This action might not be possible to undo. Are you sure you want to continue?

# Finite element analyses for limit state design of concrete structures

**Posibilities and limitations with SOLVIA
**

Masther’s Thesis in the International Master’s programme Structural Engineering

**FERNANDO LOPEZ FEJSAL AL-HAZAM
**

Department of Civil and Environmental Engineering Division of Structural Engineering Concrete Structures CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden 2005 Master’s Thesis 2005:17

MASTER’S THESIS 2005:17

**Finite element analyses for limit state design of concrete structures
**

Posibilities and limitations with SOLVIA

Master’s Thesis in the International Master’s programme Structural Engineering

FERNANDO LOPEZ FEJSAL AL-HAZAM

Department of Civil and Environmental Engineering Division of Structural Engineering Concrete Structures CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden 2005

Finite element analyses for limit state design of concrete structures Posibilities and limitations with SOLVIA Master’s Thesis in the International Master’s programme Structural Engineering

FERNANDO LOPEZ FEJSAL AL-HAZAM

©FERNANDO LOPEZ, FEJSAL AL-HAZAM 2005

Master’s Thesis 2005:17 Department of Civil and Environmental Engineering Concrete structures Division of Structural Engineering Concrete Structures Chalmers University of Technology SE-412 96 Göteborg Sweden Telephone: + 46 (0)31-772 1000

Cover: The cracking and crushing behaviour of a simply supported reinforced concrete beam, where a half of the beam were modelled, using symmetrical boundary conditions. Name of the printers / Department of Civil and Environmental Engineering Göteborg, Sweden 2005

.

using plane stress. certain combination of concrete parameters β and κ was selected for each model. it was concluded that non-linear analysis performed in SOLVIA can be used for limit state analysis of simply supported reinforced concrete beams. In general. finite element method.Finite element analyses for limit state design of concrete structures FERNANDO LOPEZ FEJSAL AL-HAZAM Master’s Thesis in the International Master’s programme Structural Engineering Division of Structural Engineering Concrete Structures Chalmers University of Technology ABSTRACT During the last decades there have been many attempts at defining robust and efficient models that can represent the cracking and crushing behavior of concrete structures. displacements and stresses in stirrups vary with the different combinations of β and κ. Among the most popular ones are the models based on the so-called “smeared crack approach”. The present study considers non-linear finite element analyses performed on five different FE-models implemented in SOLVIA FE-system. crack pattern. All these features make the smeared crack approach particularly well suited for implementation in finite element context. smeared crack approach. fixed orthogonal cracks.5-4.2. The reason for the parametric investigation was to examine how the ultimate loads. analyzed for shear type of failure. displacement. Master’s Thesis 2005:17 I . while κ could be calculated from the fracture energy of concrete and increased with respect to the tension stiffening effect. it was found that. The crack pattern doesn’t need to be known in advance. ultimate load. In order to get stable analysis and a good numerical prediction of the FE-results. It was concluded that the parameter β had to be chosen between 0. ultimate crack strain. non-linear FE-analysis. CHALMERS Civil and Environmental Engineering.85 and 0. ultimate failure capacity. depending on the FE-model. β and the parameter which defines the ultimate crack strain κ. in order to get stable analysis with good prediction of the results. It was found that this parameter had to be multiplied with factors within range 1. The focus was on the following concrete parameters: the shear retention factor. The FE-results for each model were compared with the experimental results in order to evaluate the accuracy of the FE-analysis. The advantage here is that a crack solid is modeled as a continuum and cracked material can be described with stress-strain relationship. The models were created for three types of simply supported reinforced concrete beams in order to examine shear failure. stress. A parametric investigation was also performed.9. solid and truss elements. From the comparison. Key words: shear retention factor. all the models gave reasonably good numerical predictions and similar results (within a reasonable range) to the results according to experimental and theoretical investigations.

2 Reinforcing steel 5.1 1.2 The concrete material model in SOLVIA 4.2 Convergence criteria for the different methods 5 DESCRIPTION OF TESTED BEAMS 5.3 Test performance 5.2 Failure envelopes 4.1 Non-linear behavior of concrete structures 2.3 Bending 2.2 Loads and deflections II CHALMERS Civil and Environmental Engineering.1 Stress strain relationship 4.2 Compression 2.1.2.1 Concrete 5.4 Results from the experiments 5.4.2.Contents ABSTRACT CONTENTS PREFACE NOTATIONS 1 INTRODUCTION 1.2 Material properties 5. Master’s Thesis 2005:17 .1.1 Tension 2.1.1.1 General 4.2.2.2 Smeared crack approach The shear retention factor 4 THE SOLVIA FE-SYSTEM 4.1 3.3 Numerical iteration methods and convergence criteria 4.1 Type of failures 5.3.1 The BFGS method with line searches 4.3.4.2 2 Background Purpose I II IV V 1 1 1 3 3 5 8 9 10 18 18 19 22 22 22 23 27 30 30 32 33 33 35 35 36 37 37 38 38 LIMIT STATE DESIGN OF REINFORCED CONCRETE STRUCTURES 2.4 Shear 3 CRACK MODELING OF CONCRETE STRUCTURES 3.1 Test specimen 5.

4.1.2.3 Model 3 7.7 6.3 Stresses and strains in the concrete and reinforcement 7.2 Crack pattern 7.1 Load-displacement relationship 7.2 Crack pattern 7.2 Crack pattern 7.3 Stresses and strains in the concrete and reinforcement 7.1 Load-displacement relationship 7.2 Crack pattern 7.1.1 Load-displacement relationship 7.8 6.3 6.5.3 Stresses and strains in the concrete and reinforcement 7.4.3 Stresses and strains in the concrete and reinforcement 7. Master’s Thesis 2005:17 .4 6.2 6.1 Load-displacement relationship 7.5 6.1 Model 1 7.4.2.3.6 6.3 Stresses and strains in the concrete and reinforcement 7.1 Load-displacement relationship 7.10 General Coordinate system Geometry Material models Boundary conditions and loads Mesh for model 1 Mesh for model 2 Mesh for model 3 Mesh for model 4 Mesh for model 5 39 39 39 39 39 41 41 42 43 44 46 48 49 49 50 52 53 53 55 57 58 58 60 62 63 63 65 68 69 69 71 74 75 77 83 85 88 III 7 FE-RESULTS 7.4 Model 4 7.5 Model 5 7.5.9 6.2.5.3.2 Crack pattern 7.3.1 6.6 FINITE ELEMENT MODELING 6.2 Model 2 7.1.6 Evaluation of the FE results 8 9 10 PARAMETRIC INVESTIGATIONS CONCLUSIONS REFERENCES APPENDIX CHALMERS Civil and Environmental Engineering.

The examiner and supervisor was Mario Plos from the Department of Structural Engineering at Chalmers University of Technology who supervised and examined us. Fernando López. We would like to express our most sincere gratefulness to our supervisor at Vägverket Konsult. we wish to thank our families. Finally.Preface This master thesis was carried out at Vägverket Konsult as a final project of the Master program in Structural Engineering at Chalmers University of Technology. We are thankful for his guidance during the work with this project. We are also very grateful to Björn Engström professor from the Department of Structural Engineering at Chalmers University of Technology for providing us with well documented experimental results needed partly as a base for the project. cooperation and help in the use of the FE software. We are also very thankful to Mikael Hermansson. friends and people at the company who supported us during the project work period. support and outstanding encouragement throughout the development of this project. for his constant contribution. Gothenburg. We wish to thank Mats Carlson for allowing us to work at the company and Roland Olsson who provided the working place and all necessary equipment. IV CHALMERS Civil and Environmental Engineering. Master’s Thesis 2005:17 . Gothenburg. Fejsal Al-Hazam. February 2005. Costin Pacoste for his devoted guidance. Sweden.

Incremental stiffness matrix following a tensile failure. Fracture energy. Tangent modulus at which stresses normal to a crack plane cease. Reference force.Notations Roman upper case letters Aef As As’ As0 Asv C Ec Eo Es Es’ Esh Esp Et F Fc Fs Fs’ G Gf K L M Mn Mu P P* Pa Pcr Pf Pu Pv R RMNORM RNORM RTOL STOL U & U && U Effective tension area. Cross sectional area of shear reinforcement. Reference moment. Calculated ultimate load. Length. Ultimate test load. Constant mass matrix. CHALMERS Civil and Environmental Engineering. Force carried by the concrete. Force carried by the compressive reinforcing steel. Compressive force in the concrete cross section. Moment due to factored loads. Nodal point acceleration vector. Ultimate finite element load. Master’s Thesis 2005:17 V . Nodal point force vector equivalent to the element stresses. Force carried by the tensile reinforcing steel. Modulus of elasticity of tensile reinforcing steel. Modulus of elasticity of steel plates. Nominal flexural capacity. Calculated ultimate load governed by shear. Tangent stiffness matrix based on the iteration history. Area of longitudinal compressive reinforcing steel. External load vector. Vector of nodal point displacements. Area of the least reinforcement in the tensile zone extending not less than lb. Constant damping matrix Modulus of elasticity of concrete. Tolerance used to measure line search convergence. Diagonal tension crack load. Area of longitudinal tensile reinforcing steel. Unfactored concentrated load. Shear modulus of elasticity of concrete. Calculated ultimate load governed by flexure.net + d. Modulus of elasticity of compressive steel reinforcement. Tangent modulus of elasticity at zero strain of concrete. Relative force and moment tolerance used to measure equilibrium iteration convergence. Force. Strain hardening modulus of reinforcing steel. Nodal point velocity vector.

Yield strength of concrete. Clear beam span. Depth of the neutral axis. Modulus of rupture of concrete. Required anchorage length.fl fu fv fy fy’ fyv h i l lb. Coefficient which takes into account the concrete cover and the bar size on the effective concrete area. Tensile strength of concrete. Shear span. Maximum aggregate size of concrete. Shear capacity of a section with shear reinforcement. Spacing of stirrups measured along the longitudinal reinforcement. Shear transfer in the concrete compression zone. Ultimate strength of reinforcing steel. distance between concentrated load and face support. Yield strength of compressive steel reinforcement. Function of. Nominal shear force carried by shear reinforcement. Splitting tensile strength of concrete. Yield strength of tensile steel reinforcement. Shear transfer by dowel action of the longitudinal tensile reinforcement. Compressive strength of concrete. Average crack spacing Time. Ultimate compressive strength of concrete. Basic shear strength of members without shear reinforcement. Ultimate crack displacement.net lc s srm t w wu x Interface shear transfer by aggregate interlock along a diagonal crack.Va Vay Vc Vcz Vd VR Vs Vu a atp bw c d d’ def dmax f( ) fc fcu fcy ft. Level of tensile steel centroid. Width of the cross section. Master’s Thesis 2005:17 . Yield strength of shear reinforcement. Distance from the extreme top fiber to the centroid of tension reinforcement. Concrete cover.sp ft ft. Crack displacement. Vertical component of the shear transferred across a crack. Iteration. Shear force due to factored loads. Length. Distance from the extreme top fiber to the centroid of compression reinforcement. Roman lower case letters VI CHALMERS Civil and Environmental Engineering. Shear resistance of concrete without stirrups. Height.

Compressive steel strain. Ultimate strain of steel reinforcement. Shear stress with respect to the cracked xz plane.y.6⋅fc. Greek upper case letters β ∆ ϒxy’ ϒxz’ ϒyz’ τxy’ τxz’ τyz’ U Greek lower case letters α δa δmax ε εc εc1 εcu εs ε s’ εx’. Stress. Shear crack strain with respect to the yz plane. Tensile steel strain. Increment of Shear crack strain with respect to the xy plane.y’. Normal crack strains with respect to the cracked plane in the local coordinates x’. Yield strain of tensile steel reinforcement. εy’εz’ εsu εt εtu εu εy εy’ lr ν ρ ρw σ σs σ s’ σp1. Ultimate compressive strain of concrete. Shear stress with respect to the cracked yz plane. Compressive reinforcement stress.x1 x2 Coefficient which takes into account the bond properties of bars on the crack spacing. Maximum test deflection.z’. Ratio of ordinary tension reinforcement. Shear stress with respect to the cracked xy plane. Yield strain of compressive steel reinforcement. σp3 CHALMERS Civil and Environmental Engineering. Uniaxial elastic strain of concrete at 0. Tensile failure strain of concrete. Ultimate crack strain. Angle of the shear reinforcement to the longitudinal axis of the member. Density of the concrete. Poisson’s ratio. Principal stresses in x. Uniaxial ultimate compressive strain of concrete. Strain. Uniaxial compressive strain of concrete at fc. Coefficient which takes into account the form of the strain distribution on the crack spacing. σp2. Maximum deflection obtained in finite element analysis. Effective reinforcement ratio. Shear crack strain with respect to the xz plane.z direction in the biaxial o triaxial failure envelopes. Displacement vector. Principal stress ratio. Master’s Thesis 2005:17 VII . Shear retention factor. Incremental. Tensile reinforcement stress.

Master’s Thesis 2005:17 .y’. Normal stiffness reduction factor. Normal retention factor. Bar diameter. Acceleration factor obtained from a line search.z.σx’.’ Parameter which defines the ultimate crack strain. σy’ σz’ κ µ ξ ηη θ Ø Normal stresses with respect to the cracked plane in the local coordinates x’. VIII CHALMERS Civil and Environmental Engineering. Constant relating to section depth and curtailment of reinforcement.

implemented in the SOLVIA finite element system. This study considers a concrete material model with fixed orthogonal cracks. especially due to the fact that such elements need to be modeled as continua. 2. The investigation will focus on the following parameters: 1. A parametric investigation of the concrete material model implemented in SOLVIA will be performed in order to examine the behavior of concrete elements in the cracked stage. The advantage here is that a cracked solid is modeled as a continuum allowing the cracked material to be described with a stress-strain relationship. The parameters β and κ can be determined from the material response for plain concrete.1 Introduction Background For analyzing the collapse behavior of concrete structures. 1. In addition. These features make the smeared crack approach particularly well suited for implementation in a finite element context. they are fixed in their original direction. even if they at later stages become closed. However. The SOLVIA model uses a shear retention factor β and the ultimate crack strain κ. The shear retention factorβ. these parameters can often be varied within CHALMERS Civil and Environmental Engineering. but. Among the most popular ones are the models based on the so-called “smeared crack” approach. The type of element used. the need for a realistic description of material behavior becomes more evident still.2 Purpose The main purpose of the master thesis project is to examine the capability and limitations of the SOLVIA FE system for limit state analysis of reinforced concrete structures. proper constitutive and failure criteria even for the case of relative standard structural members are required. and a second crack may develop in an orthogonal plane to the already existing ones.1 1. the type of elements used in the discretization plays also a significant role. For more complex elements. 3. The cracks also continue to influence the cracking process. for reinforced concrete modeled with complete interaction between the reinforcement and concrete. the cracks. Master’s Thesis 2005:17 1 . The study will primary concentrate on the shear collapse mechanism of beam type members. According to the definition of this type of models. once they initiate. Moreover. During the last decades there have been many attempts at defining robust and efficient models that can represent the cracking and crushing behavior of concrete. which together may alleviate the possible effects of locking. The parameter κ which defines the ultimate crack strain. the crack pattern need not be known in advance. at the same time they tend to increase the convergence problems during the computation process.

Master’s Thesis 2005:17 .certain limits without any substantial influence on the overall structure behavior. 2 CHALMERS Civil and Environmental Engineering. while other combinations instead will reduce the numerical analysis problems. Certain choices of the values of the parameters β and κ will reduce the risk of shear locking in concrete elements.

the purpose of design is to ensure that under the worst loading conditions the structure be safe. et al (1999). material properties and constructional tolerances. the component forces obtained are then compared with calculated component resistances. the structure must resist the loads for which it is designed with an adequate factor of safety against collapse. According to Allen. In the case of the ultimate limit states. These resistances are calculated by means of a strength analysis of the component as a function of the specified dimensions and material properties made in the design. Accordingly. it is required to ensure that all other relevant limit states are satisfied by the results produced. In the ultimate limit state. Once the loads are factored. The two principal types of limit states are the ultimate limit state and the serviceability limit state. or a material component in a composite structure. this factor considers the probability of deviations of the load from the value specified in codes. 2.1 Non-linear behavior of concrete structures According to Reda Taha (2003).2 Limit state design of reinforced concrete structures According to Mosley and Bungey. The concrete itself has an inelastic non linear response when the stress reaches 40 – 50% of its ultimate strength while the aggregates and the cement paste have a linear elastic behavior for up to 80% of their ultimate capacity. type of failure.E (1982).1. or it may be a mode of failure such as shear or compression. Also. A component can be a connection. CHALMERS Civil and Environmental Engineering. the designer computes deformations or sectional forces in the structure. As the limit state is approached. workmanship. fatigue and fire resistance. other limit states that may be reached are excessive vibration. et al (1999). the most important features in the serviceability limit state are: deflection. etc. and during normal working conditions. cracking and durability. adequate for its intended use. The relative importance of each state may vary depending on the nature of the structure. According to Mosley and Bungey. When a load is combined with other loads. its factor is decreased by a combination that considers the reduced possibility of different loads acting simultaneously. the mechanical behavior of concrete and its components can be expressed by a stress strain diagram as shown in Figure 2. The normal procedure is to decide the most important limit state for a particular structure and base the design on this. Resistance factors are applied to the component resistances and consider the variations of material properties and dimensions. the computations are based on a theory of structural behavior which suitably represents the real response of the structure. In assessing a particular limit state for a structure it is necessary to consider all the possible variations from loads. member. D. Master’s Thesis 2005:17 3 .

the micro cracks in the transition zone start to extend and develop. Stage II: Stress levels between 30% and 50% of the ultimate stress. It is well known that micro cracks already exist in the transition zone before the application of external loads. Whereas the ascending branch of the concrete stress strain relationship depends only on cracking extent in concrete. four stages of cracking can be identified as shown in Figure 2. cracks in the cement paste zone form about 60% of the ultimate stress. as a result of thermal stresses and shrinkage forces. from Reda Taha and N. Additionally. the micro cracks in the transition zone remain stable and hence the stress strain relationship is linear. bleeding characteristics. the cement paste and the transition zone. The stress strain diagram begins to deviate from the linear behavior at an intermediate stress. When the stress is reaching a value of 75% of the ultimate load. the micro cracks in the transition become more 4 CHALMERS Civil and Environmental Engineering. In concrete under uniaxial compression. Hsu et al (1963). the descending branch is highly influenced by the stiffness of the testing machine end conditions.1 Stress strain relationship of concrete and its components. he reported that in all stages of loading the total amount of mortar cracks were significantly less than the cracks in the transition zone (known as bond cracking). Master’s Thesis 2005:17 . which results in the non linear behavior of the concrete material. Stage III: Stress levels between 50% and 75% of the ultimate stress. The number and width of these cracks in the transition zone depend primarily on wall effect.Figure 2. studied crack propagation in concrete and reported that the micro cracks development and propagation determine the form of the stress strain diagram of concrete under uniaxial compression. thermal and carbonation shrinkage and curing history. Shrive. The transition zone represents the interfacial region between the aggregates and the cement paste. micro cracks can easily propagate in the transition zone whereas the other two phases do not crack. Because of the high porosity and low strength of the transition zone. G. The non linear behavior of concrete can be explained by defining it as a three stage heterogeneous material: the aggregates.2: Stage I: Stress levels up to 30% of the ultimate stress.

