DOI QR코드

DOI QR Code

Calculation of Maximum Effective Temperature of Steel Box Girder Bridge Using Artificial Neural Network

인공신경망을 이용한 강박스거더의 유효온도 산정

  • 이성행 (부산대학교 토목공학과)
  • Received : 2018.01.04
  • Accepted : 2018.03.09
  • Published : 2018.03.31

Abstract

An analysis using a statistical method is generally used to determine the effective temperature based on the temperature design load of a bridge. In this study, the effective temperature was calculated by building an artificial neural network (ANN) capable of improving the statistical method. A Steel box girder bridge specimen was made with a width of 2.0 m, height of 2.0 m, and length of 3.0 m and 0.2 m the upper slab. Twenty one temperature gauges were attached to measure the temperature between 2014 and 2016 for three years. An ANN was learned using the data measured from 2014~2015 and the results were compared with the Euro codes. The error rate between the Euro code and statistical analysis values was analyzed to be 4.1 % for the total measurement point. The ANN was verified and the effective bridge temperatures were calculated using the temperature data measured in 2016. The results revealed an approximate 3.97 % difference from the statistical analysis values. This degree of error is considered to be acceptable in terms of engineering for the analysis of an ANN. An ANN can easily predict the effective temperature of a bridge by knowing the input values of the region's highest temperature, bridge type, and upper asphalt thickness when designing the bridge's temperature loads.

교량의 온도 설계하중의 기준이 되는 유효온도는 통계적 방법에 의한 분석이 일반적으로 사용된다. 본 연구에서는 통계적 방법을 개선할 수 있는 인공신경망을 구축하여 유효온도를 산정하였다. 강상자형교 시험체를 폭 2.0m, 높이 2.0m, 길이 3.0m, 상부슬래브 0.2m로 제작하였다. 21개의 온도 게이지를 부착하여 3년간(2014~2015) 온도를 측정하였다. 2014~2015년 측정된 온도데이터를 바탕으로 인공신경망을 학습시키고, 그 결과를 Euro code와 비교하였다. Euro code와 통계분석값과의 오차율은 전체 측점에 대하여 4.1%로 분석되었다. 2016년 측정된 온도데이터를 이용하여 인공신경망을 검증하고, 교량 유효온도를 산정하였다. 이 결과는 통계분석 값과 약 3.97%의 차이를 보였다. 이 정도의 오차율은 인공신경망에 의한 분석이 공학적인 측면에서 수용할 수 있는 크기인 것으로 판단된다. 인공신경망은 교량의 온도하중 설계 시 그 지역의 최고 대기온도, 교량 형식, 상부 아스팔트 두께 등 입력 값만 알면 교량의 유효온도를 간편히 예측해 줄 수 있다.

Keywords

References

  1. Lee, S. H., Shin, H. K., Kim, K. N., Jung, k. S., An Experimental Study for Estimation of effective temperature for design in Steel Box Girder Bridge, Journal of korean Society of Steel Construction, Vol. 28, no. 6, pp. 449-458, 2016. https://doi.org/10.7781/kjoss.2016.28.6.449
  2. Lee, S. H., Evaluation of Maximum Effective Temperature for Estimate Design Thermal Loads in Steel Box Girder Bridges, Journal of the korea Academia-Industrial cooperation Society, vol. 17, no. 8, pp. 133-138, 2016. DOI: http://dx.doi.org/10.5762/KAIS.2016.17.8.133
  3. Lee, S. H., Study of the Temperature Difference between the Top and Bottom Web of Steel Box Girder without Concrete Slab by using Gauge Measurement, Journal of the Korea Academia -Industrial cooperation Society, Vol. 15, no. 12, pp. 7350-7356, 2014 https://doi.org/10.5762/KAIS.2014.15.12.7350
  4. British Standards Institution, Steel, Concrete and Composite Bridges, Part 2. Specification for loads, appendix E. BS5400: Part2:1978, pp. 20-23, 2004.
  5. Lee, J. H., Kim, S. G., Chang, S. P., Frequency Domain Pattern Recognition Method for Damage Detection of a Steel Bridge, Journal of Korean Society of Steel Construction, vol. 17, no. 1, pp. 1-12, 2005.
  6. Choi, B. H., Lim, J. H., Design of Steel Structures Using the Neural Networks with Improved Learning, Journal of Korean Society of Steel Construction, vol. 17, no. 6, pp. 661-671, 2005.
  7. Park, M. S., Jo, B. W., Lee, J. W., Kim, S. K., Development of Steel Composite Cable Stayed Bridge Weigh-in-Motion System using Artificial Neural Network, Korean Society of Civil Engineers, vol. 28, no. 6A, pp. 799-808, 2008.
  8. Kim, K. N., Lee, S. H., Jung, K. S., Prediction on the Fatigue life of Butt-Welded Specimens Using Artificial Neural Network, International Journal of Steel and Composite Structures, vol. 9, no. 6, pp. 557-568, 2009. DOI: https://doi.org/10.12989/scs.2009.9.6.557
  9. Yoo, C. S., Jung, K. S., Estimation of Area Average Rainfall Amount and Its Error, Journal of Korea Water Resources Association, KCI 1738-9488, vol. 34, no. 4, pp. 317-326, 2005.