STARTCON19

STARTCON19 - International Doctorate Students Conference +

Lab Workshop in Civil Engineering
Volume 2020

Conference Paper

Softened Variable Angle Truss Model

(RA-STM): Model Description and
Refinement/Optimization Proposals
Benedito Filho1 , Luís Bernardo1 , and Bernardo Horowitz2
1
C-MADE - Centre of Materials and Building Technologies, University of Beira Interior, Covilhã,
Portugal
2
Department of Civil Engineering, Federal University of Pernambuco, Brazil

Abstract
This article presents a recent softened truss model with variable angle, namely the
refined RA-STM (Rotating-Angle Softened Truss Model), to model the behaviour of
structural concrete plates under pure shear. The equations of the model, as well
as the solution procedure, are summarized. Some predictions from the RA-STM are
also presented, discussed and compared with experimental results available in the
literature. It is shown that the refined RA-STM still needs to be refined. In addition, the
Corresponding Author: need to generalize the RA-STM for more general loading cases is also discussed as well
Benedito Filho as the need to optimize the solution procedure in order to facilitate its computational
benedito.madian.filho@ubi.pt implementation.

Received: 7 January 2020

Accepted: 21 April 2020
Published: 3 May 2020

Publishing services provided by

Knowledge E
1. Introduction
Benedito Filho et al. This
article is distributed under the It is a common practice in structural engineering to discretize complex structures as a
terms of the Creative Commons
combination of simpler elements. As shown in Fig. 1 (a), some structures can be idealized
Attribution License, which
permits unrestricted use and as a set of two-dimensional structural concrete elements (panels, plates or membranes),
redistribution provided that the
with the reinforcement rebars usually arranged in orthogonal directions. Such elements
original author and source are
credited.
are only subjected to in-plane stresses, which generate internal membrane forces. The
full behaviour of such elements can be predicted with softened truss models (STM)
Selection and Peer-review under
the responsibility of the considering rotating angles (RA) for the struts. Such models are called softened variable
STARTCON19 Conference angle truss models or rotating angle – soft truss models (RA-STM).
Committee.
Simple computational tools can be created from such models, which can be very
useful to help for the design of structural concrete structures, namely to check the local
safety of membrane elements.
The implementation of these nonlinear models is usually made in computers by using
softwares with programming languages, such as MATLAB [1]. This allows to assess,
calibrate and modify such models more easily.

How to cite this article: Benedito Filho, Luís Bernardo, and Bernardo Horowitz, (2020), “Softened Variable Angle Truss Model (RA-STM): Model
Description and Refinement/Optimization Proposals” in STARTCON19 - International Doctorate Students Conference + Lab Workshop in Civil Page 36
Engineering, KnE Engineering, pages 36–48. DOI 10.18502/keg.v5i5.6909
 STARTCON19

2. Description of the RA-STM Model

The RA-STM model was proposed by Belarbi and Hsu in 1994 [2] and Pang and Hsu in
1995 [3]. Recently, this model has been refined by Silva and Horowitz in 2015 [4], Silva
in 2016 [5], Cerquido in 2017 [6] and Bernardo et al. in 2018 [7, 8].
The resistance mechanism of RA-STM is assumed to be a plane truss, in which
the concrete resists to compressive stresses and the reinforcement resists to tensile
stresses (Fig. 1 (b)). The non-linear behaviour and the softening effect of concrete are
key features of this model. This model is based on the three Navier’s principles of
material mechanics, namely: stresses equilibrium, strains compatibility and constitutive
relationships for the materials.

(a)

(b)

Figure 1: Membrane elements to analyse concrete structures [5, 6].

In the RA-STM, the equations derived from the stresses equilibrium and strains
compatibility are obtained through the analysis of the Mohr´s circle for stresses (𝜎)
and strains (𝜖), while the constitutive relationships for concrete and steel are assumed
and based on average 𝜎 − 𝜖 relationships experimentally calibrated and found in the
literature. For this reason, the relationship for concrete in compression (𝜎𝐷 − 𝜖𝐷 ) in
the principal direction of stresses has two main characteristics: a nonlinear relationship
between stresses and strains and the influence of the softening effect in concrete due to
transverse tensile stresses and diagonal cracking. For this latter, a softening coefficient
𝜁 is incorporated to the constitutive relationship for concrete, which is experimentally
calibrated and based on proposals that can be found in the literature.

