Construction and Building Materials

Construction and Building Materials 26 (212) 172 179 Contents lists available at ScienceDirect Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat Compressive strength of micropile-to-grout connections J. Veludo a, E.N.B.S. Júlio b,, D. Dias-da-Costa c a ICIST, Department of Civil Engineering, Polytechnic Institute of Leiria, Campus 2, Morro do Lena, Alto do Vieiro, 2411-91 Leiria, Portugal b ICIST, Department of Civil Engineering, Instituto Superior Técnico, Technical University of Lisbon, Lisbon, Portugal c INESCC, Department of Civil Engineering, University of Coimbra, Coimbra, Portugal article info abstract Article history: Received 28 October 21 Received in revised form 1 May 211 Accepted 13 June 211 Available online 12 July 211 Keywords: Bond Strength Confinement Retrofitting Interface Micropile Grout Push-off tests Strengthening foundations with micropiles is progressively being used, due to the major advantages that this technique presents. Nevertheless, the influence of some relevant parameters in the overall behavior of the retrofitted foundations has not yet been studied. Generally, micropiles are installed in holes drilled through the existing RC footing, which are then filled with grout. The efficiency of the load transfer mechanism depends on the bond strength of both the micropile grout and the concrete grout interfaces. This paper describes an experimental study performed to specifically study the influence of the following parameters on the bond strength between micropile grout interface: hole diameter; embedment length of the micropile; and level of confinement of the grout mass. Thirty micropile grout specimens were submitted to monotonic push-off tests, until failure. Bond strength was found to increase with a decrease of the hole diameter and with an increase of the confinement level. Ó 211 Elsevier Ltd. All rights reserved. 1. Introduction RC footing strengthened with steel micropiles is today a widespread method. However, in practice, it is empirically applied. In reality, although the parameters that influence the connection capacity are identified, these have not been quantified in a comprehensive way. In general, micropiles are applied to the existing RC footings through predrilled holes. After the micropile is installed, the hole is filled with non-shrink grout. It is known that the efficiency of the connection depends on the bond strength of both the micropile grout and the grout concrete interfaces but this has not been characterized so far. To increase the bond strength of the micropile grout interface, it is common practice to weld steel rings or a steel spiral around the perimeter of the casing. To improve the bond strength of the grout concrete interface, sometimes grooves are chipped into the wall of the hole. Detailing depends mainly on the required capacity [1 3]. Attention must also be paid to the reinforcement of the existing concrete footing since this can be inadequate for the strengthened situation. In this case, the lateral and the axial confinement of the existing footing should be increased to improve the connection capacity. Corresponding author. Tel.: +351 218 418 258; fax: +351 218 497 65. E-mail addresses: joao.veludo@ipleiria.pt (J. Veludo), ejulio@civil.ist.utl.pt (E.N.B.S. Júlio), dcosta@dec.uc.pt (D. Dias-da-Costa). As mentioned, few experimental studies have been conducted on this subject. Gómez et al. [4] performed push-off tests to study the connection capacity of smooth micropiles, grouted in reinforced concrete footings through holes drilled using a jack hammer. The authors concluded that the connection capacity is controlled by adhesion and friction at the micropile grout interface and that the residual capacity of the connection is entirely frictional and dependent on the confinement provided by the footing reinforcement. It is also concluded that the connection capacity increases with the decrease of the hole diameter and that the embedment length has little influence on the bond strength. Contrarily to micropile footing connections, the bond mechanism between reinforcing bars and concrete is well studied and it is generally accepted that it is controlled mainly by three different parameters: (1) chemical adhesion between concrete and steel; (2) friction between concrete and steel; and (3) bearing of the rebar ribs on the surrounding concrete [5 8]. Two different bond failure mechanisms can be observed: splitting failure or pull-out failure. Various studies have also been performed in grouted ribbed rebar or cable bolts. Moosavi et al. [9] studied the bond of cement grouted reinforcing bars under constant radial pressure. These authors state that the properties of both grout and confinement play an important role in developing the bond capacity. Hyett et al. [1] performed several pull-out tests of grouted cable bolts using a modified Hoek cell and concluded that the bond strength at the cable bolt grout interface is more related to friction rather than to chemical adhesion. The authors also state that the 95-618/$ - see front matter Ó 211 Elsevier Ltd. All rights reserved. doi:1.116/j.conbuildmat.211.6.7

