Assessment of Soil Liquefaction Potential Based on SPT Values at Some Ground Profiles in the North Central Coast of Vietnam

The North Central Coast of Vietnam has a wide distribution of loose sand which is often exposed on the surface. The thickness changes from a few meters to over ten meters. This sand with the loose state can be sensitive to the dynamic loads, such as earthquakes, traffic load, or machine foundations. It can be liquefied under these loadings, which might destroy the ground and buildings. The Standard Penetration Test (SPT) is widely used in engineering practice and its values can be useful for the assessment of soil liquefaction potential. Thus, this article presents some ground profiles in some sites in the North Central Coast of Vietnam and determines the liquefaction potential of sand based on SPT and using three parameters, including the Factor of Safety against Liquefaction (FSLIQ), Liquefaction Potential Index (LPI), and Liquefaction Severity Number (LSN). The research results show that the FSLIQ, LPI, and LSN values depend on the depth of sand samples and the SPT values. In this study, the sand distributed from 2.0 to 18.0m with (N1)60cs value of less than 20 has high liquefaction potential with FSLIQ<1, LPI is often higher than 0.73, and LSN is often higher than 10. The results also show that many soil profiles have high liquefaction potential. These results should be considered for construction activities in this area.


Introduction
Vietnam is the country having a long coastline and includes three main regions: the Northern, the Central, and the Southern regions. In the Northern and Southern coastal areas, soft clay soil is mainly distributed in the deltas with the thickness varying from a few meters to more than 30-50 meters, which usually needs to be treated before construction 1-7. The North Central Coast of Vietnam, including Thanh Hoa, Nghe An, Ha Tinh, Quang Binh, Quang Ngai, and Thua Thien Hue provinces, have complicated stratigraphy with different types of soft clay soil and loose to medium sand layers. The thickness of loose sand changes from a few meters to over ten meters. In this region, the demands for infrastructure development, such as building of roads and railway systems are on the rise. In particular, the loose sand layers are often distributed at the shallow depth and sensitive to the dynamic loads, such as earthquakes, traffic loads, and machine foundations. They can be liquefied under these loads and damage the buildings and constructions. Therefore, the liquefaction potential of sand in this region needs to be considered and evaluated. To evaluate the liquefaction potential of sand, there are two methods that include deterministic and probabilistic approaches 8. The deterministic method, or stress method, has been developed by Seed and Idriss 9 and modified several times. Seed  four field tests for evaluation of liquefaction resistance, including -CPT, SPT, shearwave velocity (Vs) measurements, and the Backer Penetration Test (BPT) for gravelly sites. These authors also showed the advantages and disadvantages of each method in the evaluation of soil liquefaction. The main advantages of CPT were the abundance of data and high-quality control. The main advantage of SPT was also the occurrence of plentiful data. Besides, for SPT, sand samples could be taken to determine the fine content and other grain characteristics and then used to evaluate the liquefaction potential. It was also shown that the fine content (grain size distribution) affects the liquefaction potential of soil. Cetin et al. 12 recommended a new method with a combination of probabilistic and deterministic approaches for assessing the likelihood of liquefaction initiation. Idriss and Boulanger 13-15 presented and updated the examination of SPT-based liquefaction triggering procedures for cohesionless soils. Boulanger and Idriss 16 proposed the re-examination of CPT-based and SPTbased liquefaction triggering procedures for cohesionless soils. From the literature review, since the SPT and CPT values were abundant and popular, the data from SPT and CPT have been widely used to evaluate the liquefaction potential 17. The liquefaction potential of soils can be evaluated by three parameters, including Liquefaction Evaluation Procedure -LEP 16, 18, Liquefaction Potential Index -LPI 19, 20 and Liquefaction Severity Number -LSN 11. LEP can be used to predict the soil liquefaction potential through the Factor of Safety against Liquefaction (FS Liq ). One of the main advantages of the FS Liq is that it can be used to classify soils, in which the soil will be liquefied if FS Liq is less than 1.0 16. By contrast, the disadvantage of FS Liq is that if it is greater than one, it does not confirm the safety against soil liquefaction 22. LPI was proposed to the thickness of liquefiable and nonliquefiable soil layers as well as the value of the factor of safety against soil liquefaction (FS Liq ). The advantage of LPI is providing a unique value for the entire soil column instead of several safety factors at different layers and using the SPT data to classify the liquefaction potential of geological units 20, 23. LSN reflected the more damaging effects of shallow liquefaction on residential lands and foundations 21. Besides, LSN considered the volumetric densification strain within soil layers as a proxy for the severity of liquefaction land likely damage at the ground surface. Dixit et al. 24 used the LPI value to predict the potential of liquefaction of soil distributed in Mumbai city and discovered that the majority of the sites in the city have a high potential of liquefaction. Previous studies indicated that the liquefaction potential of soil can be evaluated by several methods. However, there are limitations in using the three parameters of FS Liq , LPI, and LSN to evaluate the liquefaction potential of sandy soil. Moreover, in Vietnam, the Standard Penetration Test (SPT) is widely used in site investigation. The data of SPT are available and mainly used for design foundation. The use of SPT values for evaluating the soil liquefaction potential is still limited. Therefore, the main objective of this study is to evaluate the sand liquefaction potential by three parameters (FS Liq , LPI, and LSN) in the North Central Coast of Vietnam based on SPT values. The relationship between SPT values and FS Liq , LPI and LSN, and the variation of the latter three parameters with depth will be clarified.