DOI 10.18502/keg.v5i5.6909 Page 37

 STARTCON19

2.1. Equations for RA-STM

Tables 1 to 3 summarize the set of equations of RA-STM. Details about the derivation
of the presented equations for the original RA-STM and for reinforced concrete (RC)
panels can be found in previous studies [2, 3, 5–7]. The meaning of all parameters can
be found in the notation list at the end of this article. Some parameters can also be
visualized in Fig. 1 (b).
TABLE 1: Equilibrium and compatibility equations [2, 3].

Equilibrium equations (stresses)

⎡ 𝜎𝐿 ⎤ ⎡ 𝜎𝐿𝑐 ⎤ ⎡ 𝜌𝐿 𝑓𝐿 ⎤
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢𝜎 ⎥ = ⎢ 𝜎𝑐 ⎥+⎢ 𝜌 𝑓 ⎥ (1)
⎢ 𝑇 ⎥ ⎢ 𝑇 ⎥ ⎢ 𝑇 𝑇 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ 𝑐 ⎥ ⎢ ⎥
⎣ 𝜏𝐿𝑇 ⎦ ⎣ 𝜎𝐿𝑇 ⎦ ⎣ 0 ⎦
⎡ 𝜎𝐿𝑐 ⎤ ⎡ cos2 (𝛼𝐷 ) sin2 (𝛼𝐷 ) 2 sin (𝛼𝐷 ) cos (𝛼𝐷 ) ⎤ ⎡ 𝜎𝐷 ⎤
⎢ ⎥ ⎢ ⎥⎢ ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥
⎢ 𝜎𝑐 ⎥ = ⎢ sin2 𝛼 cos2 (𝛼𝐷 ) − 2 sin (𝛼𝐷 ) cos (𝛼𝐷 ) ⎥⎥ ⎢⎢ 𝜎𝑅 ⎥ (2)
⎢ 𝑇 ⎥ ⎢ ( 𝐷) ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥
⎢ 𝑐 ⎥ ⎢ ⎥⎢ ⎥
⎣ 𝜏𝐿𝑇 ⎦ ⎣ − sin (𝛼𝐷 ) cos (𝛼𝐷 ) sin (𝛼𝐷 ) cos (𝛼𝐷 ) cos2 (𝛼𝐷 ) − sin2 (𝛼𝐷 ) ⎦ ⎣ 0 ⎦
⎡ 𝜎𝐿 ⎤ ⎡ cos2 (𝛼𝐷 ) sin2 (𝛼𝐷 ) 2 sin (𝛼𝐷 ) cos (𝛼𝐷 ) ⎤ ⎡ 𝜎𝐷 ⎤ ⎡ 𝜌𝐿 𝑓𝐿 ⎤
⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢𝜎 ⎥ = ⎢ sin2 𝛼 cos2 (𝛼𝐷 ) − 2 sin (𝛼𝐷 ) cos (𝛼𝐷 ) ⎥ ⎢ 𝜎 ⎥ + ⎢ 𝜌 𝑓 ⎥ (3)
⎢ 𝑇 ⎥ ⎢ ( 𝐷) ⎥⎢ 𝑅 ⎥ ⎢ 𝑇 𝑇 ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥
⎣ 𝜏𝐿𝑇 ⎦ ⎣ − sin (𝛼𝐷 ) cos (𝛼𝐷 ) sin (𝛼𝐷 ) cos (𝛼𝐷 ) cos2 (𝛼𝐷 ) − sin2 (𝛼𝐷 ) ⎦⎣ 0 ⎦ ⎣ 0 ⎦
𝜎𝐿 = 𝜎𝐷 cos2 (𝛼𝐷 ) + 𝜌𝐿 𝑓𝐿 (4)

𝜎𝑇 = 𝜎𝐷 sin2 (𝛼𝐷 ) + 𝜌𝑇 𝑓𝑇 (5)

𝜏𝐿𝑇 = −𝜎𝐷 sin (𝛼𝐷 ) cos (𝛼𝐷 ) (6)

