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.11 No.4 pp.50-58

Structural Performance Change of Steel H-beam with the Application of the Total Reinforced Member Method

Lim Chhoung1, Yeong-Seok Jeong2, Kwang-Sik Nam3, Min-Ho Kwon4
1Ph.D. Student, Department of Civil Engineering, Gyeongsang National University, Jinju, Korea
2Ph.D. Candidate, Department of Civil Engineering, Gyeongsang National University, Jinju, Korea
3Bachelor Student, Department of Civil Engineering, Gyeongsang National University, Jinju, Korea
4Professor, Department of Civil Engineering, Gyeongsang National University, Jinju, Korea

본 논문에 대한 토의를 2020년 09월 30일까지 학회로 보내주시면 2020년 10월호에 토론결과를 게재하겠습니다.

Corresponding author:Kwon, Min-Ho Department of Civil Engineering, Gyeongsang National University, 501, Jinju-Daero, Jinju-si, Gyeongsangnam-do, Korea. Tel: +82-55-772-1796, Fax: +82-55-772-1799 E-mail:
July 24, 2020 August 10, 2020 August 14, 2020


This paper investigates structural performance improvement of a steel H-beam reinforced by the Total Reinforced Member (TRM) method. The TRM method is one of various methods applied to enhance the structural performance by stress induction. The applicability of the TRM method was confirmed through analytical simulation. Initially, a preliminary load to trigger prestress was determined by numerical analysis, and the structural performance changes of three different members (the typical H-beam, composite beam without prestressed made of H-beam and L-shaped steel welded at the bottom flange, and prestressed composite beam) were also calculated by numerical analysis. The three-point bending test was experimentally performed on two members, the typical H-beam and prestressed composite H-beam, to corroborate the analytical results and the feasibility of the TRM method. As a result, the study experimentally confirmed that the generated stress and maximum deflection at the lower flange of the TRM steel H-beam decreased 20.60% and 16.87%, respectively, compared to a conventional beam without reinforcement.

TRM 공법을 적용한 H-Beam의 구조성능변화

Lim Chhoung1, 정 영석2, 남 광식3, 권 민호4
1경상대학교 토목공학과 박사과정
2경상대학교 토목공학과 박사수료
3경상대학교 토목공학과 학부생
4경상대학교 토목공학과 교수


본 연구에서는 H-Beam 부재에 TRM(Total Reinforced Member)공법의 적용을 통한 부재의 구조성능개선에 대하여 분 석하였다. TRM공법은 가교의 성능개선을 위해 적용되는 여러 공법 중 프리스트레스 도입을 통해서 H-Beam 부재의 구조성능을 개선하는 공법이다. 우선 본 연구에서는 해석적인 시뮬레이션을 통해서 TRM공법의 적용 가능성을 확인하였다. 1차적으로 수치 해석을 통해서 프리스트레스 도입을 위한 선행하중을 결정하고, 순수 H-Beam 부재와 L형 강재가 용접된 부재 그리고 TRM공법 이 적용된 3가지의 부재에 대한 구조성능 변화를 해석을 통해서 확인하였다. 그리고 해석적인 결과의 검증 및 공법의 현장적용 가능성을 평가하기 위하여 순수 H-Beam 부재와 TRM공법이 적용된 2가지의 부재에 대한 3점 굽힘 시험을 수행하였다. 실험을 통하여 TRM공법이 적용된 부재에서 순수 H-Beam 부재대비 하부 플랜지의 발생응력 및 최대처짐이 각각 20.60%, 16.87% 감소 하는 결과를 확인하였다.

    National Research Foundation of Korea


    Temporary bridges are temporarily used for short-term work, especially when existing bridges are difficult to access due to construction. The temporary bridges are mostly seen near the construction sites such as the construction of the bridge, dam, and road (Kim, et al., 2013). The purpose of temporary bridge is to commute as well as to the transportation of work vehicles or materials which are needed for the construction site (Lee et al., 2019). With these purposes, the temporary bridges should be available to operate without demanding for any maintenances or repair until the completion of the construction. The Temporary bridge is an economical solution to construction site access instead of diverting or filling over a culvert in road construction.