2.1 Tension According to Heidkamp (2000). When the stress level reaches a specific value. the response of concrete in tension is well treated with the introduction of fracture mechanics. Master’s Thesis 2005:17 5 . according to the cracking stages: first. This macro crack will propagate until the stress reaches its critical stage. CHALMERS Civil and Environmental Engineering. The stress strain curve reaches a maximum value and a descending branch means that a failure starts.3. The performance of concrete can be described at two levels. Stage IV: Stress levels between 75% and 100% of the ultimate stress. mortar cracks become continuous and propagate quickly.stable and cracks of the cement paste develop rapidly. One example to illustrate this theory is to consider the fracture development of a plain concrete specimen of length L subjected to uniaxial tension under displacement control.1. the micro cracks grow together into a macro crack. due to its continuous deformation localization and failure of its components. One relevant aspect in concrete in tension is that it is a heterogeneous material which exhibits a strain softening behavior. Figure 2. from Mehta and Montrerio (1993).2 Relationship between concrete behavior and the range of cracking. Stabilized propagation of this macro crack results in a strain softening mechanism observed for concrete. shown in Figure 2. micro cracks are formed or enlarged under low stress levels and are distributed randomly. This is because it seems justifiable to deal with cracks in a smeared sense. The stress strain curve in this region is strongly non linear. High strains develop and cracks develop under sustained loads. (see Hirschhausen (1999)).

4.3 Fracture process in a plain concrete specimen. loaded uniaxially under displacement control. Finally. the micro-cracks from the former stage start to connect each to other and to localize into a zone of deformations at the weakest section of the specimen. the fracture zone is modeled as a fictitious crack with zero width before real cracking is initiated. decrease and the total deformation is governed by the material outside the damaged zone and the crack width w. in the loading stage before the maximum stress is reached. the final crack is formed but not open yet (w = 0) and the behavior of the specimen can be described by means of a stress strain relationship with increasing load. In this scheme. Hillerborg (1976). when the micro-cracks lead to separation of the specimen w ≥ wu. as the crack opens. from Plos (1995). Furthermore. After crack initiation. After further loading. a stress crack opening relation represents the softening behavior provided by the damaged process zone. Master’s Thesis 2005:17 . as it seen in Figure 2. According to Plos (1995).a) σ =0 b) σ < ft c) σ = ft ( w = 0) L L+ε L L+ε L d) σ = f(w) (0 < w < w u ) e) σ =0 _ wu ) (w > L+ ε L+w L+w w Figure 2. 6 CHALMERS Civil and Environmental Engineering. In this stage. no further stress can be transferred across the crack. local micro-cracks start to form at weak zones through the specimen. the stresses that can be transferred. A widely accepted model to describe the fracture behavior of concrete is the “Fictitious Crack Model”.

The term “Tension stiffening” is often used to represent the tension contribution of concrete after cracking. are necessary parameters to describe the crack formation besides the fracture energy. concrete will still contribute in the tension carrying capacity of the section by transferring the tensile stresses between cracks. Figure 2.5. CHALMERS Civil and Environmental Engineering. Moreover. which can be deemed as constant for a certain concrete strength class and maximum aggregate size. The area under the stress crack opening is called fracture energy.Figure 2. from Reda Taha (2003). ft. Gf and represents the dissipated energy during the fracture process. This adds to the complexity of the problem of predicting the deformations of concrete structures accurately.4 Stress-strain relation prior to initiation of cracking and stress-crack opening relation for the post cracked state for plain concrete. the shape of the stress crack opening curve and the concrete tensile strength. the tension stiffening effect. 1990). Plos (1995).5 Effect of cracked concrete in the load displacement response of a tensioned reinforced concrete prism. which in turn is the most relevant of fracture mechanics for concrete. Tension stiffening accounts for the fact that even when cracked. Master’s Thesis 2005:17 7 . (CEB-FIB. see Figure 2. from Hillerborg (1976).

Moreover.1. Favre et al. the linear elastic domain in the range of about 40% to 45% approximately with a modulus of elasticity Ec and stresses below the yield strength fcy. the tension stiffening effect can be reasonably determined by modifying the strain values of the stress strain tensile relation for concrete. Ghali et al. cyclic or sustained). 2.6. structural concrete in compression exhibits localization of deformations after its maximum strength. (1985). This means that the strain softening branch can be represented by a continuous stress strain relationship.as it is shown in Figure 2. for stresses between the yield strength and the compressive strength fc. ABAQUS 2000). Eurocode (1991). CEB-FIP (1990). adopted from Heidkamp (2000) 8 CHALMERS Civil and Environmental Engineering. (1994). Ghali et al. quality of bond of the reinforcing steel and the level of tensile stress applied. The effect of the density of reinforcement on tension stiffening should also be incorporated when tension stiffening is used for modelling reinforced concrete elements using the finite element method (Belarbi et al.4. Master’s Thesis 2005:17 .6 Uniaxial stress strain diagram of concrete under compression. Figure 2. (1999) and Gilbert (2001). The reason is that the stresses in a bar embedded in cracked concrete sections are partly transferred to the concrete elements in between the cracks.. The curve can then be divided into three parts. 1994. (1994). There are some important parameters that affect the tension stiffening of concrete for instance: the type of loading (monotonic. the strain hardening domain. A number of variations between these models exist.. but they all are based on interpolation between the uncracked and the cracked analyses. and a descending branch which represents the strain softening domain for stresses up to the ultimate compressive strength fcu.According to Plos (1995).2 Compression According to Heidkamp (2000). it is reflected in the structural response if the reinforcement is modeled with separate elements. This leads to a reduction in convergence problems when using finite element techniques. as the fictitious crack model from Hillerborg shown in Figure 2. Other comprehensive models to account for concrete tension stiffening in reinforced concrete structures include those reported by Favre et al.

The first cracks develop first at mid-span and. under these conditions. it propagates at an angle of 45°. Stage 2. Stage 3. Figure 2. the steel bars and concrete resist tension and compression forces respectively. A sudden collapse of the beam occurs by splitting of the concrete in the tensile zone. In the beam with vertical reinforcement. In that section. The neutral axis moves upwards and the flexural cracks reach the neutral axis. One of the beams has longitudinal reinforcement in the tensile zone. similar diagonal parallel cracks appear in regions with high shear stresses and the stirrups are effective in resisting principal CHALMERS Civil and Environmental Engineering. As the load is applied gradually from zero. At a load level beyond P1. parallel to the direction of principal compressive stresses and continues horizontally until the support. The maximum bending moment occurs at the mid-span of the beam. say P1.1. This results in linear elastic behavior with no cracking. Master’s Thesis 2005:17 9 . To increase the flexural capacity of a concrete beam for instance. the crack width increases and additional cracks form near the supports. the combined action of shear forces and bending moment causes a diagonal crack.3 Bending According to Nadim Hassoun (2002). from Nadim Hassoun (2002). cracks form at the bottom fibers for values of the tensile stresses in the concrete greater than its tensile strength. Stage 4. The concrete still carries minimum tensile stresses but the steel reinforcement is effective in resisting these forces. the tensile stresses in concrete at the bottom fibers are equal to its tensile strength. in which two simply supported beams are loaded to failure. This type of cracking is initiated at a distance of approximately 1. During the different loading stages the following observations can be made: Stage 1. when the maximum tensile stress in concrete reaches its tensile strength at a certain loading stage. see Figure 2. At this stage in the beam without stirrups.7 Shape of the beam 1 at shear failure. At a certain load level. longitudinal reinforcement is used.7. both the tensile stresses at the bottom fibers and the compressive stresses at the top fibers are less than the allowable strength values of concrete. A noticeable deflection of the beam can be observed at this stage. whereas the other has both compressive and tensile steel reinforcement and stirrups as well.2. The flexural behavior of a reinforced concrete beam can be illustrated by means of a simple laboratory test. the beams resist the current bending moment and shear forces at any section. cracks occur. in the top fibers.5 times the height of the beam from the support. as the load increases.

For small load increments. from MacGregor (1997). 10 CHALMERS Civil and Environmental Engineering. Master’s Thesis 2005:17 . At this stage. Figure 2.4 Shear The factors influencing inclined cracks and shear strength are so complex and numerous that it is difficult to establish a definitive conclusion regarding the mechanisms that govern such failures. and cracks widen sharply. because the shear failure is normally sudden and brittle and the inclined cracks that form are significantly larger than flexural cracks.1. The behavior of reinforced concrete beams under shear failure is different from their behavior at flexure.8.stresses. Principally three types of failures occur in concrete beams without shear reinforcement: flexural failure.9 Anchorage failure and arch action behavior of a simply supported beam. and due to the brittle manner in which cracks occur. the deformations in the steel reach the yielding strains. the steel strains develop abruptly. Regarding the design of members for shear. adopted from Nawy (2003). 2. The stirrups prevent at this load level.8 Shape of the beam 2 at bending moment failure. Types of failure of concrete beams without shear reinforcement. diagonal tension failure and shear compression failure. The compressive strains in the concrete as well as the deflections increase. Stage 5. with the steel reinforcement developing strain hardening and the concrete crushing at the top compressive fibers. No longitudinal splitting along the tensile reinforcement can be developed. the collapse of the beam. see Figure 2. Figure 2. from Nadim Hassoun (2002). the engineer has to consider sections that are strong enough to resist the external factored loads without reaching their shear strength capacity. but still the concrete strains are still lower than the allowable value. The collapse of the beam occurs for a small load increment.

failure occurs in a ductile manner and it is initiated by the yielding of the tensile reinforcement.11 Flexural failure pattern as a function of beam slenderness. These cracks result from very small shear stresses and a dominant flexural stresses. As the external load increases. The diagonal tension failure starts with the formation of few vertical flexural cracks at the mid-span.Figure 2. Figure 2. CHALMERS Civil and Environmental Engineering. cracks are mainly vertical. However. This is called flexural failure. Nawy (2003). In regions of large bending moments. the steel does not yield nor give failure warning prior to collapse.5 ≤ a ≤ 2. As the inclined cracks stabilize. 0.0 .12. If the beam is under-reinforced. the bond between the reinforcing steel and the surrounding concrete is destroyed at the support by longitudinal splitting.5 to 2d distance from the face of the support.11. from d MacGregor (1997). see Figure 2. Few inclined cracks form at about 1. if the beam is over-reinforced. Thereafter. The initial cracks widen and extend deeper towards the compressed fibers causing a marked deflection of the beam. they become wider and form a principal diagonal extending to the top compression fibers. weakening the anchorage of the reinforcement. Master’s Thesis 2005:17 11 . from Nawy (2003).10 Modes of failure of short shear span beams. additional cracks develop in the same region of the beam. see Figure 2.

In the shear compression failure few flexural cracks start to develop at the mid-span and stop propagating as destruction of the bond occurs between the longitudinal bars and the surrounding concrete at the support region. Internal forces in a beam without stirrups in shear failure. Thereafter. The final failure occurs suddenly when the inclined crack joins the crushed concrete zone.13 Shear compression failure pattern as a function of beam slenderness.14. 12 CHALMERS Civil and Environmental Engineering.12 Diagonal tension failure pattern as a function of beam slenderness. Master’s Thesis 2005:17 . Figure 2.Figure 2. an inclined steep crack suddenly occurs and continues to propagate towards the neutral axis. see Figure 2. from Nawy (2003) .13. The internal forces transferring shear across an inclined crack are shown in Figure 2. from Nawy (2003). The rate of its progress is reduced with crushing of the concrete and redistribution of stresses in the top compression fibers.

Right after crack initiation. and Vd. from MacGregor (1997). Vc = 2 f c b w d ( 2.0075 ≤ ρw ≤ 0. the vertical component of the shear transferred across the crack on both of its faces by interlock of the aggregate particles. the dowel action of the longitudinal reinforcement. Tensile concrete strength. Vay. A reinforced concrete beam without web reinforcement fails when inclined shear cracking occurs and hence its shear capacity is equal to the inclined cracking shear. Inclined cracking occurs at a principal tensile stress of one third of the tensile strength on uncracked sections. the shear is transferred across the crack trajectory ABC by Vcz. Factors affecting the shear strength of reinforced concrete beams without stirrups.1) CHALMERS Civil and Environmental Engineering.025 . when flexural cracks which appear before diagonal failures delay the elastic stress field developed at such stage. from MacGregor (1997). When this two shear forces disappear. This causes a splitting crack in the concrete along the reinforcement and Vd and Va to drop to zero. The cracking load is frequently related to the split cylinder test in biaxial state of stresses.14 Internal forces in a cracked beam w Figure 2.2. As the crack opens with increasing the shear force Vcz and the dowel force Vd. According to MacGregor (1997). when the steel ratio ρw=As /bwd varies between 0. Longitudinal reinforcement ratio. The following Equation expresses the shear strength of reinforced concrete beams without vertical reinforcement. the transferred shear Va decreases. Master’s Thesis 2005:17 13 . all the shear and compression are transmitted through the portion AB above the crack causing crushing of this section. about 40% to 60% of the total shear is carried by Vd and Vay.14 Internal forces in a cracked beam without stirrups. the shear in the compression zone.

inclined cracking occurs earlier. Shear span to depth ratio.15. The maximum moment that occurs correspond to the moment capacity Mn of the cross section. 14 CHALMERS Civil and Environmental Engineering. Figure 2. very short shear spans 0 > a/d >1. Master’s Thesis 2005:17 . According to Figure 2.16. Figure 2.17). This quotient is the relation between the shear span and the depth of the tensile reinforcement. It is represented as a/d. As a result.15 Geometry and loading of a simply supported beam with symmetric loads. flexural cracks extend higher into the beam and open wider than would be the case for large values of ρw . from MacGregor (1997). Moments and shears at diagonal cracking in a rectangular beam without web reinforcement can be represented as a function of this factor as shown in Figure 2. develop inclined cracks as links between the loads and supports.16 Shear at cracking and failure as a function of a/d ratio. from MacGregor (1997). and has a certain effect on the shear resistance.When the steel ratio ρw is small. destroying the transfer of shear flow from the longitudinal steel to the compression zone due to arch action (see Figure 2.

Figure 2. splitting failure.19 Figure 2.20. as is shown in Figure 2. The final failure for this beam is a bond failure. The most common type of failure in a beam with this shear to span ratio is anchorage failure at the ends of the tension tie. Master’s Thesis 2005:17 15 . the beam fails at the inclined cracking load shown in Figure 2. from MacGregor (1997). or dowel failure along the tensile reinforcement.18 Shear tension failure.18.5 > a/d >6. the steel acts as a tension tie of a tie arch with uniform tensile force.5. Short shear spans 1 > a/d >2.17 Arch action in a beam.19 Shear compression failure. For beams with slender shear spans 2. from MacGregor (1997). develop inclined cracks and after a redistribution of internal forces. are able to carry additional load partly by arch action. Subsequently. Splitting crack or by crushing of the compressive fibers over the crack. CHALMERS Civil and Environmental Engineering. Figure 2. from MacGregor (1997). see Figure 2. when the equilibrium is disrupted by such discontinuities.

from MacGregor (1997).21 below.16 and Figure 2. leading to a reduction of the aggregate interlock across the crack and hence earlier splitting. The shear stress at diagonal cracking tends to decrease for given values of a/d. the shear required to form an inclined crack exceeds the flexural failure load and therefore the beam fails in flexure before prior the formation of inclined cracks. crack widths in the tensile zone increase with increasing overall depth. see Figures 2. fc and ρw as the overall depth of the section increases. the shear reinforcement contributes very little to the shear resistance. failure due to yielding of the stirrups may be expected. For beams with very slender shear spans a/d < 6. The internal forces transferring shear across an inclined crack are shown in Figure 2. Size of the beam.20 Moments at cracking and failure. According to Bresler and Scordelis (1961).Figure 2. Master’s Thesis 2005:17 .14. Prior the development of diagonal cracks. When the amount of shear reinforcement is too small. but if the amount is too high a shear compression failure can occur. 16 CHALMERS Civil and Environmental Engineering.20. Behavior of beams with web reinforcement. the shear capacity of a reinforced concrete beam is increased by using vertical reinforcement. The vertical reinforcement increases the shear capacity of the beam after diagonal cracking and a redistribution of internal stresses develop in the cross section. The notations are the same as in Figure 2. Also.