Compatibility equations (strains)

⎡ 𝜖𝐿 ⎤ ⎡ cos2 (𝛼𝐷 ) sin2 (𝛼𝐷 ) 2 sin (𝛼𝐷 ) cos (𝛼𝐷 ) ⎤ ⎡ 𝜖𝐷 ⎤

⎢ ⎥ ⎢ ⎥⎢ ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥
⎢𝜖 ⎥ = ⎢ sin2 𝛼 cos2 (𝛼𝐷 ) − 2 sin (𝛼𝐷 ) cos (𝛼𝐷 ) ⎥⎥ ⎢⎢ 𝜖𝑅 ⎥ (7)
⎢ 𝑇 ⎥ ⎢ ( 𝐷) ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥
⎢ ⎥ ⎢ ⎥⎢ ⎥
⎣ 𝛾𝐿𝑇 ⎦ ⎣ −2 sin (𝛼𝐷 ) cos (𝛼𝐷 ) 2 sin (𝛼𝐷 ) cos (𝛼𝐷 ) 2 cos2 (𝛼𝐷 ) − sin2 (𝛼𝐷 ) ⎦ ⎣ 0 ⎦
𝜖𝐿 + 𝜖𝑇 = 𝜖𝐷 + 𝜖𝑅 (8)

𝛾𝐿𝑇 = 2 (𝜖𝑅 − 𝜖𝐷 ) sin (𝛼𝐷 ) cos (𝛼𝐷 ) (9)

It should be noted that the initial estimates to initialize the solution procedure are
computed using the Mohr compatibility truss model (MCTM) [9].

DOI 10.18502/keg.v5i5.6909 Page 38

 STARTCON19

TABLE 2: Constitutive relationships for the materials [5–7].

Concrete in compression

⎧ ′ 𝜖𝐷 𝜖𝐷 2
⎪ 𝜁𝑓𝐶 2 ( − , para 𝜖𝐷 ≤ 𝜁𝜖0
⎪ [ 𝜁𝜖0 ) ( 𝜁 𝜖0 ) ]
⎪ 𝜖 2
𝜎𝐷 = ⎨ ⎡ ⎛ 𝐷 − 1⎞ ⎤ (10)
⎪ ′ ⎢ 𝜁𝜖
⎜ 0 ⎟⎥
⎪ 𝜁𝑓𝐶 ⎢1 − ⎜ 4 ⎟ ⎥ , para 𝜖𝐷 > 𝜁𝜖0
⎪ ⎢ ⎜ −1 ⎟ ⎥
⎩ ⎣ ⎝ 𝜁 ⎠⎦
‵
𝑅(𝑓 )
𝜁= (11)
400𝜖𝑅
1+
√ 𝜂′
𝜌𝑇 𝑓𝑇 𝑦
𝜂= (12)
𝜌𝐿 𝑓𝐿𝑦
′
⎧ 𝜂≤1⇒𝜂 =𝜂
⎪
⎨ (13)
⎪ 𝜂 > 1 ⇒ 𝜂′ = 1
⎩ 𝜂
′ 5, 8
𝑅(𝑓𝑐 ) = ≤ 0, 9 (14)
′
√𝑓𝑐 (MPa)
Reinforcement in tension

⎧ 𝐸𝑆 𝜖𝑆 for 𝜖𝑆 ≤ 𝜖𝑛
⎪
𝑓𝑆 = ⎨ (15)
𝜖
⎪ 𝑓𝑆𝑦 [(0, 91 − 2𝐵) + (0, 02 + 0, 25𝐵) 𝑆 ] for 𝜖𝑆 > 𝜖𝑛
⎩ 𝜖𝑆𝑦

𝜖𝑛 = (0, 93 − 2𝐵)𝜖𝑆𝑦 (16)

1,5
1 𝑓𝑐𝑟
𝐵= (17)
𝜌 ( 𝑓𝑆𝑦 )
‵
𝑓𝑐𝑟 = 0, 311√𝑓𝑐𝑟 (MPa) (18)

2.2. Solution procedure

Based on recent works, which aimed to refine the RA-STM, an efficient algorithm was
successfully developed and proposed [4–7]. The associated flow chart is illustrated in
Fig. 2. This algorithm needs to be implemented in a computer. In previous studies,
MATLAB packages were used to implement the solution procedure [6, 7].

