International Journal of Advanced Engineering T echnology
E-ISSN 0976-3945
Research Paper
DEVELOPMENT OF LOW COST SHAKE TABLES AND INSTRUMENTATION SETUP FOR EARTHQUAKE ENGINEERING LABORATORY C. S. Sanghvi 1, H S Patil2 and B J Shah 3
1
Address for Correspondence
Applied Mechanics Department, L.D. College of Engineering, Ahmedabad, Gujarat, India 2 Department of Applied Mechanics, S V National Institute of Technology, Surat, Gujarat, India 3 Applied Mechanics Department, L.D. College of Engineering, Ahmedabad, Gujarat, India ABSTRACT For the development in the field of earthquake engineering, experimental study is required. To study the effects of earthquake, laboratory facilities are needed. The development has reached to a stage where earthquake simulation is achieved in laboratory. Shake table is used to provide earthquake simulation and to test the prototype and scaled model of the structure. In order to reproduce actual earthquake data, a six-degree of freedom electro-hydraulic shaking table is essential. They are very expensive and require high maintenance and operational costs. There exists a need to develop suitable teaching and learning aids to augment the classroom teaching. One of the most effective ways to achieve this is to develop simple experimental setup with suitable shake table. Development of shake table for the Earthquake Engineering laboratory to test models is a challenge. Single translation (horizontal) degree of freedom shaking tables is useful for laboratory laborator y testing to study behavior of structures. From this perspective, low cost uni-axial shaking tables were designed & fabricated at L.D College of Engineering. These low cost shake tables will be used to study behavior of structure through models under harmonic as well as random excitation. The cost of shake table is very high and it is difficult for the institutes to acquire such facilities. Based on this fact, an effort has been made to fabricate two low cost shake tables with required specifications to test models in Earthquake Engineering Laboratory along with a LVDT based instrumentation setup. The instrumentation instrumentat ion setup comprises of LVDT and Data Acquisition System. Response of models studied through shake table testing. Shake table with servo motor control & shake table with 1.0 HP motor is costing around Rs 3, 50,000/- & 80,000/respectively. The cost of instrumentation for such set up is only Rs 20,000/-. This effort will fulfill the basic need of the Earthquake Engineering laboratory in form of low cost shake table & required instrumentation, to study behavior of structure through models by shake table testing. KEYWORDS: Earthquake Engineering laboratory, shake tables, experimental study, simulation, Single degree of freedom, Accelerometer, Data acquisition, Uni-axial Shake table
1. INTRODUCTION The development in the field of earthquake engineering requires experimental study. Laboratory testing of components and structures as physical models is an effective way to study the complex phenomena. Correlation of results from laboratory experimentation and analytical modeling will increase the confidence of the researcher. A Shake table can be used to test the model of the structure which may be scaled or prototype to seismic shaking. 2. REQUIREMENT OF UNI-AXIAL SHAKE TABLE In order to reproduce actual earthquake data, a sixdegree of freedom shaking table is essential. Shake table is a very complex electro-hydraulic system, which is very expensive and requires high maintenance and operational costs. As per Clause: 6.1.1 IS 1893-2002 [3], the random earthquake ground motions, which cause the structure to vibrate, can be resolved in any three mutually perpendicular directions, i.e. two horizontal and one vertical direction. The predominant direction of shaking is usually horizontal. The vertical direction is ½ to 2/3 rd of the horizontal vibration. The self weight of structure, i.e. gravity loads, compensates the effect of vertical accelerations. Movement of shaking table means application of strong ground motions (accelerographs) to model of the structure to study their behavior. Simulation of earthquake ground motion in all six directions of consideration (i.e ±X, ±Y, ±Z) is complicated and costly. The effect of horizontal ground motion is significant on structure compared to the vertical motion which is almost 1/2 rd to 2/3 of the horizontal acceleration. Thus, ground IJAET/Vol.III/ Issue I/January-March, 2012/46-49
