Site: Yokohama National University
Division of Civil Engineering, Architecture & Marine Technology
156, Tokiwadai, Hodogaya-ku
Yokohama City, Japan
Date Visited: October 1996
Hosts: Professor Shoji Ikeda, Chairman Technical Committee, CFRRA, and Professor, Yokohama National University
Masayoshi Okoshi, Deputy Manager, Tonen Corporation
Takahiro Yamaguchi, Research Assistant, Yokohama University
Summary: Research on seismic retrofit strategies.


Professor Ikeda is a leading Japanese expert in the area of seismic design of reinforced concrete structures. He was one of the principle authors of the 1986 Standard Specification for Design and Construction of Concrete Structures, contributing Chapter 9, Seismic Design. He serves as the Chair of the Technical Committee of the CFRRA and is responsible for tests and for the review of draft guidelines and standards. Professor Ikeda has a close working relationship with the CFRAA manufacturing members, especially the Tonen Corporation, and serves as a technical advisor on CFRRA projects.


Due to litigation related to the failure of the Hanshin Expressway, Professor Ikeda is conducting tests on scale columns to assess the extent of damage and its location based on the position and presence of terminations in longitudinal reinforcement. Figure B.17 shows the damage accruing from the use of continuous bars on the outside in the longitudinal direction with terminating bars on the inside. This configuration leads to movement in the plastic hinge region from the base to the termination zone resulting in outward buckling of the continuous longitudinal reinforcement (Fig. B.18). The use of continuous bars without terminations (Fig. B.19) was seen to restrict the failure zone to the bottom hinge. Although the initial strength of the second configuration was lower than that of the first, it was considered to be a better design due to the failure mode. Current plans are to test the efficiency of carbon fiber sheet forms for the retrofit in the first case, and for strengthening in the second.

Fig. B.17. Failure mode of scale column with terminations.

Fig. B.18. Close-up of the termination zone after failure.

Fig. B.19. Failure mode of scale column with continuous longitudinal reinforcement.

A series of tests also were conducted on 25% scale rectangular bridge piers (having the same ratio of tensile reinforcement and lateral ties) built to pre-1980 specifications. Two aspect ratios of 1.5 and 3 were used (Fig. B.20) with all corners rounded to a 30 mm radius. Professor Ikeda felt that between 4 and 8 layers of carbon fiber sheet are needed in the circumferential direction and 1 to 2 layers are needed in the longitudinal direction. The longitudinal reinforcement (carbon fiber) is needed to prevent the circumferential sheets from being damaged by the flexural cracking of concrete (i.e., serving as a barrier to concrete spall induced carbon fiber breakage), and to develop a smooth stress transfer from the column to the circumferential carbon fiber reinforcement.

Fig. B.20. Specimen configuration for the aspect ratio tests.

Fig. B.21. Post-test damage state for the as-built column.

Based on static and pseudo-dynamic testing it was verified that carbon fiber sheets in the longitudinal-circumferential configuration provide an efficient means of retrofit with the capability of increasing the specific ductility level to 7. Despite such test results, the current code restricts the use of carbon fiber reinforcement to aspect ratios up to 1.5, with exceptions being made on a case-by-case basis. Damage in the specimens with a 1.5 aspect ratio without additional carbon fiber reinforcement, with one layer of longitudinal and one layer of transverse carbon fiber reinforcement, and with 2 layers of transverse reinforcement are shown in Figures B.22 and B.23, respectively. In the case of the carbon fiber sheet wrapped specimens, the layers were carefully peeled off after testing to observe damage. Yield displacement of the first specimen was 13.9 mm and failure occurred at 4 dy. In the case of the first retrofit configuration (Fig. B.22) swelling of the jacket began at 5 dy and failure occurred on the compressive cycle at 6 dy. In the case of the second retrofit configuration (Fig. B.23), no failure was seen until 7 cycles at 7 dy when overlapping regions in the sheets started to separate.

Fig. B.22. Post-test damage state for the specimen retrofitted with one layer of longitudinal and one layer of transverse carbon fiber reinforcement.

Fig. B.23. Post-test damage state for the specimen retrofitted with two layers of transverse carbon fiber reinforcement.

Current research is being conducted in three main areas:

  1. The use of a unique dynamic visualization technique to assess the effect of real time seismic events through the use of pseudo-dynamic testing combined with post-test computer controlled video recordings.
  2. The testing of large hollow columns with reinforcement terminations at a number of levels, retrofitted with longitudinally oriented carbon fiber sheets. The actual columns are 60 m high and 6 to 7 m in diameter and exist on the Tomei Expressway. Retrofit by steel jackets is considered uneconomical and technically unsound as well as logistically difficult due to the large weight of steel needed at the design height and diameter. It is envisaged that if the scale tests prove successful, the actual columns would be retrofit with carbon fiber with each column needing as much as 2.5 to 3.5 tons of carbon fiber.
  3. Testing the viability of a single layer of carbon fiber sheeting used in combination with post-tensioned aramid tendons positioned normal to the longer sides of the rectangular columns, in order to retrofit larger aspect ratio rectangular columns using composites.

Published: October 1998; WTEC Hyper-Librarian