Report on Civil Engineering Disasters (Multiple topics)
(By FAZAL-E-JALAL, student number: 018010990014, Geotechnical engineering Prof. Xu Yongfus Lab)
(For the lecture of Disaster Prevention and Mitigation in Civil Engineering)
Disaster can either be natural and manmade. In recent years, Earthquakes, Typhoons and Floods are the most significant disasters, according to latest research and media reports. It is noteworthy that the overall death toll, caused by such disasters, lessened over past one century (1900 and onwards), with developing countries witnessing myriad damage (life, property, and resources) than developed countries.
Unlike this, the economic loss went soaring during the mentioned period. By definition, a civil engineering disaster is caused by the failure of civil engineering work because of technical issues. As a civil engineer, we need to know the reasons of occurrence, mechanisms, various characteristics and respective preventive measures.
In this report, the disasters due to earthquake have been put forth. Being a Geotechnical engineer, I have majorly focused on the aspects of Dynamic properties of soils, Liquefaction, Response analysis & stability analysis, and Introduction to equations of motion pertaining to earthquake scenario, in the preceding section.
Most of the material is what Prof. Dr. Deng Jianliang had discussed in the class. An effort has been made to streamline and organize those concepts herein.
Why is soil important? We need to keep soil at controlled stress level, with limiting deformation and such that Pore water pressure (PWP) will not increase. It is noteworthy that the typical tunnel construction in Shanghai costs approx.
1 Billion ? / 30 km stretch. The Clays, Sands, Rocks all exhibit very different properties from each other.
Geotechnical seismic problems: A Schematic diagram (Prof. Tatsuoka lecture)
Liquefaction changes subsoil to liquid
Subsidence of heavy foundation, Floating of
Light embedded structure, lateral flow of soil,
The parameters of dynamic properties of soil are:
Of the 5 relatively important parameters the first set comprises Shear Modulus (G), Poissons ratio (?), Modulus of elasticity (E) which have two independent parameters each. Whereas in the second set, Damping ratio (?) and Coefficient of energy loss (?) exhibit one independent parameter each. Other parameters are Settlement (?), Angle of internal friction (?) etc.
Various field and lab tests evaluate these parameters. The values of E and ? can be determined using either Triaxial Compression (TC) tests or Resonant Column (RC) tests.
The main factors related to dynamic strength are: Strain rate i.e. Frequency, Amplitude i.e. Stress Ratio, and Number of cycles (For instance in case of road/subway these are 10 Million approximately). The three (3) states, in which soil is assumed to behave as are:
Figure 3 Stress strain behaviour of soil (Pure plastic, Pure elastic, and Elasto plastic respectively)
?o i.e. Strain rate, is defined as the ratio of
?? to ? time.
While constructing tunnels / subways, Creep
(i.e. Stress is constant, and strain changes with time)
There is no strain effect in case of sands. The strain rate effect is actually a viscous property, so there is no creep e.g. the effect of 0.5% strain increment for a 10 meter thick sand layer would be 10m * 0.5% = 50 mm, which is unacceptable.
Figure 6 Depicting positive and negative plasticity
Simulations based on models are presented in literature (Tatsuoka et al. 2008). One such model is introduced, wherein quantification of viscosity properties is stated. Deng et al. (2006 and 2012) reported the effect of Saturation degree and particle size on ? values (directly related to each other). The effect of amplitude (i.e. Stress ratio) and Nf (Number of cycles) is also reported. If the stress ratio is equal to zero, it indicates no cyclic loading. Consolidation stress ratio, Kc, is the ratio of major and minor principle stress (in case of Triaxial Compression test).
While studying the dynamic deformation of soil, it is notable that, the loose soil undergoes contraction whereas the dense soils dilate, during shearing process. In shear deformation, shape of specimen changes and in case of compression deformation the volume of specimen changes.
In case of no drainage, ?c = u ; So for sands: c = 0, And ? = 0, Therefore, ?=0
Figure 7 Discussion on dynamic PWP of soils
Moreover, the effect of following parameters on the dynamic strength of soils is important.
It is the transformation of sand into rock over (for instance) 3 Million years period. The negative ageing effect is vice versa. The addition of cement to sand is cementation, which is a type of Positive ageing effect. The sand deposits layer by layer and exhibits inherent isotropy i.e. different deformation and strength of each layer. Dilatancy is the change in volume of sample when shearing occurs. With increase in effective confining pressure, the shear strength increases. In addition, the strain is directly related to softening. At small strains, soils behave elastically. The deformation is almost zero. (e.g. in case of sands, ?? = 0.001%, so the deformation (for an 8cm height sample) = 8cm * 0.001 = 0.008cm. This is very small deformation, which is extremely difficult to calculate.)
A CD Triaxial test was conducted on cement mixed kaolin, and the creep for one day noted was 0.2% (which is very large). The material was again loaded and it showed stiff behaviour. During creep, deformation increases with time. One method of controlling creep is by Pre-loading and then Re-loading.
