Study regarding typical liquefaction damage during the 2021 Maduo Ms7.4 earthquake in China

Yuan Jinyuan, Wang Yunlong, Ma Jiajun, Zhan Beilei, Yuan Xiaoming, Wang Lanmin and Wu Xiaoyang

1. School of Civil Engineering, Heilongjiang University of Science and Technology, Harbin 150027, China

2. Key Laboratory of Earthquake Engineering and Engineering Vibration, Institute of Engineering Mechanics, China Earthquake Administration; Key Laboratory of Earthquake Disaster Mitigation, Ministry of Emergency Management, Harbin 150080, China

3. Qinghai Earthquake Agency, China Earthquake Administration, Xining 810001, China

4. Lanzhou Institute of Seismology, China Earthquake Administration, Lanzhou 730000, China

Abstract: The most important method of understanding liquefaction-induced engineering failures comes from the investigation and analysis of earthquake damage. In May 2021, the Maduo Ms7.4 earthquake occurred on the Tibetan Plateau of China. The most representative engineering disaster caused by this earthquake was bridge damage on liquefied sites. In this study, the mutual relationships between the anti-liquefaction pre-design situation, the ground motion intensity, the site liquefaction severity, and the bridge damage state for this earthquake were systematically analyzed for typical bridge damage on the liquefied sites. Using field survey data and the current Chinese industry code, simulations of the liquefaction scenarios at typical bridge sites were performed for the pre-design seismic ground motion before the earthquake and the seismic ground motion during the earthquake. By combining these results with post-earthquake investigation results, the reason for the serious bridge damage resulting from this earthquake is revealed, and the necessary conditions for avoiding serious seismic damage to bridges built in liquefiable sites is presented.

Keywords: seismic liquefaction; Maduo Ms7.4 earthquake; bridge damage; seismic code

1 Introduction

Damage to engineered structures caused by earthquakes forms the basis of earthquake disasters. Soil liquefaction can cause a reduction or loss in soil bearing capacity and can lead to massive amounts of damage to various engineered structures and infrastructures(Ishihara, 1993; Youdet al., 2001). As a result, soil liquefaction has become one of the most typical damage types caused by earthquakes, and has been a forefront research topic in the earthquake engineering field around the world (Berrill and Yasuda, 2002; Elgamalet al.,2002; Motamed and Towhata, 2010; Chenet al., 2021;Zhanget al., 2021).

Some liquefaction tests have been performed and the knowledge gained from these tests is very helpful for understanding the mechanism and characteristics of soil liquefaction. Notable developments in the relevant research include those presented by Wei and Yang (2019a,b), who studied the effects of fines and gradation on the liquefaction resistance of silty sands. They presented results and findings from a comprehensive experimental program that covered reasonably wide ranges of packing densities, initial effective confining pressures,fines contents, and initial static shear stress ratios.They revealed the mechanism of fines-content-induced reduction of the cyclic resistance and the mechanism of the base-sand effect. Observations from the tests were described in a sound theoretical context, which is very important for understanding the complicated effects of fines content on soil liquefaction. Chenet al. (2020)presented a new approach for assessing liquefaction triggering of saturated sandy soils. Based on a series of undrained cyclic triaxial tests on six types of saturated sandy soils, they found a simple laboratory-testingbased method for assessing the cyclic resistance and shear velocity of liquefiable sandy soils.

Post-earthquake investigation is the most direct means of acquiring liquefaction knowledge and improving existing liquefaction evaluation methods and countermeasures. Notable post-earthquake investigations include investigations into the 1964 Niigata earthquake (Ishihara and Koga, 1981), the 1976 Tangshan earthquake (Fu and Tatsuoka, 1984; Liu,2002), and the 1999 Chi-Chi earthquake (Yuanet al.,2003). In this century, the proportion of liquefaction disasters to the total number of earthquake disasters has been increasing, and some new liquefaction phenomena have appeared. For instance, large-scale liquefaction of gravelly soils occurred during the 2008 Wenchuan earthquake in China (Chenet al., 2009; Houet al., 2011). Liquefaction was the dominant cause of damage to buildings and infrastructures during the 2011 Christchurch earthquake in New Zealand (Cubrinovskiet al.,2011). Liquefaction caused huge economic losses and heavy casualties by producing long-distance flow-slides of liquefied soils during the 2018 Sulawesi earthquake in Indonesia (Kiyotaet al., 2020). These analyses of real seismic liquefaction events provide an opportunity to update existing knowledge, methods, and mitigation measures regarding the liquefaction issue.