J. Veludo et al. / Construction and Building Materials 26 (212) 172 179 173 bond strength depends on the pressure generated at the cable bolt grout interface caused by dilatancy during bond failure. In a previous study, Hyett et al. [11] indicate that the parameters with a major influence in cable bolt capacity are: the grout properties (w/c ratio); the cable bolt embedment length; and the radial confinement acting on the outer surface of the grout mass. Yahia et al. [12] studied the bond strength of cement grout anchors cast in rock under dry and submerged conditions. These authors concluded that the main mechanism to mobilize the bond strength is the friction developed at both grout rebar and grout rock interfaces. It is also concluded that friction depends on: the mechanical properties of both rock and grout; the geometry of the hole and of the bar; and the roughness of the drilled hole surface. Malvar [13] and Noghabai [14] performed pull-out tests with deformed bars cast in concrete cylinders under different radial confinements. Both authors conclude that the confinement is the most important factor at the slip phase and observed the increase of bond stress with the growth of the level of confinement. From the studies previously referred to, it can be assumed that the load transfer mechanism between a strengthening steel micropile and the existing reinforced concrete footing is controlled by three different parameters: (1) chemical adhesion at the steel grout and the grout concrete interfaces; (2) friction at the steel grout and the grout concrete interfaces; and (3) bearing of the welded steel rings/spirals at the steel grout interface and of the concrete grooves at the grout concrete interface. Design codes for RC structures, namely ACI 318 [15], EuroCode 2 [16] and the CEB-FIP ModelCode 199 [6], specify expressions for the design of the bond strength and development length of bars embedded in concrete. However, none of these codes presents expressions to compute the development and the bond stress of bars grouted in holes predrilled in concrete. Furthermore, using current codes expressions to determine the development length of a smooth micropile in a predrilled hole filled with grout would be too conservative. Moreover, this would require having a significantly deep foundation, which is not observed in practice. The study herein described focuses specifically the behavior of the steel grout interface with smooth micropile inserts aiming to quantify the influence of the following parameters: the hole diameter, the embedment length; and the confinement of the grout mass. 2. Research significance Strengthening existing RC footings with grouted steel micropiles is currently one of the most used retrofitting techniques of foundations. However, in practice, this is performed empirically since the behavior of the micropile footing connection has not been characterized so far. The experimental study herein described contributes to the knowledge of the behavior of the micropile grout interface. It is demonstrated that: the capacity of the micropile footing connection increases with the decrease of the hole diameter; and the bond strength strongly depends on the radial confinement; but it is not significantly influenced by the insert embedment length. 3. Experimental investigation Aiming to study the influence of the insert embedment length, the hole diameter and the confinement of the grout mass, 15 different situations were defined and, for each one, two specimens were tested in compression. In order to enable different radial confinement levels of the grout mass, steel tubes, PVC tubes and RC blocks were used. First, 3 smooth inserts were positioned inside 1 steel tubes, 1 PVC tubes and 1 holes predrilled in RC footings. Then, an unreinforced mass of grout was used to seal the void between the inserts and the walls of the tubes/holes. Afterwards, these specimens were tested in compression until failure, to evaluate the influence of the parameters referred to on the bond strength of the micropile grout interface. Besides these, the remaining parameters were kept constant, such as: grout type and strength; concrete type and strength; micropile insert; and loading procedure. In the following paragraphs, the materials adopted in this study, the different situations considered, the geometry adopted for the specimens, the production of the specimens; and the tests set-up are described. 