motion consideration is left to major two orthogonal horizontal directions. Seismic analysis of structure means to provide equivalent distributed lateral force acting at various lumped level of structure above ground based on the ground motion at that site. Thus, instead of getting into complex nature of analysis, the behavior of structure is analyzed when horizontal ground shaking occurs. Horizontal shaking of shake table is representing the horizontal shaking of ground. By changing the orientation of test model on shake table will give behavior and reading for the other orthogonal direction too. Thus, uni-axial shake table serves the purpose. It is always a challenge to develop a low cost shake table with good instrument set up up for Earthquake Earthquake Engineering Laboratory. Uniaxial shaking tables were designed and fabricated at the institute laboratory. One shake table is of 8 ft x 4 ft in size with amplitude variation of 0 to 100 mm and frequency varying from 0 to 4 Hz. Second shake table is of 5 ft x 3 ft in size with amplitude variation of 0 to 50 mm and frequency varying from 0 to 25 Hz. This two shake tables will satisfy the basic requirement for testing models falling in this frequency range. The instrumentation setup of this Shake table comprises of LVDT and Data Acquisition System. A single degree of freedom model was tested on this shake table to evaluate the performance of Shake Table and instrumentation setup. 3. SHAKING TABLES AT EARTHQUAKE ENGINEERING LABORATORY Shake tables prepared by L.D College of Engineering (L.D.C.E) are Uni-axial Electro-mechanical Shaking tables. These shaking tables are assembly of various
International Journal of Advanced Engineering T echnology steel sections that forms a table on which a plate is supported. The movement of this plate is considered as shaking of ground due to earthquake. The term Uniaxial means movement in one horizontal d irection only. Specifications of low frequency shake table Uni-axial motorized electro-mechanical Shake table Size 8 ft x 4 ft table platfor m for fixing model Operated manually and with motor as well Motor: Single-phase D.C motor with Electronic panel board, frequency range 0 to 1500 rpm. Amplitude range of Shake table is from 0 to 100 mm Harmonic and periodic simulation Frequency of simulation - 0 to 4 Hz Payload capacity : 600 kg
E-ISSN 0976-3945
Figure 2 Line diagram of Mechanism of Shake Table Thus, the excitation is harmonic in nature. The response of structure to harmonic excitation provides insight how the system will respond to other types of forces. 2500
0 5
87.5
612.5
550
550
612.5
87.5
5 7 5
0 5 2 1
A N G L E F R A M E
5 7 5
Figure 1 Electro-Mechanical low frequency Shaking Table Mechanism of low frequency shake table The Uni-axial shake table serves the purpose of Laboratory testing of models for earthquake simulation. The movement of L.D.C.E shake table is in one horizontal direction. The top plate is connected with a shaft (S). One end o f the shaft (S) is assembled with IS MC 70 35 5 block (B). This block (B) is welded with bottom of the plate. The other end of the shaft (S) in embedded into slider (L). The head of the slider (L) is grooved into the shafting (T). The distance of shaft end assembly from the zero mark of Shafting (T) groove decides the amplitude of shaking. The Shafting (T) is connected with an axle that has a pulley (P). The pulley (P) is rotated by motor attached with the belt. Thus, when the motor is operated, the belt rotates the pulley (P). This helps the axle to rotate and push the shaft forward and backward. The shaft is connected with top plate on the other end which moves the plate in horizontal direction. The movement is eased by provision of bearings block (R). Thus, the movement is smooth without any jerks, as the angles (A) connected to plate; slides on the bearing of bearing block (R). The plate clamp (M) also helps by providing vertical restraint to plate during higher frequency. The movement of plate i.e amplitude of shaking remains fixed during single instance of testing. The amount by which the shaft in the groove of Shafting (T) is shifted away from center, gives amplitude of shaking. Figure 2 shows the arrangement of the whole mechanism. The ratio of pulley (P) diameter to pulley diameter of motor is 3.0, i.e the frequency of 1500 rpm of motor is reduced to 500 rpm. The mechanism is such that at a particular instance of testing, the amplitude remains constant and the frequency can be regulated to provide variation in shaking.
IJAET/Vol.III/ Issue I/January-March, 2012/46-49
0 5
13mmDiacircle
50mmc/c
Fig. 3 Plan View of Top plate of Shake Table Showing arrangement for Model fixing. Specifications of high frequency shake table Uni-axial motorized electro-mechanical Shake table Size 5 ft x 3 ft table plat form for fixing model Operated with motor as well Motor: Servo motor with Electronic control panel, frequency range 0 to 9000 rpm. Amplitude range of Shake table is from 0 to 50 mm Harmonic and random motion simulation Frequency of simulation - 0 to 25 Hz Payload capacity : 500 kg
Figure 4 Electro-Mechanical high frequency Uni-axial Shaking Table Model is fixed on shaking table which represents structure fixed to the ground. For making this arrangement possible, the top plate is provided with number of holes of 12 mm diameter at 50mm c/c distance. This gives the flexibility of installing any size of model and also helps in changing the orientation of the model. This arrangement also helps in fixing two models at a time to study the comparative behavior of the same. Figure 3 shows the Plan view of Top plate of shake table.
