Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 2093-5145(Print)
ISSN : 2288-0232(Online)
Journal of the Korean Society for Advanced Composite Structures Vol.13 No.2 pp.43-51

Structural Performance of Deteriorated Concrete Culverts Rehabilitated by Ultra High Performance Concrete

Mai Viet-Chinh1, Han Sang Mook2
1Ph.D. Student, Department of Civil Engineering, Kumoh National Institute of Technology, Gumi, Korea
2Professor, Department of Civil Engineering, Kumoh National Institute of Technology, Gumi, Korea

Corresponding author: Han, Sang Mook Department of Civil Engineering, Kumoh National Institute of Technology, Gumi, Korea. Tel: +82-10-2988-7614, Fax: +82-54-478-7710 E-mail:
March 21, 2022 April 11, 2022 April 17, 2022


Reinforced concrete culverts are widely used due to their economical advantages and durability. After the prolonged use, concrete culvets are deteriorated by the acid compounds of wastewater in combination with service loading. This paper investigates the structural performance of deteriorated concrete culverts rehabilitated by Ultra High Performance Concrete (UHPC) technology. Test variables of the specimens are the thickness of the upper slab and wall of the UHPC culvert. Four kinds of UHPC culverts are monitored to investigate the loading capacity until the ultimate loads. Compared to normal concrete cross sections, the relatively small thickness of UHPC culverts minimizes the loss of rehabilitated section by chipping the deteriorated internal parts. The test results indicate that the new thin UHPC structures utilized for rehabilitating deteriorated concrete culvert fulfill the service load and demonstrate the feasibility in real engineering applications.

초고성능 콘크리트로 갱생한 열화된 콘크리트 하수관거의 구조적 성능

비엣 징 마이1, 한상묵2
1금오공과대학교 토목공학과 박사과정
2금오공과대학교 토목공학과 교수


철근 콘크리트 하수관거는 경제적 장점과 내구성으로 인해 많이 사용되고 있다. 오랜 공용시간 이후 하수관거는 하 수의 산화물과 공용하중에 의해 열화가 진행된다. 본 논문은 초고강도 콘크리트로 갱생한 열화된 콘크리트 하수관거의 구조적 성능에 대해 연구하였다. 실험변수는 초고강도 콘크리트 관거 상부슬래브와 벽체의 두께의 조합으로 구성하였다. 네가지 종류의 초고강도 콘크리트 관거가 극한하중을 받을 때까지 내하력을 파악하기 위해 실험을 진행하였다. 보통 강도 콘크리트에 비해 상 대적으로 작은 두께의 초고강도 콘크리트 관거는 열화된 내부 부분을 치핑하고 갱생함으로써 단면의 손실을 최소화할 수 있다. 실험결과에 의하면 열화된 콘크리트 관거를 갱생하는데 적용된 새로운 얇은 두께의 초고강도 콘크리트 구조물이 재하하중을 만 족하고 실제 공학적 적용에 용이성을 있음을 나타내고 있다.


    Rectangular reinforced concrete culverts are utilized as sewage construction or passageways over roads in expressway construction since the mid-19th Century (Chung et al., 2003). After the long lifetime, many of these structures are damaged which requires expensive replacements or alternatively rehabilitation processes. In some situations, the repair and rehabilitation of these structures is the optimal solution since it minimizes the economic losses caused by transportation network disruption and user delays. Some researches have been carried out to investigate effective solutions for rehabilitating damaged reinforced concrete culverts. In the USA, concrete invert liner is one of the most popular reinforced concrete culvert rehabilitation methods. This can be constructed by shotcrete or cast-in-place concrete. D5 steel reinforcements are placed before pouring concrete in order to strengthen the loading capacity of the rehabilitation culvert. The cost-effective solution to rehabilitate large size culverts is an advantage of this method. However, the rehabilitation structure has a short life expectancy (Hartley, 2014). Geospray application is also a popular method for rehabilitating box culverts of larger dimensions which are up to 2.6m. This method consists in applying a minimum 3.9cm of geopolymer mortar to recover and protect the damaged area of the culvert as well as fill in voids, cracks (David Keaffaber, 2019). In Japan, the SPR (Sewage Pipe Renewal) method is commonly utilized to rehabilitate old culvert systems (Sekisui, 2014). In Europe, culvert rehabilitation methods like cured-in-place patch, lining by sprayed, trowed or cast-in-place material or repair by injection as mentioned in the standard BS EN 15885:2010, are popular (Almeida et al., 2015;BS EN 15885:2010, 2010).