3.1. Materials The micropile inserts were produced using smooth 6 mm API N8 steel tube grade 562/73 MPa, with 6 mm thickness. In tests performed with concrete specimens, the tube was reinforced with a 16 mm grade 5/6 MPa bar which was first welded to the center of a 15 15 2 mm 3 steel plate. Finally, the insert was fully grouted. An average roughness of 1 mm for the tube surface was measured with a laser roughness analyzer [17]. A grout with a measured compressive strength of 53.4 MPa and a Young s modulus of 14.2 GPa, at 28 days, and a water cement ratio of.4 was adopted. The mix proportions per cubic meter are the following: 1327 kg of type I:42.5 R Portland cement, 53 l of water, 13.27 kg of modified polycarboxylate admixture (high range water reducer); and 13.27 kg of expansive admixture. At each age, a set composed by six specimens obtained from the flexure test performed on three prismatic specimens with 4 4 16 mm 3 were used to evaluate the average compressive strength at 1, 3, 7, 14, 21, 28, 56, and 9 days of age [18,19] (see Fig. 1a). Other set composed by three prismatic specimens with the dimensions previously referred to were used to evaluate the corresponding Young s modulus [2] (see Fig. 1b). The specific gravity of the grout was 19.2 kn/m 3. The following properties were measured, according to European standards [19]: a flowability of 11 s; a volume change of %; a bleed of.45%; and an air content of 2%. These results were considered acceptable, also according to European standards [21]. Unconfined Compressive Strength, fcm (MPa) Young s Modulus, Ecm (GPa) 7 6 5 4 3 2 1 53.4 32.5 Concrete Grout 14 28 42 56 7 84 98 112 Curing Time (days) 45 4 35 3 25 2 15 1 35.2 14. (a) 5 Concrete Grout 14 28 42 56 7 84 98 112 Curing Time (days) (b) Fig. 1. Mechanical properties of concrete and grout: (a) unconfined compressive strength and (b) Young s modulus.

174 J. Veludo et al. / Construction and Building Materials 26 (212) 172 179 Table 1 The complete push-off test program. Specimen For the PVC confinement, a PN1 pressure class tube with three different diameters (see Table 1) was used, presenting according to the manufacturer a Young s modulus of 3. GPa and a Poisson ratio of.3. For the steel confinement, a St 52 BK + S grade 52/6 MPa tube with three different diameters was used (see Table 1). For the RC blocks, a concrete mix with a measured compressive strength of 32.5 MPa and a Young s modulus of 35 GPa, at 28 days, was used. The following mix proportions per cubic meter of concrete were adopted: 28 kg of type II:42.5 R Portland cement, 18 l of water, 25 kg of washed siliceous sand with 2.56 fineness modulus, 71 kg of siliceous sand with 3.71 fineness modulus, 88 kg of limestone crushed aggregates with 6.35 fineness modulus and 2.8 kg of a water reducing admixture. Three cubes with 15 mm were produced to evaluate the average compressive strength [22,23] at each of the following ages: 1, 3, 7, 14, 21, 28, 56 and 9 days (see Fig. 1a). The corresponding Young s modulus was also measured [2], using 15 15 6 mm 3 specimens (see Fig. 1b). The concrete blocks were reinforced at the bottom face with an 8 mm S4 grid, with five bars in each direction, and presenting a 5 mm of nominal cover. 3.2. Test specimens Diameter of grout (D g ) (mm) Embedment length (L e ) (mm) Confinement D ext (mm) t (mm) Type P1-P2 11 2 11 4.5 PVC P3-P4 11 275 11 4.5 PVC 11 35 11 4.5 PVC P7-P8 81 35 9 4.5 PVC P9-P1 119 35 13 5.5 PVC S1-S2 1 2 11 5. Steel S3-S4 1 275 11 5. Steel S5-S6 1 35 11 5. Steel S7-S8 8 35 9 5. Steel S9-S1 12 35 13 5. Steel C1-C2 12 2 45 174 Concrete C3-C4 12 275 45 174 Concrete C5-C6 12 35 45 174 Concrete C7-C8 82 35 45 184 Concrete C9-C1 122 35 45 164 Concrete D ext external diameter; t thickness. In Fig. 2 is presented the geometry of the specimens adopted for the push-off tests: (a) specimens confined using steel/pvc tubes and (b) specimens confined using RC blocks. In Table 1 the different 15 situations considered are presented, including the diameter of the grout mass, the insert embedment length, and the type and geometry (steel, PVC or concrete) of the confinement of the grout mass. This is part of a wider experimental program aiming to assess the most significant parameters related to the connection capacity between micropiles and existing RC footings. For few set-ups, the load applied to the RC blocks could exceed the strength of the steel tubes, thus justifying the rebar represented in Fig. 2b. Since expected failure occurs at the steel grout interface, their influence is neglected. Tube External Confinement Grout Steel Plate 5 mm Dg (a) Le Steel Plate Drill Hole Tube Dywidag Bar Polystirene Insert 15 mm Dg Le 45x45 mm 2 Steel Plate Fig. 2. Geometry of the confined specimens using: (a) steel/pvc tubes and (b) RC blocks. 