Liquefaction evaluation is the top priority for ensuring engineering safety and protection against liquefaction(Beroyaet al., 2009; Papathanassiouet al., 2015).The term “simplified procedures” is based on standard penetration tests (SPTs) and liquefaction survey data from historical earthquakes; these “simplified procedures” are most widely used for evaluating the seismic liquefaction resistance of soils in engineering fields. Seed and Idriss (1971) proposed a basic simplified procedure for evaluating the seismic liquefaction resistance of soils. In 1996, a group of 21 experts reviewed the developments and came to a consensus regarding further improvements to the simplified procedure (Youd and Idriss, 1997; Youdet al., 2001). After this, some simplified liquefaction prediction methods have also been presented (Juanget al., 2003; Idriss and Boulanger, 2006; Mosset al., 2006;Shenet al., 2016).

China is a country that experiences serious earthquake and liquefaction hazards, and Chinese experts have been attempting to establish a simplified procedure to evaluate the seismic liquefaction resistance of soils on the Chinese mainland since the 1966 Xingtai earthquake. At present,different industries on the Chinese mainland, including the building, highway, railway, water conservancy, and nuclear power industries, have their own seismic design specifications containing site liquefaction evaluations.For instance, in the highway industry, the JTG B02-2013 (2013) is currently used to address the seismic design problem for highways on the Chinese mainland.In this standard, a simplified procedure for evaluating the liquefaction potential during highway construction is included. Moreover, the seismic codes used on the Chinese mainland are mandatory standards, so they have a decisive influence on the seismic safety of engineering constructions.

In May 2021, the MaduoMs7.4 earthquake occurred on the Tibetan Plateau of China. One of the most important features of this earthquake was large-scale soil liquefaction. Due to the scarcity of people in the earthquake-stricken area, this liquefaction did not cause extensive damage, but typical liquefaction-induced bridge damage occurred. First, the sites containing the most important bridges near the epicenter all experienced severe liquefaction; some of these bridges severely collapsed, some were partially damaged, and some were slightly damaged. Second, all these bridges were located in high seismic intensity areas, and the seismic designs for all the bridges were completed according to the industry code. Third, these bridges were all located on a high-altitude region approximately 4,200 m above sea level, in which construction is expensive. There have also been instances of bridge failure in liquefaction sites on the Chinese mainland in the past, but unfortunately,there is a lack of detailed investigation data to support a complete analysis of the effectiveness of the code.Therefore, it is very important for the research regarding and development of liquefaction disaster mitigation technology to explore the relationship between site liquefaction and bridge damage and to identify the problems in existing seismic codes that are exposed during earthquakes.

2 Basic information regarding the Maduo earthquake

2.1 Location of the earthquake

On May 22, 2021, aMs7.4 earthquake occurred at 34°59′ N, 98°34′ E in Maduo County, Qinghai Province,China. The earthquake-stricken area is located on the Tibetan Plateau, which has the highest average altitude in the world, with an average elevation of more than 4,000 m.

The Maduo earthquake had a focal depth of 17 km,and its microscopic epicenter was located in Huanghe Township, Maduo County. The seismogenic fault with a sinistral strike-slip rupture trended in the north-westerly direction, resulting in a surface rupture of approximately 100 km (Xuet al., 2021).

Maduo County is an important ecological barrier and an important transportation hub on the Tibetan Plateau. The earthquake-stricken area is deep in the hinterland of the inland plateau, and the plateau has a typical continental semi-dry early climate. There are many rivers in this area, and the months of May through September comprise the season of concentrated rainfall.

2.2 Macroscopic liquefaction phenomena

According to an investigation, large-scale liquefaction occurred during the MaduoMs7.4 earthquake; 40 liquefaction zones were distributed throughout a region 174 km long and 48 km wide. The farthest liquefaction zone was 107 km away from the earthquake epicenter.

The typical liquefaction phenomena that occurred during the Maduo earthquake are illustrated in Fig. 1.Figure 2 shows that this earthquake included all the typical liquefaction phenomena that could be found in previous earthquakes.

The area stricken by the MaduoMs7.4 earthquake lies in the Yellow River source region. The strata are quaternary Holocene strata with a generally flat valley and lacustrine basin, and the strata structure is simple and stable. The shallow surface soil is primarily composed of sand and silt in a generally non-dense state; moreover,the month of May, when the earthquake occurred, is part of a season characterized by concentrated rainfall.All these aspects formed the conditions that led to the large-scale liquefaction produced by the MaduoMs7.4 earthquake.

