Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1598-4095(Print)
ISSN : 2287-7401(Online)
Journal of The korean Association For Spatial Structures Vol.19 No.2 pp.63-72
DOI : https://doi.org/10.9712/KASS.2019.19.2.63

Comparison Between Performance of Wireless MEMS Sensors and an ICP Sensor With Earthquake-Input Ground Motions

S. T. Mapungwana*, Jong-Ho Lee**, Sung-Won Yoon***
*Dept. of Architecture, Seoul National University of Science and Technology
**Dept. of Architecture, Seoul National University of Science and Technology
교신저자, 정회원, 서울과학기술대학교 건축학부 교수, 공학박사 School of Architecture, Seoul National University of Science and Technology Tel: 02-970-6587 Fax: 02-979-6563 E-mail: swyoon@seoultech.ac.kr
April 24, 2019 May 15, 2019 May 15, 2019

Abstract


Wireless sensors are more favorable in measuring structural response compared to conventional sensors in terms of them being easier to use with no issues with cables and them being considerably cheaper. Previous tests have been conducted to analyze the performance of MEMS (Micro Electro Mechanical Systems) sensor in sinusoidal excitation tests. This paper analyzes the performance of in-built MEMS sensors in devices by comparing with an ICP sensor as the reference. Earthquake input amplitude excitation in shaking table tests was done. Results show that MEMS sensors are more accurate in measuring higher input amplitude measurements which range from 100gal to 250gal than at lower input amplitudes which range from 10gal to 50gal. This confirms the results obtained in previous sinusoidal tests. It was also seen that natural frequency results have lower error values which range from 0% to 3.92% in comparison to the response spectra results. This also confirms that in-built MEMS sensors in mobile devices are good at estimating natural frequency of structures. In addition, it was also seen that earthquake input amplitudes with more frequency contents (Gyeongju) had considerably higher error values than Pohang excitation tests which has less frequency contents.



지진 입력 진동대를 이용한 무선 MEMS 센서와 ICP 가속도계의 성능 비교

마 푼과나 시부시시웨*, 이 종 호**, 윤 성 원***
*학생회원, 서울과학기술대학교 건축과, 석사과정
**학생회원, 서울과학기술대학교 건축과, 박사과정

초록


    Seoul National University of Science and Technology

    1. Introduction

    Structural Health Monitoring of buildings during and after Earthquakes is important because it gives an understanding on the behavior of the structures under earthquake load. This provides knowledge on the state of the infrastructure that affects safety of people. However, SHM (Structural Health Monitoring) systems have practically and financially related challenges when it comes to instrumentation and implementing it. Traditional SHM systems are generally expensive due to the requirements of high sensitivity and accuracy of the sensors. As a result, MEMS sensors have been used in SHM systems because they are inexpensive and easy to install1).

    MEMS sensors can be used in estimating the response acceleration of the structures. A series of shaking table tests have been carried out to confirm the performance of in-built MEMS sensors in mobile devices. An application called Vibration App was used in measuring the vibrational response from sinusoidal input motions in shaking table tests2). The range of input amplitude was 0.01~1.0g whereas the range of input frequency was 0.5~10.0Hz. It was seen that as amplitude and frequency increased, the accuracy of MEMS sensors in comparison to the reference sensors also increased.

    The performance of MEMS sensors with earthquake–wave excitation tests have been conducted3). The acceleration response of the system was compared in terms of acceleration time history, frequency-domain plot and 5% damped-acceleration response spectra. The results only compared the responses for earthquake scaled at 200gal. Only the peak amplitudes from the time history graphs were considered in comparing the performance of the MEMS sensors to the reference sensor. The structural responses at lower peak amplitudes were not analyzed.

    Previous studies have not compared the peak amplitudes at low amplitude values in earthquake-wave excitation tests. This paper analyzes the performance of built-in MEMS sensors in earthquake-wave excitation tests by comparing the performance of the MEMS sensor to the performance of an ICP sensor for low and high peak amplitudes. The shaking table was excited by two earthquake excitations, Gyeongju Earthquake of magnitude 5.8ML on the Richter magnitude scale that occurred on 15 November 2017 and Pohang Earthquake that occurred on 12 September 2016 of magnitude 5.4ML on the Richter magnitude scale4).