5 mm Table 2 Wall stiffness of the confining tubes and RC footings used in tests. Confinement E cm (GPa) m d o (mm) d i (mm) K r (MPa/mm) PVC 3 a.3 9 81 8.2 PVC 3 a.3 11 11 5.4 PVC 3 a.3 13 119 4.7 Steel 2 a.3 9 8 613.4 Steel 2 a.3 11 1 41.3 Steel 2 a.3 13 12 282.9 Concrete 37.1 b.2 45 82 714.8 Concrete 37.1 b.2 45 12 557.9 Concrete 37.1 b.2 45 122 449.7 a b Values assumed according to manufacturer. At time of tests (17 days after casting concrete blocks). One of the main objectives of this investigation was to compare the bond strength at the micropile grout interface using different confinement levels of the grout mass. The PVC tube represents the lowest confinement level and was mainly used to prevent splitting of the grout mass during the hardening process. The steel tube was adopted to apply a moderate confinement. For the highest confinement level, a concrete block with dimensions 45 45 5 mm 3 was adopted. The radial stiffness (K r ) of the different types of confinement considered can be estimated from the thick wall cylinder theory, according to the following equation [24]: 8 9 K r ¼ 2E < ð1 þ vþ d 2 o = d2 i h i ð1þ : d i ð1 2vÞd 2 i d 2 ; o where E is the Young s modulus and v is the Poisson ratio of tubes, and d i and d o are inside and outside diameters of the tubes, respectively. Computed values of K r for PVC tubes, steel tubes and RC footings are listed in Table 2. In Fig. 3 are shown the different stages of the production of the 2 specimens confined using steel and PVC tubes: (a) each micropile insert was first glued to a steel formwork in the center of the respective steel/pvc confinement tube and (b) then, the micropile inserts and the holes between these and the confinement tubes were filled with grout. In Fig. 4, four stages of the production of the 1 specimens confined with RC blocks are presented: (a) first, steel and wooden formwork were assembled to execute the 45 45 5 mm 3 RC blocks; (b) then, concrete was cast and vibrated; (c) after 28 days of indoor curing, the blocks were drilled using a diamond coring system for insert application and (d) lastly, the prefabricated micropiles inserts were positioned into the holes and these were sealed using grout. 3.3. Test set-up The test set-up (Fig. 5) was planned for both situations: (a) specimens confined using steel/pvc tubes and (b) specimens confined using RC blocks. In the first case, a steel plate with a central hole was placed on bottom of each specimen to allow the slippage of the inserts during push-off tests. In the second case, a compressible polystyrene plate, capable of assuming significant compressive strains with negligible compressive stresses for the expected maximum vertical displacement, was placed on bottom of each hole. Specimens confined with steel/pvc tubes were tested using a compressive testing machine with a capacity of 6 tons (Fig. 6a and b). Specimens confined with RC blocks were tested using a compressive testing machine with a capacity of 5 tons (Fig. 6c). In both situations, the relative displacement between the insert and the grout mass was measured with two displacement transducers (LVDT) placed between the loading plate and the surface of the grout mass. The applied load was monitored using the machine pressure gauge as well as an external load cell placed on top of the loading plate. A data logger was used to record data from the load cell and from the LVDTs. All tests were conducted until failure with displacement control, considering a load rate of.25 mm/s, in order to also obtain the residual connection capacity. 4. Results and discussion In Table 3, the average values of the peak load, the corresponding bond strength and the axial displacement, obtained with the push-off tests, are presented. The average bond strength, f b (MPa), is the peak load, P u (N), divided by the nominal surface area of the embedment length, l b (mm), of the micropile insert. Two different types of failures were observed. For tests performed with PVC confinement, failure occurred by radial fracturing