2.3 Estimation of the PGA distribution

Due to the sparsity of strong-motion observation stations in the earthquake-stricken region, few farfield strong-motion recordings were collected during this earthquake. To evaluate the spatial distribution of the ground motion intensity, Liet al. (2021) assumed 4,461 virtual monitors around the epicenter of the Maduo earthquake. The three-component acceleration time history series for these 4,461 virtual points was given using the stochastic finite-fault method for three-dimensional ground motion simulation based on various source rupture models, a fault slip model,and source rupture stochastic models. Comparisons of the peak ground acceleration (PGA) and peak ground velocity (PGV) values from the far-field strongmotion observations and the ground motion prediction equations showed that the peak ground motions obtained from the simulation method agreed well with the median predictions and could describe well the trend of attenuation with distance.

Based on the simulation results discussed above, the PGA distributions in the Maduo earthquake liquefaction zones were estimated. The values ranged from 80 gal to 660 gal, and 88% of the PGA in the liquefaction zones was between 100 gal and 500 gal, with the most (33%)falling between 100 gal and 200 gal. The smallest PGA in the liquefaction zones occurred at the westernmost point and was 107 km away from the epicenter, while the location of the largest PGA in the liquefaction zones was 3.8 km from the epicenter. Based on the simulation, the estimated PGA values at the sites of the Yematan No. 1 bridge, the Yematan No. 2 bridge, and the Wuermeigang bridge in the MaduoMs7.4 earthquake region were approximately 0.45 g, 0.40 g, and 0.25 g, respectively.Although the three bridges were all approximately 30 km from the epicenter of the earthquake, considering that the source of the Maduo earthquake was a linear source,the Wuermeigang bridge was farthest from the fault zone and, as a result, the PGA at the site of the Wuermeigang bridge was smallest.

Fig. 1 Macroscopic liquefaction phenomena from the Maduo Ms7.4 earthquake

3 Post-earthquake investigation into the bridge damage in the liquefaction zones

Maduo County is located in a herding region where most of the houses are made of bricks and wood.Maduo County is in a seismic high-risk region, and the local government has been strengthening the seismic capacities of the residential buildings. As a result, only a few houses sustained damage during the MaduoMs7.4 earthquake.

The damage to bridges represents the most significant engineering damage resulting from the MaduoMs7.4 earthquake. It caused traffic disruptions and significantly affected the post-earthquake disaster relief. The most remarkable damage was sustained by three bridges,the Yematan No. 1 bridge, Yematan No. 2 bridge, and Wuermeigang bridge, and liquefaction occurred at all three bridge sites during the MaduoMs7.4 earthquake.

An isoseismic map of the MaduoMs7.4 earthquake(China Earthquake Administration, 2021) and the locations of the three bridges are shown in Fig. 2. For simplicity, the Yematan No. 1 bridge, Yematan No. 2 bridge, and Wuermeigang bridge are referred to hereafter as the Y1B, the Y2B, and the WB, respectively.

3.1 Post-earthquake situation of the Y1B

The Yematan No. 1 bridge (Y1B), with a length of 507.4 m and an altitude of 4,219 m, is located on the Gongyu Expressway. The Y1B is divided into upper and lower lines with 25 spans. The span of the bridge is 20 m and the width of the deck is approximately 9 m. It has a double-column pier substructure.

The post-earthquake investigation shows that the Y1B sustained extremely serious damage, as shown in Fig. 3(a). The road surface at the end of the abutment slab on the north side of the Y1B was extruded and arched,and the north side of the first span girder hit the abutment back wall, fell off, and moved northward approximately 1.2 m. The northward displacement of the southernmost main girder led to an opening of approximately 0.8 m in the expansion joint of the south abutment and main girder. Eighteen main girders from the upper line of the bridge and 17 main girders from the lower line collapsed as a result of this earthquake, and the collapsed girders accounted for 70% of the total number of girders. Even though the other main girders did not fall, the actual displacements of the main girders were very near the support width limit.

The post-earthquake investigation also shows that severe liquefaction occurred throughout the entire Y1B site, as shown in Figs. 3(b) and 4. Severe liquefaction occurred on the north ground of the Y1B site, and sandblasting covered an area of approximately 300 m × 300 m. Severe liquefaction was also found in the southern area of the bridge, the visible sand ejecting region covered approximately 100 m × 100 m, and the maximum sand ejecting hole diameter was 3.0 m.