When the diagonal crack opens. the inclined crack opens more rapidly causing Vay to decrease and Vcz and Vd to increase rapidly as well. CHALMERS Civil and Environmental Engineering.21 Internal forces in a cracked beam with stirrups. Before flexural cracks all the shear is carried by the concrete. the stirrups crossing the inclined crack yield and Vs remain constant for higher shear forces. Master’s Thesis 2005:17 17 . Between the flexural and inclined cracking stages. When the stirrups yield. from Macgregor (1997). The shear transferred by tension in the stirrups is Vs. As a consequence T2 becomes less than T1 and this difference depends on the amount of shear reinforcement. Vay and Vcz.Figure 2. the compressive force C1’ and the shear force Vcz’. this action continues until either a splitting failure or crushing in the top fibers occurs due to a combination of shear and compression forces (from MacGregor (1997)). the external shear is carried by Vd. act always on the part of the beam below the crack. Eventually.

the use of non-linear finite element methods is necessary. to analyze properly the behavior of concrete structures in the ultimate limit state. In the case of softening materials. reliable finite element crack models must have successful numerical treatment methods as a prerequisite. This allows the element size to be similar as that of the specimens from which the constitutive relations and fracture criteria were obtained. The smeared crack approach is based on the development of appropriate continuum material models. governing its failure mechanism. Structural engineers must have basic knowledge in the use of finite element programs and methods due to the constant and rapid development of such techniques.1 Smeared crack approach One of the fundamental features required in order to analyze reinforced concrete structures in a realistic way. Cracking represents the key feature in modeling the nonlinear behavior of reinforced concrete.3 Crack modeling of concrete structures According to Plos (1995). this can lead to strong mesh size dependence and no energy dissipation in the finite element calculations. cracks are smeared over a certain finite element area corresponding to a Gauss point of the finite element. type of elements. but also on a number of other options such as material models. iteration method and convergence criteria. modeling of applied loads and supports and so on. The smeared crack models are usually formulated in stress strain space. Master’s Thesis 2005:17 . rotating crack models and plasticity 18 CHALMERS Civil and Environmental Engineering. fixed non-orthogonal crack models. special considerations are necessary to introduce some of the energy concepts of fracture mechanics. The main issue in this approach is the modification of the stiffness properties and equilibrium conditions at integration points of cracked areas. Choosing a linear method can lead to incorrect results due to the strongly non linear material behavior caused by cracking. The smeared crack models are classified into four categories: fixed orthogonal crack models. 3. The description of cracking and failure within finite element analysis of quasi-brittle materials such as concrete has led to two fundamentally different approaches: the discrete and the smeared one. Smeared crack models are nowadays widely used by structural engineers for most practical analyses since these models give reasonable and accurate results at lower costs than in the discrete crack models. A computational advantage in this technique is that no mesh modifications are required throughout the analysis of the structure. Jirásek (1998). Therefore. Pavlovič (1995). The quality of the analysis depends on one hand on the choice regarding the level of detailing on which the analyst models the problem. lies on a suitable material model. Cervenka Consulting (2000). Traditional smeared-crack models are known to be susceptible to stress locking and possible instability at late stages of the loading process. This is the case when the concrete crushes in the compressive zone and the reinforcement yields in the cracked tensile zones. Consequently.

1 Shear transfer mechanism in a simply supported reinforced concrete beam. According to Pavlovič (1995). aggregate interlock. structural members subjected to shear forces. etc. the cracking response of concrete structures can be described by continuum constitutive models that account for the strain softening normal to the cracked plane.1 which has a flexural shear failure that propagates from the extreme compression fiber to the tensile reinforcement. dowel action and aggregate interlock. from Basu (2003).models. 3. Master’s Thesis 2005:17 19 . in finite element analyses. Va and dowel action of the longitudinal tension steel. The latter is defined as the portion of aggregates from one side of a crack into open spaces for the transfer of compressive and shear loads through the crack trajectory with parallel directions between the crack surfaces. Additional cracks can form at the same point but only in a plane which is orthogonal to the existing ones. Shear stresses in concrete members may produce inclined cracks across the section as well as a sliding type of failure along a well defined plane due to external tension. finite element models consider the reduction in shear stiffness along the cracked plane at integration points CHALMERS Civil and Environmental Engineering. In the fixed orthogonal crack models once a crack is initiated it is locked in its original direction. Upon the application of shear forces the problem of quantifying the shear stress transfer across the crack sections arises. In this project. Also. Now. it is important to point out the parameter called Shear Retention Factor (β) in numerical analysis. shrinkage. perpendicular to the principle tensile stress at the initiation point. In addition. axial forces and sometimes torsion. Vd. Consider the beam in Figure 3. The general shear transfer mechanisms are: through uncracked concrete. arch or truss analogy. fixed orthogonal crack models have been used. The shear is transferred across the crack by a combination of the following three actions: direct shear transfer by the uncracked concrete in the compression zone Vcz.2 The shear retention factor According to Buyukozturk (2004). when the concept of aggregate interlock has been clearly outlined. Figure 3.e. once a crack is formed it is never removed and continues to influence the failure process even if at later stages during the loading history it is completely closed. i. resist also the effects of flexure. load history.

µ is a normal retention factor. A constitutive relation at each integration point may be expressed in the form: ⎡ ∆σ x ' ⎤ ⎡2G + µ ⎢ ∆σ '⎥ ⎢ ⎢ y⎥ ⎢ µ ⎢ ∆σ z ' ⎥ ⎢ 0 ⎥=⎢ ⎢ ⎢∆τ xy '⎥ ⎢ 0 ⎢ ∆τ xz '⎥ ⎢ 0 ⎥ ⎢ ⎢ ⎢ ⎣ 0 ⎦ ⎢ ⎣∆τ yz '⎥ 0 ⎤ ⎡ ∆ε x ' ⎤ ⎢ ∆ε ' ⎥ 2G + µ 0 0 0 0 ⎥ ⎥⎢ y ⎥ 0 0 0 0 0 ⎥ ⎢ ∆ε z ' ⎥ ⎥ ⎥⎢ 0 0 G 0 0 ⎥ ⎢∆γ xy '⎥ 0 0 0 β G 0 ⎥ ⎢ ∆γ xz '⎥ ⎥ ⎥⎢ 0 0 0 0 βG ⎥ ⎦⎢ ⎦ ⎣∆γ yz '⎥ µ 0 0 0 ( 3. ∆γ xy . Master’s Thesis 2005:17 .1) Where G is the shear modulus of concrete. are the increments of normal stresses in the global coordinates. This means that the axes coincide with the principal stress directions with reference to a local coordinate system (x’. ∆ε z are the increments of normal crack strains in the global coordinates. ∆τ xy .y’) across and along the crack. ∆τ yz are the increments of shear stresses in the global coordinates. Parameter β represents the shear retention factor. ∆γ yz are the increments of shear crack strains in the global coordinates. a constitutive C-matrix becomes anisotropic and the material coordinate system rotates. Before cracking. the stress state changes from biaxial to uniaxial. ∆σ x .1). ∆σ y . When cracking occurs. ∆ε x . ∆τ xz .in an element. Therefore. ∆σ z . the constitutive relations at each integration point can be expressed in the form: ⎡ ∆σ x ⎤ ⎡2G + µ ⎢ ∆σ ⎥ ⎢ ⎢ y⎥ ⎢ µ ⎢ ∆σ z ⎥ ⎢ µ ⎢ ⎥=⎢ ⎢∆τ xy ⎥ ⎢ 0 ⎢ ∆τ xz ⎥ ⎢ 0 ⎢ ⎥ ⎢ ⎢∆τ yz ⎦ ⎥ ⎢ ⎣ 0 ⎣ µ 2G + µ µ 0 0 0 0 ⎤ ⎡ ∆ε x ⎤ ⎢ ∆ε ⎥ 0 0 0⎥ ⎥⎢ y ⎥ 2G + µ 0 0 0 ⎥ ⎢ ∆ε z ⎥ ⎥ ⎥⎢ 0 G 0 0 ⎥ ⎢∆γ xy ⎥ 0 0 G 0 ⎥ ⎢ ∆γ xz ⎥ ⎥ ⎥⎢ 0 0 0 G⎥ ⎢∆γ yz ⎦ ⎥ ⎦⎣ µ µ 0 0 ( 3. These points can be defined as the points in an element used for the numerical integration of stiffness matrices and internal force vector. Consequently. but within the local coordinate system. ∆ε y .2) Where the notations in Equation (3. ∆γ xz . As a consequence. 20 CHALMERS Civil and Environmental Engineering. strains develop orthogonally to the direction of the maximum principal tensile stresses. material properties of finite element meshes need to be evaluated at these points.2) are the same as in Equation (3.

It should be pointed out that the shear retention factor is a parameter associated with the concept of aggregate interlock.This parameter quantifies the decrease in the incremental shear stiffness that occurs due to cracking. since a crack propagates in the direction of the maximum principal compressive stresses and opens in the orthogonal direction. By performing a finite element analysis to determine the ultimate load it is noticed that these conditions cause a drop of about 30% in the load carrying capacity of the beam in comparison to the analysis of the same beam without preformed cracks. A widely held assumption states that aggregate interlock plays an important role in transferring shear across cracked areas. Aggregate interlock is not in agreement with the fundamental failure mechanisms of concrete.1) and even more when β is reduced from 0. CHALMERS Civil and Environmental Engineering.2) leads to either zero values in the diagonal elements or to ill conditioning. which means that when cracking occurs the remaining stiffness in the post cracking concrete is negligible. it is common to have the whole concrete finite elements in tensile zones so that all their integration points may crack. experimental evidences point out contrary conclusions. In practical situations. the modeling of brittle materials requires minimizing this possibility by the assignment of a non zero value to β. Master’s Thesis 2005:17 21 . Therefore. What is the value that should be assigned to β? The absence of a shear retention factor and for an element with all the integration points cracked in parallel planes.1 to 0. This is the case of a simply supported reinforced concrete beam subjected to two symmetrical point loads with cracks preformed in the region below the neutral axis. However. the stiffness matrix C in Equation (3. The number of mechanisms increases when the structure goes from the uncracked to the cracked stage (β = 0. This is because aggregate interlock does not have any effect on the shear resistance.

These models should be able to predict the non linear load deformation characteristics of the fracture process. von Hirschhausen (1999). A general overview of the process in generating models and obtaining different results is presented. Master’s Thesis 2005:17 . from SOLVIA Engineering AB (1999). SOLVIA Engineering AB (2000). The analyses models are defined by SOLVIA-PRE which is free format and is given as commands. SOLVIA and SOLVIA-POST. The nodal and element results can be displayed. see Figure 4. parameters and data lines. incremental equilibrium equations are used to solve nonlinear static displacement and stress analyses with different iteration methods based on time dependent variables. In the SOLVIA-POST database the results from the analyses are stored.2 The concrete material model in SOLVIA To consider non linear deformations in reinforced concrete structures. For the analysis of a reinforced concrete beam. stresses and temperatures under static or dynamic conditions. In the SOLVIA stage.1 The three stages of SOLVIA. 22 CHALMERS Civil and Environmental Engineering. listed and scanned with components referenced to a user selected coordinate system.4 4. 4. it is necessary to define suitable material models to represent this behavior.1 The SOLVIA FE-System General SOLVIA is a powerful finite element system for linear and nonlinear analysis of displacements. Figure 4.1. the modeling should consider apart from a concrete model. a material model able to describe the behavior of the reinforcing steel. It consists of three processors SOLVIA-PRE. This is where all input data is generated and stored as well as built up in a logical sequence.

1) σ / fc = ( Eo / Ec )(ε / εc ) 1 + A(ε / εc ) + B(ε / εc ) 2 + C (ε / εc ) 3 ( 4. concrete. A common uniaxial stress strain curve relation for the concrete material model in SOLVIA is shown in Figure 4.2) Eo[1 − B(ε / εc ) 2 − 2C (ε / εc ) 3 ] E= [1 + A(ε / εc ) + B (ε / εc ) 2 + C (ε / εc ) 3 ] 2 ( 4.2 Uniaxial stress strain relationship for concrete The material is in tension when ε > 0 and the stress strain curve is linear with an elastic Young’s modulus Eo until the maximum tensile strength is reached. and εc > ε ≥ εu .2.3) CHALMERS Civil and Environmental Engineering. Master’s Thesis 2005:17 23 . plastic and elastic to simulate the collapse behavior of the different beams. σ = Eo ⋅ ε When the material is in compression ε < 0 the following relations are employed ( 4.2. which shows three phases: ε ≥ 0.1 Stress strain relationship The concrete material model in SOLVIA is considered as plastic based on a non linear uniaxial stress strain relation including strain softening. σ ft 1 Et ⋅ηη εu εc Εc Εo 1 κ ⋅ εt εt = Εo ft ε fcu fc Figure 4.e. 0 > ε ≥ εc.In the SOLVIA finite element analyses three different material models were used i. 4. These features allow the analyst to model tensile failures (cracking) using the smeared crack approach and compression failures (crushing) by failure envelopes.

Master’s Thesis 2005:17 . see Figure 4.3. The parameters Eo .A= ( Eo / Eu ) + ( p 3 − 2 p 2 )( Eo / Ec ) − (2 p 3 − 3 p 2 + 1) p ( p 2 − 2 p + 1) 2 Eo − 3 − 2A Ec Eo +A Ec ( 4. a comparison among different stress strain curves for concrete is presented by means of several plots. However.4) B= ( 4. Ec. and Eu = fcu / εu can be obtained from uniaxial tests. fcu. 24 CHALMERS Civil and Environmental Engineering. εc. fc.5) ( 4. εu. because sometimes is difficult to determine accurate values. In order to define the mentioned parameters of concrete.6) C = 2− The stress strain relationship in the above equations assumes monotonic loading conditions. p = εu / εc . they can be assumed reasonably by the analyst within a base range.

5 0.0 -2.7) The compressive strain εc at fc is found from the parabolic stress strain curve by the following relation: CHALMERS Civil and Environmental Engineering.0 -3 -5 OEHLERS & JOHNSON BBK SOLVIA -10 -15 -20 -25 -30 -35 -40 Figure 4.5 3. Master’s Thesis 2005:17 25 .0 -3.0 -0.3 are calculated according to Oehlers (1995) and Johnson (2004).0 0.10 Stress [MPa] 5 Strain [10 ] 0 -4.3 Uniaxial concrete stress strain relations based on different approaches. From the CSA Code (Ref 3-7) the initial tangent modulus Eo is assumed to be about 10 percent greater than the secant modulus: Eo = 5500 fc (MPa units) ( 4.5 -1.5 -2. The principal parameters in the curve with rectangular symbols in Figure 4.5 4.0 1.0 2.0 -1.0 3.5 1.5 -3.5 2.

adopted from BBK (1994).12) ( 4.11) The non linear branch εc1 < ε < εc is approximated by σ ⎛ εc − ε ⎞ = 1 − 0.4 Uniaxial stress strain relationship for concrete.2. Figure 4.85 ⋅ fc ( 4.6 fc Ec ( 4. The linear portion of the curve is defined as εc1 = 0.10) Where f ' c = fcu The curve with triangular symbols was calculated using the equations given in BBK.10 which represents the unloading portion of the curve in Figure 4.4⎜ ⎟ fc ⎝ εc − εc1 ⎠ k ( 4.5⎜ − 1⎟ ⎝ εc1 ⎠ 26 CHALMERS Civil and Environmental Engineering. giving the post-peak region of the stress strain diagram.14) εc = 0.εc = 2 fc Eo ( 4.0020 ⎛ εc ⎞ k = 1.9) The ultimate compressive strain εu can be calculated with Equation 4. 1+ εu = εc ⎜ ⎜ ⎜ ⎝ ⎛ ⎜ f 'c ⎟ ⎟ ⎟ fc ⎟ ⎠ ⎞ ( 4.8) The ultimate compressive stress fcu can be determined by fcu = 0.13) ( 4. Master’s Thesis 2005:17 .