J. Veludo et al. / Construction and Building Materials 26 (212) 172 179 175 Fig. 3. Production of specimens confined using steel and PVC tubes: (a) micropile inserts and steel and PVC Tubes and (b) grouting of the specimens. with radial displacement of grout wedges. For higher confinement levels (steel/concrete confinement), a mixed failure mechanism was registered, consisting of splitting-induced push-off of the insert. In both cases, three to four resulting splitting wedges were observed which is also in agreement with Malvar [13]. Furthermore, in tests performed in RC footings no visible cracks were observed in the concrete block. In these cases the mechanical properties of the grout, the micropile surface and the diameter of the grout mass are the critical parameters. The load displacement curves obtained in the tests are shown in Fig. 7a c for each confinement situation considered. It can be stated that the obtained load displacement responses have similar shape for all the tests. In fact, each curve consists of a linear branch until 7 8% of the failure load is reached, followed by a non-linear branch from this point onwards. A sudden drop of the load carrying capacity appears just after reaching the peak load followed by a ductile post-failure response. In all tests, significant residual bond stress after a slip of 15 mm is identified, representing approximately 5% of the bond strength (at peak load). This results corroborate those obtained by Gómez et al. [4]. From Fig. 7d, it can be seen that for the same diameter of grout mass, 1 mm, and the same embedment length, 35 mm, the connection capacity increases with the confinement level. It can also be observed that the beginning of non-linearity also depends on this parameter, increasing with the level of confinement. In addition, the initial stiffness increases with the increase of the radial confinement. After the peak load, for different radial confinement, the curves gradually evolve towards a similar rate, revealing the same friction mechanism. The bond strength at the micropile grout interface for smooth micropile inserts depends on two components: chemical adhesion Fig. 4. Preparation of RC footings: (a) formwork, (b) casting concrete footings, (c) drilling holes on footings and (d) grouting of micropiles inserts into the holes. and friction mechanism. The effect of chemical adhesion is transitory since it is destroyed immediately after slippage of the

176 J. Veludo et al. / Construction and Building Materials 26 (212) 172 179 Load Cell LVDT Load Cell LVDT LVDT LVDT (a) (b) Fig. 5. Tests set-up for the specimens confined using: (a) steel/pvc tubes and (b) RC blocks. micropile insert in the beginning of the push-off test. After this stage the insert starts to slip relatively to the grout mass and from this point onwards the bond strength is controlled by friction. Simultaneously, the shear displacement causes normal dilatation and consequently normal stresses are developed. The dilation is dependent on the confinement level and grout quality. The mechanical properties of the grout and of the confinement material, the surface irregularities of the insert, and the diameter of the grout mass, influence the magnitude of these stresses. According to Moosavi et al. [9] grouts with lower quality generate smaller dilatation and, consequently, reduced bond strength at micropile grout interface. In all push-off tests performed, the failure of the connection was observed after a deflection of less than 2. mm of the insert head has occurred (see Table 3). In Fig. 8 the relationship between embedment length of the insert and the displacement at the peak load is shown for specimens with the same diameter of the grout mass, 1 mm. It can be seen that for the same confinement level, the displacement at peak load increases with the increase of the embedment length of the insert. Furthermore, the displacement at the peak load seems to be proportional to the embedment length of the insert. For specimens with the same type of failure mechanism (steel/concrete confinement) the displacement at peak load decreases with the increase of the level of confinement. In the case of PVC confinement, the displacement is lower due to the longitudinal cracking of the grout allowed by the radial deformation of the confinement tube. Based on the results of push-off tests performed with smooth micropiles inserts grouted in predrilled holes in RC footings, Gómez et al. [4] presented the following