column design in Poland – Jazowa case study

. This paper presents the ‘CPT 2012’ model incorporated into the AFNOR NF P94-262:2012-07, French standard for pile design fully compatible with Eurocode 7, to the wider Polish audience. The bearing capacity of three reference columns for Vistula Marshlands have been calculated according to ‘CPT 2012’ model and AFNOR recommendations. Then, the design resistances have been compared with ultimate column bearing capacity measured during static load tests conducted on reference columns. The results of comparison are discussed and the discrepancies between measured and calculated bearing capacities are shortly commented.


Introduction
Pile design using field geotechnical investigation involves using direct or indirect methods.In direct methods the soil properties are estimated using field tests and then these properties are used to calculate the pile unit base and shaft resistances.In direct methods the results of the field testing, e.g.Cone Penetration Test (CPT), are directly used to assess the pile unit base and shaft resistances.The CPT is a standard site investigation tool widely used in Polish geotechnical practice.There are many pile design methods where CPT results are used [1].One of the recent is 'CPT 2012' [2] included in AFNOR NF P94-262:2012-07 standard for pile design.The AFNOR standard is fully compatible with Eurocode 7, which is additional advantage of 'CPT 2012' model.In this paper, three representative columns constructed in Vistula Marshlands are chosen as reference ones for bearing capacity calculation according to 'CPT 2012' model and AFNOR standard.Then, the results are compared with the results of column static loading tests.The aim of this paper is to present the 'CPT-2012' model to the Polish audience and to show its application to local conditions.The discrepancies occurred are shortly discussed and the general performance of the 'CPT 2012' model within the Polish site conditions is commented.

'CPT 2012' model
The AFNOR NF P94-262:2012-07, revised due to Eurocode 7 recommendations [3], offers two pile calculation methods based on in-situ investigation, i.e., new pressuremeter model (PMT 2012) and new penetrometer model (CPT 2012), both fully compatible with Eurocode 7. The basic concept in new French standard is to distribute piles into 8 classes and 20 categories, see Table 1.The 'CPT-2012' model uses only cone resistance qc (or corrected cone resistance qt).The sleeve friction fs is omitted due to possible high variability of this reading [4].The unit pile base resistance qb is calculated using the following equation [5]: (1) where kc is a function of the soil type and pile class and qce is equivalent cone resistance.The qce averaging method in the base neighbourhood suggested by NF P94-262:2012-07 standard is modified version of Bustamante and Ginaselli [6] recommendations.The equivalent cone resistance qce is defined as [5]: (2) where a and b are the characteristic lengths, see Figure 1, qcc is a corrected cone resistance profile which is obtained by elimination of extreme values (higher than 1.3 times of average qc) and D is pile length measured from the surface level.The kc factor depends on effective embedded depth Def [5]: (3) where Def is defined as the effective embedded length in pile end bearing layer, hD is equal to 10B (B is the pile diameter).When Def/B ≥5 then kc=kcmax , where kcmax is the maximum bearing resistance factor, see Table 2.When Def/B≤5 the following equations are valid: for clays and silts, (5) for intermediate soils, (6) for sands and gravels and (7) for chalk, marl and rock.The pile unit shaft resistance qs is given by formula [5]: (8) where α is empirical coefficient depending on soil type and pile category, see Table 3 and fsol is soil-depended function [5]: where a, b, c are soil-type parameters, see Table 4.The unit shaft resistance expressed by equation ( 8) has to fulfill condition of qs≤qsmax where qsmax is maxiumum pile unit shaft resistance (see Table 5).The calculated pile bearing capacity in compression Rc (or in tension Rt) is as follows [3,5]: where qb is pile unit base resistance, Ab is the pile base area, qsi is pile shaft resistance corresponding to i layer and Asi is pile shaft area corresponding to the i layer.In AFNOR standard the design approach 2 and 'ground model' procedure is used.Consequently, the design value of pile resistance can be expressed as follows [3,5]: for compression and (13) for tension, where Rd,c is design pile bearing capacity in compression, Rd,t is design pile bearing capacity in tension, γRd is model factor, γgm is second model factor and γt is the resistance factor.The values of factors γRd are defined within the AFNOR and they are presented in table 6.The ground model safe factor γgm is equal to 1.1 and the resistance factor γt on the total characteristic resistance is equal 1.1 for compression piles and 1.15 for tension piles.
The above presented procedure is relatively straightforward in terms of pile shaft resistance calculation.However, the embedded length in pile end bearing stratum (see h distance in Fig. 1) is problematic to interpret.In AFNOR standard the h distance is defined as a embedded length in bearing soil strata.In terms of friction pile or pile that is based on bearing strata, the definition of h distance is not clear.