    Based on the above brief literature review, the findings of these studies suggest that in many cases, rehabilitating the damaged structure may be more effective in direct monetary costs and downtime than constructing new structures (Simpson et al., 2016;Meegoda & Zou, 2015). An important goal which needs to be achieved after the rehabilitation process is the high loading capacity and durability. Conventional rehabilitation methods cannot guarantee the long lifetime of rehabilitated structures. The objective of the present article is to investigate the loading capacity of deteriorated reinforced culverts rehabilitated with the UHPC, a material possess superior characteristics in mechanical property and high durability (Kim et al., 2012). Thin thicknesses of the upper slab and wall of UHPC rehabilitated culverts minimize the loss of section due to chipping or grooving. The results obtained from this study open up a new approach of using UHPC material in combination with the aluminum formwork in rehabilitating deteriorated culverts which can potentially bring immense benefits.


    It is impractical to repair old reinforced concrete culvert systems placed underground in the big cities by open cut method. In this research, we developed an aluminum formwork system, which can be installed in a narrow space.

    Fig. 1 shows the aluminum formwork used in this research. It contains small aluminum formwork plates which can be transported inside the closed area of old culvert system and connected together. Before installing the formwork, the surface of old culverts can be chipped to ensure strong bond between the old surface and the UHPC layer. Horizontal and vertical reinforcements are also inserted to enhance the loading capacity of the structure. Then, UHPC material which has good flow can be poured into this section without any clogging (Fig. 2). This method is suitable for a narrow space. It reduces the increment of the cross section of rehabilitation structures, and provides high durability due to the dense structure. This under ground retrofitting method also does not interfere with the flow of traffic on the road. Apparently, thin thicknesses of upper slab and wall of UHPC rehabilitated culverts result in minimizing the loss of section by the chipping or grooving.


    3.1 Material Property

    The components of the UHPC mix are shown in Table 1, including cement, water, binder, silica flour, fine sand, and superplasticizer, in which the water/binder ratio is 0.2. High strength steel fibers with a volume fraction of 1.5% are distributed evenly to improve the strength and ductility of UHPC. Readers are referred to the research (Jang & Han, 2021) for properties of fibers. Compressive strength tests of UHPC specimens were conducted using the machine with capacity of 2,000kN. The average compressive strength of all specimens is 130MPa.

    The steel type of SD400 with 19mm diameter is used for reinforcement in the UHPC culvert. Steel bars are characterized with the yield stress of 400MPa, the Young modulus of 200.000MPa and the maximal elongation of 25%. The steel bars are placed in the upper slab, wall and lower slab of UHPC culvert. Fig 3 shows the reinforcements and cross section of the rehabilitated UHPC culvert T10W7. Layers of the reinforced steel bar in culverts T10W10, T7W7, and T7W5 are similar to the specimen T10W7.

    Table 2 summarizes the configuration of the rehabilitated UHPC culverts. T10W10 specimen has the overall size 2mx1.5mx1m, while the other specimens are of the same overall size 2mx1.7mx1m. The cross section of specimens differs in the thickness of the upper slab and the wall. For instance, T10W7 means the thicknesses of the upper slab and, the wall are 100mm and 70mm, respectively.

    3.2 Testing setup program

    UHPC culvert tests are carried out by four-point bending, as shown in Fig. 4 under an electro-hydraulic servo pressure testing machine. This machine has a loading capacity of 1,000kN. The two applied loads are transferred by two steel beams which are located at mid segment with 0.6m distance. During the test, the loading was applied with an increment of 0.05mm/sec until the failure occurred. The load produced by the actuator and the displacement of the actuator are measured continuously throughout the entire test.