    Since the temporary bridge is one of the most important things which help transportation construction goes smoothly, its serviceability and its performance capacity are also needed to be ensured. The conventional temporary bridges have a narrow gap between the bridge pillars, so when a natural disaster such as flood occurs, floating objects like a tree accumulates and hits the bridge pillars, so that it can cause the bridge to collapse under high water flow. Due to this frequent occurrence of these accidents, the need for a temporary bridge with a longer span was brought up, and longer span with lower girder is also needed when the space under the bridge is not sufficient (Kim and Yhim, 2020). To this end, several various methods such as the prestressing method, thermal deformation method, edge modification method, and truss method were developed.

    Pre-stressed beam with internal and external tendons for the concrete bridges has been widely used around the globe due to its ability to have a longer span and shorter height of the girder with great improvement in load-carrying ability (Park et al., 2010). Since the concrete pre-stressed beam is quite popular, the studies of it have also been made to a great extent (Ren et al., 2018), yet for the researches of the pre-stressed steel beam with tendons are relatively limit, and they are even much rare for the beams which are pre-stressed by other elements besides tendons like the experiment in this paper (AASHTO 1994 & Federal Highway Administration 1986). Hence, this paper is going to present about the improvement of the structural performance of steel H-beam members, pre-stressed by L-section beam using the Total Reinforced Member (TRM) method experimentally and numerically.


    The properties of elements which were used to make simulation are shown as follows and in Table 1. The H-section beam has 6m length and 588mm height. The width of the cross-section, thickness of flange and web are 300mm, 20mm, and 12mm, respectively. The L-section beam is used as reinforcement to the H-section beam. The length of L-section beam is 5.4m and the height, the width, and the thickness of the L-shaped cross-section are 100mm, 100mm and 10mm, respectively.

    In numerical analysis, ABAQUS, general purposed commercial finite element software, was used for the numerical analysis (ABAQUS, 2016). The beam is modeled as 3D S4R shell element which has 4 nodes with reduced integration and fine mesh to prevent shear locking and hourglass mode, respectively. The mesh of beam model is shown in Fig. 1. Before the analysis of the prestressed beam, determining the preload for making prestress was attempted first. The boundary conditions was defined using a reference point in the middle of the section at the location of the support. Then, the boundary were set to restrain the displacement in vertical and out plane direction of both top and bottom flange with respect to the reference point, so that the section can rotate like a roller support. The boundary condition for both preloading step and the mid-span loading step were the same.

    Two points loads were applied vertically upward from the bottom of non-prestressed H-section beam specimens at the distance of 1.425m (L/4) from the center of each side support with the boundary condition mentioned above. Three different two-points load, considered as pre-loads, 70kN, 110kN, and 190kN, were attempted. The stresses and displacements due to those forces are summarized in Table 2. Then, fourteen different min-span loads were tried from 225kN to 550kN, which is around the maximum load of universal testing machine (UTM), in order to find how much stress on the bottom flange of the nonprestressed beam exceed the allowable when the maximum load is applied. The stress and displacement that came from those fourteen loads are arranged in Table 3. Since the maximum load of UTM is only 500kN and the allowable stress of the beam is 140MPa, 110kN pre-loading was chosen as real pre-load which produces stress around 30% of the allowable stress on the beam for the experiment.

    To find the generated stress of the prestressed element whether it exceeds the allowable or not due to 3-points bending test, the procedures of analysis followed the actual process from assembling the specimen to testing it. The analysis procedures were as follows. Three steps of analysis were created. Initially, H-section beam was introduced, and the boundary condition was the same as mentioned above by using the reference point to make the roller support. In the first step, two preloading of 110kN were applied at the L/4 position from the center of each side support. In the second step, two L-shaped steel were introduced but as straight elements to the deflected beam on its bottom flange, then, the preloading was gradually taking out. Then, a prestressed beam was formed. Finally, a concentrated load of 500kN was applied at the mid-span of the prestressed beam to check the generated stress and the displacement. At that time, 3-points bending was also performed to a normal H-section beam without being reinforced to compare the results which is shown in Table 4 to check for the improvement of TRM method.


