1 Introduction
Urban space resources above the ground are becoming increasingly tight, and people are beginning to focus on the development and utilization of underground space. Land resources are nonrenewable. To avoid repeated excavation of roads, the method of laying pipelines has been changed from direct burial to a centralized arrangement in underground comprehensive utility tunnels. Utility tunnels are used for distribution of electric power, communication, gas, and other pipelines, and act as lifelines of the city. Once a fire occurs, the utility tunnel will collapse, and the entire city may be paralyzed, causing huge economic losses. Therefore, fire resistance research and fire protection in underground utility tunnels have attracted significant attention.
In recent years, numerous studies have been conducted on utility tunnel fires. Beard and Carvel [
1], Ingason and Li [
2] were the first to summarize the problems of fire detection and various types of active and passive fire protection strategies from an engineering practice perspective. Kim et al. [
3] and Ko [
4] compared the effects of various factors, such as the form of the tunnel cross-section, on the smoke flow characteristics of tunnel fires, providing detailed guidance for the design of utility tunnel fire protection and engineering contingency measures.
However, early studies on utility tunnel fires were limited to engineering experience. In recent years, many scholars have explored fire temperature responses using experiments and simulations. Beji et al. [
5], Huang et al. [
6,
7], and Zavaleta et al. [
8] conducted in-depth studies on the smoke temperature field distribution of utility fires in specific confined space scenarios. Through real-time monitoring, they obtained the characteristics of the flame spread, influencing factors of the mass loss rate and heat release rate, and distribution of the temperature field under various conditions. Liang et al. [
9] analyzed the propagation process of a cable fire in a T-shaped-cross-section pipe gallery. Zhang and Zhao [
10] confirmed the necessity of closing a fire door from the perspective of the temperature field. Ring et al. [
11,
12] studied the fire behavior of underground concrete frame structures with rectangular sections in a combination of experiments and numerical simulations to measure and analyze parameters such as internal concrete temperature distribution, reinforcement temperature, concrete burst depth, and crack distribution. However, most current studies focus on the fire protection design and fire spread characteristics, and there are few studies on the structural response of utility tunnels.
Regarding the study of mechanical properties of utility tunnels, Marshall and Haji [
13] theoretically derived and analyzed the interaction between underground utility tunnels and soil from factors such as burial depth of the tunnel, soil stiffness, and force transfer methods. Bennett et al. [
14] studied the mechanical properties of a double-cabin utility tunnel structure based on test results. Dasgupta and Sengupta [
15] designed underground tunnel models and conducted experiments to analyze changes in the mechanical properties and stiffness of underground tunnel structures under static loads. These studies have provided constructive evidence and opinions. However, they are limited to room temperature conditions, and the current materials of utility tunnels are limited to ordinary concrete. Zhang et al. [
16–
19] proposed an estimation method capable of quickly assessing the spalling risk of fire-loaded tunnel linings and proposed a numerical procedure for quasi-brittle fracture and dynamic fracture. Zhou et al. [
20] proposed a model to simulate complex crack patterns. The proposed models and optimization procedures simplified crack analysis under complex conditions. However, research on the fire performance of utility tunnels mostly focuses on finite elements, and there is a lack of fire test research under multivariable and multi-working conditions.
Many factors affect the study of utility tunnel fire tests including difficulty of practical operation, high cost of platform construction, and post-disaster irreparability. Therefore, the current research on multi-hazard utility tunnels is mostly focused on the level of temperature response, such as material properties, numerical simulation, and fire design, whereas the study of multivariable and whole process structural response is lacking. To address the aforementioned problems, this study had the main objectives as follows. 1) A mix proportion of full lightweight concrete was designed. 2) The entire fire test process was carried out on a utility tunnel. 3) The temperature response at the heating and cooling stages was summarized, and the influences of construction methods and concrete materials on the structural response of the utility tunnel were analyzed.
2 Test design
2.1 Material properties of concrete
According to “Specification for Mix Proportion Design of Ordinary Concrete” JGJ55-2011 [
21], and “Technical Standard for Application of Lightweight Aggregate Concrete” JGJ/T12-2019 [
22], the mix proportion design of concrete was carried out. Data for C40 and LC40 concrete are shown in Tab.1 and Tab.2, respectively. The data includes material compositions and mix ratios.