3. Comparative Analysis with Some Experimental Results

Table 4 presents the main properties of two RC panels tested under shear and found
in the literature [10, 11]. Some of the experimental results of these panels are compared
below with those computed with the refined RA-STM [6, 7].

DOI 10.18502/keg.v5i5.6909 Page 39

 STARTCON19

TABLE 3: Set of equations for the refined RA-STM [5–7]

Proportional loading

𝜎𝐿 = 𝑚𝐿 𝜎1 (19) 𝜎𝑇 = 𝑚𝑇 𝜎1 (20)

𝜏𝐿𝑇 = 𝑚𝐿𝑇 𝜎1 (21)

𝜎𝐿 + 𝜎 𝑇 𝜎𝐿 + 𝜎 𝑇 2
𝜎1 = + + 𝜏𝐿𝑇 2 (22)
2 √( 2 )

𝑚𝐿 𝜎1 − 𝜌𝐿 𝑓𝐿 = 𝜎𝐷 cos2 (𝛼𝐷 ) (23)

𝑚𝑇 𝜎1 − 𝜌𝑇 𝑓𝑇 = 𝜎𝐷 sin2 (𝛼𝐷 ) (24)

𝑚𝐿𝑇 𝜎1 = −𝜎𝐷 sin(𝛼𝐷 ) cos(𝛼𝐷 ) (25)

1 ′
𝜎1 = ′ (𝐵 ′ ± √𝐵 ′2 − 4𝐴′ 𝐶 ) (26)
2𝐴

𝐴′ = 𝑚𝐿 𝑚𝑇 − 𝑚𝐿𝑇 2 (27)

𝐵 ′ = 𝑚𝐿 𝜌𝑇 𝑓𝑇 − 𝑚𝑇 𝜌𝐿 𝑓𝐿 (28)

𝐶 ′ = 𝜌𝑇 𝑓𝑇 𝜌𝐿 𝑓𝐿 (29)

Relations between stresses and strains

𝑚𝐿 + 𝑚𝐿𝑇 cot(𝛼𝐷 )
𝑓𝐿 = 𝜎1 (30)
𝜌𝐿
𝑚 + 𝑚𝐿𝑇 tan(𝛼𝐷 )
𝑓𝑇 = 𝑇 𝜎1 (31)
𝜌𝑇
−𝑚𝐿𝑇
𝜎𝐷 = 𝜎 (32)
sin(𝛼𝐷 ) cos(𝛼𝐷 ) 1
𝑚 + 𝑚𝐿𝑇 cot(𝛼𝐷 )
𝜖𝐿 = 𝐿 𝜎1 (33)
𝐸𝑆 𝜌𝐿
𝑚 + 𝑚𝐿𝑇 tan(𝛼𝐷 )
𝜖𝑇 = 𝑇 𝜎1 (34)
𝐸𝑆 𝜌𝑇
−𝑚𝐿𝑇
𝜖𝐷 = 𝜎 (35)
𝐸𝑐 sin(𝛼𝐷 ) cos(𝛼𝐷 ) 1

Residual function for MCTM

𝜖𝐿 − 𝜖𝐷
𝐹𝑀𝐶𝑇 𝑀 = − tan2 (𝛼𝐷 ) = 0 (36)
𝜖𝑇 − 𝜖𝐷

Residual function for RA-STM

𝜖 − 𝜖𝐷
⎡ 𝜎𝐷 𝑇 − 𝑚𝐿 𝜎1 + 𝜌𝐿 𝑓𝐿 ⎤ ⎡0⎤
⎢ 𝜖𝑅 − 𝜖𝐷 ⎥ ⎢ ⎥
𝐹𝑅𝐴−𝑆𝑇 𝑀 =⎢ ⎥=⎢ ⎥ (37)
𝜖 − 𝜖𝐷
⎢ 𝜎𝐷 𝐿 − 𝑚𝑇 𝜎1 + 𝜌𝑇 𝑓𝑇 ⎥ ⎢ ⎥
⎣ 𝜖𝑅 − 𝜖𝐷 ⎦ ⎣0⎦

𝜌𝑇 𝑓𝑇 𝑦
∗𝜂 =
𝜌𝐿 𝑓𝐿𝑦

DOI 10.18502/keg.v5i5.6909 Page 40

 STARTCON19

Figure 2: Flow Chart.