relationship for the average bond strength, f b in MPa, related to the width of grout-filled annular space around the insert, a in mm: 4:14 :54 a 6 f b 6 4:83 :54 a ð2þ This expression is valid for values of a between 6.35 mm and 5.8 mm, a cement grout with a compressive strength higher than 27.6 MPa at 28 days of age, a Young s modulus higher than 6.9 GPa and it considers the confinement provided by the grout-filled annular space around the insert (parameter a ) for a constant high steel reinforcement of 1% of volume of the concrete block. A thoroughly discussion of Eq. (2), including an approximate procedure to estimate the effect of the reinforcing of the concrete block on the confinement, can be found in [4]. In Fig. 9, the results of the research herein described are shown, in terms of the relationship between the bond strength and the Fig. 6. Push-off tests with smooth micropile inserts: (a) confined with PVC tube, (b) confined with steel tube and (c) grouted in predrilled holes in concrete footings. confinement level. A high correlation coefficient is observed (.96). Furthermore, this relationship seems to be linear and follows a Mohr Coulomb criterion [13,14]. Based on this, the authors propose the following relationship to estimate the bond strength at micropile grout interface, f b in MPa: f b ¼ :11K r þ 1:214 where K r is the radial stiffness in MPa/mm, given by Eq. (1). It must be stated that Eq. (3) is valid for smooth micropiles, grouted in holes ð3þ

J. Veludo et al. / Construction and Building Materials 26 (212) 172 179 177 Table 3 Summary of test results. Specimen Peak load (P u ) (kn) Bond strength (f b ) (MPa) Displacement at peak load (mm) Confinement P1-P2 4.3 1.7.31 PVC P3-P4 46.5.9.346 PVC 7.5 1.7.528 PVC P7-P8 45.8.69.868 PVC P9-P1 91.8 1.39.69 PVC S1-S2 16.7 4.26.952 Steel S3-S4 244.9 4.72 1.92 Steel S5-S6 345.4 5.23 1.414 Steel S7-S8 396.3 6.1 1.774 Steel S9-S1 245.5 3.65 1.338 Steel C1 a -C2 281.3 7.46.759 Concrete C3 a -C4 42.2 (391.2 b ) 7.76 1.886 (1.11 b ) Concrete C5-C6 485.5 7.36 1.287 Concrete C7-C8 549.8 8.33 1.341 Concrete C9-C1 41.9 6.23 1.184 Concrete a b Discarded values Instrumented inserts. Very low peak load. Load and displacement at the onset of failure (see Fig. 7c). drilled in existing RC footings, using a cement grout with a compressive strength higher than 53 MPa at 28 days of age, a Young s modulus higher than 14 GPa and it assumes that the concrete block has a very low steel reinforcement ratio (most common situation in rehabilitation works). To estimate the bond strength at micropile grout interface using Eq. (3), the radial stiffness needs to be computed first using Eq. (1). For this purpose, both E and v need to be measured. This can be done either through cores extracted from the existing footing or non-destructive testing (NDT). In case the existing foundation is found to be cracked, external prestress must be provided. It should also be mention that, to use Eq. (3) for design purposes, a suitable safety factor must be adopted. For identical hole s geometry and materials mechanical properties, Eqs. (2) and (3) give similar results. This can imply that, for smooth micropiles and uncracked footings (situation observed both in tests herein discussed and in tests performed by Gómez et al. [4]), the contribution of the footing s reinforcement for the confinement level is not relevant. Connection Capacity (kn) Connection Capacity (kn) 1 9 8 7 6 5 4 3 2 1 Smooth Micropiles Grouted into PVC pipes Tests P1-P2: Le=2mm; Dg=1mm Tests : Le=35mm; Dg=1mm Tests P9-P1: Le=35mm; Dg=12mm Displacement (mm) P9-P1 P7-P8 P1-P2 P3-P4 5 1 15 2 45 4 35 3 25 2 15 1 5 Tests P3-P4: Le=275mm; Dg=1mm Tests P7-P8: Le=35mm; Dg=8mm Smooth Micropiles Grouted into Steel pipes Tests S1-S2; Le=2mm; Dg=1mm Tests S5-S6: Le=35mm; Dg=1mm Tests S9-S1: Le=35mm; Dg=12mm (a) Displacement (mm) S7-S8 S5-S6 S3-S4 S9-S1 S1-S2 5 1 15 2 (b) Tests S3-S4: Le=275mm; Dg=1mm Tests S7-S8: Le=35mm; Dg=8mm Connection Capacity (kn) Connection Capacity (kn) 6 5 4 3 2 1 Smooth Micropiles Grouted into RC Footings C4 5 1 15 2 Tests C1-C2: Le=2mm; Dg=12mm Tests C5-C6: Le=35mm; Dg=12mm Tests C9-C1: Le=35mm; Dg=122mm 6 5 4 3 2 1 Displacement (mm) (c) (d) C7-C8 C5-C6 C9-C1 C2 Tests C3-C4: Le=275mm; Dg=12mm Tests C7-C8: Le=35mm; Dg=82mm Embedement Length: Le=35mm Diameter of Grout Mass: Dg=1mm C5-C6 S5-S6 5 1 15 2 Displacement (mm) PVC Confinement Steel Confinement RC Footings Fig. 7. Results of tests performed with different radial confinement: (a) PVC confinement, (b) steel confinement, (c) RC footings and (d) comparison of results.