Fig. 2 Locations of the Yematan No. 1 bridge (Y1B), Yematan No. 2 bridge (Y2B), and Wuermeigang bridge (WB) in the seismic intensity distribution map

In this study, the liquefaction index and grade represent the severity of ground failure caused by liquefaction. The larger the liquefaction index and grade,the more serious the ground failure, and the more severe the earthquake damage caused by liquefaction. The specific meaning of the liquefaction grade used herein is explained next. A slight liquefaction grade indicates that there were sporadic sand boiling holes throughout the site; these had only a small influence and essentially did not change the surface morphology of the site. A moderate liquefaction grade indicates that there were many sand boiling sites and that the area covered by sand composed a large portion of the total site area, for example, more than 20%. A severe liquefaction grade indicates that serious ground subsidence or serious changes in the original surface morphology were caused by dense boiling sites or a large amount of sandblasting in the site.

3.2 Post-earthquake situation of the Y2B

The Yematan No. 2 bridge (Y2B) is also located on the Gongyu Expressway. The Y2B is divided into upper and lower lines, with 44 spans in the upper line and 45 spans in the lower line. Its basic structure is the same as that of the Yematan No. 1 bridge, but its length is 887 m.

The post-earthquake investigation shows that the northward displacements of the main girders at the northern ends of the upper and lower lines were relatively small; however, the main girder deformation in the bridge gradually increased from north to south.As shown in Fig. 5(a), seven main girders on the south side of the lower line bridge fell. No main girders fell in the southern portion of the upper line, but many of the rubber supports in the southern portion of the upper line had slipped, and all the main girders nearly fell off.

The post-earthquake investigation shows that, unlike the total damage and the severe liquefaction throughout the entire Y1B site, the Y2B damage can be divided into two parts according to the site liquefaction severity. First,the northern part of the bridge was lightly damaged,the displacements of the main girders were small, and there was no liquefaction in the northern site. Second,in the southern part of the bridge, the displacements of the main girders in the severe liquefaction region increased significantly and were accompanied by bridge pile foundation damage and continuous local falling of the main girders; in addition, the ground in the southern portion was severely liquefied, as shown in Fig. 5(b).

The relationship between the damage and site liquefaction for the Y2B obtained from the survey is shown in Fig. 6. The figure shows that the site experienced heavy liquefaction and the upper bridge was then heavily damaged, and vice versa. There is a good correlation between the damage severity of each part of the Y2B and the liquefaction severity of the site under the Y2B.

3.3 Post-earthquake situation of the WB

The length of the Wuermeigang bridge (WB) is 307.0 m, and it is located on the Xili Expressway. It is separated into upper and lower lines, and it has a doublecolumn pier substructure.

The post-earthquake investigation shows that the upper and lower bridges remained standing and that some of the retaining blocks were cracked and the supports had certain residual displacements. As shown in Fig. 7(a), however, each pier column was visually in good condition, and the bridge maintained a good passable state after the earthquake. As shown in Figs. 7(b) and 8, the entire WB site experienced severe liquefaction, visible ground cracks, and site lateral displacement.

Fig. 3 Post-earthquake damage and site liquefaction for the Yematan No. 1 bridge

Fig. 4 Post-earthquake relationship between damage and site liquefaction for the Yematan No. 1 bridge

Fig. 5 Post-earthquake bridge damage and site liquefaction for the Yematan No. 2 bridge

Fig. 6 Post-earthquake relationship between bridge damage and site liquefaction for the Yematan No. 2 bridge

Fig. 7 General view after the earthquake and site liquefaction for the Wuermeigang bridge

4 Anti-liquefaction pre-design

According to the Chinese seismic codes, the Y1B,the Y2B, and the WB are all class B bridges with key fortifications. Before the earthquake, the three bridge sites were investigated in detail, liquefaction problems were analyzed, and corresponding anti-liquefaction measures were taken. The liquefaction analyses of the three bridge sites were completed according to the current Chinese Specification of Seismic Design for Highway Engineering (SSDHE) (JTG B02-2013, 2013).

4.1 SSDHE liquefaction analysis method

The SSDHE liquefaction prediction formula used in this study was formed on the Chinese mainland in the 1970s. The formula was obtained by statistical regression of the double linear model from 156 groups of data,which were acquired from liquefaction survey data from the Chinese mainland, especially the liquefaction data from the 1976 Tangshan earthquake.