15) Where ρ is the density of the concrete in (kg/m³) assumed to be 2400 for this case. material failures envelopes are employed to model the material failures of concrete in tension and compression in two and three dimensional analyses as well as to account for multiaxial conditions in the stress strain diagram.2 Failure envelopes According to SOLVIA Engineering (1995). from the SOLVIA concrete model. 4.0035 and 0. 0.0035. As it is seen the material tensile strength decreases with the introduction of compressive stresses in any of the two principal directions.3 + 0.εcu = ⎜ 0. but not with tensile stresses.95 fc . from SOLVIA Engineering (1995). which gives εcu = 0. the principal parameters in the last curve without symbols can be defined with the ranges: 0.0031 ≤ εcu ≤ 0.5 Biaxial failure envelope. A biaxial failure envelope in two directions is shown in Figure 4.0022 . Master’s Thesis 2005:17 27 . Figure 4.0020 ≤ εc ≤ 0.8 fc ≤ fcu ≤ 0. Finally.0035 2400 ⎠ ρ ⎞ ( 4. The biaxial failure envelope is a special case of the triaxial failure envelope for crushing and splitting failures for two dimensional analyses. CHALMERS Civil and Environmental Engineering.7 ⎛ ⎝ ⎟ ⋅ 0.5.2.

and σp1 > σp2 = σp3. The triaxial tensile failure envelope is a general case of the biaxial envelope in two dimensional analyses (see Figure 4. Biaxial and triaxial compression loads delay the formation of bond and mortar cracks because the period of stable crack propagation is longer and concrete becomes more ductile. 28 CHALMERS Civil and Environmental Engineering. sections parallel to the plane σp2 . Figure 4. from SOLVIA Engineering (1995). the compressive failure envelope is defined by 24 points in terms of discrete ratios as shown in Figure 4.7 Triaxial failure envelope for concrete in compression.6 Triaxial failure envelope for concrete in tension.7 These input points should cover all the expected principal stress values (σp1 ≥ σp2. In this region.6). σp3.7.Figure 4. In this plane. σp1 > σp2 = β σp3 (with 0 < β < 1). from SOLVIA Engineering (1995).σp3 are input as well.σp3 stresses. However. Master’s Thesis 2005:17 . only one section is required. where β = σp2/σp3 and represents the principal stress ratio in Figure 4. This is defined by σp1 = σp2 > σp3 to σp1 > σp2 = σp3. For six different input values of σp1. where σp3 is the minimum principal stress). a bilinear failure envelope is obtained from the input values of σp3 at σp1 > σp3 = σp2. For values of σp1 in between the input points. linear interpolation is used to define a biaxial failure surface for the σp2 .

This gradient Et can also be realistically estimated by Et = -70Ec/(57+ft). The factor β is also referred to as the shear retention factor. Bazant (1984). η n is the normal stiffness reduction factor. Master’s Thesis 2005:17 29 . The normal and shear stiffness in a crack plane are reduced by two factors denoted as η n and β respectively. the parameter κ governs the gradient of the descending branch beyond the cracking strain of concrete. with an increase in strains and decrease of stiffness and stress across the crack plane respectively. Etηη Ec y Etηη νEc Ec m 0 0 0 Gβ m 0 0 0 0 Gβ 0⎤ 0⎥ ⎥ 0⎥ ⎥ 0⎥ 0⎥ ⎥ G⎦ ⎥ ( 4. Ec is the modulus of elasticity of concrete.16) G = Ec/2⋅(1+ν) is the shear modulus of elasticity.Tensile failure A tension failure is assumed to occur if the tensile stress in a principal stress direction exceeds the tensile strength.17) The concrete tensile behavior at a given section is determined using a bilinear stress strain relationship that allows the concrete to retain certain amount of tensile stress beyond the cracking strain. CHALMERS Civil and Environmental Engineering.2.7. The shear stress in an active crack plane is limited by σ ≤ β ⋅ ft ( 4. by specifying the strain value at which the concrete tensile stress is assumed to be zero. This leads to crack formation orthogonal to the principal stress directions. In the smeared crack approach. which gives a value of κ = 7. β is the shear stiffness reduction factor. the concrete tensile stresses are averaged over a representative length to span cracks. The incremental stress strain matrix. As it is seen in Figure 4. following a tensile failure is as follows: ⎡ Et ⎢ ⎢ 1 ⎢S C= ⎢ 1 −ν 2 ⎢ ⎢ ⎢ ⎢ ⎣ Where Et is the stiffness across a crack plane.

U is a vector of nodal point displacements.3 Numerical iteration methods and convergence criteria In SOLVIA. In this project the BFGS method was used. All these iteration methods can be used with or without a line search procedure. The method is suited for softening materials. and with smaller errors than the Full Newton. Modified Newton and the BFGS method. This type of methods can predict the softening behavior more effectively than the Modified Newton method. Master’s Thesis 2005:17 . ∆U is a incremental displacement vector. The following equations are established in iteration (i) to define the algorithm of the method: t + ∆t K ( i −1) ∆U ( i ) = t + ∆t R ( i ) − t + ∆t F ( i −1) ( 4. (Hirschhausen (1999)) .19) ( 4. θ is an acceleration factor obtained from a line search in the direction of ∆U .20) ∆U ( i ) = t + ∆t U ( i ) − t + ∆t U (i −1) t + ∆t U (i ) = t + ∆t U (i −1) + θ (i ) ∆U (i ) Where K is a tangent stiffness matrix based on the iteration history. F is a nodal point force vector equivalent to the element stresses. The suitable choice of an iteration method depends directly on the treated problem. 4.Compression failure Crushing failure occurs when the stresses σ p 2 and σ p 3 obtained by interpolation from σ p1 in the triaxial compressive failure envelope lie outside the biaxial failure envelope. This secant matrix is then continuously updated. two basic methods are implemented: Full Newton. resulting in isotropic material properties with strain softening in all directions. According to Klaus-Jürgen Bathe (1996). the method consists in the establishment of a secant stiffness matrix from a previous solution.3. R is an external load vector. from Hirschhausen (1999).1 The BFGS method with line searches This method belongs to the family of Quasi Newton schemes developed for iteration on nonlinear system of equations.18) ( 4. 30 CHALMERS Civil and Environmental Engineering. 4.

i. the expense of the iteration increases. The stiffness matrix is then updated with the following relation ( t + ∆t K −1 ) ( i ) = A ( i )T ( t + ∆t K −1 ) ( i −1) A ( i ) Under the condition that c (i ) ( 4. the line search may prevent divergence of the iterations. The final value of θ that satisfies Equation 4. Klaus-Jürgen Bathe (1996).25) where A (i ) = I + v (i ) w (i )T is an n x n matrix with vectors v ( i ) = −c (i ) ⋅t + ∆t K ( i −1) ∆U ( i ) −( t + ∆t R (i ) − t + ∆t F ( i −1) ) w (i ) = t + ∆t ( 4. CHALMERS Civil and Environmental Engineering.. (∆U T )( t + ∆t R − t + ∆t F ( i −1) ) is satisfied for a convergence tolerance STOL with the following Equation: (∆U T )( t + ∆t R ( i ) − t + ∆t F ( i ) ) ≤ STOL ⋅ (∆U T )( t + ∆t R ( i ) − t + ∆t F ( i −1) ) ( 4. With the line search performed within iteration (i). Define the direction of a displacement vector increment ∆U =( t + ∆t K −1 ) ( i −1) ( t + ∆t R (i ) − t + ∆t F ( i −1) ) ( 4.23 determines the solution of Equation 4.23) 4.e.26) ( 4.21) 2. but less iteration may be needed for convergence. Evaluate the equilibrium of the displacement vector increment ∆U with a line search t + ∆t U (i ) = t + ∆t U (i −1) + θ (i ) ∆U (i ) ( 4. In addition. 5.27) ( 4. The acceleration factor θ ( i ) is varied until the component of the out of balance loads in the direction of the displacement vector increment. Master’s Thesis 2005:17 31 .24) ⎛ (∆U (i )T ) ⋅( t + ∆t R (i ) − t + ∆t F ( i −1) ) ⎞ =⎜ ⎜ (∆U (i )T ) ⋅( t + ∆t K (i −1) ∆U (i ) ) ⎟ ⎟ ⎝ ⎠ 1/ 2 < TOL ( 4. This increased robustness is the major reason why a line search can in general be effective.22) 3.28) ∆U (i ) (∆U ( i ) ) ⋅( t + ∆t R ( i ) − t + ∆t F ( i −1) ) T K ( i −1) ∆U ( i ) = θ ⋅( t + ∆t R ( i ) − t + ∆t F ( i −1) ) A line search is an important feature for many iterative solution procedures.The procedure is as follows: 1.22.

30) Where R and F have the same meaning as in Equation (4. RNORM is a reference force. U C is a constant damping matrix. the following criteria were used: Force and Moment convergence criteria For the translational degrees of freedom: t + ∆t && (i −1) − C t + ∆tU & ( i −1) − t + ∆t F ( i −1) R − M t + ∆tU RNORM ≤ RTOL ( 4. M is a constant mass matrix. RTOL is a relative force and moment tolerance used to measure equilibrium iteration convergence.2 Convergence criteria for the different methods The different iterative solution methods in SOLVIA use the same convergence criteria in every iteration step to define equilibrium. In this project.29) For the rotational degrees of freedom: t + ∆t && (i −1) − C t + ∆tU & ( i −1) − t + ∆t F ( i −1) R − M t + ∆tU RMNORM ≤ RTOL ( 4. & is a nodal point velocity vector. RMNORM is a reference moment. Master’s Thesis 2005:17 . 32 CHALMERS Civil and Environmental Engineering. U && is a nodal point acceleration vector.3.18).4.

loading and cross section for beam OA-3 CHALMERS Civil and Environmental Engineering. Two of the test beams considered by Bresler and Scordelis were included in this report.30 Figure 5.433 0. 0.5 5.090. These are denoted as OA-3 and A-1. while beam B-14 was loaded with two symmetrical concentrated loads. The dimensions of the loading and support steel plates can be found in table 5. Beams OA-3 and A-1 were subjected to a concentrated load acting at mid span. In addition one beam tested at Chalmers University of Technology and denoted as B-14 was also analyzed.1-5.1 Geometry.06 0.09 0.06 0. Master’s Thesis 2005:17 33 .56 2x0.1. The dash lines in the figures indicate the location of the longitudinal and vertical reinforcement.1 Description of tested beams Test specimen Bresler and Scordelis investigated experimentally the shear strength of 12 simply supported reinforced concrete beams. Figures 5.064 0.3 show the geometry of the tested beams with reinforcement and applied concentrated load (the dimensions of the beams are expressed in meters).

075 0.23 Figure 5. loading and cross section for beam B-14.039 0.040 0.040 0. 34 CHALMERS Civil and Environmental Engineering.040 0.2 Geometry. Master’s Thesis 2005:17 .039 0.222 0.0750.34 0.0.

4 30 25.2 5.8 230 304.0. CHALMERS Civil and Environmental Engineering.2.4 30 25.1 Dimensions of the support and loading steel plates for the tested beams.06 0.38 0.8 230 304. loading and cross section for beam A-1.18 0.3 Geometry.56 0. Table 5.128 0.05 0.4 5. Master’s Thesis 2005:17 35 .2 (mean values).8 25.1 Material properties Concrete The material properties of all the beams investigated experimentally are shown in Table 5.4 304.8 100 304.8 304. Beam l [mm] Support plate b [mm] h [mm] l [mm] Loading plate b [mm] h [mm] OA-3 B-14 A-1 304.8 127 100 127 25.06 0.30 Figure 5.

The values of the compressive strength fc. for beam OA-3 and A-1.62 29.54-cm beams at the third points of an 45. while ten specimens were used.24x2.14 3. Using the compressive strength fc. The average values of modulus of rupture ft.fl.fl [MPa] 37. for beam B-14 the values for Ec and ft for the same beam were calculated.2 24.2. The following stress strain relationship shown in Figure 5. In all cases.4 for the steel reinforcement model was considered to define the strain hardening modulus in the FE analyses. σ fu fy Ε sh Εs εy Figure 5.86 30. 36 CHALMERS Civil and Environmental Engineering. Master’s Thesis 2005:17 .72-cm span.2 Concrete material properties for tested beams(mean values) Ec [MPa] - Beam OA-3 B-14 A-1 fc [Mpa] ft [MPa] ft. were obtained from tests on 15x30-cm cylinders at an age of 28 days.2 Reinforcing steel The reinforcing steel was modeled using a bilinear elastic-plastic model with isotropic strain hardening based on the Von Mises criteria and monotonic loading conditions. the longitudinal and vertical reinforcing steel was modeled by using truss elements. Table 5. obtained by loading 15.24x15. can also be found in table 5.09 2.4 εsu ε Stress strain relationship for reinforcing steel. to get the same parameter for beams OA-3 and A-1.2.4 - 5. The mean compressive strength for beam the B-14 was obtained from a test series which included three specimens.78 - 4.

The load was reapplied in increments of 45 kN to the load near failure and then it was increased in increments of 22.32 253. 5.72 628 201576. CHALMERS Civil and Environmental Engineering. The beam was first loaded at about 30% of the ultimate load and in two or three increments and then the load was removed. and deformation control was performed in the tests.81 2565.68 207000 345. The beam was gradually loaded to the failure stage and unloaded stepwise afterwards.4 Results from the experiments A brief summary of the overall behavior.25 kN until failure occurred.3 and 5.86 628 Table 5.The steel plates were modeled using an isotropic linear elastic material model with plane stress elements and solid elements.4 Material properties and cross section area for the stirrups.34 56.84 325. As [mm²] Es [MPa] fyv [MPa] Beam A-1 B-14 63.21 1005.3 Test performance Beams OA-3 and A-1 were tested under center point load at an age of 13 days. Master’s Thesis 2005:17 37 . cracking load and ultimate strength of designed series of test beams is presented below.4 Table 5. A s[mm²] As ’[mm²] Es [MPa] fy [MPa] Es ’[MPa] fy ’[MPa] Beam OA-3 A-1 B-14 3847. Beam B-14 was tested under two point loads using deformation control.96 555.35 603 205718.3 Material properties and cross section area for the longitudinal reinforcement.84 5.62 207000 552.841 207000 325.5 189.66 218144. The properties of the reinforcing steel for the beams tested experimentally and the dimensions of the support and loading steel plates can be seen in Tables 5.

4. Shear compression failure occurred in beam A-1.Diagonal tension failure .1-7. 7. This beams failed shortly after the formation of the “critical diagonal tension crack”. The reason for the failure is the longitudinal splitting in the compression zones.7.56 311. Pcr [KN] Pu [KN] Beam OA-3 A-1 B-14 δmax [mm] 27.76 267.22 38 CHALMERS Civil and Environmental Engineering. Failures developed without propagation of the flexural cracks.5 Summary of the experimental results.3.21 in sections 7. 7. Final failures occurred due to splitting in the compression zone.3. Master’s Thesis 2005:17 .94 14.57 467.2 Loads and deflections The ultimate and cracking loads together with the maximum deflection obtained from the experiments can be found in Table 5. 5. In this case the diagonal tension cracks were formed at approximately 60 percent of the ultimate load.64 118. The crack pattern for all tested beams can be seen in figures 7.5.17 in Sections 7. The load-deflection diagrams for all tested beams together with corresponding analysis results are shown in Figures 7. near the load point and horizontal splitting along the tensile reinforcement near the bottom edge of the beam.4.14 and 7. Table 5.3.1 Type of failures According to experimental results two different types of failures were obtained for the tested beams: .1-7.22 23.Shear compression failure Diagonal tension failure occurred in the beams OA-3 and B-14.22 90 378. but without splitting along the tension reinforcement.10 and 7. indicating that the mechanism of failure was shear-compression.5.

see Figure 4. with isotropic strain hardening based on the Von Mises criteria and monotonic loading conditions. A detailed explanation of the concrete material model in SOLVIA can be found in section 4. the longitudinal and vertical reinforcing steel was modeled by using truss elements. 3-node parabolic truss elements with an elastic-plastic material model for the steel reinforcement (stirrups and longitudinal reinforcement). Five different models were used to examine the capabilities and limitations of the SOLVIA FEsystem. The BFGS iteration method with LINE SEARCHES was employed for the FE-analysis. The z-axis was point along the height of the cross section in the north direction. • • 6.1 Finite element modeling General The finite element program SOLVIA was used for the FE-analysis.2 Coordinate system The finite element models were defined in a Cartesian coordinate system with a y-z plane and an x-axis pointing outwards. The following stress strain relationship shown in Figure 6. The following types of elements were chosen: • 27-node solid elements for the concrete material models in three dimensions and 8-node plane stress elements for the concrete models in two dimensions. The reinforcing steel was modeled using a bilinear elastic-plastic model. based on an uniaxial stress-strain relationship.1 CHALMERS Civil and Environmental Engineering. The AUTO-STEP method with different force norms and tolerances was used to control the load increment up to the ultimate load.6 6. 6.2 in Section 4.4 Material models The concrete was modeled using the SOLVIA concrete material models.1.2. The y-axis was pointing along the length of the beam in the east direction. Master’s Thesis 2005:17 39 .2. 6. In all the cases. 8-node plane stress elements for the support and loading plates in three dimensions and 8-node plane stress elements for the same plates in two dimensions.3 Geometry Due to symmetry only half of the beams were modeled.