    2. MEMS sensor properties

    Vibration App is a spectrum analyzer of vibration measurements that uses built in accelerometers and gyroscope inside an iPhone, iPod Touch, and iPad5). As shown in <Fig. 1>, real-time measurements can be performed in three axes directions of x, y, z (two-directional horizontal acceleration and one-direction vertical acceleration).

    An Application called Vibration App was used in measuring the Earthquake Response of the MEMs sensor. Details of the Vibration App have been explained2). The acceleration response spectra was obtained by using Matlab. Damping ratio of 5% was considered because it is adequate in considering the damage of a structure due to seismic events.

    3. Earthquake-wave excitation in shaking table tests

    Shaking-table tests with Gyeongju and Pohang excitation were carried out. By considering the North South (NS) and East West (EW) directions of the two earthquakes, a total of four tests were carried out as shown in <Table 1>.

    <Fig. 2> shows the equipment used. <Table 2> shows the specifications of the shaking table and details of the dynamic signal analyzer used. The m+p Vibpilot was used in this experiment. It is used in vibration controlling, dynamic signal analysis and data acquisition applications7).

    3.1 Sensors used

    An ICP sensor was used as the reference sensor to compare results with the MEMS sensor. The ICP sensor interface with data acquisition, signal analysis and recording instruments. Regardless of the cable type used, these sensors have a fixed voltage sensitivity. In addition, they have low noise, voltage-output signal compatible with standard readout, signal analysis, recording and data acquisition equipment10).

    Four mobile devices with built-in MEMS sensors were used. The MEMS sensor is ultra-small, triaxial, low-g acceleration sensor with digital interfaces, aiming for low-power consumer electronics applications. It allows very low-noise measurement of accelerations in 3 perpendicular axes. In addition, it senses tilt, motion, shock and vibration in cellular phones, computer peripherals, man-machine interfaces, virtual reality features and game controllers11). <Table 3> shows a summary of the characteristics of the sensors used.

    3.2 Experimental setup

    <Fig. 3> shows layout of the specimen on the shaking-table test. Five specimen were used for this test. The APS 400 uniaxial shaking table was used to produce sin earthquake-wave excitation for the experiment and the m+p Vibpilot was used in this experiment for vibration controlling, data acquisition and dynamic signal analysis. A series of 4 tests with Gyeongju and Pohang excitations were carried out as described earlier. Input sampling rate of the earthquake data was 100Hz and the sampling frequency for the MEMS sensors was set to 100Hz. The time duration for each measurement was 40 seconds. The input amplitude from the earthquake data was not re-scaled.

    4. Shaking-table test results

    Performance of the MEMS sensors was analyzed by comparing the time series, fourier amplitude and response spectra with the reference sensor.

    4.1 Time series

    Time history curves were plotted for the two earthquake-input amplitudes and the higher and lower peak values were analyzed.

    Time response of the sensors were compared as shown in <Fig. 4> and <Fig. 5>. The time history response also shows zoomed in curves for the highest peak and lowest peak values to have a clearer image of the comparisons. It can be seen from <Fig. 4> and <Fig. 5> that the MEMS sensors have considerably lower amplitude values in comparison to the reference sensor at lower peak accelerations than at higher peak amplitudes.

    There is also an observed time lag in time history curve for the mobile MEMS sensors in comparison to the ICP sensor. This might be due to the deviation in sampling frequency from the actual frequency of 100Hz.

    4.2 Fourier amplitude

    Fourier amplitude helps in obtaining the natural frequency of a system and the power or energy of the structure with respect to frequency contents. Fourier amplitude were plotted for the two earthquake-input amplitudes as shown in <Fig. 6> and <Fig. 7>. It can be seen from <Fig. 7> that the Gyeongju Earthquake has various frequency contents. The performance of the MEMS sensors is generally similar to the reference sensor in fourier amplitude plots.