For horizontally layered sandy and silty soils with buried depths less than 15 m, the SSDHE liquefaction discrimination formula (JTG B02-2013, 2013) can be expressed by Eq. (1):

whereNcrrepresents the liquefaction critical value of the penetration blow count,dwis the groundwater depth (m),dsis the soil depth (m), andρcis the percentage of the clay content in the soil layer.ρcis set equal to 3 whenρcis less than 3, andN0= 8.65lnamax+ 26.5, whereamaxrepresents the peak horizontal acceleration at the ground surface (g).

The site liquefaction index can be calculated using Eq. (2):

whereILErepresents the site liquefaction index,nis the total number of liquefied layers at the site,Niis the penetration blow count of theith soil layer,Hiis the thickness of theith soil layer (m), andWiis the weight function value of the hazard influence represented by the unit thickness of theith soil layer (m-1).Wican be expressed by Eqs. (3-1) and (3-2):

The site liquefaction grade,GLE, can be determined from Table 1, and the anti-liquefaction measures for class B bridges stipulated in the SSDHE (JTG B02-2013, 2013) are listed in Table 2.

4.2 Anti-liquefaction pre-design of the Y1B

The engineering geology boring logs for the Y1B site, obtained from the engineering geological survey report, are shown in Fig. 9. The bridge site is primarily composed of quaternary Holocene alluvial and diluvial strata, which include silt, silty sand, fine sand, gravel sand, and gravel, and its formation structure is simple.As shown in Fig. 9, the shallow soil layers at the Y1B site can be divided into four engineering geological layers, beginning from the top:

(1) Fine sand: tawny brown, uniform, contains a small amount of silt, slightly dense. Silty sand: tan,relatively uneven, slightly dense.

(2) Silty soil: grayish-brown to grayish-black,relatively uniform, contains a small amount of fine sand,slightly dense.

(3) Gravelly sand: tan, has good roundness, relatively uneven, moderately dense, contains silt and breccia.

(4) Gravel: tan, has good roundness, moderately dense and dense, contains silt and breccia, has large pebbles in local regions.

Based on the boring logs shown in Fig. 9, the liquefiable layer information for the north and south sides of the Y1B site is listed in Tables 3 and 4, respectively.From JTG B02-2013 (2013), the anti-liquefaction predesign level for the Y1B site is prescribed for a moderate earthquake, and the PGA of the liquefaction fortification for the Y1B is 0.20 g. With the ground motion parameter,the liquefaction index determined for the soil layers and the site liquefaction grade for the Y1B site obtained from JTG B02-2013 (2013) are shown in Table 5. Based on these results, the anti-liquefaction measure for the Y1B site can be acquired from Table 2.

Table 1 Corresponding relationships between liquefaction grade (GLE) and liquefaction index (ILE)

Table 2 Anti-liquefaction measures for class B bridges stipulated in JTG B02-2013 (2013)

Fig. 8 Site liquefaction for the Wuermeigang Bridge

Fig. 9 Engineering geology boring logs for the Yematan No. 1 bridge site

4.3 Anti-liquefaction pre-design of the Y2B

The engineering geology boring logs for the Y2B site, which were obtained from the survey, are shown in Fig. 10. The bridge site is primarily composed of quaternary Holocene alluvial and diluvial strata, and its formation structure is simple. As shown in Fig. 10,the shallow soil layers in the Y2B site can essentially be divided into four engineering geological layers,beginning from the top:

(1) Silty sand: tan, relatively uniform, contains approximately 10% gravel, slightly dense.

(2) Gravelly sand: tan, has good roundness, relatively uneven, moderately dense, contains approximately 15%gravel.

(3) Silty soil: yellowish-brown, uneven soil,approximately 30% gravel, slightly wet, moderately dense.

(4) Round gravel: mixed in color, 0.2–2.0 cm in diameter, has good roundness, moderately dense and dense.

Based on the boring logs shown in Fig. 10, the liquefiable layer information for the Y2B site is listed in Table 6. Note that there is a silty sand layer approximately 1.6 m thick at the north side of the Y2B site; however, it is all above the water table, and as a result it is not listed here. According to the specified anti-liquefaction predesign level in JTG B02-2013 (2013), the PGA of the liquefaction fortification for the Y2B site is 0.15 g. The liquefaction index determined for the soil layers and site liquefaction grade for the Y2B site are shown in Table 7,and no anti-liquefaction measure is required, according to Table 2.