    3.3 Test Result

    3.3.1 Load-deflection Curve

    Load-deflection curves of the UHPC culvert tests are shown in Fig. 5. It can be seen that each curve contains three stages. The first stage starts at 0 and finishes at point Ai (A, A1, A2, A3). The second stage is from point Ai (A, A1, A2, A3) to Bi (B, B1, B2, B3) and the third stage is from point Bi (B1, B2, B3) to Ci(C1, C2, B3). In the first stage, P-U curves exhibit essentially the un-cracked elastic behavior. This is expected since they have properties similar to conventional reinforced concrete. The flexural loading capacity of the culvert depends on the un-cracked UHPC and steel bars, leading to a linear load– deflection curve with a constant slope. As the applied load exceeds elastic stage (points A, A1, A2, A3), some micro-cracks occur in the inner surface of culvert. Initial cracks in culvert T10W10, T10W7, T7W7, T7W5 are observed at the load of 349kN, 295kN, 155kN and 95kN, respectively, which correspond to the deflection of 16mm, 32mm, 22.5mm and 21.3mm in Fig. 5. Then, the number and length of cracks gradually increase. In the linear stage, the loading capacity of T10W7 (A1-295kN) is remarkably larger (190%) than that of T7W7 (A2-155kN). The loading capacity of T7W7 (A2-155kN) is 163% larger than T7W5 (A3-95kN). These results indicate the significant influence of the thicknesses of the upper slab and the wall on the loading capacity of UHPC culverts.

    In the second stage, load-deflection curves enter into the nonlinear stage, where the strain-hardening behavior starts and multiple cracks propagate in the UHPC culvert. Visible cracks are observed in the mid-span of the upper slab and the outer face of the wall. The slope of load–deflection curves in Fig. 5 shows decreasing trend corresponding to the reduction in the bending stiffness of the UHPC culvert. As the testing load increases to the value of B (600kN), B1 (370kN), B2 (280kN), B3 (206kN), load-deflection curves in Fig. 5 enter the third stage of yield strengthening. In this stage, load–deflection curves show plasticity behavior. The testing load increases slowly while the value of the deflection increases considerably at a higher rate. Furthermore, some cracks in this stage propagate rapidly, following an increment in the crack width. When the load in the curves of T10W7, T7W7, and T7W5 specimens reaches to the point C1, C2, C3, the UHPC culvert fail due to the crushing of the concrete material. In the test of T10W10 specimen, as the testing load increases to 600kN, it remains almost unchanged. The value of 600kN can be considered as the maximum loading capacity of T10W10 culvert. Peak load of T10W10 UHPC culvert is therefore 600kN, which is 152%, 193% and 283% larger than that of T10W7 (393kN), T7W7 (310kN) and, T7W5 (212kN), respectively. The width of UHPC culverts in this test is 1m. The loading capacity of the specimens T10W10, T10W7, T7W7 and T7W5 in the elastic stage is 349kN, 295kN, 155kN, and 95kN, respectively. In the case of a 12m length of truck and culvert with dead load and live load, the loading capacity of 4,188kN, 3,540kN, 1,860kN, and 1,140kN can be obtained. This rehabilitation method can provide structures with sufficient loading capacity to resist service loading.

    3.3.2 Failure Mode and Cracking Pattern

    Figs 6, 7, 8 and 9 show the crack distribution and final damage state of UHPC culverts after the test. It is observed that the failure mode of all culvert tests is the same and it is typically that of a flexural failure. Based on Figs 6, 7, 8 and 9, less cracks were observed in the surface of the upper slab and the wall of T10W10 in comparison with the other culverts. No crack was found in the inner surface of the wall for all specimens. The crack that developed at the mid-span section of the upper slab was deeper than cracks in other areas. In addition, no sign of failure due to the buckling in the wall of UHPC culvert was observed, even in the T7W5 specimen with 50mm thickness of wall. This result indicates that all UHPC culverts possess sufficient capacity to sustain testing loads for the examined configuration.