0031 0.for the steel reinforcement model was considered to define the strain hardening modulus in the FE analyses.62 -29.2 -24.021 -0. Table 6.1 ⋅ 10 ¨ −3 respectively. Master’s Thesis 2005:17 .887 0.82 40 CHALMERS Civil and Environmental Engineering.78 3.4. The properties of the reinforcing steel for the beams tested experimentally and the dimensions of the support and loading steel plates can be seen in Tables 5.002 εu 0.2.353 -24. for the models Model 1 2 3 4 5 κ 14 13 18 14 16 β 0.89 0.8 0.0031 0. where the summary of the material properties is given. The chosen and calculated values of the mentioned concrete parameters used in the FE-analysis for all the models can be found in Table 6. The steel plates were modeled using an isotropic linear elastic material model with plane stress elements or solid elements depending on the model. and stiffness reduction factor (described as ‘stifac’ in the SOLVIA code) were set to be 0. The parameters which are not given in the tables. 3 and 5 the average values of modulus of rupture. εu.002 -0. such as E0.1 εsu ε Stress strain relationship for reinforcing steel.002 -0.9 33734 30400 4.1. Principal stress ratio β.3 and 5.2-5.89 0.0031 0. fcu.828 -32.14 2. In the FE-analysis for model 1.1.86 4.353 -24.0031 33734 30400 26994.78 -37. as the tensile strength of concrete were used.0031 0.82 -22.2 -32.62 -29.4. The material parameters of concrete and reinforcing steel used in the finite element analyses can be found in Tables 5. σ fu fy Ε sh Εs εy Figure 6.887 Eo [Mpa] ft [Mpa] fc [Mpa] fu [Mpa] εc -0.14 2.75 and 0.09 -37.1 The concrete parameters used in FE-analysis.002 -0. εc were chosen and calculated according to Section 4.

this type of boundary conditions resulted in good numerical predictions.03 ⋅ E s in accordance with Figure 5. For model 1 and model 3 the initial load were set to be concentrated. The initial applied load was chosen as: concentrated for the models created with plane-stress elements in 2D (model 1-3) and pressure for the models made with solid elements in 3D ( model 4 and model 5 ). when the forces of 1000 N and 500 N were divided with the area of the plate surface. 6.261 N/m2 respectively. and defined boundary conditions can be seen in Figures 6. while the strain hardening modulus for reinforcement was calculated as E sh = 0. while the steel reinforcement consists of 46 parabolic truss elements. is given for each model in Table 5. in the analysis for the 3D models.6 Mesh for model 1 The concrete section for model 1 representing beam OA-3. The calculation of the pressure loads was performed in order to obtain load-deflection curves comparable to the curves from the experimental results. A disadvantage of the modeling of pinned boundary conditions using this approach is that. was chosen Esp = 210 GPa. while the nodes along the symmetry line were fixed in y-direction. The models 1-3.2.5 Boundary conditions and loads Two different approaches were used to define the boundary conditions and the loads on the FE models based on the type of elements: • The boundary conditions for the models created with plane stress elements in 2D were defined according to the following assumptions: the middle nodes at the bottom of the supporting plate were locked in z-direction. For all the models the location of the applied initial load. equal to 10714. A-1 and B-14 respectively. were created using plane stress elements in 2D for beams OA-3. In the following section a detailed descriptions of the FE-models can be found.4 in section 5. where Es.8. Master’s Thesis 2005:17 41 . it generates a small moment reaction as in clamped boundaries. These pressure loads were obtained from the calculations.3. Steel plates were modeled with one plane-stress element. where the pressure load acts. The pressure was applied as initial load for the models 4-5.The modulus of elasticity for steel plates.2-6. However. 6. The boundary conditions for the models created with solid elements in 3D were defined as following: The surface of the cross section in the mid-span of the beam was fixed in the y-direction and the displacements of all the nodes in the surface were prevented. • The initial applied load varied depending on the model. The horizontal surface at the bottom of the supporting plate was locked in z-direction. equal to 500 N. while the same load were set to be 1000 N for model 2. CHALMERS Civil and Environmental Engineering. was defined using 161 plane-stress elements.2.668 N/m2 and 43478. while model 4 and model 5 were created with solid elements in 3D and represent beams OA-3 and B-14 respectively.

one for inner and outer concrete covers and another for remaining plane-stress elements. the element height in the z direction was determined by the location of the longitudinal reinforcement.The length of the plane-stress elements was chosen to be equal. Master’s Thesis 2005:17 . see Figure 6. Figure 6. as well as for the elements under the loading plate.2 Mesh of the concrete and reinforcing steel for model 1. The plane-stress elements have the same length except elements 73-78 and 85-90. 6. Similar to model 1. applied loads and defined boundary conditions are shown in Figure 6. The mesh of the concrete and reinforcing steel. for the elements above the supporting plate.3. with element numbers. The steel plates were modeled using one single plane stress element.7 Mesh for model 2 The concrete section for model 2 representing beam B-14 was modeled using 114 plane stress elements while the longitudinal steel reinforcement and the stirrups were modeled using 54 and 28 parabolic-truss elements respectively. The element height in the z direction was adjusted by the location of the reinforcement. 42 CHALMERS Civil and Environmental Engineering. two different element heights were defined.2. The length of all truss elements was defined in the same way as the plane elements. Therefore.

Steel plates were modeled with 2 equal plane-stress elements. 203 plane-stress elements were used while the longitudinal steel reinforcement and stirrups consists of 54 and 28 parabolic truss elements respectively. see Figure 6.8 Mesh for model 3 To model the concrete of beam A-1.Figure 6. affected by the location of the longitudinal and vertical reinforcement. The dimensions of the plane-stress elements and truss elements were as in the previous models.4. CHALMERS Civil and Environmental Engineering. Master’s Thesis 2005:17 43 .3 Mesh of the concrete and reinforcing steel for model 2. 6.

6. Master’s Thesis 2005:17 . 44 CHALMERS Civil and Environmental Engineering. together with the element numbers can be seen in Figures 6. The mesh for model 4. 6.Figure 6. One solid element was chosen to model each steel plate.9 Mesh for model 4 The concrete section for three dimensional model of beam OA-3.5 and 6.4 Mesh of the concrete and reinforcing steel for model 3. The element dimensions for the longitudinal reinforcement were chosen according to the same approaches as for model 1. was created using 154 solid elements.

Figure 6.5 Mesh of the concrete with shown location of the reinforcement for model 4. Master’s Thesis 2005:17 45 .Figure 6.6 Mesh of the concrete for model 4. CHALMERS Civil and Environmental Engineering.

7 Mesh of the concrete with shown location of the reinforcement for model 5. see Figures 6. Figure 6.6.7 and 6. The details for FE mesh and number of elements used to model the longitudinal reinforcement and stirrups for model 5 can be found in Appendix A. The element length for the longitudinal reinforcement was chosen according to the same principle as they were done for model 2.10 Mesh for model 5 The mesh of the concrete section for model 5 made in 3D for beam B-14 consists of 120 solid elements. Each steel plate was modeled with one solid element.8. 46 CHALMERS Civil and Environmental Engineering. Master’s Thesis 2005:17 .

8 Mesh of the concrete with shown cross-section and element numbers for model 5. CHALMERS Civil and Environmental Engineering.Figure 6. Master’s Thesis 2005:17 47 .

1 The location of integration points for plane and solid elements. Two different types of plots were also obtained according to FEanalysis and these are: contour plots of cracks and displacement and vector plots of the same items.2 The location of integration points for truss elements.5. For all the models the force tolerance during the iteration steps was chosen to be one percent. Master’s Thesis 2005:17 . For load-deflection diagrams the points at mid-span were chosen as a reference points. these are: load-displacement relationship. Figur 7. Five types of diagrams were obtained from the FE-analysis. stress-strain relationship and stress variation as well as strain variations over the concrete sections and reinforcing bars. To plot the load-deflection diagrams and the stress-strain diagrams certain points were chosen.1 and 7.2 the location of the integration points for plane stress elements. including plots of the predicted crack pattern for each of the models are presented in detailed in sections 7. 48 CHALMERS Civil and Environmental Engineering.1-7. The reference points selected to plot the stress-strain relationships for the concrete and longitudinal reinforcement are integration points for certain elements in the structures. In Figures 7. The chosen points represent critical points in the structure were the maximum values are expected to occur.7 FE-results The FE-results. Figur 7. solid and truss elements is shown.

A load of 347.1. CHALMERS Civil and Environmental Engineering.90 kN and a mid-span deflection of 22. To plot the stress and strain variations over the cross-section and along reinforcing steel. In the FEanalysis the concrete parameters shear retention factor β. which defines the ultimate crack strain were set to be 0. while the values in contour and vector plots of cracks and displacement were calculated in the last load step. see Figure 7.3. certain cross-sections and bars were chosen. These sections were critical.1 Load-displacement relationship To examine how the deflection varies with increasing load. 7.89 and 14 respectively for this model. and were obtained in the last load step. The lines with triangular symbols represent experimental results. Master’s Thesis 2005:17 49 .The values in both the load-deflection and the stress-strain diagrams were calculated in each load step. Figure 7. 7. and the parameter κ.1 Model 1 Model 1 is the plane stress model for beam OA-3. A force norm of 180 kN was chosen. while the lines without symbols are according to two-dimensional FE-analysis performed for beam OA-3. tested experimentally. a plot of the load deflection curve was used.3 Load-deflection relationship.10 mm were obtained from the FE-analysis in the last load step.

calculated in the last load step are presented.6 illustrates an interpretation of the vector crack normal. several flexural cracks can be noted.1.2 Crack pattern A plot of the state of the concrete for the last load step is shown in Figure 7.In Figure 7. The initiation of the critical diagonal tension crack can also be observed in the compression zone.6. 7. Master’s Thesis 2005:17 .5. according to the FE-analysis.4 the vertical displacement for model 1 as a contour plot and the vertical displacement of all the nodes in the model as a vector. 50 CHALMERS Civil and Environmental Engineering. see Figure 7. Figure 7. The crack pattern in Figure 7. near the load point. while the elements close to the support plate have very small displacement. In the crack pattern. obtained from the FE model. The crushing of the concrete at the loading plate was a result of local high stress concentrations. Over the support plate displacement was prevented. together with the splitting in the tension zone. The plot shows the crushed concrete in the compression zone and crack pattern predicted by the FE-model. It can be observed that the elements close to the location where the load was applied have the largest vertical displacement.4 Contour and vector plots of the displacements in the last load step for model 1.

Figure 7. Master’s Thesis 2005:17 51 .7 Crack pattern for beam OA-3 according to the experimental results.normal.Figure 7. according to the performed tests can be seen in Figure 7.5 Contour plot of the state of concrete including the values of the maximum ultimate load. The real crack pattern for beam OA-3. CHALMERS Civil and Environmental Engineering. The plots were obtained in the last load step. (upper plot) together with the plot of vector crack.5.6 Crack trajectories according to the vector plot of crack normal in Figure 7. for Model 1. Figure 7. for Model 1.7.

7.1.3

Stresses and strains in the concrete and reinforcement

Figure 7.8 shows the diagrams of the stress and strain variations along the height of the selected cross section and the outer reinforcement layer. The values in these diagrams were calculated in each load step. The depth of the neutral axis is approximately 0.30 meters, according to the diagrams.

Figure 7.8

Stress and strain variations over selected cross section and along the longitudinal reinforcement for Model 1, calculated in each load step.

The stress-strain relationship for concrete and reinforcing steel obtained in each integration (load) step can be found in Figure 7.9. To plot the stress-strain relationship for concrete in tension, integration point 9 in element 7 was chosen. To obtain a diagram for the stress-strain relationship for the concrete section in compression, element 8 and integration point 9 were chosen. Point 2 in element 44 was selected to plot stress-strain relationship for reinforcing steel. All these elements with the integration points can be found in the diagrams for stress strain relationship, where ‘EL’ and ‘P’ denote elements and integration points respectively.

52

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

Figure 7.9

Stress-strain relationships in the concrete and reinforcing steel calculated in each load step, for Model 1. The first and second diagrams from the left to the right represent compressive and tensile behavior of the concrete, while the diagram on the right shows the tensile behavior of reinforcing steel.

7.2

Model 2

Model 2 is the one created for test beam B-14, using plane stress elements. The concrete parameters β and κ for this model were chosen to be 0.887 and 13 respectively. The force norm is equal to 80 kN in this case.

7.2.1

Load-displacement relationship

In this case, the maximum load reached according to the FE-analysis was 117.90 kN, while the reached mid-span deflection was 15.26 mm, see Figure 7.10. These items were obtained in the last load step.

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

53

Figure 7.10 Load-deflection relationships. The lines with triangular symbols represent the experimental results, while the lines with rectangular symbols are according to two-dimensional FE-analysis for beam OA-3.

In the plot of the vertical displacements for the model, the zone on the right side of the loading plate has the largest vertical displacements. The vertical displacements decrease in the zone closer to support plate. Over the support plate the vertical displacement were locked, see Figure 7.11. The plots of the displacements as a vector and as a contour were reached in the last load step.

54

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

11 Contour and vector plots of the displacements in the last load step for model 2.6 the behavior and crack pattern of the Model 2 was illustrated.2. obtained in the FE-analysis.12 indicates crushing of concrete on both sides of the plate.13. These crack trajectories shown in Figure 7. represent the interpretation of the vector plot of crack normals. This is due to high stress concentrations in the zone near the load point.Figure 7. The crack pattern according to FE model shows flexural cracks in the tension zone and the formation of a ‘diagonal tension crack’ in both the compression and tension zones. CHALMERS Civil and Environmental Engineering. 7. Master’s Thesis 2005:17 55 .2 Crack pattern Similarly as in Figures 7. The initiation of the diagonal tension crack can be observed in both the compression zone near the loading plate and in the tension zone close to the support. Figure 7.5 and 7.

Figure 7.14 Crack pattern for beam B-14 according to the experimental results. 56 CHALMERS Civil and Environmental Engineering.Figure 7. performed for beam B14 can be found.12. for Model 2. for Model 2.13 Crack trajectories according to the vector plot of crack normal in Figure 7. Figure 7. In Figure 7. Master’s Thesis 2005:17 .14 the crack pattern according to the tests. The plots were obtained in the last load step.normal.12 Contour plot of the state of concrete including the values of the maximum ultimate load. (upper plot) together with the plot of vector crack.

16 shows the stress-strain relationship for concrete and reinforcing steel for the critical elements and integration points. Master’s Thesis 2005:17 57 . Figure 7. The particular cross-section and reinforcement used to construct the diagrams are shown in the same figure. The values in the diagrams were calculated in each load step.3 Stresses and strains in the concrete and reinforcement In Figure 7. The chosen elements with the integration points can be found in the diagrams and the values in these diagrams were calculated in each load step. CHALMERS Civil and Environmental Engineering.2. Figure 7.15.7.15 Stress and strain variations over selected cross section and along the longitudinal reinforcement for Model 2. the diagrams of the stress and strain variations along the height of the selected cross section and the outer reinforcement layer are shown. calculated in each load step.

3. for Model 2.16 Stress-strain relationships in the concrete and reinforcing steel calculated in each load step. while the diagram on the right shows the tensile behavior of reinforcing steel. For this model. Master’s Thesis 2005:17 .Figure 7.8 and κ = 18 were used in this case. see Figure 7. The concrete parameters β = 0.17. 7.1 Load-displacement relationship A maximum load of 478 kN and a mid-span deflection of 9. The first and second diagrams from the left to the right represent compressive and tensile behavior of the concrete.3 Model 3 Model 3 was created for experimentally investigated beam A-1. 7. 58 CHALMERS Civil and Environmental Engineering.053 mm were reached for this model in the last load step. a force norm was set to be 100 kN.

shown as a contour and as a vector. can be found in Figure 7.Figure 7.18. calculated in the last load step. while the lines without symbols were obtained in two-dimensional FE-analysis performed for beam A-1. Master’s Thesis 2005:17 59 . The lines with triangular symbols are according to the experimental results. CHALMERS Civil and Environmental Engineering.17 Load-deflection relationships. The vertical displacement for model 3.

Master’s Thesis 2005:17 .19 shows the crushing of the concrete in the zone under the loading plate and the vector plot of crack normals.18 Contour and vector plots of the displacements in the last load step for model 3.Figure 7.2 Crack pattern Figure 7.3. calculated in the last load step. 60 CHALMERS Civil and Environmental Engineering. 7.

normal. Master’s Thesis 2005:17 61 . The crack trajectories in this Figure illustrate the interpretation of the vector plot of crack normals.20 Crack trajectories according to the vector plot of crack normal in Figure 7. for Model 3.Figure 7.20. Figure 7.21. The only difference compared with model 1 is the ‘critical diagonal tension crack’. according to the FE-model. see Figure 7. A small part of the ‘critical diagonal tension crack’. CHALMERS Civil and Environmental Engineering.19 for Model 3. initiated at the loading plate in the compression zone. The real crack pattern for beam A-1 obtained in the experiments is shown in Figure 7. The plots were obtained in the last load step. The crack pattern for model 3 is very similar to the crack pattern for model 1. (upper plot) together with the plot of vector crack. can be observed.19 Contour plot of the state of the concrete including the values of the maximum ultimate load.

3 Stresses and strains in the concrete and reinforcement In Figure 7. the diagrams for stress and strain variations over the selected crosssection and along the reinforcement layer are presented. The stress-strain relationship for concrete and reinforcing steel in each load step can be found in the diagrams shown in Figure 7.21 Crack pattern for beam A-1.22 Stress and strain variations over selected cross section and along the longitudinal reinforcement for Model 3.3. The values for these diagrams were obtained in each load step. 62 CHALMERS Civil and Environmental Engineering. calculated in each load step. 7. according to the experimental results. Master’s Thesis 2005:17 .Figure 7.23. Figure 7.22.