International Journal of Advanced Engineering T echnology 4. INSTRUMENT SETUP Instrumentation setup is a data logging system which records measurements continuously, in a number of digital electrical devices. These data are required to be processed to give data in the required format. This data processing unit forms the data acquisition system. Based on availability of LVDT as a sensing unit, a 4 channel data acquisition system was developed. This includes a LVDT and a circuit that processes the data and provides in word format with data logging rate of average 15 samples per second. The hardware and software components used for this instrumentation setup are as follows: Hardware [6]: 1. Atmel AT89C52 Microcontroller [4], 2. ADC0808CCN, 3. Vibration Sensor (LVDT), 4. MAX 232, 5. TL084 Quad Op-Amp 6. MCT2E Optoisolator 7. Power Supply (-5V to +5 V): Power Supply (-15V to +15V) 8. Other circuit Components (resistors, capacitors etc...). Software: Keil Version 3, Visual Basic 6
Figure 5 Main circuit and hardware component for data processing 5. WORKING OF INSTRUMENTATION SETUP Three LVDT’s are considered for circuit design as sensing component. Here vibration sensor is nothing but a Linearly Varying Displacement Transducers is having a stroke length of 100 mm. the LVDT gives signal in range of 0 to 5 volts. This range is used for providing displacement result of 0 to 100 mm range. The circuit is designed to log 3 LVDT data and a frequency counter data. 4 ADC (analog to digital convertor) channels, 3 for the vibration sensor and one for the frequency are used. This sensor will give 0 to 5 volts for 0 to 100 mm displacement. The output of the sensors is given to 3 different channels of ADC. These pulses are given to the 2 nd order low pass filter to convert it into DC (direct current). T he output of the filter is given to the ADC. Now the channels of the ADC are rotated by the AT89C52 software. These channels are scanned by the software program. The scanning time depends on the Visual Basic (VB) software. Minimum scanning time is 1 ms. The output of the ADC is in hex. So it is converted into BCD (binary coded decimal) by the software. The appropriate range is set by the software itself. i.e. 0 to255 is converted into (-50 to +50 mm) for the sensors and (0 to 20Hz) for the frequency input. Data are transmitted serially to the computer’s serial terminal port using RS232 communication. This data will be handled by VB program and log file will be created for testing. Figure 6 shows software screen of VB program showing screening of data for sensor A.
IJAET/Vol.III/ Issue I/January-March, 2012/46-49
E-ISSN 0976-3945
Figure 6 Software screen of program developed in VB 6. TEST MODEL AND SETUP FOR PERFORMING EXPERIMENT: The single degree of freedom model, made from A304 stainless steel [1], was tested on uni-axial shaking table and response at top of model was measured using LVDT. To verify the results logged by LVDT, uniaxial accelerometer was also installed on model. Thus, the experimental setup consists of a SDoF model, LVDT and its circuit, uniaxial accelerometer and 16 channel vibration analyzer, supporting stand for LVDT and shake table. (See Figure 6). The model consists of a top plate 390 x 390 mm size supported by four 6mm square rods. The rods are fixed in bottom plate of 490 x 490 mm with check nuts. The model fabrication is done in such a way where check nuts are used instead of welding the model. This is done to avoid error during welding and to achieve better response of model. The natural frequency calculation for the model is given in Table I.
Figure 7. Experimental setup including model and instrumentation setup. Table: I Natural frequency calculation for as build model 1
2 3 4 5 6 7 8 9 10
11 12 13 14
Vertical rods a) width(x) b) Depth(z) c) Ixx d) Izz Top Plate thickness Density of steel A304 Elasticity E = Height of structure Lateral Stiffness K =4*(3EI/L^3) N/m Dimension in X-dir Dimension in Z-dir No of Columns Mass of Structure Top plate : Vertical rods: Extra for nuts/bolts: Total lumped mass m Omega ω ω = (K/m)^0.5 Frequency f f = ω / 2*Pie Natural period (T)
6 6 108 108 5 7407.41 210000 490
Mm Mm mm^4 mm^4 Mm Kg/m:^3 M pa Mm
2313.32 390 390 4
N/m Mm Mm Nos
5.63 0.64 0.28 6.55
Kg Kg Kg Kg
18.79
rad/s
2.99 0.334
Hz Sec