    4.3 Acceleration response spectra

    An acceleration response spectrum is a plot of the peak acceleration response of a series of oscillators of a variety of natural frequency that are forced into motion by one shock13). It is important to consider earthquake response spectrum in structures because seismic analysis of structures depend on the frequency content of the ground motion2). Response Spectra were plotted for the two earthquake-input amplitudes as shown in <Fig. 8> and <Fig. 9>. Reference sensors have considerably higher response acceleration values than the MEMS sensors.

    Two distinct peaks were observed for Pohang Response Spectra as shown in <Fig. 8> at time period of about 0.1s and 5.5s. This means that vibrations at frequencies (1/0.1s) and (1/5.5s), would lead to very high acceleration responses so these frequency values should be considered when designing the structures in SHM.

    5. Analysis of shaking-table test results

    5.1 Analysis of error values in time history curves

    As explained earlier, lower peak amplitudes of range 10gal to 50gal and higher peak amplitudes above 100gal were considered in data analysis.

    To get a clearer analysis of the accuracy of the MEMS sensors in comparison to the reference sensor, the acceleration ratio, (α) was calculated using equation (1), where Amems is the average peak amplitude for each MEMS sensor and Aref the average peak amplitude for the reference sensor.

    α = A m e m s A r e f
    (1)

    <Fig. 10> and <Fig. 11> show the acceleration ratio values at lower acceleration values and higher acceleration values. At lower acceleration values, acceleration ratio is not close to one but at higher acceleration, the acceleration ratio is closer to one. This confirms the previous study results in shaking table tests that MEMS sensors are more accurate at higher amplitude values2).

    5.2 Analysis of error values in fourier amplitude

    To analyze the accuracy of the MEMS sensors in comparison to the reference sensor, in terms of natural frequency, the error values (ψf) was calculated using equation (2), where Fmems is the average natural frequency obtained from the fourier amplitude graphs for each MEMS sensor and Fref the average peak acceleration amplitude for the reference sensor. <Table 4> shows average natural frequency values for each sensor and their corresponding error values.

    ψ f = F r e f F m e m s F r e f × 100
    (2)

    The error values are considerably lower, ranging from 0% to 2.99% except for two measurements from iPhone 1 and iPhone 2, which were 3.92% and –3.36% respectively.

    5.3 Analysis of error values in response spectra

    To analyze the accuracy of the MEMS sensors in comparison to the reference sensor, the error values (ψa) was calculated using equation (3), where RAmems is the average peak response acceleration of each MEMS sensor and RAref the average peak response acceleration amplitude for the reference sensor. <Table 4> shows average response acceleration for each sensor and their corresponding error values. The error values in response acceleration are considerably higher, ranging from 0.45% to 6.38%. This might be due to the limitations of MEMS sensor in accurately measuring input accelerations with various frequency contents.

    ψ a = R A m e m s R A r e f A r e f × 100
    (3)

    6. Conclusion

    From this experiment, the performance of the in-built MEMS sensors in mobile devices was confirmed by a series of shaking table test with earthquake input amplitudes. It was seen that MEMS sensors are more accurate in measuring higher input amplitude values, for ranges 100gal to 250gal than at lower amplitude values, for ranges from 10gal to 50gal. This confirms the previous studies that have been carried out in sinusoidal excitation tests. In addition, both MEMS and the reference sensor obtained lower error values of ranges 0% to 3.36% in natural frequency results whereas they obtained considerably higher error values of ranges 0.45% to 6.38% for the response spectra.

    Although the results of response spectra in terms of response acceleration were not that accurate due to high error values, it can be seen from the graphs that MEMS sensors are capable of measuring response spectra because the shape of the trend of the graphs is similar to that of the reference sensor and the error values in terms of period were very low.

    In addition, it was seen that error values for Gyeongju earthquake were generally higher than that of Pohang. This might be due to the limitations of MEMS sensor to accurately measure input amplitudes with various frequency contents. Another limitation in using MEMS sensors is that they are less accurate at lower input amplitudes, that is amplitudes below 10gal. The maximum sampling frequency that can be set for built-in MEMS sensors is 100Hz which limits their capability to take measurements at higher sampling frequencies

    Overall, the averaged error values for all MEMS sensors ranging from 0.45% to 4.89% are considerably low. This shows that MEMS sensors are generally capable of being used in earthquake measurements.