Materials and methods
As reported from site investigation, the ground profiles in the North Central Coast of Vietnam are mostly loose sand and exposed on or near the surface 25-30. To evaluate the liquefaction potential of sand in the North Central Coast of Vietnam ( Figure-1), SPT values and the samples from the boreholes were collected. The soil samples were used to determine particle size and classify the soil. The SPT was conducted in the boreholes with an interval of 1.52m in depth. The geological cross section in 10 sites is plotted in Figure-2. In general, the soil profiles in all studied sites include two layers: the upper layer is sand (1) and the lower layer is clay soil (2). The depth of distribution and the SPT values for the sand layer are shown in Table- ) with loose to medium state. These deposits are often exposed on the surface with a thickness that ranges from a few meters to ten meters 25-30.  In these profiles, 58 boreholes were used for SPT tests. The SPTs were conducted under ASTM D1586. The samples were collected from standard penetration tests and used for particle size distribution test under ASTM D422.
According to the results of SPT, the standard resistance (N) values change from 1 to 38 blows with the average value of 9 blows. From experimental results, it can be seen that the SPT value is normally smaller than 15 blows and sand is in a loose state. The highest value of SPT is found at Site 10, and the four sites with SPT values of more than 15 blows are Sites 1,8,9,10 (Figures-2,3 and Tables-1, 4).  The variation of fine contents of soil (<0.075 mm) in different locations in the studied area is shown in Figure-4. The fine contents of soil (<0.075mm) change from 0.3% to 45.3% with the average content of 9%, showing significant changes from site to site. The highest fine content in the soil is found at Site 8. The smallest fine content in sand is at Sites 9 and 4. The soils almost belong to SP, SP-SM (Poorly graded sand, poorly graded sand with silt). In this study region, the maximum of ground surface acceleration values found in Ha Tinh, Nghe An, Quang Binh, Thanh Hoa, Quang Tri, and Thua Thien Hue are 0.1172g, 0.1102g, 0.095g, 0.062g, 0.1439g, and 0.0573g, respectively (TCVN 9386:2012). For the protection of construction, this region has the highest ground surface acceleration of 0.1439g and the value of the Importance Factor ( i ) is 1.25. Thus, the maximum horizontal ground surface acceleration value is a max = 0.180g, which is equivalent to a moment magnitude of earthquake value of M=7. In the past, in Vietnam, the highest earthquake occurred in Dien Bien province along the Ma river fault with the magnitude of 6.75. As predicted, the maximum earthquake magnitude of 7 occurs in Vietnam within a return period of 123 years 31. Thus, the earthquake magnitude of 7 will be chosen for this investigation.
In the present study, the SPT-based LEP, as reported in Boulanger & Idriss 16, will be used for assessing the factor of safety against liquefaction (FS Liq ): (1) where CSR -The earthquake-induced cyclic stress ratio; CRR -The cyclic resistance ratio, as computed by Boulanger & Idriss 16. If the factor of safety is less than 1.0, the soils will be liquefied. If the factor of safety is greater than 1.0, liquefaction will be unlikely to occur. Then, the LPI, as reported in Iwasaki et al. 19 and modified by Sonmez 20, will be calculated. The equation of LPI is presented as follows: ∫ ( ) (2) where W(z)= 10-0.5z, F 1 =1-FS Liq for FS Liq <1.0, F 1 =0 for FS Liq >1.0 and z is the depth below the ground surface (m). From the LPI, the potential liquefaction can be classified as shown in Table-2.