    The crack development for the UHPC culvert test can generally be divided into three stages corresponding to three stages of load-deflection curves in Fig. 5. The first stage is defined from the beginning to the initial cracking. The initial cracks were observed in culvert specimens at points Ai of load-deflection curves (Fig. 5). The second stage corresponds to the section from Ai to Bi in the load-deflection curves. This stage is characterized by the rapid development of numerous cracks as the applied load increases. The visible cracks also extended through the height of the culvert and gradually penetrated into the cross section of the upper slab. Larger number of hairline cracks were observed at the mid-span tension zone and the intersection area between the wall and the upper slab. The general direction of the cracks is perpendicularly from bottom to the top of the side face. In the third stage, when the neutral axis of the cross section of UHPC culverts rises upward due to the increment of applied load at point Ci in Fig. 5, cracks at the mid-span of the upper slab and in the intersection area develop to the failure state. The main failure state of UHPC culvert includes the yield effect of steel bars in the tensile zone and the crushing effect of UHPC element in the compressive zone.

    3.3.3 Strain Test Result

    Electrical resistance strain gauges were used to capture the strain behavior on the surface of the culvert. The strain gauge type is PL-60-11-1L with a length of 60mm and the gauge factor is 2.07±1%. The concrete strain gauges were used in this study to measure compressive and tensile strains.

    Figs 10a and 10b show the strain gauges location and strain value of T10W7 specimen. Table 3 summarizes strain values of some locations in the T10W7 culvert. The chosen locations in Table 3 exhibit larger strain values than elsewhere. Only at the position of gauge 3 we recorded a minimum strain value of -780με in compression. The other gauges recorded positive strain values. This indicates that the neutral axis of the wall section is located inside the center line, near the inner face of the wall. Positive strain values at position 2 and 5 are larger than at the position 4. The maximal strain value observed at the position 5 is 1543με, corresponding to the maximum load of 393kN. The strain value at positions 2, 4, and 5 shows that positive bending moment in the outer face of the wall is larger than positive bending moment in the middle of upper slab.

    In order to further understand the strain of concrete element at other locations of the wall, in the test of T7W7 and T7W5, the number of strain gauges was increased from 6 to 10. Strain profile of the T7W7 culvert is shown in Figs 10c, 10d and Table 4. Gauges at position of 2, 4, 5, 7, 9 show negative values (compression). The minimum value observed at position 5 is -1600με which corresponds to load of 310kN. On the contrary, gauges at position of 1, 3, 6, 8, 10 recorded positive values (tension) where the maximum value is 5100με at position 8. Therefore, the neutral axis of the wall section is located between cross-sections 7-8 and 9-10, 1-2 and 3-4.

    Figs 10e and 10f show the strain gauges location and strain values of the T7W5 specimen. Table 5 summarizes strain values of some locations in the T7W5 culvert. In specimen T7W5, strain gauges at positions 2, 4, 5, 7, 9 show negative values of compressive elements. Minimum strain at the position 5 is nearly -1,500με. Strain gauges at positions 1, 3, 6, 8, 10 exhibit positive values of tensile elements, where maximum value of 4,285με is observed at position 3. Similarly to T10W7 and T7W7 culverts, strain values at positions 3 and 8 in specimen T7W5 are larger than at the position 6. Obviously, at the failure point, the positive moment value at section 3-8 of the wall is larger than at section 6 in the mid-span of the upper slab.

    Under the vertical load of the UHPC culvert test, the upper slab is a fixed end flexural element whereas, the wall is a member subjected to compression and bending forces. According to the strain results, the neutral axis of the wall is always located in the cross-section, tending to be near the inner surface. The maximum strain in the outer surface of the wall is larger than the position at mid span of the upper slab. With an increment in the applied load, the concrete strain increased accordingly along with an upward movement of the neutral axis. However, no sign due to buckling in the wall was observed. These results indicate that UHPC material for thin wall of culvert still ensures the loading capacity according to design requirements. In addition, even though the steel reinforcements are placed in the tension zone of thin upper slab and thin wall, the presence of fibers in UHPC are necessary to enhance the tensile strength of the cross section and eliminate brittle failure.


    UHPC culvert tests were conducted to demonstrate the feasibility of the rehabilitation of deteriorated culvert systems. Each test specimen was loaded incrementally until failure. The following conclusions can be drawn from this study:

    • (1) The load-deflection curve of all UHPC culvert specimens exhibited typical flexural behaviour. The loading capacity of UHPC culvert T10W10 is 152%, 193% and 283% larger than T10W7, T7W7 and T7W5, respectively. Decreasing the thickness of upper slab and wall significantly reduces the loading capacity of UHPC culverts.