Reference site
The reference testing site is located in Jazowa, northern Poland.Jazowa site is a part of the Vistula Marshlands and it lies within the S7 highway, currently under construction.The 4 testing fields have been constructed in close distance to the highway.The site geotechnical investigation consists of 15 CPTu soundings, presented in Fig. 2 with corresponding soil profile.The soil layers are distinguished according to European Soil Classification System (ESCS) [7].The 69 CMC columns, 0.4m in diameter, have been constructed and 27 columns have been proof-tested.For the purpose of this paper the 3 characteristic columns (tested in compression) with embedded lengths 11m, 14.5m and 17m have been chosen to analysis.The results of static tests of selected columns are presented in Fig. 3.The total force which acts on column head was measured (QSLT) and the ultimate column capacity (Qult) was defined as force that involves column settlement of 10%B.
The CMC columns drilled with full displacement auger correspond to pile class C3 and category 7, see Table 1.The soil layers are matched with those used in AFNOR (see Tables 2,3,4).Fig. 4 presents the columns with the corresponding (the closest in the pile neighborhood) CPTu sounding result, where also the averaging length for cone resistances has been shown.
The column bearing capacity has been calculated using the procedure described in section 2. The bearing capacity factors kc for columns with embedded length of 11m, 14.5m and 17m are equal to 0.500, 0.222 and 0.500, respectively.The maximum column unit shaft resistance is equal to 130kPa for soft soils and intermediate soil, while it corresponds to 200kPa for sands, see Table 3.The installation factor α is equal to 0.95, 1.15 and 1.45 for soft soil, intermediate soils and sand layers, respectively.The fsol has been determined according to the equation (9).Then, the pile shaft capacity has been calculated and finally, the total pile resistance has been obtained as a sum of pile shaft resistance and pile base resistance.The design value of resistance is retrieved after safety factors application.

Results and Discussion
The calculated (Rc) and design (Rd) value of bearing capacity for selected columns is presented in Table 7, where also the column base and column shaft components have been shown.The comparison between calculated and design resistance and the measured column bearing capacity is summarized in Table 8.The calculated values of column bearing capacity are in good agreement with the measured ones for 11m and 14.5m length column.However, the bearing capacity of the 17m length column is overestimated.Application of the safety factors changes the situation.As one can see, the design value of column bearing capacity almost perfectly fits to the measured bearing capacity for the column 17m in length (end bearing column).The result for friction column (11m length) is also satisfactory.However, the bearing capacity of the 14.5m length column which is only based on the bearing soil layer (column toe is not embedded in the competent soil strata) is almost 2 times underestimated.The reason is due to 'CPT 2012' model, where French site conditions and field tests have been used for model calibration.Consequently, the revision of empirical factors used in 'CPT 2012' model may be needed for fitting the 'CPT 2012' model to local Polish conditions.

Conclusions
In this paper the case study of column bearing capacity calculation for Jazowa testing site has been presented where 3 representative columns are selected to analysis.The column bearing capacities have been calculated according to 'CPT 2012' model, fully compatible with Eurocode 7. The results were compared with static load tests measurements.The 'CPT 2012' model by AFNOR recommendations provides safe estimation of column capacity.However, some discrepancies are observed.The column design resistance of the end bearing column embedded in hard soil layer almost perfectly fits to the ultimate bearing capacity obtained from static load test results.In the remaining cases a significant underestimation of column design resistance is observed.In order to verify Jazowa testing site conclusions and general performance of 'CPT 2012' model, larger case study database is needed.Consequently, the calibration of the 'CPT 2012' model due to local conditions in Poland are recommended if similar results as for Jazowa site will be achieved.
The research is supported by the National Centre for Research and Development grant PBS3/B2/18/2015.

Table 8 .
Calculated and design column resistance versus measured pile bearing capacity Fig. 3. Results of static load test for Jazowa reference columns.