4.4 Anti-liquefaction pre-design of the WB

The engineering geology boring logs for the WB site, which were obtained from the survey, are shown in Fig. 11. The bridge site is mainly composed ofquaternary Holocene alluvial and diluvial strata, aeolian strata, and diluvial strata. The aeolian layer consists primarily of fine sand and the alluvial layer consists primarily of silt and gravel, and the formation structure is simple. As shown in Fig. 11, the shallow soil layers of the WB can be divided into four engineering geological layers, beginning from the top:

Table 3 Liquefiable layer information for north side of the Yematan No. 1 bridge site

Table 4 Liquefiable soil layer information for south side of the Yematan No. 1 bridge site

Fig. 10 Engineering geology boring logs for the Yematan No. 2bridge site

Table 5Liquefactionindex(ILEi) of soil layersand site liquefaction grade(GLE) evaluated usingJTGB-02-2013(2013) for theYematan No. 1bridge underpre-design earthquake conditions

(1) Fine sand: tan, uniform, slightly dense.

(2) Silty soil: tan, relatively uniform, moderately dense, contains a small amount of breccia.

(3) Gravel: mixed in color, uneven soil, approximately 30% silt, moderately dense.

(4) Silty soil: tan, relatively uneven, moderately dense, tan, contains 25% gravel.

Based on the boring logs shown in Fig. 11, the liquefiable layer information for the east and west sides of the WB site is listed in Tables 8 and 9, respectively.According to the specified anti-liquefaction pre-design level from JTG B02-2013 (2013), the PGA of the liquefaction fortification for the WB is 0.20 g. The liquefaction index determined for the soil layers using JTG B02-2013 (2013) is shown in Table 10, and the liquefaction grade for the entire WB site is severe. Based on these results, the anti-liquefaction measure from Table 2 for the WB is that for severe liquefaction.

5 Simulations of the liquefaction that occurred during the Maduo earthquake

Based on the estimated PGA values presented above for the Y1B, Y2B, and WB sites during the MaduoMs7.4 earthquake, simulations of the liquefaction issues for the three bridge sites were performed using JTG B02-2013(2013).

5.1 Liquefaction simulation for the Y1B site

The liquefaction index (ILEi) of the soil layers and the site liquefaction grade (GLE) evaluated using JTG B02-2013 (2013) for the Y1B site under the MaduoMs7.4 earthquake conditions are listed in Table 11.

Comparisons of Table 11 with Table 5 show that both the site liquefaction grade and the thickness of the liquefied soil layers obtained for the estimated earthquake loading are greater than those obtained for the pre-design earthquake loading. At the north side of the Y1B site, the liquefaction layers change from Y1-N-1 to Y1-N-1 and Y1-N-2, and the liquefaction grade increases from slight (4.7) to severe (18.1). At the southside of the Y1B site, the liquefaction layers change from Y1-S-1 and 1-S-2 to Y1-S-1, Y1-S-2, and Y1-S-3, and the liquefaction grade increases from slight (5.8) to severe (28.5). These facts indicate that the liquefaction situation of the Y1B site under the estimated earthquake loading is quite different from that under the pre-design earthquake loading.

Table 6 Liquefiable soil layer information for south side of the Yematan No. 2 bridge site

Fig. 11 Engineering geology boring logs for the Wuermeigang bridge site

Table 7 Liquefaction index (ILEi) of soil layers and site liquefaction grade (GLE) evaluated using JTG B-02-2013 (2013) for the Yematan No. 2 bridge under pre-design earthquake conditions

5.2 Liquefaction simulation of the Y2B site

The liquefaction index (ILEi) of the soil layers and the site liquefaction grade (GLE) evaluated using JTG B02-2013 (2013) for the Y2B site under the MaduoMs7.4 earthquake conditions are listed in Table 12.

Comparisons of Table 12 with Table 7 also show that part of the site liquefaction grade and the thicknesses of the liquefied soil layers obtained for the estimated earthquake loading are significantly different from those obtained for the pre-design earthquake loading. At the north side of the Y2B site, no liquefaction occurred for the estimated seismic ground motion or the designed seismic ground motion. At the south side of the Y2B site, however, there are three liquefaction layers and the liquefaction grade is severe (26.3). These facts indicate that, compared with the pre-design situation, there is a significantly increased potential for liquefaction severity at the south side of the Y2B site under the estimated earthquake loading.