A force norm of 80 kN and a force tolerance of one percent were chosen in the FE-analysis. for model 4 were 369. The first and second diagrams from the left to the right represent compressive and tensile behavior of the concrete.Figure 7. In this case parameters β and κ are equal to 0.28 mm respectively.4 Model 4 Model 4 is the solid model for tested beam OA-3.4. for Model 3.1 Load-displacement relationship The maximum load and mid-span deflection obtained in the FE-analysis in the last load step.89 and 14 respectively.23 Stress-strain relationships in the concrete and reinforcing steel calculated in each load step. while the diagram on the right shows the tensile behavior of reinforcing steel. see Figure 7.24. CHALMERS Civil and Environmental Engineering.84 kN and 24. 7. 7. Master’s Thesis 2005:17 63 .

while the lines without symbols are according to the FE-analysis performed in 2D and in 3D respectively.25. obtained in the last load step can be found in Figure 7.24 Load-deflection relationships for beam OA-3. Master’s Thesis 2005:17 . The vertical displacement for Model 4 as a contour plot and the vertical displacement of all nodes of the model as a vector.Figure 7. The larger load and displacement were obtained in 3D analysis. The lines with triangular symbols represent experimental results. 64 CHALMERS Civil and Environmental Engineering. obtained in the last load step.

Figure 7.25 Contour and vector plots of the displacements in the last load step for Model 4.

7.4.2

Crack pattern

Figure 7.26 indicates the crushing zone at the loading plate for model 4, created with solid elements in 3D. It can be observed that the crushing zone for model 4 is much smaller than the same zone for model 1, created for the same beam, but as a 2D model. The plots of the cracking and crushing concrete zone were obtained in the last load step.

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

65

Figure 7.26 The contour plot of the crushing and cracking zone in 3D for Model 4.

The plot of the crushed zone in the concrete and vector plot of the crack-normal in y-z plane, reached in the last load step are shown in Figure 7.27.

66

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

Figure 7.27 Contour plot of the state of concrete including the values of the maximum ultimate load, (upper plot) together with the plot of vector crack- normal, for Model 4. The plots were obtained in the last load step.

The crack pattern for the model is very similar to the crack pattern obtained for model 1; the crack trajectories are the interpretation of the vector plot of crack normals from the FE-analysis, see Figure 7.28.

Figure 7.28 Crack trajectories according to the vector plot of crack normal in Figure 7.27 for Model 4.

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

67

Master’s Thesis 2005:17 .29 Crack pattern for beam OA-3.3 Stresses and strains in the concrete and reinforcement In Figure 7. according to the experimental results. 68 CHALMERS Civil and Environmental Engineering. The values in these diagrams were obtained in each integration step. calculated in each load step. for model 4 are presented. Figure 7.30 Stress and strain variations over selected cross section and along the longitudinal reinforcement for Model 4. Figure 7. the stress and strain variations over the cross-section in the mid-span and along the length of the outer reinforcement layer. 7.29.30.The real crack pattern for beam OA-3 according to the experimental studies is shown in Figure 7.4.

32. CHALMERS Civil and Environmental Engineering. for Model 4.5.887 and 16 respectively.30 kN and 17. 7.95 mm respectively.31.5 Model 5 Model 5 was created for beam B-14.Stress-strain relationship in the concrete and reinforcing steel for model 4 can be found in Figure 7. Master’s Thesis 2005:17 69 . while the diagram on the right shows the tensile behavior of reinforcing steel. where the values in the diagrams were obtained in each load step.1 Load-displacement relationship According to the FE-model the maximum load and the mid-span deflection reached in the last load step are: 110. see Figure 7. the parameters β and κ were chosen to be 0. 7. Figure 7. The first and second diagrams from the left to the right represent tensile and compressive behavior of the concrete. In the FE-analysis for this model. The force norm used in the FE-analysis was equal to 80 kN.31 Stress-strain relationships in the concrete and reinforcing steel calculated in each load step. using solid elements.

shown as vector and calculated in the last load step are presented in Figure 7.32 Load-deflection relationships for beam B-14. respectively. obtained in the last load step.Figure 7. Master’s Thesis 2005:17 . The vertical displacement for model 5 as a contour plot and the same displacement for all the nodes in the model.33. The line with triangular symbols is according to the experimental results. The lines with rectangular and without symbols are according to FE-analysis for beam B-14 performed in 2D and in 3D. 70 CHALMERS Civil and Environmental Engineering.

CHALMERS Civil and Environmental Engineering.2 Crack pattern The crushing zone at the loading plate for model 5. reached in the last load step can be seen in Figure 7. The size of the crushing zone is almost the same for the two dimensional and three dimensional models created for beam B-14. created with solid elements in 3D. Master’s Thesis 2005:17 71 .34. 7.5.Figure 7.33 Contour and vectors plot of the displacements in the last load step for Model 5.

72 CHALMERS Civil and Environmental Engineering.35 shows the plot of the crushing zone and vector plot of crack-normal in y-z plane. Figure 7.34 The contour plot of the crushing and cracking zone in 3D for Model 5. obtained in the last load step. Master’s Thesis 2005:17 .Figure 7.

35.Figure 7.35 Contour plot of the state of concrete including the values of the maximum ultimate load. The only difference is the ‘critical diagonal tension crack’ which can be seen more pronounced in model 5.36 Crack trajectories according to the vector plot of crack normal in Figure 7.normal. see Figure 7. Master’s Thesis 2005:17 73 . for Model 5. The crack pattern for model 5 is similar to crack pattern for model 1. CHALMERS Civil and Environmental Engineering. The crack trajectories in this figure represent the interpretation of the vector plot of crack normals.36. Figure 7. The plots were obtained in the last load step. for Model 5. (upper plot) together with the plot of vector crack.

37. The stress-strain relationships for concrete and steel elements in selected points obtained in each load step can be found in Figure 7. Figure 7.3 Stresses and strains in the concrete and reinforcement Stress and strain variations over the selected cross-section and along the length of the outer reinforcement layer calculated in each integration (load) step are shown in Figure 7. 74 CHALMERS Civil and Environmental Engineering.38.39. 7. Figure 7. Master’s Thesis 2005:17 .38 Stress and strain variation over selected cross section and along the longitudinal reinforcement for Model 5.5.37 Crack pattern for beam B-14. according to the experimental results. calculated in each load step.The real crack pattern for beam OA-3 according to the experimental results can be seen in Figure 7.

while the diagram on the right shows the tensile behavior of reinforcing steel.3% greater than the same load obtained in the experiments. In the crackpattern according to the FE-analysis. In this case the ultimate load for the FE-model was 2. given in table 6. These results were compared with the same results obtained from the FE analysis to evaluate the quality and accuracy of the numerical predictions. for the chosen combination of parameters β and κ. Master’s Thesis 2005:17 75 .1 where 1%-9% less than the collapse loads obtained during the experiments. for Model 5. 7. The comparison between the experimental and the FE-results is summarized below: • Regarding load-deflection relationship. Concerning the shear type of failure. The deflection according to FE-calculation is 13%-47% less than deflection in experimentally tested beams. The first and second diagrams from the left to the right represent tensile and compressive behavior of the concrete. it was found that the maximum loads obtained from the FE-analyses. the flexural cracks initiated at the bottom and the ‘critical diagonal tension cracks’ which cause the failure of • CHALMERS Civil and Environmental Engineering.Figure 7.6 Evaluation of the FE results In the experimental studies. The only exception was model 3.39 Stress-strain relationships in the concrete and reinforcing steel calculated in each load step. in all FE models very similar crack trajectories as in the experimental studies where obtained. load-deflection relationships and shear types of failures for the beams were examined.

sp. In FE-analysis for model 1 and model 3-4.fl. The stress and strain variations over the critical concretesections and along the reinforcing bars have similar trends as the ones in the theoretical studies. and parameters κ. Based on the comparison between the FE results and the experimental results. were used as the tensile strength . the model was slightly stiffer. The non-linear three dimensional FE-analyses gave better prediction of the deflection than the same analysis performed in 2D. A detailed calculation of the ultimate load for the tested beams can be found in APPENDIX B. Master’s Thesis 2005:17 . a new FE-analysis was performed for model 1. 76 CHALMERS Civil and Environmental Engineering. In order to see the difference in the ultimate loads and deflections using the modulus of rapture compared to the splitting tensile strength of concrete ft. the measured average values of modulus of rupture ft.the structures can be observed. although due to the boundary conditions. It was observed. it was concluded that all the models gave reasonably good numerical predictions and similar results (within a reasonable range) to the results according to experimental and theoretical investigations. The similar behavior of the beam can be observed in the experimental studies. • The stress-strain relationships in the critical points of the beam. The difference between deflections was 11%. A calculation of the fracture energy of concrete Gf. for both concrete and reinforcing steel. calculated from the fracture energy can be seen in APPENDIX C. using splitting tensile strength of concrete was 13% less than the same load using modulus of rapture. were very similar to those shown in the theoretical studies. that the ultimate load obtained in the analysis.

beam-deflection and stresses in stirrups. the ultimate load Pu. a parametric study was carried out. Master’s Thesis 2005:17 77 .1-8. The load according to the FE-analysis corresponds to the half of the load obtained in the experiments for Model 1 and Model 3-4. because of the symmetrical boundary and loading conditions. κ (called ‘kappa’ in the SOLVIA code).1-8.5. A summary of the variation of β and κ is presented in tables 8. the number of iterations was not possible to increase. while in the some of the models the restart files increased the number of the load steps and improve slightly FE analysis. In table 8. β (defined as ‘shefac’ in the SOLVIA code). For each of the models. where ‘R’ in the tables of the parametric investigations denotes restart files. In some of the analysis. and the maximum deflection δmax reached in the experiments can be also found. CHALMERS Civil and Environmental Engineering. In order to increase the number of the load steps. affects the ultimate load.8 Parametric investigations In order to evaluate how the certain combination of the concrete parameters such as shear retention factor. The first combination includes large values for both β and κ. The second and third are: large κ with small β and large β with small κ. and the parameter which defines the ultimate crack strain. three main combinations were performed and presented. restart files were used for some of the cases in the parametric studies.5.

656 0.493 0.753 0.11 94.950 20.07 79.922 20.789 0.642 0.10 0.430 103.Table 8.62 103.747 Time 101.22 Step 102 105 106 105 177.758 9.96 58.038 20.356 0.362 102.649 53.90 56.358 0.55 180.7 0.1 149.40 100.953 0.09 0.987 23.95 17.45 176.731 0.17 103.4 102.261 11.403 0.08 98. δmax=27.09 0. where ‘R’ denotes restart files Pu=378.001 11.934 17.937 0.855 85.877 δa / δmax 0.133 0.954 Large κ and small β 40 40 45 45 35 30 35 0.35 93.95 0.00 Pa [kN] 2⋅ Pa Pu δa [mm] 21.42 143.35 58.10 0.750 0.97 87 86 147 R R 78 CHALMERS Civil and Environmental Engineering.5 0.886 0.325 18.3 0.57 kN .1 Parametric investigation for Model 1.539 0.410 0.98 1.943 11.417 0.10 82 78 80 82 83 131 81 82 R R R R R Small κ and large β 10 10 5 0.644 0.59 180.498 0.788 0.45 179.94 mm Large κ and β κ 30 45 58 65 β 0.95 1 1 149.545 0.948 0.641 18.64 0.10 0.35 53.116 138. Master’s Thesis 2005:17 .198 85.749 0.90 0.518 0.467 10.

955 11.2052 3.493 86.546 0.004 1.67 68.725 12.5778 -0.875 0.995 Large κ and small β 35 35 40 30 35 0. δmax=23.613 0.664 -0.9312 -0.853 12.30 93.448 -1.42 122.8633 10.828 1 1.4479 -0.077 60.474 -0.104 22.850 11.00 1.36 91 138 195 128 167 180 R CHALMERS Civil and Environmental Engineering.4955 1.42 0.442 14.522 8.95 1 1 103.005 0.20 0.475 0.89 93.30 96.Table 8.602 0.607 stirrup-force [kN] max min 6.85 73.849 14.727 0.7168 -0.556 0.56 mm Large κ and β κ 30 45 40 45 β 0.65 79.433 0.125 -1.32 76.00 0. where ‘R’ denotes restart files Pu=118.7627 2.180 13.90 1.619 0.47 67.48 97.2976 15.2 Parametric investigation for Model 2.72 122.0492 1.418 0.009 9.99 1.95 0.22 kN .2 118.5265 Time 92.9745 9.29 86 90 74 77 89 R R R R Small κ and large β 10 7 7 10 6 6 1.897 0.20 0.309 δa / δmax 0.20 Step 93 95 100 107 117.5448 4.728 0.20 0.197 15.672 0.00 1.172 0.843 118.30 0.89 92.02 87.659 0.211 12.5002 2.00 Pa [kN] Pa / Pu δa [mm] 14.654 15.178 7.510 0.191 11.7201 -1.036 1.518 0.88 62.664 0.21 96.7125 10.156 85.110 86.90 0.35 77.40 57.6348 -1.370 0.26 68.49 67.588 0.912 81.0874 -0.45 118.0276 9.3138 -0.3264 2.0052 -0.735 10.741 0.3929 -1.651 0. Master’s Thesis 2005:17 79 .036 -0.

901 181.714 2.235 δa / δmax 0.90 Pa [kN] 2⋅ Pa Pu δa [mm] 5.3672 1.4348 3.6325 0.419 0.1548 0.00 0.777 0.3261 1.0024 -1.8482 -0.866 0.74 67.9405 -1.964 Large κ and small β 30 30 35 35 35 30 30 35 0.19 112.62 194.87 80.76 61.Table 8.10 0.48 1.56 Step 204 235 216 265 190.242 0.3255 -1.352 0.5857 4.29 99.39 129 183 135 138 132 143 155 179 R R R Small κ and large β 5 5 10 1.4054 -1.6777 59.14 83.20 0.6 125.2395 3.576 0.833 2.1441 5.512 0.7177 1.9464 -1.52 122.2358 5.37 225.2411 48.686 0.8611 5.75 160.29 64.78 0.20 0.9761 3.9373 -0.850 0.370 0.00 1.9941 0.151 0.81 67.368 0.252 0.00 1.2857 3.8279 -0.8922 0.11 0.509 stirrup-force [kN] min -1.35 119.263 5.90 1.396 -0.2344 3.422 -1.3787 -1.25 kN .10 0.7862 3.227 0.375 0.8981 -0.00 1.61 0.581 0.555 0.31 163 349 220 R 80 CHALMERS Civil and Environmental Engineering. Master’s Thesis 2005:17 .2334 1.28 101.95 90.22 mm Large κ and β κ 45 30 40 40 β 0.30 0.1491 1.9592 0.9326 -0.62 134.00 97.816 0.1105 -1.412 0.268 0.537 0.274 Time 95.012 5.55 202.30 0.29 166. δmax=14.5682 4.228 0.8044 5.46 129.6293 0.30 62.3 Parametric investigation for Model 3.58 198.209 0. where ‘R’ denotes restart files Pu=467.524 0.563 max 1.48 135.5829 3.326 7.10 0.81 97.

20 0.119 144.1 0.984 17.343 25.30 0.840 0.725 37.004 0.339 0.1843 19.621 0.507 0. Master’s Thesis 2005:17 81 .399 64.848 0.619 0.947 23.32 115.725 1. Pu=378.90 0.δmax=27.898 46.458 22.35 141.281 20.11 191.24 0.13 Step 66 65 70 70 191.009 1.95 66.91 194.171 14.835 Time 65 65 65.809 14.910 0.821 0.35 111.35 0.010 1.23 190.259 0.972 10.509 0.475 17.857 5.029 Large κ and small β 35 35 35 30 30 35 35 0.10 48 52 49 51 48 26 52 48 52 R R R R R Small κ and large β 5 10 10 1 1 0.186 0.11 0.433 25.745 0.10 0.10 0.09 193.763 0.389 0.693 23.81 49.20 0.4 Parametric investigation for Model 4.43 32.04 186.92 1.57 kN .87 39.9 137.35 149 82 68 R CHALMERS Civil and Environmental Engineering.92 96.588 0.11 153.024 1.Table 8.79 21.612 0. where ‘R’ denotes restart files.90 1 1 Pa [kN] 2⋅ δa [mm] 23.323 δa / δmax 0.697 0.67 64.94 mm Large κ and β Pa Pu κ 30 45 30 45 β 0.64 63.536 0.