    To analyze typical and prestressed beams, some conditions are assumed. All procedure in the analysis follows Hooke’s law, and infinitesimal deformation is considered in the section. Local buckling and lateral local buckling are ignored. The moment stress due to external forces is:

    σ = M I y

    M is the moment from an external force. I is the principal moment inertia of the element. y is the distance from the centroid to the top or bottom side of the beam.

    The results of the stresses of a typical and prestressed beam are shown in Table 5. The first row is the stress of the typical beam from 500kN load at the mid-span. The second row is also the stress of the typical beam at the mid-span, but it is from two 110kN preloading at quarter points. The last row is the stress of the composite beam without prestressed from the 500kN load at mid-span.


    In this study, two types of 3-points bending experiments which are under a static load of both pre-stressed by L-section beam and non-prestressed beam were performed. Two pre-loading of 110kN, which is at L/4 distance from each center of supports from both sides, were applied to the prestressed element (Fig. 2). The L-shaped steels were used as the strengthening elements to make prestress on the bottom flange of the beam. The H-section beams, without being reinforced, were used as control specimens, and they are used to compare with the reinforced ones for its enhancement in stress generating and its displacement at the mid-span.

    4.1 Experimental Overview

    The main goal of this research is to investigate the enhancement of the H-section beam in stress generating at the tension side due to the flexural bending from the external load when the member is strengthened by two external L-section beams using TRM method. The experiment is mainly carried out in quasi-static state. The L-section beams were welded on both sides of the web of the H-section beam in the bottom side when the beam was being under pre-loading to bring the prestress on the flange of the beam. With this approach, compressive stress on the bottom flange were introduced (Fig. 3). A H-section beam whose height and width are 588mm and 300mm, respectively, were used with 20mm of the thickness of the flange and 12mm of the thickness of the web (Fig. 4 (a)). Meanwhile, the L-section beam, used to introduce the prestressed, were made of 100mm in height and width with the 10mm of the thickness (Fig. 4 (b)). The properties of members used in this research are shown in Table 1.

    4.2 Fabrication of Steel H-beam

    Two types of the full scale of H-section beams were tested, and the span length was 6m. One of them was a traditional H-section beam, used as a control specimen. The other one was the strengthened one, which was tested to find the improvement of the flexural capacity as well as its displacement improvement by TRM method. The strengthened beam was made of H-section beams and two L-shaped steels welded on the bottom side of the beam after pre-loading. Fig. 5 demonstrates the procedure of making prestressed Steel H-beam using TRM method. A pre-loading, at the location of L/4 from each center of supports inward, was applied upward to generate prestressed on the bottom flange of the beam. The pre-loading was chosen to be 110kN at each position as explained previously in Section 2. While the H-section beam is being under pre-loading, 2L-shaped steel are welded to that deflected beam follow the same curvature as the beam. After done welding, the pre-loading was removed, and the beam comes back with the existed compressive stress on the bottom flange.

    4.3 Loading and Deflection Measurement

    A three-point bending test was carried out for the specimen to measure the improvement in stress generating due to flexural bending. An actuator with the load cell of 1,000kN was operated to apply loading at the middle point of each beam. Vertical displacement was measured during the test by two linear variable differential transformers (LVDT) which were installed at the quarter-point (L/4) and at the mid-span (L/2) of each beam. Strain gauges were attached also on the web to use in calculating shear strain, on top and bottom flange of the beam, and on the L-section beam to trace the tensile and compressive strain. All LVDT and strain gauge is shown in Fig. 6 and 7.