5.3 Liquefaction simulation of the WB site

The liquefaction index (ILei) of the soil layers and the site liquefaction grade (GLE) evaluated using JTG B02-2013 (2013) for the WB site under the MaduoMs7.4 earthquake conditions are listed in Table 13.

Comparisons of Table 13 with Table 10 also show that both the site liquefaction grade and the thicknesses of the liquefied soil layers obtained for the estimated earthquake loading are nearly equal to those obtained for the pre-design earthquake loading. On the east and west sides of the WB site, neither the liquefaction layers nor the liquefaction grade changes. These facts indicate that compared with the pre-design situation, there are no significant changes in the liquefaction situation at the WB site for the estimated earthquake loading.

6 Comprehensive analysis

Since liquefaction played an important role in the bridge damage during the MaduoMs7.4 earthquake,it is important to analyze the relationship between the liquefaction and the damage to the Y1B, the Y2B, and the WB to mitigate or prevent further liquefaction disasters.The mutual relationships between the anti-liquefaction pre-design, the earthquake liquefaction state, and the earthquake damage state are shown in Table 14 for the three bridges. According to JTG B02-2013 (2013), the anti-liquefaction pre-design levels for the three bridges all are prescribed for a moderate earthquake (i.e.,seismic ground motion with a probability exceeding 10% in 50 years) and the PGA values of the liquefactionfortification for the Y1B, Y2B, and WB sites are 0.20 g,0.15 g, and 0.20 g, respectively. In these cases, the predicted site liquefaction states for the Y1B, Y2B, and WB sites are slight, none, and severe, respectively. The anti-liquefaction pre-design and the anti-liquefaction measures for the three bridges were determined according to JTG B02-2013 (2013).

Table 8 Liquefiable soil layer information for east side of the Wuermeigang bridge site

Table 9 Liquefiable soil layer information for west side of the Wuermeigang bridge site

Table 10 Liquefaction index (ILEi) of soil layers and site liquefaction grade (GLE) evaluated using JTG B-02-2013 (2013) for the Wuermeigang bridge under pre-design earthquake conditions

Table 11 Liquefaction index (ILEi) of soil layers and site liquefaction grade (GLE) evaluated using JTG B-02-2013 (2013) for the Yematan No. 1 bridge site under Maduo Ms7.4 earthquake conditions

Table 14 shows, however, that the estimated PGA values for the Y1B and Y2B sites during the MaduoMs7.4 earthquake are approximately 0.45 g and 0.40 g,respectively. Note that according to the Ground Motion Parameter Zonation Map of China (GB 18306-2015,2015), the 0.45 g that the Y1B bridge suffered and 0.40 g that the Y2B bridge sustained are approximately equivalent to a rare earthquake (i.e., seismic ground motionwith a probability exceeding 2% in 50 years). Severe liquefaction occurred at the Y1B site corresponding to the estimated 0.45 g, i.e., what was designed to be slight site liquefaction was severe site liquefaction, and as a result the bridge was seriously damaged during theearthquake. It is particularly necessary to point out the relationship between the damage state of the Y2B and the site liquefaction state below the Y2B. As shown in Table 14, for a design ground motion of 0.15 g, both the north and south sides of the Y2B sites were determined to be non-liquefied, and the soil and foundation should not require anti-liquefaction treatments. Under the estimated 0.40 g that occurred during the MaduoMs7.4 earthquake, there is no liquefaction at the north side of the bridge site, but there is severe liquefaction at the south side of the bridge site. This result corresponds well to the liquefaction state, in which the south side of the bridge site has serious bridge damage, while there is only slight damage on the north side of the bridge site. These facts demonstrate that a good correspondence exists between the location and extent of the bridge damage and the occurrence and extent of the site liquefaction under the bridge, indicating that bridge failure is closely related to the liquefaction of the site.

Table 12 Liquefaction index (ILEi) of soil layers and site liquefaction grade (GLE) evaluated by SSDHE for the Yematan No. 2 Bridge site under Maduo Ms7.4 earthquake conditions

Table 13 Liquefaction index (ILEi) of soil layers and site liquefaction grade (GLE) evaluated using JTG B-02-2013 (2013) for the Wuermeigang bridge site under Maduo Ms7.4 earthquake conditions

Table 14 Comparison of anti-liquefaction pre-design, earthquake liquefaction state, and earthquake damage state for three bridges during the Maduo Ms7.4 earthquake

By contrast, due to the relatively large distance from the fault, the PGA suffered by the WB site is 0.25 g,which is approximately equivalent to the 0.20 g adopted in the anti-liquefaction pre-design for the WB. Severe liquefaction occurred at the WB site under the estimated 0.25 g; however, the bridge was originally designed to prevent severe liquefaction, and as a result the WB only suffered slight damage during the earthquake.

