Professional Documents
Culture Documents
Structures
journal homepage: www.elsevier.com/locate/structures
Decoupling of arch action and truss action in deep beams by strain energy T
⁎
Zhi-Qi He , Tian Xu, Zhao Liu
School of Civil Engineering, Southeast University, Nanjing 211189, China
A R T I C LE I N FO A B S T R A C T
Keywords: It is commonly accepted that deep beams transfer shear forces through a combination of arch action and truss
Deep beam action, but there is disagreement on how to estimate the fraction transferred by each assumed action. This study
Truss action provides a theoretical method to decouple the two shear-carrying mechanisms from the perspective of strain
Arch action energy. Through multiple-level least-energy solutions, a two-panel truss consisted of parallel compression di-
Strut-and-tie model
agonals and orthogonal tension ties is obtained for the minimum strain energy. Assuming that vertical ties of the
Strain energy
truss model and diagonal ties of the combined model have the same energy, the fraction of load transferred by
the truss action is decoupled as 2λ/3–0.5 (where λ is the shear span-depth ratio). It means that the entire shear is
carried by the arch action when λ ≤ 0.75, while the entire shear is carried by the truss action when λ ≥ 2.25. Six
deep beams were tested with instruments placed along the assumed direct struts and indirect struts, from which
the contributions of the two load-transfer mechanisms can be separated experimentally. The theoretical method
agrees well with the experimental results of the current study and the others collected from the literature.
1. Introduction truss action for a/d ≥ 2 and the entire shear is transferred by the arch
action for a/d ≤ 1/2. Foster and Gilbert [13] suggested that the frac-
Deep beams can be classified as members in which a significant tion of load resisted by truss action can be determined by (√3a/d – 1)/2.
portion of the shear force is directly transferred from the load to the It means that the entire force is carried by truss action for a/d ≥ 1.73,
support (Fig. 1). It is now commonly accepted that deep beams transfer while the entire shear is transferred by arch action for a/d ≤ 0.58. The
shear through two major force paths [1–6]: arch action (or direct-strut analysis of strain energy by Brown and Bayrak [14] indicated that for
action) and truss action (or beam action). The shear transfer mechanism beams with a/d less than 2, the preferred mechanism for resisting loads
in deep beams is strongly influenced by the shear span-depth ratio (a/ is a single direct strut between the load point and the support.
d). Members with lower a/d (approximately 1) are dominated by the Few analytical approaches have been proposed to decouple the arch
arch action in which the tension in the longitudinal tie is almost con- action and the truss action in deep beams. Matamoros and Wong [15]
stant along the full length of the beam. Members with larger a/d (ap- used the stiffness method to solve the statically indeterminate model
proximately 3) are dominated by the truss action in which the vertical that is a combination of these two mechanisms, by assuming that the
ties are representations of the stirrups and the tension in the long- modulus of elasticity and the cross-sectional area are constant and
itudinal tie changes along the length of the beam. Beams with inter- equal for all members. From the sectional compatibility condition of
mediate a/d transfer shear through a combination of these two me- shear deformation, Kim and Jeong [16] developed a model for evalu-
chanisms, but there is disagreement on how to estimate the fraction ating the contribution of arch action to shear resistance in shear-critical
transferred by each assumed model [7–9]. reinforced concrete beams. He et al. [17] derived the contribution of
Guidance on the fraction of force transferred by either arch or truss each mechanism from a criterion for maximizing the strength capacity
action is sparse. The ACI Code [10] and AASHTO LRFD Specification of the combined model. Nakamura et al. [18] developed a numerical
[11] require a minimum angle of 25 degrees between struts and ties to Rigid-Body-Spring-Method (RBSM) to decouple the arch and beam ac-
control the use of shallow angle struts. It means that truss members tions in reinforced concrete beams.
with vertical ties are required when a/d is greater than 2.14. In ac- Some experimental methods have been developed to measure the
cordance with the 1999 FIP recommendations [12], the fraction of load two aforementioned contributions separately. Through the measure-
resisted by the truss action depends on the shear span-depth ratio as ment of concrete deformation along direct struts and indirect struts,
given by (2a/d – 1)/3. It means that the entire shear is carried by the Breña and Roy [7] developed an experimental method to quantify the
⁎
Corresponding author.
E-mail address: z.he@seu.edu.cn (Z.-Q. He).
https://doi.org/10.1016/j.istruc.2020.04.020
Received 31 December 2019; Received in revised form 23 March 2020; Accepted 10 April 2020
Available online 16 April 2020
2352-0124/ © 2020 Institution of Structural Engineers. Published by Elsevier Ltd. All rights reserved.