Fig. 3 shows the 𝜎𝐷 − 𝜖𝐷 curves for concrete in compression (in the principal direction
of compressive stresses) for panel VA1 [10]. Two curves are presented, the experimental
one and the theoretical one computed from the refined RA-STM.
For the theoretical curve, a first stage with a quasi linear behaviour is observed. In
this stage, as the compressive stress increases the corresponding strain also increases
but at smaller rates. When the reinforcement yields, a strong nonlinear behaviour is

DOI 10.18502/keg.v5i5.6909 Page 41

 STARTCON19

TABLE 4: Properties of tested RC panels under shear.

observed until the peak stress is reached. After this, the strains highly increase with
a small variation of the compressive stresses. The panel continues to behave in this
way until the criterion which define the theoretical failure is reached (at least one of the
materials reach its conventional strain failure).
In the experimental graph, a first stage is also observed in which the compressive
stress increases linearly with the corresponding strains, this latter at smaller rates.
When the compressive stress reaches approximately 4,4 MPa, a slight decrease of
the stiffness, as well as a sudden increase of the strain, is observed. This point of the
graph corresponds to the cracking of the panel. After this, a second stage is observed,
for which the compressive stress increases with the corresponding strains, this latter
again at smaller rates, until the yielding of the reinforcement. In the third stage the strains
highly increase and the compressive stress slowly decreases until failure is reached.
In general, the main features of the compressive concrete behaviour are well captured
by the theoretical model, as well as the peak stress of concrete. However, the theoretical
model is not able to predict the transition from the cracked stage to the uncracked stage.
This is because the refined RA-STM neglects the influence of concrete in tension in the
perpendicular direction to the principal compressive stresses (𝜎𝑅 = 0, see Fig. 1 (b)),
The previous results show that concrete in tension is important to be considered for
low loading levels. In fact, the refined RA-STM considers that the plate is already fully
cracked since the beginning of the loading. This simplification constitutes a drawback
of the model, which needs to be solved because the transition from the cracked stage
to the uncracked stage is important to be correctly predicted in order to assess the
panel for loading services.
Fig. 4 shows the shear (𝜏𝐿𝑇 ) versus shear strain (𝛾𝐿𝑇 ) curves, both experimental and
theoretical, for panel B2 [11]. The dashed curve corresponds to the prediction from the
refined RA-STM, while the continuous curve corresponds to the original RA-STM from
Pang and Hsu in 1995 [11]. This latter accounts for the tensile strength of concrete,

DOI 10.18502/keg.v5i5.6909 Page 42

 STARTCON19

Figure 3: Comparison between 𝜎𝐷 − 𝜖𝐷 curves for panel VA1 [10].

in opposition to the refined RA-STM. Fig. 4 shows that the original RA-STM captures
better the behaviour of the panel for low loading levels. This shows the importance
for the model to include the influence of the concrete in tension. For the ultimate
stage, the refined RA-STM shows better agreement with the experiment curve. In Fig.
4, both versions of the theoretical model capture well another feature of panel B2.
This panel is not reinforced symmetrically (see parameter 𝜂 in Table 3), its transverse
reinforcement ratio is lesser that its longitudinal reinforcement ratio. For this reason,
both reinforcements (longitudinal and transverse) didn´t yield for the same strain. This
is shown in the graphs by two distinct points for which it is possible to observe
two consecutive losses of stiffness, corresponding to the yielding of the transverse
reinforcement (firstly) and longitudinal reinforcement (latter). This shows that RA-STM
model is able to capture well the nonlinear behaviour under shear of asymmetrically
reinforced panels.