    In this section, the flexural behaviors of prestressed H-beam using TRM method were investigated experimentally. During the test, the strain on the middle of the section at the quarter-point and the deflection at the quarter-point and mid-span point are measured using strain rosettes. The tensile and compressive stress are estimated using measured strains based on elastic theory. The experimental results were used to compare with theoretical solutions in order to verify whether the experimental procedures are applicable to the real construction field. A three points loading test was conducted in order to investigate the improvement of stress generating at the tensile side of the beam due to the flexural bending and the maximum deflection of the beam at the mid-span. As expected, the results illustrated that the tensile stress generating was noticeably decreased in prestressed composite section (H-section and L-section) when it compares with the non-prestressed H-section beam which used as a control beam.

    5.1 Comparison between Experimental and Theoretical Solutions

    The strain results obtained from the top and bottom flange of prestressed or non-prestressed H-section beam under static load were compared to the theoretical solutions as well as the result obtained from ABAQUS in order to verify whether the experimental data is acceptable. In order to make a comparison of the result of those three, the relationship between load and strain was plotted as shown in Fig. 8. All the specimen was applied with a concentrate load of 500kN at the mid-span. Fig. 8 (a), and (b) shows the bottom and top strain of the conventional H-section beam, respectively. The experimental, theoretical, and FEA results are compared. Even there are some discrepancies, it shows those results are very similar to each other and it also shows same tendency. Such discrepancies may be caused by the assumption made in theoretical computation and boundary conditions in finite element modeling. As depicted in Fig. 8 (c) and (d), the value of strain from theory calculation in those figures have the different condition compared to the other two. The differenceis mainly due to the composition action in the experiment. It is worth to note that H-section beam is deflected but the L-section beam is not deflected when the L-section beam is welded after preloading. Such process is very difficult to model in the finite element simulation and theoretical calculation. Moreover, the calculating of the prestress on the bottom of the beam after applying preloading and introducing L-shaped steel is also tough in manual calculation using theory approache. So, the strain result from theory calculation in Fig. 8 (c) & (d) are from normal composite beam of H-section and L-section without any prestresses. They are purposely used to show the improvement of TRM method. Although there exist some discrepancies between experimental and analytical approaches, they show how the TRM method improves the structural performance of the beam.

    Fig. 8 (c) illustrates the bottom strain of the H-section beam composited by two L-section with the 2 points pre-loading of 110kN. The strain from simulation and experiment are similar though the difference is getting slightly bigger as load increases. It is because of differences in assembling L-shaped steel in experiment and simulation. In the experiment, the L-section was welded in curve shape as the deformed shape of the beam due to pre-loading, yet in the simulation, the L-section was introduced in the straight shape regardless of the deformed shape of the beam. If compared to the value from theory calculation of non-prestressed composite beam, the result of experiment and simulation is much better which means that the TRM method does help improve the structural performance of the beam. Fig. 8 (d) demonstrate the top strain of reinforced beam whose condition is the same as the one in Fig. 8 (c). In this figure, the results of the experiment and simulation are quite big different. The strain value from the experiment is much smaller than the one from the simulation. It is because of the same reason as the reason for the differences of the bottom strain in Fig. 8 (c) which is because of the procedure of introducing L-shaped between experiment and simulation. The error between experiment and simulation are shown in Table 6.

    5.2 Comparison of the Stress on the Bottom Flange at Mid-Span

    The strains of the bottom flange at mid-span were measured to evaluate the improvement of the flexural capacity of the externally prestressed H-section beam by L-section with two-point preloading using TRM method. The comparison of obtained strain from the experimental for both prestressed and non-prestressed beam is shown in Fig. 9. It can be seen that the prestressed beam had more stiffer loading capacity compared to the typical one. For the conventional beam, with a concentrate load of 500kN at the mid-span of the beam, the tensile stress on the bottom flange is 166.95MPa while the tensile stress of prestressed beam which was under the same load condition shows only 132.51MPa which is 20.6% less than the typical one. It is because the pre-loading produced the pre-compressive stress on the bottom flange of the beam which was under the tensile stress due to the point load at the mid-span. As those two pre-loadings produced opposite stress on the bottom flange, the capacity of the beam is improved.