In wet tests, the soil in powdered form is incorporated in the Oedometer (Height 2cm, Diameter 6cm). It is loaded to 50 KPa. Then, the powdered sample is inundated with water. Sometimes, while sending water to a dry sample in order to make it partially saturated, deformation (as large as 10%) is experienced. It is associated to the fact that, because the rod does not touch the surface of specimen and thus ?v approaches zero. Slowly, it moves downwards and ?v starts increasing. At the start of adding water, the stress drop and then increases. Then, creep occurs over a large stress range due to cementation. This is associated with ageing effect. In addition, in order to study the inherent anisotropic behaviour, the sand specimens in inclined layer are taken. The specimens are frozen and put in the Triaxial Compression (TC) machine. A significant difference in peak strengths and no significant effect on residual strengths is revealed. The External Displacement Transducer (EDT) is helpful in determining strain (?). ?=?H/H
Figure 8 Characteristics of LDT and Gap sensor, for measuring deformation
A review: Settlement of foundation of Akashi Strait Bridge in Japan
The earthquake (Magnitude 7.2) effect on bridge with 2km span was investigated, and it not only survived but also was serviceable. The foundation across the span comprised Sedimentary soft rocks on LHS, Granite on RHS and Gravels (beneath Pier 2P, ? = 70m) on LHS upper region. During construction, almost 10mm settlement occurred which is somehow natural. The Youngs Modulus i.e. E (kgf/cm2) was determined using Static tests (Unconfined Compression Test, Primary loading test etc.) and Dynamic tests (Shear wave velocity test). The difference between values of both methods are huge. Therefore, it was checked by using LDT and the results obtained using Dynamic tests (i.e. Edynamic = 20,000 kgf/cm2; Estatic = 1,000 kgf/cm2) were found to be more accurate.
It is a phenomenon wherein the strength and stiffness of soils is reduced by earthquake vibrations or other rapid loading. This is also referred to as Redefect, due to occurring repeatedly with time. Saturated soils are most prone to such phenomenon. The exertion of pressure by water on the soil particles increases to an extent where soil particles start moving relative to each other. Earthquake shaking often triggers this increase in water pressure, but construction related activities such as blasting can also cause an increase in water pressure. The problem originated from Japan in 1960s. In 1964, researchers in Japan vowed to heed towards this problem. When liquefaction occurs, the strength of the soil decreases and, the ability of a soil deposit to support foundations for buildings and bridges is reduced .
In addition, as ?=c^’+?ctan?’, it is obvious that Confining pressure (?c) is directly related to strength of soil. In case of sands, ?=?ctan?’, because c^’=0. If ?=0, the Bearing capacity is decreased thereby causing Liquefaction.
Mechanism: As, ?’= ?-u . In case of sands,?=tan?’, because c^’=0.
In extreme case, ?=u. After complete loss of effective stress, sand has no shear strength and develops large deformations even under minor shear stress.
During liquefaction, energy is dissipated. This is due to rupture and rebound in the subduction zone. There is very gentle motion. The dissipation of energy is similar to the vibration felt when a piece of chalk is broken. In addition, the different types of faults are normal fault (PULLING phenomenon), Reverse fault (PUSHING phenomenon), Strike slip fault (// to strike) and Combination faults.
Out of these, the REVERSE FAULT is most dangerous of all. (e.g. 2011 Japan earthquake, 2015 Nepal earthquake). Most of the earthquakes occur on faults. Some faults are Active in nature, where earthquakes either occurred in recent past or will occur in near future. However, there is always a High degree of uncertainty.
Factors of Liquefaction, Liquefaction Tests
Liquefaction changes subsoil to liquid
Subsidence of heavy foundation, floating of light embedded structure, lateral flow of soil, Consolidation settlement.
Torsion test is used to simulate the effect of earthquake on soil. The cyclic loading simulates it. Whereas, the Shaking Table (ST) test utilizes a sand model to be tested under shaking amplitude. The ST test is time consuming, expensive and the results are affected (e.g. boundary conditions and size effect. Also, the scale of model is different from that of real structure).
The cyclic loading Triaxial Compression (TC) test is also employed. In Hollow torsion (HT) test, one end of sample is fixed whereas shear force is applied at the other end to induce moment.TC tests controls ?1 and ?3 only, whereas the HT test controls ?1, ?2, and ?3. In addition, in TC tests the disturbance of material in the lab are different from those in the field.
Shear box (SB) or Direct shear (DS) tests are also used. They yield in plots between shear stress and shear strain. The sample is subjected to cycles of loading and unloading, and it is observed that almost 10% strain is witnessed in case of Loose sands , whereas, the strain change is quite small in case of dense sands.
In cyclic triaxial test, pressure is decreased and deviator stress follows a sinusoidal pattern. In case of loose sand, pressure keeps on decreasing approaching zero i.e. confining pressure is zero. For sands, ?=???_c tan?’, if ?=45 degrees, ?=0 (Because ???_c=0). Whereas, in case of dense sands, these cannot be fully liquefied. The liquefaction resistance will increase. Triaxial Compression tests are mostly used to study liquefaction.