In addition, according to JTG B02-2013 (2013),the Y1B, the Y2B, and the WB all are class B bridges,which belong to the key fortification category.According to JTG B02-2013 (2013), the seismic design of the bridge superstructures is based on rare earthquake fortification, so the bridge superstructures should have had a relatively sufficient seismic margin for the MaduoMs7.4 earthquake. According to JTG B02-2013 (2013),however, only fortification against moderate earthquakes was adopted for the liquefaction issue, which is the cause of the Y1B and Y2B failures. If liquefaction fortification for rare earthquakes had been conducted, it is likely that serious damage to the Y1B and the Y2B could have been avoided during the earthquake. In the authors′ opinion,there are two reasons for the underestimations in the PGA predictions for the pre-design phase. First, the problem of liquefaction was considered in the seismic design. However, there is no clear understanding of how the liquefaction degree changes for different PGA values while there is a good understanding of how the superstructure damage degree changes for different PGA values. Second, the cost of liquefiable soil treatment is very high, and it is not readily accepted in engineering practice. Because of this, the liquefaction and superstructure states are inconsistent for design ground motion in JTG B02-2013 (2013). The seismic design of the bridge superstructure was selected for the relatively high-level ground motion PGA values of approximately 0.3–0.4 g, while the anti-liquefaction design was selected for the relatively low-level ground motion PGA values of approximately 0.15–0.2 g. The anti-liquefaction level adopted in the design was significantly lower than that of the superstructure. Then, the anti-liquefaction margin was seriously inadequate. The discrepancy between the superstructural seismic design and anti-liquefaction design led to the bridge damage during the Maduo earthquake.

Meanwhile, according to the WB survey results,the severe liquefaction experienced by the site during the Maduo earthquake did not lead to severe damage to the bridge, primarily because the pre-design had taken countermeasures against the predicted severe liquefaction. In other words, although severe liquefaction occurred in the site during the earthquake, the antiliquefaction engineering measures taken according to JTG B02-2013 (2013) caused the bridge to avoid severe damage.

7 Conclusions

Investigation and analysis of the typical liquefaction damage during the 2021 MaduoMs7.4 earthquake led to five primary conclusions:

(1) The most representative engineering disasters during the 2021 MaduoMs7.4 earthquake were bridge damage and significant site liquefaction. The shallow surface soils at the bridge sites are composed of nondense saturated sand and silt in quaternary Holocene strata, and thus formed the subjective conditions for the notable liquefaction at the bridge sites during this earthquake.

(2) Anti-liquefaction pre-design for all the bridges was performed using JTG B02-2013 (2013). The seismic designs of the bridge superstructures were selected for relatively high-level ground motion with PGA values of approximately 0.3–0.4 g so that the superstructures would be sufficient to resist this earthquake, while the anti-liquefaction designs were selected for relatively lowlevel ground motion with PGA values of approximately 0.15–0.2 g.

(3) The post-earthquake investigation shows that for the portion of the bridge that was seriously damaged,the soil layer below the damaged portion of the bridge was accompanied by severe liquefaction. The sites with severe bridge damage are predicted to have no liquefaction or very low liquefaction grades in the predesign because of the use of lower-level ground motion conditions for the predictions.

(4) The post-earthquake investigation also shows that there was an instance where the bridge was not seriously damaged despite severe site liquefaction. The reason for this scenario is that the site of the bridge was judged to be severely liquefied in the pre-design and strict anti-liquefaction measures were adopted according to the industry code.

(5) The method of liquefaction prediction and the anti-liquefaction measures in JTG B02-2013 (2013)itself are essentially reliable; however, there is an obvious problem in the coordination between the antiseismic pre-design standards and anti-liquefaction predesign standards for superstructures. The relatively lowlevel ground motion adopted in the anti-liquefaction predesign made the anti-liquefaction margin of the bridge seriously insufficient for this earthquake, and, at the very least, the design should be upgraded to high-level ground motion to be consistent with the superstructure seismic design in future engineering practice.

Acknowledgment

This research is jointly supported by the Natural Science Foundation of Heilongjiang Province(ZD2019E009), Key Project of National Natural Science Foundation of China (U1939209) and the Research Start-Up Fund for High-Level Talents of Heilongjiang University of Science and Technology.