According to “Lightweight Aggregates and Test Methods” GB/T 17431.2-2010 [
23], the average relative numerical tube pressure index of the aggregates was measured in a numerical tube pressure test on ceramic pellets, as shown in Fig.1. The average numerical tube pressure of the aggregate was 7.23 MPa.
Cubic compressive strength tests and axial compressive strength tests were carried out on ordinary concrete and full lightweight concrete. The results are shown in Tab.3 and Tab.4, respectively. The test showed that the failure surface of full lightweight concrete was smooth, and the failure mode was aggregate cracking failure, and this was obviously different from the failure mode of equal-strength ordinary concrete.
With respect to the high-temperature performance of concrete, Peng et al. [
24] have found that lightweight aggregate concrete members have greater fire endurance periods than members made with ordinary concrete. The superior performance is due to a combination of lower thermal conductivity, lower coefficient of thermal expansion, and inherent thermal stability.
2.2 Specimen design and production
According to the research results of NFPA92 [
25], when the geometric ratio of the scaled model is less than one-eighth, the validity of the fire model test is not reliable. Taking account of the existing laboratory conditions and the test objective, the scale ratio of the utility tunnel model was chosen to be 1 : 6. According to the “quasi-dimensional analysis method”, a similar proportional relationship between the model and the prototype was obtained, as shown in Tab.5.
The test used a single-silo rectangular section utility tunnel as the research object. The design length was 1200 mm, outer dimension of the cross-section was 500 mm × 500 mm, and wall thickness was 50 mm. The design parameters and reinforcement diagrams are presented in Tab.6 and Fig.2, respectively.
Cast-in-situ and prefabricated utility tunnel models were developed. Cast-in-situ utility tunnels were formed in one pour. The production process of the prefabricated utility tunnels was to prefabricate four unit slabs, assemble them using a steel lap, and then pour concrete. The production processes of cast-in-situ utility tunnels and prefabricated utility tunnels are shown in Fig.3 and Fig.4, respectively.
2.3 Loading system
The entire fire process included four stages: overburden load application, heating, cooling, and static load after the fire.
When underground utility tunnels suffer from fire, they also continue to bear the external forces of the soil, vehicles, and other loads. According to the buried depth of the utility tunnel to simulate the buried depth of the utility tunnel, based on a typical depth of 6 m, the vertical load value was applied on the utility tunnel at 85 kN. The test was carried out by four-point loading of the specimen through a secondary distribution beam. Four rigid blocks were placed at the bottom of the tunnel such that the top and bottom plates were subjected to the same force in the opposite direction. A schematic of the equivalent loading of the soil load is shown in Fig.5(a).
After vertical loading to 85 kN, a heating plate was used to heat from both ends of the tunnel in opposite directions. A fire loading system with a rated temperature of 1200 °C was used to heat up according to the ISO834 standard heating curve to simulate a fire scenario. Equation (1) is an expression of the international standard warming curve of ISO 834 [
26]
where is the initial temperature (ambient temperature) in degrees Celsius (°C); T is the temperature of the heating plate after time , in degrees Celsius (°C); is the temperature rise time in minutes (min).
Heating was performed for 2h. The specimen was then cooled via ventilation. The heating connection and loading devices are shown in Fig.5(b) and Fig.5(c), respectively.
After the temperature dropped to the ambient temperature, a vertical load was applied to the top plate until the ultimate damaged state was reached.
2.4 Monitoring plan
As shown in Fig.6, several thermocouples were installed on the top and btom plates of the tunnel as well as the wall plates. In this way, the temperature field changes in the tunnel were measured during the fire process, and the temperature distribution patterns along the longitudinal direction of the tunnel and across its cross-section were analyzed. Concrete strain gauges were arranged on the top, bottom, and wall panels of the utility tunnel to analyze variations in the concrete strain. The arrangement of the concrete strain gaugesot is illustrated in Fig.7. As shown in Fig.8, displacement gauges were arranged at both ends and at the mid-span of the top and wall plates of the specimen to measure the structural deformation of the utility tunnel during the fire.