    • (2) Although the steel bars are placed in the thin cross section of the upper slab and the wall, steel fibers in UHPC show superior effect in strengthen loading capacity and eliminate brittle behavior.

    • (3) Typical flexural failures with ductile behavior of the crushing of concrete in the compressive zone, localized cracking of concrete in tension zone were observed in all culvert specimen tests. No sign of failure due to buckling phenomenon of the wall in UHPC culverts was observed. In real structures, new thin UHPC culvert is surrounded by old culvert, so buckling almost never occurs.

    • (4) The optimization of UHPC culvert design demonstrates that T7W7 and T7W5 culvert types can provide structure with sufficient capacity to resist the dead load and live load of the road. This rehabilitation method with slender cross sections, which minimizes the loss of section by the chipping and a larger spanning capacity, can be an efficient method to replace old culverts without open cut processes.


    This paper was supported by Research Fund, Kumoh National Institute of Technology (2021).


    Aluminum Formwork inside the Old Culvert
    Rehabilitated Shape of Culvert After Pouring UHPC
    Cross Section of UHPC Culvert T10W7
    Test Setup of UHPC Culvert
    Load-deflection Curve of UHPC Culvert Note: Point Xi (Load-kN, Deflection-mm)
    Crack and Failure in UHPC Culvert T10W10
    Crack and Failure in UHPC Culvert T10W7
    Crack and Failure in UHPC Culvert T7W7
    Crack and Failure in UHPC Culvert T7W5
    Diagram of Strain Gauges and Strain of UHPC Culvert


    UHPC Mix Composition
    Types of Culvert Specimen
    Strain Value of Some Locations in T10W7 Culvert
    Strain Value of Some Locations in T7W7 Culvert
    Strain Value of Some Locations in T7W5 Culvert


    1. Almeida, M. , Covas, D. , and Beceiro, P. (2015), Rehabilitation of Sewers and Manholes: Technologies and Operational Practices, Trust Manual of Best Practice, Vol. 6.
    2. BS EN 15885:2010 (2010), “Classification and Characteristics of Techniques for Renovation and Repair of Drains and sewers,” British Standard.
    3. Chung, W. S. , Yu, M. J. , Koo, J. Y. , Kim, I. T. , and Lee, H. D. (2003), “Determination of Optimal Rehabilitation Time for Types of Cast Iron Pipes of Small Pipe Diameter on Water Distribution System in Metropolitan,” World Water and Environmental Resources Congress, pp. 3495–3501.
    4. David Keaffaber, P. E. (2019), “Advanced Geopolymer Mortar System for the Structural Rehabilitation of Large Diameter Culvert and Sewer Infrastructure,” The Innovations in Geotechnical and Materials Engineering: The 2019 TAC-ITS Canada Joint Conference, Halifax, NS.
    5. Hartley, N. (2014), Culvert Rehabilitation Guidance, State of Maine Department of Transportation, USA.
    6. Jang, H. B. and Han, S. M. (2021), “Flexural Behavior of Ultra-High-Performance Concrete Deckplate used in Permanent Form,” Journal of the Korean Society for Advanced Composite Structures, Vol. 12, No. 4. pp. 1-8.
    7. Kim, J. J. , Noh, H. C. , and Reda Taha, M. M. (2012), “Flexural Failure Design Criteria for Retrofitted RC Slabs using FRP-UHPC Hybrid System,” Journal of the Korean Society for Advanced Composite Structures, Vol. 3, No. 2, pp. 11–18.
    8. Meegoda, J. N. and Zou, Z. (2015), “Long-Term Maintenance of Culvert Networks,” Journal of Pipeline Systems Engineering and Practice Vol. 6, No. 4. pp. 1-11.
    9. Sekisui (2014), Outstanding in The World, In Sekisui Chemical Co. Japan.
    10. Simpson, B. , Moore, I. D. , and Hoult, N. A. (2016), “Experimental Investigation of Rehabilitated Steel Culvert Performance under Static Surface Loading,” Journal of Geotechnical and Geoenvironmental Engineering, Vol. 142, No. 2. pp. 1-12.