40-50
A major expression of liquefaction, undulation and damage to the ground surface, severe total and differential settlement of structures

>50
Severe damage, extensive evidence of liquefaction at surface, severe total and differential settlements affecting structures, damage to services

Results and discussion
From the SPT value, the normalized standard penetration resistance values (N1) 60cs was calculated according to Boulanger and Idriss 16 as follows: (N 1 ) 60cs = (N 1 ) 60 +(N 1 ) 60 (5) where (N 1 ) 60 is normalized to an overburden pressure of approximately 1 atm (approximately 101.3kPa) and a hammer energy ratio of 60 percent; FC is fine content (%). (7) (N 1 ) 60 = C N . N 60 . (8) where are the overburden correction factor, hammer energy ratio, borehole diameter correction, sampler correction, and rod length correction, respectively, and N M is field measured N values. N 60 is standardized, energy-corrected, and denoted as (9) Nu et al. where E m is hammer efficiency; N is the measures SPT value. The overburden correction factor, C N , can be denoted as 14: √( ) (11) with (N 1 ) 60cs values are limited to 46 values for the use in these expressions. The relationship between (N 1 ) 60cs , CRR, and FS Liq is shown in Figure-5 and the relationship between FS Liq and (N 1 ) 60cs is plotted in Figure-6. The variation of FS Liq is also shown in Figure-7. These figures show that the liquefaction potential depends on the depth of soils as well as (N 1 ) 60cs values. For the moment magnitude of the earthquake of 7.0 and peak ground acceleration of 0.180g, the liquefaction probability is almost smaller than 1.0; it is considered that the soil layer will be liquefied under this cyclic loading. There are 190 of 264 SPT values with FS LiQ < 1.0, with the soil having a high potential of liquefaction. Ahmad et al. 32 indicated that the liquefaction potential was affected by different soil conditions, the validity of case history data, and calculation methods.    Table-4 show the variation of LPI with depth at different sites. At sites 1-4 and 6-7, LPI ranges from 0.0 to 43.5. There is nonliquefaction at the depth of under 2.0 m. It is consistent with the FS LiQ , if the FS LiQ >1, where the liquefaction does not occur. At these sites, liquefaction potential is from low to high. At Site 5, the LPI changes from 2.1 to 54.8. It seems that the potential of liquefaction is from low to high. However, at this Site, LPI is almost higher than 15, so there is a high potential of liquefaction. At Sites 8-10, the LPI varies from 0.0 to 47.5. There are many LPI values that are equal to zero, especially in Site 10. This indicates that the potential of liquefaction is very low or nonliquefaction. It is believed that the FS LiQ >1 and the (N 1 ) 60cs in these Sites are higher than those in other Sites. The variation of LSN with depth is presented in Figure-9. It can be seen that the LSN changes from 0 to 42.4 and LSN increases with increasing depth. It is also shown that at the depth is less than 2 m, the LSN is smaller than 10, and liquefaction potential is low.

Conclusions
Based on the results obtained from this study, some conclusions can be drawn. The thickness of the sand layer in the North Central Coast of Vietnam varies from a few meters to more than ten meters and it is often exposed on the surface. The SPT values change from 1 to 38 with an average value of 9. The fine contents (<0.075 mm) change from 0.3% to 45.3%. The potential liquefaction of sand layers in this region depends on the SPT values and the distribution depth. Sandy soils have different potential of liquefaction. The highest potential of liquefaction is found at Site 5 and the lowest potential is at Site 10. The sand distributed from 2.0 to 18.0m with (N1) 60cs value of less than 20 has liquefaction potential with FS LIQ <1 , LPI that is often higher than 0.73, while LSN is often higher than 10.
The results of the present study can be used to predict the liquefaction potential of soil for building construction in the North Central Coast of Vietnam.