3 Test failure phenomena
3.1 Specimen damage process
(1) Cast-in-situ utility tunnels (UT1 and UT3)
During the application of vertical load, long vertical cracks with a small thickness appeared in the middle of the wall plate. Cracks extended and the number of cracks increased as the vertical load was increased. After heating for 15 min (heating plate temperature: 677 °C), new symmetrically distributed cracks appeared in the wall panel. After heating for 40 min (heating plate temperature: 823 °C), water vapor overflowed from the top of the utility tunnel, water seepage appeared on the surface of the roof and wall panels, and the surface color changed from gray−cyan to gray−white. After heating for 2 h (heating plate temperature: 982 °C), the heating plates on both sides were removed. The concrete at the end was burned red and cracked, and full-length cracks appeared in the inner walls of the top and bottom plates. When the temperature of the specimen was lowered to room temperature, the maximum crack widths of UT1 and UT3 were measured to be 0.58 and 0.68 mm, respectively. The test failure phenomena of the cast-in-situ utility tunnels are shown in Fig.9.
The load and fire conditions were symmetrical, and thus the cracks were symmetrically distributed. For the top plate, owing to the combined action of the negative bending moment and thermal stress, the tensile stress at the edge of the top plate near the axillary angle was relatively large. Therefore, the edge of the top plate is a weak area of the tunnel that should be strengthened. Under the load action, the bending moment at the mid-span of the top and bottom plates was the largest, and the inner sides of the top and bottom plates were tensioned, resulting in a number of full-length cracks along the longitudinal direction of the utility tunnel. The cracks in the wallboard were evenly distributed. In addition, the crack spacing was similar to the reinforcement spacing, indicating that the reinforcement ratio had a certain influence on the distribution of cracks in the utility tunnel.
(2) Prefabricated utility tunnels (UT2 and UT4)
After the vertical load reached 40 kN, cracks began to appear on the walls of the specimen, and the number of cracks increased with increasing load. When the load reached 85 kN, the load was kept constant and the heating stage was started. In the early stages of heating, the cracks extended and increased in number. When the temperature was increased for 36 min (heating plate temperature: 797 °C), water vapor flowed from the top and end of the specimen. After heating for 2 h (heating plate temperature: 976 °C), the heating was stopped. Many cracks were observed in the bottom plate. The assembly seams were clear, and the concrete at the end was burned red and cracked. The test failure phenomena of the prefabricated utility tunnels are illustrated in Fig.10.
Compared with the failure phenomenon of the cast-in-situ utility tunnel, cracks at the junction of the old and new concrete of the prefabricated utility tunnel were apparent. Insufficient adhesion at the interface between the old and new concrete resulted in cracks under the action of thermal-mechanical coupling. Therefore, the prefabricated utility tunnel was prone to damage to integrity and thermal insulation, and the bonding performance of the new and old concrete should be enhanced.
3.2 Failure mode
After cooling to room temperature, the specimens were loaded vertically to the point of ultimate damage.
Cast-in-situ utility tunnels UT1 and UT3 had shear damage along 45° at the axillary corners, with diagonal cracks in the axillary corners connected to through-length cracks in the inner walls of the top and bottom slabs.
Step-type diagonal cracks appeared along the junction of the precast and post-cast concrete layers in prefabricated utility tunnels UT2 and UT4, resulting in shear damage to the top or bottom slab. The weak location was the junction area between the precast concrete layer and the post-cast concrete layer. The inner walls of the top and bottom plates were found to be rich in mid-span cracks by observing the distribution of cracks on the top and bottom surfaces of the specimens. The inner wall surface was under tension, and the tensile stress exceeded the tensile strength of the concrete, resulting in cracking.
Damage to the prefabricated specimen occurred at the junction of the prefabricated components and the post-cast concrete, whereas the cast-in-situ specimen had no concrete-to-concrete interface. Therefore, the cast-in-situ utility tunnel had better integrity than the prefabricated utility tunnel.
4 Temperature response
4.1 Longitudinal temperature distribution
To study the effect of different construction methods on the temperature distribution of the utility tunnel sections, the temperature–time curves of end sections A-A and middle sections B-B of specimens UT3 and UT4 were plotted and are shown in Fig.11.
In the heating stage, the temperature of each section was maintained at room temperature for the first 15 min, so the temperature response was small during the early stage. When the temperature of each section reached approximately 100 °C, there was a gentle step in the temperature–time curve. This was because the evaporation of crystalline water within the concrete absorbed heat, resulting in some stagnation of temperature change. For the mid-span section away from the heating plate, the water vapor evaporated more slowly; therefore, the step of the temperature–time curve in the midspan section was more obvious than that in the cross-section near the heating plate. The heating rate at each section measurement point increased sharply after this step. From comparison of the temperatures of the measurement points at different positions in each section, it can be seen that the temperatures of the measurement points on the top plate were higher than elsewhere, and this phenomenon was more obvious at the end of the specimen.