    Acknowledgement

    This paper was supported by scholarship funding from Seoul National University of Science and Technology. The writers are grateful to the authorities for their support.

    Figure

    KASS-19-2-63_F1.gif

    iPad Mini 4 axial direction6)

    KASS-19-2-63_F2.gif

    Equipments used (APS 1459) and VibPilot7))

    KASS-19-2-63_F3.gif

    Experimental setup (APS 400 with specimen used)

    KASS-19-2-63_F4.gif

    Pohang time history

    KASS-19-2-63_F5.gif

    Gyeongju time history

    KASS-19-2-63_F6.gif

    Pohang fourier amplitude

    KASS-19-2-63_F7.gif

    Gyeongju fourier amplitude

    KASS-19-2-63_F8.gif

    Pohang response spectrum

    KASS-19-2-63_F9.gif

    Gyeongju response spectrum

    KASS-19-2-63_F10.gif

    Peak acceleration ratio for Pohang

    KASS-19-2-63_F11.gif

    Peak acceleration ratio for Gyeongju

    Table

    Shaking table experiment overview

    Apparatus specifications8),9)

    List of laboratory equipments3),10),11),12)

    Error calculations for natural frequency results

    Reference

    1. Girolami, A., Brunelli, D., & Benini, L. (2017). Low-cost and distributed health monitoring system for critical buildings. Proceedings of the 2017 IEEE Workshop on Environmental, Energy, and Structural Monitoring Systems (EESMS), Italy
    2. Mapungwana, S., Jung, Y. S., Lee, J. H., & Yoon, S. W., "Comparison Between Performance of a Wireless MEMS Sensor and an ICP Sensor in Shaking Table Tests", Journal of Korean Association for Spatial Structures, Vol.18, No.4, pp.49~59, 2018
    3. Shrestha, A., Dang, J., & Wang, X., "Development of a smart‐device‐based vibration-measurement system: Effectiveness examination and application cases to existing structure", Structural Control Health Monitoring, Vol.25, No.3, 2017
    4. Kim, H. S., Sun, C. G., & Cho, H. I., "Geospatial Assessment of the Post- Earthquake Hazard of the 2017 Pohang Earthquake Considering Seismic Site Effects", International Journal of Geo- Information, Vol.7, No.9, 2018
    5. Vibration Version 2 [Software]. (2013). Diffraction Limited Design LLC. Retrieved from http://www.dld-llc.com
    6. iPad Mini 4 [Apparatus and software]. (2014). Apple Inc. Retrieved from https://web.archive.org/web/20141016183214/http://www.apple.com/ipad-mini-3/
    7. Vibration Controller and Dynamic Signal Analyzer (Vibpilot). [Apparatus]. (2019). Retrieved from https://www.mpihome.com/files/pdf/vibpilot_us.pdf
    8. APS 400 ELECTRO- SEIS [Apparatus]. (2018). APS Dynamics Inc. Retrieved from https://sensores-de-medida.es/catalogo/generador-de-vibraciones-electrodinamico-aps-400/
    9. APS 145 Amplifier [Apparatus]. (2018). APS Dynamics Inc. Retrieved from https://apsdynamics.com/en/products/power-amplifier.html
    10. ICP Sensor [Apparatus]. PCB Piezotronics MTS Systems Corporation. Retrieved from http://www.pcb.com/products/model/333b50
    11. Bosch Sensortec [Apparatus]. (2018). Bosch. Retrieved from https://www.bosch-sensortec.com/bst/products/all_products/bma280
    12. Clover, J. (2014, September 26). iPhone 6 and 6 Plus Equipped With Two Accelerometers for Power Management, Improved User Experience [Web log post]. Retrieved from https://www.macrumors.com/2014/09/26/iphone-6-6-plus-two-accelerometers/
    13. Inman, D. J., "Engineering Vibration Response", 4th Edition, Pearson, pp.1~697, 2013.