Comparing consolidation in Laboratory and field, K0 consolidation is generally taken (Fully drained condition assumed) in field whereas in laboratory, it is assumed isotropic (and may be partially drained condition). For design purposes, the stress ratio is generally assumed as 20 cycles. In dynamic loading, the plot between major principal stress and time follows a uniform sinusoidal trend. However, the same graph is irregular shaped in field.
Figure 10 Explanation of liquefaction curve
Generally at DA=20%, the specimen can be defined to as have Fully Damaged. Or, at times maybe DA=10% can also be set the threshold. Because, both depict quite a large deformation.
Physical meaning of Liquefaction curve:
If earthquake intensity is stronger, the cyclic stress ratio should be more. In addition, if the duration of earthquake is more, number of cycles are more. The failure maybe either due to one of these two possibilities.
Table 2 Various effects of different parameters on Liquefaction
In the plot between q and p i.e. the Stress path figure, because p does not approach 0, therefore Partial liquefaction occurs only (According to Laboratory test results). P = (?_1+?2??_3)/3. Sieve analysis on 20 liquefiable samples was carried out, and it was found that D50 concentrated from 0.1 to 0.2 mm. D50 = 1.0mm (Gravel) are also Liquefiable. It is also possible that might be sand have travelled everywhere and has taken gravels along with it.
Standard Penetration Test (SPT) is very popular in Liquefaction. If N< 15; Liquefaction may occur [This conclusion is still valid for so many earthquakes hitting parts of Japan].
Densify the sands, by using sand compaction piles. The size of sands should not be very small
Rotation during drilling is done. (Because the noise of the vibro-hammer is too large)
If an airport is constructed in the sea, dynamic consolidation is used (weight may upto 5 Tonns). Noise is very large, but does not matter when dealing in the sea.
Compaction grouting (very popular nowadays). Grout is injected into foundation. Sand is compressed from loose into dense state.
Gravel Drain (In order to dissipate the PWP i.e. Pressure should not reach zero. Gravels are used, because have large value of coefficient of permeability)
Lowering GWT. In unsaturated sand layer, no liquefaction occurs.
When large sized tanks are to be placed over ground, then in order to prevent liquefaction Deep wells are provided beneath the ground surface.
Response Analysis and Stability Analysis
For instance, in order to see the effect of earthquake on a Dam, it is considered that Earthquake is a force applied to it and the response is studied. This is a simple approach. In case of foundations, even a small-applied force is detrimental. The effect of that force (i.e. earthquake) depends on (1) How much is the magnitude of earthquake? (2) How far is the epicenter? In addition, the acceleration time history gives further details about it.
The Pseudo-static method is applicable on Restraining wall design, which is done using Mononobe-Okabe method. Where, Weight W is splitted as KhW and KvW. A Kh = 0.15 represents very strong earthquake.
The relative displacement for centre point of building at bottom (on GSL) is zero, whereas that at top is very large. Ah (t) or ah (t) is used to calculate Amax, Vmax, and Dmax. The point is, how to calculate?
Figure 14 The dynamic model and most important equation of dynamics
x ?(t)=d_x/d_t , x ?(t)=?d^2?_x/?d_t?^2 , x_0=Absolute displacement , x=Relative displacement
The response of earthquake to structure greatly depends on ? and ?. As, ah (t) is used to calculate Amax, Vmax, and Dmax. Whereas, ? and ? are different for different structures. Displacement during earthquake is calculated using Linear superposition method. Then, the displacement, velocity and acceleration response spectra, respectively, are calculated as:
? x?_max=? x?_max (?,?)
(x ?+x ?_0)=?(?Ssin(?t-?))?_max
Where, (?=2?/T=Natural frequency of vibration)
Figure 15 Response spectrum plot for calculating Displacement, Velocity and Acceleration
Stability analysis of slope resisting earthquake
During earthquake, ah(t) is determined. Vmax and Dmax are to be determined. Newmarks method is used for this purpose. If earthquake force is very small, the block on the slope stands still. Moreover, if the earthquake force is very large, the block will move downslope immediately. V and S are calculated as;
V= ???(K-K_y )g.d_t ? S= ???V.d_t ? (k= Seismic coefficient, ky = Assumed as constant)
Nevertheless, in case of liquefaction, the resistance decreases and even after earthquake stops the movement of slope continues. Therefore, in reality, ky is not constant. (Because during liquefaction, shearing resistance is zero). Yokowatashi Earthquake investigation case study reveals that in real state, the angle of internal friction (?) was 39 ?. It was very stable and FOS > one. In Liquefaction, even earthquake stopped but the slope slided below as ? had reduced to 16 ? (due to the earthquake). The excess PWP largely affects the displacement. In addition, Ground water level (GWL) significantly affect the residual displacement.
Experiments on soils with small deformations
For elastic model, we need E: Elastic Modulus, G: Shear Modulus, ? or D: Poissons ratio (0.15 0.3), and ?: Damping ratio (It is a viscous property. i.e. Soil absorb energy and decays with time). Soil may also act like liquid (e.g. in Liquefaction). The shear strength is very small or deformation is very large i.e. Double Amplitude (DA) = 10%. How to evaluate these Parameters?