The heating was stopped after 120 min. A rapid ventilation method was adopted for cooling treatment. Owing to the thermal inertia of the concrete material, the temperature of each section of the utility tunnel specimen at this time did not drop immediately when the ambient temperature began to fall. Instead, it continued to rise for some time before falling. It was found that the temperature of each measurement point of the utility tunnel specimen reached the maximum temperature in the cooling stage after heating was stopped rather than in the heating stage. Therefore, the utility tunnel structure might be damaged or collapse during the cooling stage after the fire was extinguished.
The trends of the temperature–time curves of the prefabricated and cast-in-situ utility tunnels were consistent. The temperature of the top plate at the end section (A-A section) was higher than that of the wall and bottom plates in the whole heating and cooling stage. Here, the maximum temperature difference between the top and bottom plates was 226 °C. The temperature of the top plate in the middle section (B-B section) was similar to that of one wall plate. In this case, the maximum temperature difference between the top and bottom plates was 78 °C. As the temperature longitudinal propagation distance increased, the temperature difference between the top and bottom plates decreased.
4.2 Temperature gradient variation
To study the temperature gradient distribution of the top plate and wall plate sections of the utility tunnel, measurement points 13, 5, and 15 on the top plate at 0, 25, and 50 mm, respectively, from the inner wall were selected. Further, measurement points 14, 6, and 16 on the wall plate at 0, 25, and 50 mm, respectively, from the inner wall were selected. Because the temperature gradient changes in the top and wall panels of each specimen section were similar, the temperature–time curves of UT3 were used as an example, as shown in Fig.12.
An obvious temperature gradient was observed when the temperature had been increased for 20 min. The temperature difference between the inner and outer wall measurement points was large. As the heating time increased, the temperature gradient along the plate thickness direction rapidly increased, and the temperature difference gradually increased. When the temperature reached 100 °C, the crystalline water in the concrete evaporated and this process absorbed heat, resulting in a stagnation of temperature rise and a plateau in the curve. The farther away from the fire surface, the slower the water evaporation, and the more obvious the platform phenomenon. After heating for 120 min, the temperature difference between the inner and outer walls of specimen UT3 reached 531.7 and 508.6 °C, respectively.
Fig.12 shows that the temperature response was significantly different along the cross-sectional diameter direction of the utility tunnel section. During and after heating, the temperature inside the utility tunnel rapidly increased. However, the heating rate outside the utility tunnel was slow because of the thermal inertia of the concrete. The farther away from the inner wall along the cross-sectional diameter direction, the longer the temperature stagnation time. With an increase in the heating time, the temperature difference between the inside and outside of the utility tunnel gradually increased. The large temperature difference inside and outside caused a large temperature stress inside the utility tunnel, which deteriorated the concrete properties along the thickness direction of the utility tunnel.
5 Structural response
5.1 Concrete strain variation
The concrete strain variation can be used as an important basis for analyzing the structural force performance of a utility tunnel. The strain variation pattern for each specimen differed slightly before the vertical load reached the overburden load. Fig.13 shows the load–strain curve for each measurement point of specimen UT1.
The variation range of the strain value at each measurement point on the top and bottom plate was −600με–400με. Specimen measurement points 2 and 14 showed significant growth in compressive strain, and measurement point 13 showed a significant tensile strain. Each measurement point of the wall plate was in tension, and the span-wise strain along the length of the wall plate was slightly greater than that at both ends. When the load increased to 50 kN, the concrete strain of the wall plate increased significantly. When the load reached the design value of 85 kN, each measurement point was not yielded. After the specimen was heated for 15 min, the concrete strain gauge failed.
5.2 Displacement variation of the utility tunnel in the heating and cooling stages
The displacement change of the utility tunnel in the heating and cooling stages reflected the deformation law of each part of the utility tunnel under the joint action of a high temperature and load. As shown in Fig.14, displacement gauges were arranged at measurement points 1, 2, and 3 to measure the vertical displacement of the top plate, axillary angle of the utility tunnel, and horizontal displacement of the wall plate, respectively. Fig.15 shows the displacement–time curves of each measurement point of tunnels UT1, UT2, UT3, and UT4 in the heating and cooling stages.