International Journal of Advanced Engineering T echnology 7. RESULTS OF MODEL TEST LOGGED BY LVDT: LVDT based instrumentation setup is capable of logging data in terms of displacement only. The average sampling rate is 15 S/s (Samples per second). The testing was performed for 47 seconds. Figure 8 shows the data logged by LVDT in graphical format as Displacement v/s Time. Samples in terms of 1/15 th of second for 720 samples on X axis and values of displacement logged for each 1/15th of second are plotted on Y. The maximum displacement is at 9 th second i.e at 115 th and 117th sample. The value is +34.8 mm and -35.6 mm respectively. The shake table amplitude during testing was set to 15mm. Thus, maximum relative displacement will be 35.6 mm less 15 mm to give peak value of displacement at 9th second as 20.6 mm. 8. RESULTS OF MODEL LOGGED BY ACCELEROMETER In experimental setup for model testing, 4 accelerometers were used. Acc-1 was mounted on top plate of shake table which provides an input motion value for experimental model. Acc-2 was mounted on top of experimental model to provide response of the
E-ISSN 0976-3945
model due to shaking. Acc-3 was mounted on top of LVDT stand to verify that the stand is stiff enough to provide correct results of LVDT . Acc-4 was mounted on base plate of model to verify that shake table shaking and base plate shaking are same. Figure 8 shows graphical representation of all four accelerometers [5]. As shown in figure 9, FFT curve [5] analyzed for Acc-2 i.e for accelerometer mounted on model to compare the results with LVDT results. The displacement values of peak displacement for Acc 1 to 4 obtained from FFT curve are as below: Peak displacement for Acc-1: 15.03 mm Peak displacement for Acc-2: 33.83 mm Peak displacement for Acc-3: 15.74 mm Peak displacement for Acc-4: 14.68 mm
From the above values we can say that Acc-1, Acc-3 and Acc-4 are almost having same displacement values. Thus, shake table amplitude, base plate of model and top of LVDT wooden stand all have same displacement value. As shown in figure 10, the displacement value of model top is 33.83 mm. Thus, relative displacement of model is 33.83 less 15.03 i.e 18.8 mm.
TIME (s) Figure 8 Graphical representation of Displacement v/s Time logged by LVDT
Figure 9 Graphical representation of as acceleration v/s Time logged by accelerometers
IJAET/Vol.III/ Issue I/January-March, 2012/46-49
International Journal of Advanced Engineering T echnology
E-ISSN 0976-3945
. Figure 10 FFT curve for Acc-2 i.e model response. Table:II Comparison of parameters of model testing on low frequency shake table Parameters Displacement, mm Natural freq. Hz
LVDT 20.6 2.99
Acc-2 18.8 3.15
% diff. w.r.t Acc-2 9.57 5.08
Table:III Comparison of parameters of model testing on high frequency shake table Parameters Displacement, mm Natural freq. Hz
LVDT 21.2 3.09
Maximum displacement of model sensed by LVDT and Acc-2 accelerometer are 20.6 mm and 18.8 mm respectively. The natural frequency calculated from the physical properties of model was 2.99 Hz and that sensed by accelerometer was 3.15 Hz. (see table II). Thus, the values of displacement and natural frequency are comparable and the nature of curve is also matching (see figure 8 & 9). 9. CONCLUSION ON EXPERIMENTAL STUDY: 1. The data sampling rate of accelerometer is 640 Samples/sec and the sampling rate of LVDT is 15 Samples/sec. With this sampling rate we are able to achieve acceptable results with maximum variation of 9.57%. The cost of LVDT based instrumentation setup is around Rs. 20000=00 only while that of vibration analyzer with accelerometer is 22 lacs Rs. 2. The cost of developing 8 ft x 4 ft size low frequency uni-axial shake table is Rs 100000=00 along with LVDT set up which is very simple, cost effective and yet provides acceptable results of displacement for the peak frequency of 3 Hz. 3. The cost of developing 5 ft x 3 ft size high frequency uni-axial shake table is Rs 370000=00 along with LVDT set up which is very simple, cost effective and yet provides acceptable results of displacement for the peak frequency of 25 Hz. 4. This low cost shake table and instrumentation setup is the first step for the development of earthquake engineering laboratory which will provide the vision to
IJAET/Vol.III/ Issue I/January-March, 2012/46-49
Acc-2 19.8 3.24
% diff. w.r.t Acc-2 6.42 4.62
experience the subject of dynamics practically. It will help: To develop understanding of dynamic response of structures to undergraduate students; To reinforce theoretical concepts through the use of “hands-on” laboratory experiments; To provide an opportunity to use modern engineering tools including sensors and data acquisition/analysis equipment. REFERENCES •
•
•
1.
2.
3.
4. 5. 6.
ASTM Designation A 276-87, Standard Specification for Stainless and Heat Resisting Steel Bars and Shapes. Harry G. Harris, Drexel University and Gajanan M. Sabnis, Howard University, “Structural modeling and Experimental Techniques”, second edition, CRC Press. IS 1893 (Part 1): 2002 “Criteria for Earthquake Resistant Design of Structures, Part-1 General Provisions and Buildings (Fifth Revision)”, Bureau of Indian Standards, New Delhi. Mohammad Ali Mazdi, “The 8051 Microcontroller and Embedded Systems”. OROS 3-Series/NVGate User’s Manual - for V3.10 – January 2006. www.alldatasheet.com