990 27.074 0. Master’s Thesis 2005:17 .142 74.2644 -1.22 kN .9197 Time 37.56 mm Large κ and β κ 30 30 35 40 45 β 0.85 13.8781 6.045 0.606 15.6048 0.1164 -0.1126 -0.1054 2.192 0.0995 0.450 0. where ‘R’ denotes restart files.235 0.30 0.5599 -0.37 13.90 1 1 1 Pa [kN] Pa / Pu δa [mm] 10.4510 -0.228 0.266 1.674 0.220 0.630 Large κ and small β 40 40 30 30 35 35 0.1160 -0.1073 -0.20 0.907 103.8815 0.171 0.366 stirrup-force [kN] min -0.113 -0.1170 -0. Pu=118.446 25.00 22.7340 2.7481 2.1704 0.420 0.8396 0.350 0.098 0.63 36.0818 11.52 133 R 143 148 Small κ and large β 10 5 1.Table 8.83 35.37 15.0928 0.3549 0.35 26.492 72.00 1.7697 0.6279 δa / δmax 0.074 0.653 20.9662 13. δmax=23.063 31.80 Step 380 529 349 361 373 R R R R 76.475 0.356 0.47 135 140 12.6661 0.250 0.22 34.0680 1.3911 8.99 51.118 -0.013 124 R 12.089 0.2165 0.0984 0.66 70.596 0.276 24.5 Parametric investigation for Model 5.651 0.2982 1.6187 max 1.1114 -0.040 0.20 0.610 0.993 27.035 -0.877 0.872 8.229 0.90 0.1318 0.5129 0.252 116 R 10 100 R 82 CHALMERS Civil and Environmental Engineering.207 0.2514 8.9602 0.10 0.35 0.

with flexural cracks formed in the tension zone were also observed from the crack patterns. The shear retention factors. The deflections were in all cases less than the deflections obtained from the experiments. where the initiation and formation of a critical diagonal tension cracks with flexural cracks in the tension zone were observed. Stresses and strains in the stirrups and in the longitudinal reinforcement were greater or lower. The type of elements used. Master’s Thesis 2005:17 83 . The deflections obtained in the FE-analysis were less than the deflection of the tested beams. The study was on shear failure for beam type members. The initiation of the diagonal tension cracks. but still within a reasonable range.9 Conclusions In this master thesis project. The crack patterns for some of the models for certain combinations of κ and β were very similar to the real crack patterns. The stress-strain relationships. which defines the ultimate crack strain in tension (defined as ‘kappa’ in the SOLVIA code). It was found that the prediction of the ultimate load as well as the crack patterns for FE analysis with both plane-stress elements and solid elements. obtained in the FE-analysis were very similar to those assumed in theoretical studies. the following conclusions can be drawn: • The combination with large values of β and κ gave very stable analyses in most of the cases. • CHALMERS Civil and Environmental Engineering. After a parametric investigation. For some of the FE models (Model 2 and Model 4) the ultimate load was greater than the load according to the experiments. 2. (depending on the model and the chosen values for κ and β) than the values obtained in combinations with large κ and β. 3. related to variation of the concrete parameters β and κ. except for model 3. were very similar and close to the analyses obtained in the experiments. These combinations have to be avoided. β (defined as ‘shefac’ in the SOLVIA code). The concrete material in the beams was modeled with plane-stress elements in two dimensions (2D) and with solid elements in three dimensions (3D). The FE-analysis performed with solid elements gave better prediction of the deflection. the capability and limitations of the SOLVIA FE-system for limit state design of reinforced concrete structures were investigated. focusing on the following parameters in the concrete material model: 1. The parameter κ. The ultimate loads and deflections from the FE analysis were much smaller than in the experiments for all the models. The combination with small κ and large β gave both stable and unstable analysis depending on the FE-model. which makes the prediction unsafe. the stress and strain variations along the critical sections and reinforcing bars. which gave a good prediction of the ultimate load.

2. with good predictions of the results. To estimate parameter κ from the fracture energy of concrete a good starting point is to calculate it. In the crack patterns. In model 1-2 and model 4 it was found that the parameter κ had to be multiplied with a factor within range 1. this value of κ must be increased because of the tension stiffening effect. However. respectively. It was found that parameter β had to be chosen between 0. certain combinations of concrete parameters β and κ were selected for each model. without initiation of a critical diagonal tension cracks.3 and 3.85 and 0. 84 CHALMERS Civil and Environmental Engineering.5-3. for the analysis of reinforced concrete (with no bond slip between the reinforcement and the concrete). Master’s Thesis 2005:17 . it was shown that non-linear FE analyses performed in SOLVIA FEsystem can be used for limit state analyses of simply supported reinforced concrete beams analyzed for shear type of failure.9.• The last combinations in the parametric studies. with large κ and small β gave very unstable analyses. only flexural cracks in the tension zone in the mid span could be seen. In order to get a stable FE-analysis. the same parameter had to be multiplied with factors 4.5. Stresses and strains in stirrups and in the longitudinal reinforcement decreased greatly for some of the models. compared with the values obtained for the combinations of κ and β mentioned above. very low ultimate loads and small deflections. In the FE-analysis for model 3 and model 5. In general.

Klaus-Jürgen. pp 3 downloadable at: http://www. Prentice Hall. 35-40.inf. 42-43.cervenka. 1999. Ontario. 1994 pp – 34. Utgåva 2: Konstruktion. Massachusetts Institute of Technology. Helge von (1999): Crack control of restrained concrete structures. Structural and Environmental Engineering.118. Department of Civil and Environmental Engineering. Sweden.bauwesen. Gothenburg. 2003. 28. 30.119.tuuenchen.mit. (2001): On the choice between discrete or smeared approach in practical structural FE analyses of concrete structures. (1996): Finite Element Procedures. 2000.tcd. D. Telford. Master thesis.84 CHALMERS Civil and Environmental Engineering. L. Digest CBD-221.ca/cbd/cbd221e. Holger (2000): Solution of non-linear equilibrium equations in FEanalysis of concrete structures. Institute for Research in Construction. Buyukozturk.pdf Eervenka.cz/Web/printarticle. Sverige. Massachusetts. pp 14-16. Stockholm. 1981. pp 24-25.054/1. Jendele. Department of Structural Mechanics. London : T. 1993 Bresler and Scordelis (1961): Shear strength of reinforced concrete beams. Ireland. 32-33. course literature. (downloadable at http://irc.pdf Bathe. Canada. University of California. Boston.A. 55-56. 1990 pp – 66. Gothenburg. 65. 49-52. J. (downloadable at www. Master’s Thesis 2005:17 85 . pp 10. Internet page: http://www. Institute of Engineering Research. Chalmers University of Technology. Series 100. pp 78. 68-69.541 Mechanics and Design of Concrete Structures. Trinity College Dublin. National Research Council Canada. Sverige. Dublin.asp?id=56 Heidkamp.. byggtjänst. Berkeley. Rexdale.pdf) Hirschhausen. June 1961. Betong-Handbok (1990): Betong-Handbok. USA. June 2001. Fourth International Conference on Analysis of Discontinuous Deformation in Glasgow. New Jersey.10 References Allen. pp 759-761.de/personen/holger/sol_nonl_fe. Department of Civil. pp 1.gc. Lecture notes. Scotland. Örebro.nrc-cnrc. U.S. Oral (2004): 1. Sweden. Department of Structural Engineering.edu/NR/rdonlyres/Civil-and-EnvironmentalEngineering/1-054Spring2004/76D87641-71B5-4432-88B65194DD14312E/0/o_07_shr_f_st_sd. 3-4 downloadable at: http://aka-ocw.html) Basu. UK. Master thesis.ie/Civil_engineering/staff/b_basu/3A2LEC13. B (2003): Structural Design 3A2. AB Svensk byggtjänst. BBK 94 (1994): Boverkets handbok om betongkonstruktioner: BBK 94 Band 1: Konstruktion. Englewood Cliffs. Chalmers University of Technology. California.E (1982): Limit States Design. issue 13. AB Svensk Comité euro-international du béton (1990): CEB-FIP model code 1990 design code.

Prentice Hall.N. Blackwell Publishing. pp 158. pp 56-59. New Jersey. London : 1995. 2002. N. February 2003. 2-31. November. March. 2002. 1997. ENGINEERING AB.page/jem97b. (1999): Reinforced concrete design.(downloadable at www.pdf) Reda Taha.redataha. United States. Internet page: http://dgcwww. James G. M (2002): Structural concrete : theory and design. Ratnamudigedara (2002): Crack spacing. Edward G. pp 190-195. 2002. 1995. pp 2-28. Heron Quay. M. pp 15-16.165418/public/02Chapter1-4.(2004): Composite Structures of Steel and Concrete (third edition). United States. Vol. 209. Nawy. 173-175. Göteborg.156. Nadim Hassoun. Reda Taha. pp 42. august. Oxford. Mario (2002): Finite element analyses of reinforced concrete structures. 2002.gu. 248-250. Piyasena. Västerås. University of New Mexico. Plos. United States. 1999.D and Pavlovic. M. Milan and Zimmermann. Queensland.Jirásek. U. pp 149. Prentice Hall. pp 38-42.pdf Plos. 1998. 164.P. 32. Sweden. Kotsovos. Austrialia. pp 7-8. 86 CHALMERS Civil and Environmental Engineering.1. Mahmoud and.K. crack width and tension stiffening effect in reinforced concrete beams and one-way slabs. 100. Faculty of Engineering. (1995): Composite steel and concrete structural members fundamental behaviour.com/Publications/CI-Fuzzy-Cracking.G (2001): Fracture Mechanics of Concrete. Deric J. Mario (1996): Finite element analyses of reinforced concrete structures. 3. Thomas (1998): Rotating crack model with transition to scalar damagel.html. Fracture of Civil Engineering Materials ENCI-617. UK (2004). Thomas Telford. Oxford. downloadable at: http://www4.au:8080/adtroot/uploads/approved/adtQGU20030606. Griffith University Gold Coast Campus.ch/WWWLSC/jirasek. ACI Structural Journal. 4446. Journal of Engineering Mechanics ASCE. 179-180. 182-185. pp 277-284. Basingstoke : Macmillan. 124 No. 2-32. Prentice Hall. Mosley. Doctoral thesis.(1995): Structural concrete : finite-element analysis for limit-state design. Larsson. 2-29. Shrive. 1995.H. New Jersey. pp 2. SOLVIA MacGregor. Lecture notes. Gunnar (1995): The concrete material model in Solvia. Concrete structures compendium 96:14. Vol. Oehlers. Sweden.edu. Göteborg. W. Sweden. New Jersey. : Pergamon. pp 1-3. Johnson R. Mahmoud (2003): A novel fuzzy-based approach for quantifying tension stiffening of reinforced concrete structures. No. 1996. (1997): Reinforced concrete mechanics and design. (2003): Reinforced concrete : a fundamental approach.epfl. 5-9. pp 6. Master’s Thesis 2005:17 .

Users manual. Sweden.PDF).46. 6.25-9. Master’s Thesis 2005:17 87 .edu/~mrtaha/Publications/ENCI-617-Notes.52. 9. Västerås. Report SE 99-1.26. pp 1.Albuquerque. 6.unm. CHALMERS Civil and Environmental Engineering. pp 1-3.2. SOLVIA ENGINEERING AB. 2001.41. New Mexico. (downloadable www. 6. at SOLVIA Finite Elment System (1999): Solvia PRE for stress analysis. 1999.

Master’s Thesis 2005:17 .APPENDIX AMesh of the reinforcement for the models created in 3D A-1 Model 4 A-2 Model 5 B- Calculation of the ultimate load B-1 Beam OA-3 B-2 Beam B-14 B-3 Beam A-1 C- Calculation of the fracture energy of the concrete for tested beams 88 CHALMERS Civil and Environmental Engineering.

1 Mesh of the steel reinforcement for Model 4. CHALMERS Civil and Environmental Engineering.AA-1 Mesh of the reinforcement for the models created in 3D Model 4 Figure A. Master’s Thesis 2005:17 1 .

A-2 Model 5

Figure A.2

Mesh of the longitudinal steel reinforcement for Model 5.

Figure A.3

**Mesh of the shear reinforcement for Model 5.
**

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

2

B

Calculation of the ultimate load

B-1 Beam OA-3

0.433 0.56

2x0.064 0.06 0.090.09 0.06 0.30

Figure B.1

Geometry, loading and cross section for beam OA-3.

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

3

Dimensions

**bw = 0.305 m d = 0.465 m
**

Materials

h = 0.56 m

L = 6.40 m

Concrete Reinforcing steel

**fc = 37.62 MPa fy = 552.96 MPa
**

Es = 205.72 GPa

ft 1 = 4.14 MPa

As = 3847.81 ⋅ 10 −6 m²

**The ultimate load due to flexural capacity
**

Flexural capacity of the section

ε cu

x h d As bw

Figure B.2

fc 0.8x Fc 0.4x

εs

Fs

Mn

Equilibrium and strain compatibility approach in the analysis of the flexural capacity of reinforced concrete beams.

Horizontal equilibrium

Fc = Fs

(B.1)

Force in the concrete

Fc = 0.8 fcbwx

Force in the steel reinforcement

Fs = σs ⋅ As

1

(B.2)

(B.3)

This corresponds to the average value of the modulus of rupture of concrete given in the experimental tests of Bresler Scordelis.

4

CHALMERS Civil and Environmental Engineering, Master’s Thesis 2005:17

62 ⋅ 10 6 ⋅ 0. (B.5) Assumptions: σs fy Εs εy Figure B.7) (B.1).3).5 ⋅ 10 −3 2) Yielding of reinforcing steel (B.96 ⋅ 10 6 ⋅ 3847.Strain in the steel reinforcement εs = d−x ⋅ εcu x (B. (B.81 ⋅ 10 −6 x = 0.8) σs = fy Substituting Equations (B. Master’s Thesis 2005:17 5 .232 m CHALMERS Civil and Environmental Engineering.6) εs ≥ εy (B.3 εs Stress-strain relationship for reinforcing steel.5) and (B.8) in Equation (B. 1) Crushing of concrete εcu = 3. the depth of the neutral axis is found: 0.2).305 x = 552.8 ⋅ 37.4) Yield strain in the steel reinforcement εy = fy Es (B.

8 fcbwx .Check of assumptions The strain at the level of the tensile reinforcement is found with Equation (B.72 ⋅ 10 9 εy = 0.96 ⋅ 10 6 εy = 205.0027 εs > εy ∴The reinforcement yields.465 and x = 0.62 ⋅ 10 6 ⋅ 0.4): εs = 0.4 x) With Equation (B.00352 The yield strain in the steel reinforcement is found with Equation (B. d = 0.232 ⋅ (0.10) 792.232 Mn = 0.4 ⋅ 0. The assumption is correct.8 ⋅ 37. Flexural capacity Mn = Fc ⋅ (d − 0.5): 552. Master’s Thesis 2005:17 .2) Fc = 0.10 kNm (B.0035 0.465 − 0.232 εs = 0.465 − 0.232 ⋅ 0.40 Pf = 495.9) Ultimate load due to flexural capacity Pf = 4 ⋅ Pf = 4 ⋅ Mn L (B.305 ⋅ 0.232) Mn = 792.06 kN 6 CHALMERS Civil and Environmental Engineering.10 6.

13) as follows: ft .135 CHALMERS Civil and Environmental Engineering. the splitting tensile strength can be considered in calculations of the shear resistance as the tensile strength ft.The ultimate load due to shear capacity Vc C Va T R Figure B.11:6 with the Equation (B. sp = 5 ⋅ ft .465 m (B.48 MPa Sectional constant relating to sectional depth and curtailment of reinforcement ξ = 1. sp = 2.12) For members subjected to bending and shear forces. Shear resistance of the concrete without stirrups Vc = bwdfv Concrete basic shear strength (B.465 = 1. fl = ⋅ 4.4 Vd Internal forces in the beam OA-3 failing in shear.30 ft Splitting tensile strength (B.14 = 2. sp 3 3 3 ⋅ ft . fl = ft .13) ft = ft .6 − d for 0.50 m d = 0. Master’s Thesis 2005:17 7 .48 5 5 (B.2 m < d ≤ 0. This parameter is defined as a function of the modulus of rupture according to Betonghandbok 11.6 − 0.14) ξ = 1.11) fv = ξ ⋅ (1 + 50 ⋅ ρw) ⋅ 0.

12): fv = 1.13).Reinforcement ratio corresponding to As 0 ρw = As 0 bwd (B.305 ⋅ 0.16) Pv = 402.50 kN The ultimate load of the beam OA-3 is the lesser of the two calculated loads.135 ⋅ (1 + 50 ⋅ 0.14) and the concrete tensile strength ft in Equation (B.0136) ⋅ 0.305 ⋅ 0.419 ⋅ 10 6 Vc = 201.0136 0. (B.48 ⋅ 10 6 fv =1.465 Substituting the values obtained from Equations (B.11): Vc = 0. Master’s Thesis 2005:17 .30 ⋅ 2. P* = 402. the shear resistance Vc can be calculated with Equation (B. and is due to shear.58) 2 4 = 1923.15) Ø=28.25 kN Vc is the maximum shear force that acts in the beam and is equal to half of the concentrated load applied in the mid span.419⋅MPa With the value of fv.58 mm.91 ⋅ 10 −6 = 0.25 (B. 3 bars form 1 layer.91⋅10 −6 ρw = 1923.50 kN 8 CHALMERS Civil and Environmental Engineering. Ultimate load due to shear resistance Pv = 2 ⋅ Vc Pv = 2 ⋅ 201.465 ⋅ 1. As0 = 3 ⋅ π (28.

040 0.040 0.B-2 Beam B-14 0.34 0. CHALMERS Civil and Environmental Engineering.039 0. loading and cross section for beam B-14.23 Figure B.039 0.5 Geometry.040 0.0750.222 0. Master’s Thesis 2005:17 9 .075 0.