During the heating stage of the fire, the roof plates of the utility tunnels were heated, and they expanded. However, owing to the effect of the overburden load, the heating-caused expansion deformation of the roof plate was not sufficient to offset the force deformation under the external load. Therefore, the roof plate of each specimen exhibited a downward deflection deformation under the action of temperature and load, and the deflection in the middle of the span was significantly larger than that at the end of the span. With the gradual deterioration of the material properties at high temperatures, the ability of the structures to resist deformation decreased, and the deflection of the top plates grew continuously. When the temperature reached the maximum, the deflection of the top plate of specimen UT1 was the smallest, and the vertical displacement in the span was 3 mm, which was 1/166 of the net span of the top plate. The vertical displacement in the span of specimen UT2 was 5.48 mm, which was 1/91 of the net span of the top plate. The cast-in-place utility tunnels were more resistant to deformation. The deflection of the top plate of each specimen varied significantly during the cooling stage, and the deflection of specimen UT4 changed the most significantly. Within 20 min of cooling, the mid-span deflection of the UT4 top plate increased by 2.08 mm. With a further decrease in temperature, the structure shrank and the displacement grew slowly. The displacement of the top plate was the largest, and the displacement of the axillary angle was the smallest.
During the heating and cooling stages of the fire, the displacement exhibited nonlinear growth. With the gradual deterioration of the material properties at high temperatures, the ability of the structure to resist deformation decreased, resulting in an accelerated displacement rate. After the temperature increase was stopped, the internal temperature of the specimen continued to increase, and the material further deteriorated, leading to structural damage. The maximum displacement rate was reached during the cooling stage.
Comparison of the displacement–time curves of the utility tunnels for different construction methods and different materials are shown in Fig.16 and Fig.17, respectively.
When the temperature was increased for 20 min, the displacements of UT2 and UT4 changed abruptly. This was owing to the significant deterioration of the old and new concrete junctions of the prefabricated tunnels at high temperatures, resulting in misalignment. The top plate displacement of UT4 was 16.8 mm, which was approximately 41.8% higher than that of UT3. The displacement of the UT4 wall plate was 6.61 mm, which was approximately 2.6 times the displacement of the UT3 wall plate of 2.5 mm. The displacement at the axillary angle of UT4 was 5.03 mm, which was approximately 4.09 times the displacement of 1.23 mm at the axillary angle of UT3.
The displacement-time curve of the utility tunnel with different materials was similar, but compared with the ordinary concrete utility tunnel, the displacement change of the full lightweight concrete utility tunnel was larger. The top plate displacements of specimens UT3 and UT4 were about 11.4% and 14.2% higher than that of specimens UT1 and UT2, respectively.
The construction method had a significant impact on the deformation of the utility tunnel. As the temperature was increased, the material properties gradually deteriorated, resulting in an increase in structural deformation. Whether at the highest temperature or after falling to room temperature, the deformation of the prefabricated utility tunnel was greater than that of the cast-in-situ utility tunnel. The concrete material also affected the deformation of the utility tunnel. In the process of heating and cooling, the displacement changes were similar, but the deformation of the utility tunnel with full lightweight concrete was larger.
5.3 Load–displacement curves of the utility tunnels in the static load stage after the fire
High temperatures can damage the microstructure of concrete, which, in turn, leads to a decrease in its structural bearing capacity. The higher the temperature, the more severely the microstructure of the concrete is damaged, and the more its structural bearing capacity is reduced [
27]. The mechanical properties of structures after fire are important criteria for judging their fire resistance. The shape and changing trend of the load–displacement curve of the specimen provide a comprehensive reflection of the macroscopic mechanical properties of the structure [
28]. The load–displacement curve of the top plate of the utility tunnel reflects the overall stiffness and load-bearing capacity change trend of the structure. The load–displacement curves of the top plate of the utility tunnel with different materials and the load–displacement curves of the utility tunnel with different construction methods are presented and compared, as shown in Fig.18 and Fig.19, respectively.