4. Future Developments for the Refined RA-STM

The previously presented results, related with panels VA1 and B2, generally show that
the refined RA-STM is able to capture well the general features of panels under shear,

DOI 10.18502/keg.v5i5.6909 Page 43

 STARTCON19

Figure 4: Comparison between 𝜏𝐿𝑇 − 𝛾𝐿𝑇 curves for panel B2 [11].

namely for the ultimate stage. Additional examples with other panels under shear can
be found in previous studies [6, 7], including for prestressed panels [8].
As referred before, the refined RA-STM neglects the influence of concrete in tension
in the perpendicular direction to the principal compressive stresses. This drawback
needs to be solved because the transition from the cracked stage to the uncracked
stage needs to be correctly predicted. This is because current codes of practice impose
to assess structural members for both the ultimate and service loading. For this latter, it
is important to check the cracking load, as well as the stress state in the materials and
the stiffness of the member after cracking.
The previous discussion only involved the behaviour of panels under shear. However,
other loading conditions, such as axial stresses (combined or not with shear stresses)
are common in structures (see Fig. 1 (a)). Few previous studies using RA-STM are focused
on such condition loadings. Some of them show that RA-STM still do not capture the
real behaviour of such panels and numerical problems are observed, which are mainly
related with errors in the convergence criteria [6]. This shows that the RA-STM model still
needs to be improved and generalized to other and more realistic loading conditions.
Finally, the intrinsic nonlinear feature of the RA-STM leads to a complex solution
procedure, as previously shown. This problem can probably be minimized if the set of

DOI 10.18502/keg.v5i5.6909 Page 44

 STARTCON19

equations are rewritten using other criteria, such as the energetic ones. Many problems
in physics are more easily formulated if they are based on energetic criteria, regardless
of how they are applied (minimization of the potential energy or principle of virtual work
for static or pseudo-static systems, least action principle for dynamics systems, etc.).
The application of energetic criteria to establish the solution procedure for the RA-STM
is still an open issue.
The previous discussion shows that the refined RA-STM still need further develop-
ments.

5. Conclusions

This article presented a recent softened truss model with variable angle, namely the
refined RA-STM, to model the behaviour of structural concrete membranes under pure
shear. The equations, as well as the solution procedure of the model, were summarily
presented. Some predictions from the refined RA-STM, related to panels tested under
shear, were also presented and compared with the corresponding experimental results
which were found in the literature. From the comparative analysis, it was shown that
the general features of RC panels under shear are well captured by the refined RA-
STM, namely for the ultimate stage. For low loading conditions, the need to refine
the model was also shown. In addition, some ideas were discussed in order to justify
future developments for the refined RA-STM. Such developments constitute the main
objectives for the Ph.D. project of the first author, which include specifically the following
ones:
- to unify the refined RA-STM for reinforced concrete and prestressed concrete
membranes;
- to consider the influence of concrete in tension in the refined RA-STM through
the incorporation of an additional average constitutive relationship in the calculation
procedure;
- to generalize the refined RA-STM to other loading conditions, including shear
combined with axial forces and also cyclic loading;
- to optimize and simplify the calculation procedure of the refined RA-STM by using
energetic criteria.
Finally, it should be referred that experimental data related with structural concrete
membranes are still scarce and are not sufficient to fully assess the reliability of the
refined RA-STM. For this reason, additional numerical results based on calibrated models
using FEM will certainly be useful.