Z.-Q. He, et al. Structures 26 (2020) 185–192
186
Z.-Q. He, et al. Structures 26 (2020) 185–192
Table 1
Geometry and strain energy index of different models.
a/d Optimal model by computer Optimal analytical Pure arch Pure truss
searching model model model
member forces, the total strain energy was calculated by summing the
member force times its length for all tension members, as expressed in
Eq. (1). The optimum solution is given by the node geometry that
produces the minimum energy. The iterative process of the algorithm is
described as follows:
1. Set Node B to a unique location near the support of the beam. Note
that Node B is always on the bottom chord of the truss.
2. Move Node D in the two-dimensional zone to all possible geome-
tries. Note that both x-coordinate and y-coordinate are variable for
Node D.
Fig. 3. Optimal topology shapes for deep beams. 3. For each new location for Node D, the strain energy of the truss is
solved and recorded.
4. Move Node B from the support to the center of the beam, and the
procedure from Steps 1 to 3 is repeated.
5. Read the minimum value of strain energy and obtain the corre-
sponding truss geometry.
Fig. 4 and Table 1 list the optimal STM geometry that produces the
minimum strain energy for each a/d. At an a/d of 1, the strain energy
index (ΣTili) of the optimal two-panel truss is 0.950, which is very close
to that of the direct-strut model (0.980). At an a/d of 2.5, the strain
energy index of the optimal truss is 10.205, which is very close to that
of the pure truss model (10.215). Therefore, the strain energy is a
reasonable index for identifying the load transferring mechanism in
deep beams.
From the optimal STM configurations shown in Fig. 4, it can be
found that: 1) the two inclined Struts AB and CD are approximately
parallel to each other; and 2) the inclined Tie BD is approximately
perpendicular to Struts AB and CD. Therefore, an analytical model with
a single variable (i.e., the inclination angle of Struts AB and CD) can be
established, as shown in Fig. 5.
187
Z.-Q. He, et al. Structures 26 (2020) 185–192
where λ = a/d.
Fig. 5. Optimal analytical model and its decomposition. The strain energy of different models is listed in Table 1 and plotted
in Fig. 7. Among that, expressions for calculating the strain energy in
The strain energy in the inclined tie is: the pure arch model and the pure truss model are as follows:
Fig. 6. Optimal strut inclination angle for minimum strain energy. γtruss = 2λ /3 − 0.5 for 0 ⩽ γtruss ⩽ 1 (14)
188
Z.-Q. He, et al. Structures 26 (2020) 185–192
where λ = a/d.
From Eq. (14), it can be seen that: 1) the entire shear is carried by
the arch action (direct-strut model) for a/d ≤ 0.75; 2) the entire shear
is carried by the truss action for a/d ≥ 2.25; and 3) both mechanisms
contribute equally to resist the shear force when a/d = 1.5. The above
conclusions are consistent with the strain-energy curves shown in
Fig. 7.
4. Experimental verification
This section solves the problem of using the testing data to quantify
the contribution of each shear-transfer mechanism. As shown in Fig. 10,
the relationship between the shear force and the strut compression in
the arch mode and the truss model can be expressed as
Fig. 10. Measurement of strut strains.
Varch = Narch sin θarch , Vtruss = Ntruss sin θtruss (15)
where Varch, Vtruss = shear forces transferred by arch action and truss the direct strut and the indirect strut, respectively.
action, respectively; Narch, Ntruss = compressions in the direct strut and The compressions in the direct strut and the indirect strut can be
the indirect strut, respectively; and θarch, θtruss = inclination angles of obtained from the measured strut strains as follows:
189
Z.-Q. He, et al. Structures 26 (2020) 185–192
Fig. 11. Development of strut strains and truss action with loading.
Table 2
Fraction of load transferred by truss action in deep beams.
Reference Specimen a/d εarch,u (με) εtruss,u (με) Proportion of truss action
Tested Proposed Eq. (14) FIP (1999) Foster and Gilbert (1998)
Alcocer and Uribe (2008) Four beams 1.49 N.A. N.A. 0.50 0.49 0.66 0.79
Breña and Roy (2009) DB1.5–0.75 1.5 N.A. N.A. 0.48 0.50 0.67 0.80
DB1.5–0.50 1.5 N.A. N.A. 0.47 0.50 0.67 0.80
DB2.0–0.75 2.0 N.A. N.A. 0.81 0.83 1.00 1.00
DB2.0–0.50 2.0 N.A. N.A. 0.75 0.83 1.00 1.00
190
Z.-Q. He, et al. Structures 26 (2020) 185–192
Narch = bWarch·Ec εarch, Ntruss = bWtruss·Ec εtruss (16) • The number of deep-beam tests measuring the truss and arch con-
where b = beam width; Ec = elasticity modulus of concrete; εarch, tributions is very limited in the literature. Such a kind of experi-
εtruss = averaged compressive strains of the direct strut and the indirect mental tests on deep beams with a wide range of shear span-depth
strut, respectively; and Warch, Wtruss = widths of the direct strut and the ratios would be helpful to validate the analytical method.
indirect strut, respectively.