40 m a=1m Materials Concrete Reinforcing steel fc = 29.8x Fs' Fc 0. Horizontal equilibrium Fc + Fs ' = Fs (B.18) (B.78 MPa fy = 628 MPa As = 1005.23 m d = 0.039 m L = 3.32 ⋅ 10 −6 m² As ' = 603.20 m fyv = 532 MPa The ultimate load due to flexural capacity Flexural capacity of the section d' h d As' x ε cu ε s' fc 0.34 m d’ = 0.50 ⋅ 10 −6 m² s = 0.19 ⋅ 10 −6 m² Es = 207 GPa Shear reinforcement Asv = 56.285 m h = 0.19) 10 CHALMERS Civil and Environmental Engineering.20 MPa ft = 2.Dimensions bw = 0.8 fcbwx Force in the compressive reinforcement Fs ' = σs '⋅ As ' (B. Master’s Thesis 2005:17 .6 εs Fs Mn Equilibrium and strain compatibility approach in the analysis of the flexural capacity of reinforced concrete beams.17) Force in the concrete Fc = 0.4x As bw Figure B.

Master’s Thesis 2005:17 11 .5 ⋅ 10 −3 2) The compressive steel behavior is described by part 1 in Figure B.27) CHALMERS Civil and Environmental Engineering.7 (B. 1) Crushing of concrete εcu = 3.24) εs ' < εy σs ' = εs ' Es 3) The tensile steel behavior is described by part 2 in Figure B.25) (B.Force in the tensile reinforcement Fs = σs ⋅ As (B.23) Assumptions: σs fy Εs εy Figure B.7 (B.26) εs ≥ εy (B.22) Yield strain in the steel reinforcement εy = fy Es (B.7 εs Stress-strain relationship for reinforcing steel.21) Strain in the tensile reinforcement εs = d−x ⋅ εcu x (B.20) Strain in the compressive reinforcement εs ' = x − d' ⋅ εcu x (B.

0017 The yield strain in the steel reinforcement is found with Equation (B. and d’ = 0.2).077 m Check of assumptions x − 0.039) Mn = 158. x = 0.4 ⋅ 0. the depth of the neutral axis is found 0.0030 εs ' < εy ∴The reinforcement does not yield.4 x) + Fs ' (d − d ' ) With Fs ' = σs '⋅ As ' and σs ' = εs ' Es (B.285 − 0.077.19 ⋅ 10 −6 (0.7): εy = 628 ⋅ 10 6 207 ⋅ 10 9 εy = 0. (B.5): εs ' = 0.20 ⋅ 10 6 ⋅ 0.039 Mn = 0.67 kNm 12 CHALMERS Civil and Environmental Engineering.0017 ⋅ 207 ⋅ 10 9 ⋅ 603.285.077 ⋅ (0.0035 ⋅ 207 ⋅ 10 9 ⋅ 603.23 ⋅ 0.29) Mn = Fc ⋅ (d − 0.32 ⋅ 10 −6 x The strain at the level of the compressive steel is found with Equation (B. (B.0035 0.2).077 ⋅ 0.285 − 0. Master’s Thesis 2005:17 .5).10). (B.23x + x = 0.039 ⋅ 0.20 ⋅ 10 6 ⋅ 0.4 x) + εs ' EsAs ' (d − d ' ) With Equation (B.035 0. The assumption 3) is correct Flexural capacity Mn = Fc ⋅ (d − 0.1).077 εs’ = 0.0094 εs > εy∴The reinforcement yields.σs = fy (B. (B.6): εs = 0.3).28) Substituting Equations (B.077) + 0.285 − 0. The assumption 2) is correct The strain at the level of the tensile steel is found with Equation (B.077 − 0.12) in Equation (B. (B.039 ⋅ 0.19 ⋅ 10 −6 = 628 ⋅ 10 6 ⋅ 1005.8 ⋅ 29. d = 0.4).8 ⋅ 29.077 εs = 0.

67 1. Master’s Thesis 2005:17 13 .67 kN CHALMERS Civil and Environmental Engineering.30) 158.Ultimate load due to flexural capacity Pf = Pf = Mn a (B.00 Pf = 158.

As0 = 2 ⋅ π (16) 2 4 = 402.0061 0.12 ⋅ 10 −6 ρw = = 0.2 m < d ≤ 0.315 m Reinforcement ratio corresponding to As 0 ρw = As 0 bwd (B.12 ⋅ 10 −6 402.30 ft (B. Shear resistance of concrete without stirrups Vc = bwdfv Concrete basic shear strength (B.23 ⋅ 0.50 m d = 0.32) Sectional constant relating to sectional depth and curtailment of reinforcement ξ = 1.34) Ø=16 mm.33) ξ = 1.6 − d for 0.31) fv = ξ ⋅ (1 + 50 ⋅ ρw) ⋅ 0.6 − 0.The ultimate load due to shear capacity Vc C s Asv f yv Asv f yv T R Figure B. 2 bars form 1 layer.8 d Shear resisted by stirrups.285 = 1.285 m (B.285 14 CHALMERS Civil and Environmental Engineering. Master’s Thesis 2005:17 .

17).16).11 + 38.2 Shear capacity of the section with shear reinforcement VR = Vc + Vs (B.0061) ⋅ 0. (B.315 ⋅ (1 + 50 ⋅ 0. Master’s Thesis 2005:17 15 .15).30 ⋅ 2.11 kN Shear reinforcement contribution Vs = Asvfyv ⋅ 0.70 ⋅ 10 6 fv =1.39 ⋅ 10 6 Vc = 91. and is due to shear.65 kN CHALMERS Civil and Environmental Engineering.39⋅MPa With the value of fv. Vs = 56. and therefore Pv = VR The ultimate load of the beam B-14 is the lesser of the two calculated loads.65 kN Ultimate load due to shear resistance VR is the maximum shear force that acts in the beam.285 0.18) and the concrete tensile strength ft in Equation (B.54 kN 0. Vc = 0.9d (sin α + cos α ) s (B. P* = 129. the shear resistance of concrete Vc can be calculated with Equation (B.50 ⋅ 10 −6 ⋅ 532 ⋅ 10 6 ⋅ Vs = 38.36) VR = 91.23 ⋅ 0.54 VR = 129.35) α = 0 ∴No inclined reinforcement. According to BBK ft≤ 2.70 MPa fv = 1.285 ⋅ 1.Substituting the values obtained from Equations (B.9 ⋅ 0.

B-3 Beam A-1 0.56 0.05 0. 16 CHALMERS Civil and Environmental Engineering.9 Geometry. loading and cross section of the beam A-1.38 0.128 0.18 0.06 0.06 0.30 Figure B. Master’s Thesis 2005:17 .

10 Equilibrium and strain compatibility approach in the analysis of the flexural capacity of reinforced concrete beams.8x Fs' Fc 0.8 fcbwx 1 (B.66 m Concrete Top reinforcing steel fc = 24.37) Force in the concrete Fc = 0.465 m Materials h = 0.35 ⋅ 10 −6 m² Bottom reinforcing steel fy = 555.72 MPa Es = 218.34 ⋅ 10 −6 m² s = 0.305 m d = 0.84 MPa The ultimate load due to flexural capacity Flexural capacity of the section d' h d As' x ε cu ε s' fc 0.050 m L = 3.86 MPa Es ' = 201. Horizontal equilibrium Fc + Fs ' = Fs (B.144 GPa As = 2565.21 m fyv = 325.38) This corresponds to the average value of the modulus of rupture of concrete given in the experimental tests of Bresler Scordelis.09 MPa fy ' = 345.86 MPa As ' = 253.56 m d’ = 0.Dimensions bw = 0.4x As bw εs Fs Mn Figure B. Master’s Thesis 2005:17 17 .576 GPa ft 1 = 3. CHALMERS Civil and Environmental Engineering.21 ⋅ 10 −6 m² Shear reinforcement Asv = 63.

40) Strain in the compressive reinforcement εs ' = x − d' ⋅ εcu x (B.39) Force in the tensile reinforcement Fs = σs ⋅ As (B. Master’s Thesis 2005:17 . 1) Crushing of concrete εcu = 3.44) Assumptions: σs fy Εs εy εs Figure B.41) Strain in the tensile reinforcement εs = d−x ⋅ εcu x (B.Force in the compressive reinforcement Fs ' = σs '⋅ As ' (B.5 ⋅ 10 −3 (B.45) 18 CHALMERS Civil and Environmental Engineering.43) Yield strain in the tensile steel reinforcement εy = fy Es (B.42) Yield strain in the compressive steel reinforcement εy ' = fy ' Es ' (B.11 Stress-strain relationship for reinforcing steel.

0035 0. Master’s Thesis 2005:17 19 .46) (B.035 0.576 ⋅ 10 9 εy’ = 0.11 εs ' ≥ εy ' σs ' = fy ' 3) The tensile steel behavior is described by part 2 in Figure B. The assumption 2) is correct The strain at the level of the tensile steel is found with Equation (B.86 ⋅ 10 6 ⋅ 253.72 ⋅ 10 6 218.47) εs ≥ εy (B.72 ⋅ 10 6 ⋅ 2565.0025 CHALMERS Civil and Environmental Engineering. (B.465 − 0.00172 εs ' > εy ' ∴The reinforcement yields.0037 The yield strain in the tensile steel reinforcement is found with Equation (B.144 ⋅ 10 9 εy = 0.49) σs = fy Substituting Equations (B.11).05 ⋅ 0.7): εy ' = 345.11 (B.2).4). (B.86 ⋅ 10 6 201.8 ⋅ 24.8): εy = 555.2) The compressive steel behavior is described by part 2 in Figure B.305 x + 345.1).35 ⋅ 10 − 6 = 555. (B.228 εs’ = 0.09 ⋅ 10 6 ⋅ 0. the depth of the neutral axis is found: 0.48) (B.228 ⋅ 0.228 − 0.228 εs = 0.0027 The yield strain in the compressive steel reinforcement is found with Equation (B. (B.21 ⋅ 10 − 6 x = 0.6): εs = 0.5): εs ' = 0.228 m Check of assumptions The strain at the level of the compressive steel is found with Equation (B.13) in Equation (B.3).

86 ⋅ 10 6 ⋅ 253.09 ⋅ 10 6 ⋅ 0.35 ⋅ 10 −6 (0.228) + 345.465 − 0.66 Pf = 586.68 kNm Ultimate load due to flexural capacity Pf = 4 ⋅ Pf = 4 ⋅ Mn L 536.228 ⋅ (0.05) Mn = 536. σs ' = fy ' Mn = Fc ⋅ (d − 0. Master’s Thesis 2005:17 .4 x) + fy '⋅ As ' (d − d ' ) (0.2).050 Mn = 0. x = 0.15) With Equation (B.68 3.465.54 kN 20 CHALMERS Civil and Environmental Engineering.4 ⋅ 0.14) With Fs ' = σs '⋅ As ' .εs > εy∴The reinforcement yields.8 ⋅ 24. The assumption 3) is correct Flexural capacity Mn = Fc ⋅ (d − 0.228. d = 0.305 ⋅ 0. and d’ = 0.4 x) + Fs ' (d − d ' ) (B.465 − 0.

11:6 with the Equation (B.17) For members subjected to bending and shear forces.30 ft Splitting tensile strength (B.19) ξ = 1.465 m (B. the splitting tensile strength can be considered in calculations of the shear resistance as the tensile strength ft. This parameter is defined as a function of the modulus of rupture according to Betonghandbok 11. Shear resistance of concrete without stirrups Vc = bwdfv Concrete basic shear strength (B. sp 3 3 3 ⋅ ft .2 m < d < 0.6 − d for 0.12 Shear resisted by stirrups. sp = 5 ⋅ ft .86 = 2.50 m d = 0.16) fv = ξ ⋅ (1 + 50 ⋅ ρw) ⋅ 0.465 = 1.18) as follows: ft .135 m CHALMERS Civil and Environmental Engineering. fl = ⋅ 3.18) ft = ft . Master’s Thesis 2005:17 21 . fl = ft .6 − 0.32 5 5 (B.The ultimate load due to shear capacity Vc C s Asv f yv Asv f yv T R d Figure B. sp = 2.32 MPa Sectional constant relating to sectional depth and curtailment of reinforcement ξ = 1.

21) α = 0 ∴No inclined reinforcement.58) 2 4 = 1282.22) VR = 203.30 ⋅ 2.32 ⋅ 10 6 fv =1.34 ⋅ 10 −6 ⋅ 325.305 ⋅ 0.35 kN VR is the maximum shear force that acts in the beam and is equal to half of the concentrated load applied in the mid span.15 MPa With the value of fv.465 ⋅ 1. Vs = 63.465 Substituting the values obtained from Equations (B.0090) ⋅ 0. (B.58 mm.13 (B. 2 bars form 1 layer. Master’s Thesis 2005:17 .21 Shear capacity of the section with shear reinforcement VR = Vc + Vs VR = 162.465 0.84 ⋅ 10 6 ⋅ Vs = 41. 22 CHALMERS Civil and Environmental Engineering.18).17): fv = 1.0090 0. As0 = 2 ⋅ π (28. the shear resistance of concrete Vc can be calculated with Equation (B.13 kN 0.9 ⋅ 0.135 ⋅ (1 + 50 ⋅ 0.20) Ø=28.22 kN Shear reinforcement contribution Vs = Asvfyv ⋅ 0.60 ⋅ 10 −6 ρw = 1282.60 ⋅ 10 −6 = 0.305 ⋅ 0.16): Vc = 0.Reinforcement ratio corresponding to As 0 ρw = As 0 bwd (B.19) and the concrete tensile strength ft in Equation (B.9d (sin α + cos α ) s (B.15 ⋅ 10 6 Vc = 162.22 + 41.

23) The ultimate load of the beam A-1 is the lesser of the two calculated loads.Ultimate load due to shear resistance Pv = 2 ⋅ Vc Pv = 2 ⋅ 203.70 Kn CHALMERS Civil and Environmental Engineering.70 kN (B.35 Pv = 406. P* = 406. Master’s Thesis 2005:17 23 . and is due to shear.

09 90.89 The ultimate crack displacement wu. From this relationship. Gf [Nm/m²] Max.89 77. and wu. dmax [mm] fc [MPa] Gf [Nm/m²] Beam OA-3 B-14 A-1 19 19 19 37. The fracture energy of concrete Gf. Table C.2.62 29. shown in Figure C. was calculated according to Hilleborg’s Fictitious Crack Model. Table C. for which the crack normal stress becomes zero. could be found as: 24 CHALMERS Civil and Environmental Engineering. the fracture energy is approximated as the area under the curve.55 69.C Calculation of the fracture energy of the concrete for tested beams According to the RILEM Draft Recommendation TC50-FMC contained in the CEBFIP Model Code 90 the fracture energy Gf is determined on notched specimens loaded in flexure. Master’s Thesis 2005:17 .1 Fracture energy Gf for different concrete grades and aggregate sizes. and are shown in Table C.1.1. is the energy required to propagate a tensile crack of unit area.2 Values of the fracture energy for tested beams. aggregate size dmax (mm) 8 16 32 C12 40 50 60 C20 50 60 80 C30 65 75 95 C40 70 90 115 C50 85 105 130 C60 95 115 145 C70 105 125 160 C80 115 135 175 The values of the fracture energy for the different beams in this thesis project were determined by interpolation of the values shown in Table C.20 24.

3) ε cu x h d Aef def a tp Figure C.2) where x1 = 0.wu = 2 ⋅ Gf ft (C. Master’s Thesis 2005:17 25 .25 − def 8(h − x) (C. was determined from the relation: srm = 50 + x1 + x 2 φ lr (C.1 Stress-crack opening relation for concrete after crack initiation.8 for high bond bars x 2 = 0.1) Figure C. The average crack spacing srm.2 Effective tension area ε2 ε1 ε CHALMERS Civil and Environmental Engineering.

2.58 16 28. Table C.3 for the different beams based on the previous Equations.5 131.057 0.08 127.1 as: εt = ft Eo (C.040 465 285 465 95 55 95 305 230 305 67710 27830 63836. 26 CHALMERS Civil and Environmental Engineering.72 The ultimate crack strain εtu for smeared crack models in the case of perfect bond between the steel reinforcement and the surrounding concrete was calculated with Equation (C.3 560 340 560 232 77 228 28.86 127 340 114.def = 2c + φ atp = h − d lr = (C.9) and are shown in Table C.10 Srm[mm] d[mm] atp[mm] bw[mm] Aef[mm²] lr 0. Srm.0602 49.21 25 42. Master’s Thesis 2005:17 .8) according to Figure 4. wu[mm] c[mm] def[mm] h[mm] x[mm] Ø[mm] As[mm²] Beam OA-3 B-14 A-1 Beam OA-3 B-14 A-1 0.3 Average crack spacing.5) (C.31 2566.6) As Aef The values of the average crack spacing are summarized in Table C.81 1005.2 in section 4.47 167.036 0. calculated for tested beams.7).7) The tensile crack strain εt for concrete was determined with Equation (C.0558 0.8) Finally the KAPPA values for the different beams were calculated with Equation (C.4. εtu = wu srm (C.4) (C.58 3847.0733 0.

4 Beam OA-3 B-14 A-1 εtu 5.59x10-4 4. for tested beams.11:6 for the experimental tested beams of Bresler Scordelis.35x10-5 8.35x10-5 9.33 29720. Master’s Thesis 2005:17 27 .38x10-4 3.48 2.59x10-4 εt 7. ft1[MPa] Eo[MPa] (C.86 1 The tensile strength of concrete ft was considered as the value of the splitting tensile strength ft.36 26994.67 4.32 33734.fl calculated according to Betong Handbok 11.78 2.59x10-5 κ 7. CHALMERS Civil and Environmental Engineering.k= εtu εt Calculated values of parameterκ.18 2.61 4.9) Table C.