    5.3 Comparison of the Deflection at the Mid-span of the Beams

    The overall deflections of the beams were measured and arranged in Fig. 10. It illustrates the comparison of deflection of both typical and prestressed beam at the mid-span. The H-shape beam shows 9.6mm deflection at the mid-span while the prestressed beam with L-section shows 7.98mm which is 16.87% less than the non-prestressed beam and it indicates the TRM method can significantly improve the deflection of the beam, so that the applicable span length will increase without change of height of girder.


    In this study, the performance of Total Reinforced Member technique for H-section beam was investigated. The traditional H-shape beam was prestressed by preloading and two L-section beams welded at the bottom side of H-shape beam. Those L-section take a role as prestressing tendon in prestressed concrete beam. Several pre-loadings were attempted using Finite element approach using ABAQUS in order to find the best preloading for the beam and for the experimental condition. As the result, 110kN was chosen as preloading for the beam at its quarter-point from both points. After making prestressing on the beam, and welding L-shaped steel to the beam, a 500kN concentrated load was applied at mid-span. Comparing to non-prestressed beam, the prestressed beam was significantly improved in both flexural capacity and deflection. In particular, the stress on the bottom flange and the deflection of the prestressed beam was reduced by 20.60% and 16.87%, respectively. In conclusion, the TRM method is beneficial and applicable in strengthening the steel H-beam because it provides ease of application and economic feasibility.


    This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT)(No. NRF-2019R1A2C1003007).


    Meshed Prestressed Beam Model for Finite Element Analysis
    Conceptional Illustrate of Prestressing Attempt
    Sectional Stress of H-beam due to Pre-loading
    Sectional Detailing of H-beam and L-beam
    The Procedure of Forming the Prestressed Beam
    Virtual Location of LVDT and Strain Gauge
    Three Points Bending of Prestressed Beam
    The Comparison of Top and Bottom Strain of the Beam
    Bottom Stress of the Typical and Reinforced Beam
    Displacement of Typical and Reinforced Beam


    Properties of Steel H-beam and L-shaped Steel
    Stresses on the bottom flange and Displacements at Mid-span due to Pre-loading on Non-prestressed Beam
    Stresses on the bottom flange and Displacements at Mid-span due to Mid-span Load on Non-prestressed Beam
    Stress and Displacement at Mid-span of the Prestressed and Non-prestressed Beam
    Stress of Prestressed and Non-prestressed Beam
    The Error Percentage between Experiment and Simulation


    1. AASHTO (1994), Standard Specification for Highway Bridge, 1st Edition.
    2. ABAQUS (2016), Commercial FE Software and Documentation, Dassault Systèmes Simulia Corporation, Johnston, RI, USA.
    3. Federal Highway Administration (1986), Highway bridge replacement and rehabilitation program, Bridge Division Office of Engineering, Washington, DC.
    4. Kim, C. S. , Kim, J. T. , and Kang, J. G. (2013), “Analysis of the Cause for the Collapse of a Temporary Bridge Using Numerical Simulation,” Scientific Research, Engineering, Vol. 5, pp. 997-1005.
    5. Kim, T. G. , and Y, S. S. (2020), “Full-Scale Model Test on the Structural Behavior of U-Shaped Composite Bridge Reinforced with Arch Edge Girders,” Journal of Korean Society for Advanced Composite Structures, Vol. 11, No. 1, pp. 22-32.
    6. Lee, H. , Lim, J. , Kang, Y. , and Kong, J. (2019), “A Comparative Study on Bridge Inspection and Performance Assessment,” Journal of Korean Society for Advanced Composite Structures, Vol. 10, No. 6, pp. 91-102.
    7. Park, S. K. , Kim, T. W. , Kim, K. S. , and Hong, S. N. (2010), “Flexural Behavior of Steel I-beam Prestressed with Externally Unbounded Tendons,” Journal of Constructional Steel, Research, Vol. 66, pp. 125-132.
    8. Ren, Y. Z. , Wang, Y. Q. , Wang, B. , Ban, H. Y. , Song, J. and Su, G. (2018), “Flexural Behavior of Steel Deep Beam Prestressed with Externally Unbonded Straight Multi-tendons,” Thin-Walled Structure, Vol. 131, pp. 519-530.