The widths of the direct strut and the indirect strut are almost equal 6. Conclusions
in the tested beams. From Eqs. (15) and (16), the load-transfer ratio
between the truss action and the arch action can be derived as This study provides a theoretical method to decouple the two shear-
carrying mechanisms in deep beams from the perspective of strain en-
Vtruss ε sin θarch ε 1 + λ2 ergy. The following conclusions can be drawn:
η= = truss = truss ·
Varch εarch sin θtruss εarch 1 + (λ /2)2 (17)
Finally, the fraction of load transferred by the truss action can be
• From multiple-level least-energy solutions, it is found that a two-
panel truss consisted of parallel compression diagonals and ortho-
obtained as gonal tension ties is the optimal model for deep beams. The in-
η clination angle is related to the shear span-depth ratio as
γtruss =
1+η (18) θ = 77°–13λ.
• The fraction of load resisted by the truss action depends on the shear
span-depth ratio, as given by γtruss = 2λ/3–0.5. The entire shear is
4.3. Results and comparisons
carried by the arch action (direct-strut action) for λ ≤ 0.75; the
entire shear is carried by the truss action for λ ≥ 2.25; and both
Fig. 11 shows the measured strains along the direct and indirect
mechanisms contribute equally to resist the shear force when
struts during the tests. There are two parallel indirect struts for a shear
λ = 1.5.
•
span in the assumed truss model. Due to the presence of diagonal cracks
The measured averaged strains along the assumed direct struts and
crossing the interior strut, only the data in the exterior strut are used to
indirect struts can be used to separate the contributions of the two
quantify the contribution of truss action [7]. The contribution of truss
shear-carrying mechanisms in deep beams, as expressed in Eqs. (17)
action is calculated using Eqs. (17) and (18) and plotted in Fig. 11. It is
and (18). In the test, it was observed that the truss action is gra-
observed that the truss action is gradually weakening with the loading.
dually weakening with the loading.
The ultimate contribution of truss action at failure is listed in Table 2.
Experimental results in the literature about the load-transfer me-
Declaration of Competing Interest
chanisms in deep beams are also collected. Alcocer and Uribe [33]
presented the test results of four deep beams with the same a/d = 1.49.
The contribution of the truss action was calculated from measuring the The authors declare that they have no known competing financial
strains in the stirrups recorded in the tests. Test results suggested that interests or personal relationships that could have appeared to influ-
both mechanisms (arch action and truss action) contributed approxi- ence the work reported in this paper.
mately equally to resist the shear forces. Breña and Roy [7] carried out
an experimental evaluation of the shear transfer mechanism in deep Acknowledgments
beams. In the test, linear potentiometers were used to measure the
average strains along direct struts in the assumed arch model and in- This work was supported by the Natural Science Foundation of
direct struts in the assumed truss model. Jiangsu Province, China (No. BK20180063); and the National Key R&D
The test results of the six beams in the current study along with the Program of China (No. 2017YFC0703402).
eight beams from the literature, are summarized in Table 2 and com-
pared with the predictions of different equations. The proposed Eq. (14) References
is in good agreement with the test results of all specimens with different
a/d (the predicted values are on average 105.8% of the experimental [1] Marti P. Basic tools of reinforced concrete beam design. ACI J 1985;82(1):46–56.
[2] ASCE-ACI Committee 445. Recent approaches to shear design of structural concrete.
values, with a standard deviation of 7.5%), while both the methods J Struct Eng 1998;124(12):1375–417.
recommended by FIP 1999 [12], Foster and Gilbert [13] are shown to [3] Reineck KH. Examples for the design of structural concrete with strut-and-tie
overestimate the contribution of truss action. models. ACI SP-208. Farmington Hill (MI): American Concrete Institute; 2002.
[4] Liu J, Mihaylov BI. A comparative study of models for shear strength of reinforced
concrete deep beams. Eng Struct 2016;112:81–9.
5. Discussions [5] Shuraim AB, El-Sayed AK. Experimental verification of strut and tie model for HSC
deep beams without shear reinforcement. Eng Struct 2016;117:71–85.
[6] Lim E, Hwang SJ. Modeling of the strut-and-tie parameters of deep beams for shear
This study provides a theoretical method to decouple the two shear- strength prediction. Eng Struct 2016;108:104–12.
carrying mechanisms in deep beams, based on the previous work by [7] Breña SF, Roy NC. Evaluation of load transfer and strut strength of deep beams with
Schlaich et al. [21], Liang et al. [28], Breña and Roy [7], and Hu et al. short longitudinal bar anchorages. ACI Struct J 2009;106(5):678–89.