The load–displacement curves of each specimen in the static load stage after the fire showed the same trend and were divided into two stages: the working stage with cracks, and the damage stage. The load–displacement curves tended to be straight in the working stage with cracks. As the load increased, cracks continued to develop. Continuing to the load, the curve (see Fig.18 and Fig.19) showed an obvious turning point accompanied by a sudden drop in structural stiffness and a sharp increase in displacement. As a result, the top plate of the utility tunnel was crushed and damaged.
The elastic modulus of ordinary concrete is larger than that of full lightweight concrete. Therefore, the ordinary concrete utility tunnel has a stronger ability to resist deformation, and the rigidity of the ordinary concrete utility tunnel is greater, as compared with the full lightweight concrete utility tunnel. The post-fire bearing capacities of ordinary concrete tunnels UT1 and UT2 were 69.67% and 14.3% higher than those of full lightweight concrete tunnels UT3 and UT4, respectively. The rigidity of the cast-in-situ utility tunnel was greater than that of the prefabricated utility tunnel. The different construction methods had a greater impact on the post-disaster bearing capacity of the ordinary concrete utility tunnel and a smaller impact on the post-disaster bearing capacity of the full lightweight concrete utility tunnel.
The ultimate bearing capacity of each specimen at the static load stage after the fire was compared with that of the specimens named UTA and UTB in the static load test at room temperature that was studied by the research group previously. The parameters of UTA and UT2 are the same, and the parameters of UTB and UT4 are the same. The ultimate bearing capacity of each specimen is listed in Tab.7.
The ultimate bearing capacities of UTA and UTB in the static load test at room temperature were 223.3 and 174.7 kN, respectively. The residual bearing capacities of UT2 and UT4 after heating for 2 h were 163.0 and 147.0 kN, respectively. Under the action of fire, the bearing capacities of the ordinary concrete utility tunnel and full lightweight concrete utility tunnel were reduced by 27% and 16.8%, respectively. Owing to the thermal inertia of concrete, there was a large temperature gradient inside and outside the utility tunnel, resulting in the deterioration of structural performance.
Compared with the ordinary concrete utility tunnel, the bearing capacity loss of the full lightweight concrete utility tunnel after the fire is small, which indicates good fire resistance of full lightweight concrete and provides more favorable conditions for rescue after fire.
6 Conclusions
In this study, fire tests of four single-section scaled-down utility tunnels were carried out. The effects of different construction methods and materials on the fire behavior of the utility tunnels were explored by analyzing the temperature and structural responses of the utility tunnels throughout the fire exposure. The results are summarized as follows.
1) Cast-in-situ utility tunnels UT1 and UT3 had shear damage along 45° at the axillary corners, with diagonal cracks in the axillary corners connected to through-length cracks in the inner walls of the top and bottom slabs. Step-type diagonal cracks appeared along the junction of the precast and post-cast concrete layers in UT2 and UT4 of the utility tunnel, resulting in shear damage to the top or bottom slab.
2) As the heating time was increased, the temperature gradient along the plate thickness direction rapidly increased. The maximum temperature difference between the inner and outer walls was 531.7 °C. In any section, the temperature of the top plate was the highest, and the highest temperature occurred in the cooling stage after stopping the heating, which provided a reference for the fire protection design and rescue relating to utility tunnels.
3) The displacement of the top plate was greater than that of the wall plate. Further, the displacement of the wall plate was greater than that of the axillary angle. The deflection of the top plate of each specimen significantly varied during the cooling stage. Within 20 min of cooling, the mid-span deflection of the top plate increased by 2.08 mm. With a further decrease in temperature, the structure shrank, and the displacement slowly increased.
4) The construction methods had a significant impact on the structural response of the utility tunnels. The displacement of the top plate of the prefabricated utility tunnel was 41.8% larger than that of the cast-in-situ utility tunnel. The displacements of the wall plate and axillary angle were 2.6 and 4.09 times higher than those of the cast-in-situ utility tunnel, respectively.
5) The bearing capacity and stiffness of the ordinary concrete utility tunnel were greater than those of the full lightweight concrete utility tunnel. However, the full lightweight concrete utility tunnel had good ductility, strong deformation capacity, and good fire resistance, which were difficult to collapse after being affected by fire. The residual bearing capacities of the ordinary concrete utility tunnel and full lightweight concrete utility tunnel after fire were 163 and 147 kN, respectively, and the bearing capacity losses were 27% and 16.8%.
PDF
(10753KB)