DOI 10.18502/keg.v5i5.6909 Page 45

 STARTCON19

Notation

𝐸𝑠 = Young’s modulus for steel

𝐹𝑀𝐶𝑇 𝑀 = residual function for MCTM
𝐹𝑅𝐴−𝑆𝑇 𝑀 = residual function for RA-STM
′
𝑓𝑐 ;𝑓𝑐𝑚 = uniaxial compressive strength of concrete
𝑓𝑐𝑟 = tensile strength of concrete
𝑓𝐿 = tensile stress in the longitudinal reinforcement
𝑓𝐿𝑦 = yielding stress of the longitudinal reinforcement
𝑓𝑆 = average tensile stress in the steel bars
𝑓𝑠𝑦 = yielding stress of the steel bars
𝑓𝑇 = tensile stress in the transverse reinforcement
𝑓𝑇 𝑦 = yielding stress of the transverse reinforcement
𝑘 = umber of solution points
𝑚𝐿 = longitudinal proportionality coefficient
𝑚𝐿𝑇 = shear proportionality coefficient
𝑚𝑇 = transverse proportionality coefficient
𝑛max = maximum number of solution points
𝛼𝐷 = angle of the principal compressive stresses in the concrete membrane element
𝛼𝑅 = angle of the principal tensile stresses in the concrete membrane element
𝜖0 = strain correspondent to the peak stress
𝜖𝑐𝑢 = ultimate strain for concrete in compression
𝜖𝑠 = average strain in the steel bars
𝜖𝑠𝑦 = yielding strain of the steel bars
𝜖𝐷 = principal average compressive strain
𝜖𝐿 = longitudinal average strain
𝜖𝑅 = principal average tensile strain
𝜖𝑇 = transversal average strain
𝛾𝐿𝑇 = average shear strain ζ = softening coefficient
𝜌𝐿 = longitudinal reinforcement ratio
𝜌𝑇 = transverse reinforcement ratio
𝜎1 = principal tensile stress in the RC membrane element
𝜎𝐷 = principal compressive strain in the concrete membrane element

DOI 10.18502/keg.v5i5.6909 Page 46

 STARTCON19

𝜎𝐿 = longitudinal normal stress in the RC membrane element

𝜎𝐿𝑐 = longitudinal normal stress in the concrete membrane element
𝜎𝑅 = principal tensile strain in the concrete membrane element
𝜎𝑇 = transverse normal stress in the RC membrane element
𝜎𝑇𝑐 = transverse normal stress in the concrete membrane element τ shear stress
𝜏𝐿𝑇 = shear stress in the RC membrane element
𝑐
𝜏𝐿𝑇 = shear stress in the concrete membrane element

References

[1] MathWorks. (2017). ”MATLAB - R2017”. Academic license.

[2] Belarbi A, Hsu TTC. Constitutive laws of concrete in tension and reinforcing bars
stiffened by concrete. Struct J Am Concr Inst. 1994;91(4): 465–474.

[3] Pang XB, Hsu TTC. Behavior of reinforced concrete membrane elements in shear.
Struct J Am Concr Inst. 1995;92(6):665–679.

[4] Silva JRB, Horowitz B. Efficient procedure to estimate the load–deformation behavior
of reinforced concrete panels under membrane forces [in Portuguese]. CILAMCE
2015: Iberian Latin American Congress on Computational Methods in Engineering;
2015; Rio de Janeiro-RJ, Brazil; vol. 36.

[5] Silva JRB. Efficient procedure for the analysis of reinforced concrete sections using
the softened truss model [unpublished master’s thesis]. Recife, Brazil: Department
of Civil Engineering, Federal University of Pernambuco; 2016.

[6] Cerquido BMD. Analysis of reinforced concrete sections with the softened
truss model [unpublished master’s thesis]. Covilhã, Portugal: Department of Civil
Engineering and Architecture, University of Beira Interior; 2017.

[7] Bernardo LFA, Cerquido BMD, Silva JRB, Horowitz B. Efficient Refined Rotating-
Angle Softened Truss Model Procedure to Analyze Reinforced Concrete Membrane
Elements. Struct Concr. fib 2018;19(6): 1971-1982.

[8] Bernardo L, Lyrio A, Silva J, Horowitz B. Refined Softened Truss Model with Efficient
Solution Procedure for Prestressed Concrete Membranes. J Struct Eng ASCE.
2018;144(6): 04018045.

[9] Collins MP. Torque-twist characteristics of reinforced concrete beams. Inelasticity

and non-linearity and non-linearity in structural concrete. Waterloo: University of
Waterloo Press, 1973; p. 211–231.

DOI 10.18502/keg.v5i5.6909 Page 47

 STARTCON19

[10] Zhang LX, Hsu TTC. Behavior and analysis of 100 MPa concrete membrane elements.
J Struct Eng ASCE. 1998;124(1):24–34.
[11] Pang XB, Hsu TTC. Behavior of reinforced concrete membrane elements in shear.
Struct J Am Concr Inst. 1995;92(6):665–679.

DOI 10.18502/keg.v5i5.6909 Page 48