[8] Kuo WW, Cheng TJ, Hwang SJ. Force transfer mechanism and shear strength of
[27]. The limitations of the current work are discussed as follows:
reinforced concrete beams. Eng Struct 2010;32(6):1537–46.
[9] Ismail KS, Guadagnini M, Pilakoutas K. Strut-and-tie modeling of reinforced con-
• The shear-carrying mechanism in deep beams is influenced by sev- crete deep beams. J Struct Eng 2017;144(2):04017216.
[10] ACI Committee 318. Building code requirements for structural concrete and com-
eral factors, including the shear span-depth ratio, the amount of
mentary (ACI 318–19). Farmington Hills (MI): American Concrete Institute; 2019.
transverse reinforcement and the compressive strength of concrete. [11] AASHTO. AASHTO LRFD bridge specifications (8th ed.). Washington (DC):
Increased amounts of web reinforcement can shift portions of the American Association of State Highway and Transportation Officials; 2017.
load to the truss mechanism [14,19]. [12] F.I.P. Commission 3. Practical design of structural concrete 1999 London, UK:
•
SETO.
When calculating the strain energy in the strut-and-tie model, only [13] Foster SJ, Gilbert RI. Experimental studies on high-strength concrete deep beams.
the tension ties are included in the current study. Some previous ACI Struct J 1998;95(4):382–90.
studies showed that: the concrete contribution to the strain energy is [14] Brown MD, Bayrak O. Design of deep beams using strut-and-tie models – Part I:
evaluating US provisions. ACI Struct J 2008;105(4):395–404.
diminished relative to the steel contribution in the two-panel truss [15] Matamoros AB, Wong KH. Design of simply supported deep beams using strut-and-
model; but for the direct-strut model, the strain energy stored in the tie models. ACI Struct J 2003;100(6):704–12.
struts is not likely to be negligible [14]. [16] Kim W, Jeong J. Decoupling of arch action in shear-critical reinforced concrete
191
Z.-Q. He, et al. Structures 26 (2020) 185–192
beams. ACI Struct J 2011;108(4):395–404. [25] Lourenço MS, Almeida JF. Adaptive stress field models: formulation and validation.
[17] He ZQ, Liu Z, Ma ZJ. Investigation of load-transfer mechanisms in deep beams and ACI Struct J 2013;110(1):71–81.
corbels. ACI Struct J 2012;109(4):467–76. [26] Tjhin TN, Kuchma DA. Computer-based tools for design by strut-and-tie method:
[18] Nakamura H, Iwamoto T, Fu L, Yamamoto Y, Miura T, Gedik YH. Shear resistance advances and challenges. ACI Struct J 2002;99(5):586–94.
mechanism evaluation of RC beams based on arch and beam actions. J Adv Concr [27] Hu Q, Ley MT, Russell BW. Determining efficient strut-and-tie models for simply
Technol 2018;16(11):563–76. supported beams using minimum strain energy. ACI Struct J 2014;111(5):1015–25.
[19] Mihaylov BI, Bentz EC, Collins MP. Behavior of large deep beams subjected to [28] Liang QQ, Uy B, Steven GP. Performance-based optimization for strut-tie modeling
monotonic and reversed cyclic shear. ACI Struct J 2010;107(6):726–34. of structural concrete. J Struct Eng 2002;128(6):815–23.
[20] Wu YF, Hu B. Shear strength components in reinforced concrete members. J Struct [29] Leu LJ, Huang CW, Chen CS, Liao YP. Strut-and-tie design methodology for three-
Eng 2017;143(9):04017092. dimensional reinforced concrete structures. J Struct Eng 2006;132(6):929–38.
[21] Schlaich J, Schäfer K, Jennewein M. Toward a consistent design of structural con- [30] He ZQ, Liu Z. Optimal three-dimensional strut-and-tie models for anchorage dia-
crete. PCI J 1987;32(3):74–150. phragms in externally prestressed bridges. Eng Struct 2010;32(8):2057–64.
[22] Zhang N, Tan KH. Direct strut-and-tie model for single span and continuous deep [31] ANSYS Release 10.0 [Computer software]. ANSYS, Inc. Canonsburg, PA.
beams. Eng Struct 2007;29:2987–3001. [32] Al-Nahlawi KA, Wight JK. Beam analysis using concrete tensile strength in truss
[23] Yang TY, Dashlejeh AA, Arabzadeh A, Hizaji R. New model for prediction of ulti- models. ACI Struct J 1992;89(3):284–90.
mate load of prestressed RC deep beams. Structures 2020;23:509–17. [33] Alcocer SM, Uribe CM. Monolithic and cyclic behavior of deep beams designed
[24] Ali MA, White RN. Automatic generation of truss model for optimal design of re- using strut-and-tie models. ACI Struct J 2008;105(3):327–37.
inforced concrete structures. ACI Struct J 2001;98(4):431–42.
192