PURPOSES : During the summer of 2018, a heat wave (temperatures > 33°C) lasted for more than 30 days, causing blow-ups at eight different locations in South Korea. The blow-up phenomenon occurred when the internal temperature of the concrete slab increased. Simultaneously, as the concrete slab expands excessively, the length of the end of the slab increases, thus resulting in a lateral compressive force; when the slab cannot withstand this force, it rises or breaks. Blow-up is caused by a variety of factors, including increased temperature and humidity, accumulation of incompressible substances inside discontinuous surfaces, alkali–silica reactions, and aging of the concrete pavement. Several researchers have presented models to forecast blow-ups, such as the A. D. Kerr and G. Yang models, which have been applied to explain the blow-up phenomenon. However, this model has some limitations. This paper discusses a method to overcome these limitations.
METHODS : Buckling, the most important theory describing the blow-up phenomenon, was reviewed, and the buckling principle was confirmed. Subsequently, the input variables of the Kerr and Yang models and the mechanism for predicting the occurrence of blow-ups were identified. The PGBA program was used to confirm the lifetime of the expansion joint and the blow-up occurrence time based on the expansion joint spacing to review the limitations of the two studied models.
RESULTS : The Kerr and Yang models did not consider variables such as the expansion joint spacing or length of the integrated adjacent slab. In other words, it is necessary to reconsider the appropriateness of blow-up time predictions in relation to changes in expansion joint spacing and slab length. The expansion joint lifetime and blow-up occurrence time were predicted using the PGBA program. It was confirmed that as the expansion joint spacing increases, the expansion joint lifetime decreases. However, the blow-up occurrence time was shown to be the same (equal to 59 years), which is a limitation of the Kerr and Yang models used in the PGBA program. This resulted in a limitation in which variables for the expansion joint spacing cannot be used.
CONCLUSIONS : Through blow-up simulation experiments and actual field data, an appropriate slab length should be determined, and a blow-up model should be developed based on the slab length. If a blow-up prediction based on concrete slab length and a blow-up model based on are developed, the blow-up prevention technology will be applied to the appropriate blow-up time and location to avoid traffic accidents and reduce human and property damage.

ABSTRACT
1. 서론
2. 블로우업 관련 이론 및 모델 고찰
    2.1. 좌굴(Buckling) 이론
    2.2. 블로우업 예측 모델 검토
    2.3. A. D. Kerr 모델
    2.4 G. Yang 모델
3. 블로우업 모델 한계점
    3.1. A. D. Kerr 모델의 한계점 검토
    3.2. G. Yang 모델의 한계점 검토
4. 블로우업 예측 모델의 한계점 검증
    4.1. PGBA(Pavement Growth and Blow-up Analysis)프로그램 개요
    4.2. PGBA(Pavement Growth and Blow-up Analysis)프로그램을 통한 블로우업 모델의 한계점 검토
5. 결론
REFERENCES

• 안희락(강릉원주대학교 토목공학과 박사과정) | Ahn Hui Rak
• 김영규(강릉원주대학교 방재연구소 연구교수 · 공학박사) | Kim Young Kyu
• 이승우(강릉원주대학교 토목공학과 교수 · 공학박사) | Lee Seung Woo Corresponding author