178 J. Veludo et al. / Construction and Building Materials 26 (212) 172 179 Displacement at Peak Load (mm) 2. 1.6 1.2.8.4 In Fig. 1, the relationship between the diameter of the grout mass and the bond strength at the micropile grout interface is illustrated. It can be observed that, for specimens confined with PVC tubes, the bond strength increases with the diameter of the grout mass around the insert. In this situation, the grout cover provides itself confinement through tensile hoop stresses prior to cracking. On the contrary, for specimens confined with steel tubes and for specimens grouted in RC footing, it can be observed that the bond strength decreases with the increase of the diameter of the grout mass around the insert. In this situation, the bond strength depends on the mechanical properties of the grout and on the confinement materials. The bond strength at the micropile grout interface versus the embedment length is presented in Fig. 11. The bond strength seems to vary slightly with the embedment length for specimens with steel confinement, whereas no variation is detected for specimens confined with PVC tubes or RC footings. These results are consistent with those obtained by Gómez et al. [4] with smooth casing inserts grouted in predrilled holes in RC footings. 5. Conclusions S1-S2.95 C1-C2.76 P1-P2.3 Diameter of the Grout Mass: Dg=1mm S3-S4 1.1 C3-C4 1.1 P3-P4.35 S5-S6 1.41 C5-C6 1.29.53. 15 2 25 3 35 4 Embedment Length (mm) PVC Confinement Steel Confinement RC Footings Fig. 8. Relationship between embedment length and the displacement at peak load for the same diameter of grout mass 1 mm. Bond Strength (MPa) 1 9 8 7 6 5 4 3 2 Embedment Length: Le=35mm 1 f b =.11K r + 1.214; R² =.9572 2 4 6 8 Radial Stiffness (MPa/mm) Fig. 9. Relationship between bond strength and radial stiffness. From the experimental study herein presented, the following conclusions can be drawn. In all tests, immediately after reaching Bond Strength (MPa) 1 8 6 4 2 C7-C8 8.33 S7-S8 6.1 P7-P8.69 Embedment Length: Le=35mm C5-C6 7.36 5.23 1.7 C9-C1 6.23 P9-P1 3.65 P9-P1 1.39 7 8 9 1 11 12 13 Diameter of the Grout Mass (mm) PVC Confinement Steel Confinement RC Footings Fig. 1. Relationship between bond strength and diameter of the grout mass. Bond Strength (MPa) 1 8 6 4 2 C1-C2 7.46 S1-S2 4.26 P1-P2 1.7 Diameter of the Grout Mass: Dg=1mm C3-C4 7.76 C5-C6 7.36 S3-S4 4.72 P3-P4.9 S5-S6 5.23 1.7 15 2 25 3 35 4 Embedment Length (mm) PVC Confinement Steel Confinement RC Footings Fig. 11. Relationship between bond strength and embedment length. the peak load, a sudden decrease on the load carrying capacity of the specimens is observed, followed by a ductile post-failure response. In this phase, the connection capacity gradually decreases until an almost constant value is reached. This residual bond strength represents approximately 5% of the maximum bond strength, which is still significant. The connection capacity is first controlled by chemical adhesion at the micropile grout interface and then by friction, being the latter proportional to the radial confining. However, after the peak load, the load displacement diagrams tend to similar rates, indicating that the post-peak phase is controlled by the same friction mechanism, independently of the level of confinement. The stiffness of the confinement materials has a major influence on the bond strength when the Young s modulus of the confinement material is smaller than that of the grout. In this case, failure occurs by radial fracturing with radial displacement of grout wedges. On the contrary, for higher confinement levels, the mechanical properties of the grout and the diameter of the grout mass are the critical parameters. The variation of the bond strength at micropile grout interface is shown to be strongly related with the confinement level (correlation coefficient of.96), therefore being this one of the most important parameters. In the case of micropiles grouted in holes predrilled in existing RC footings and for steel confinement, the bond strength increases

J. Veludo et al. / Construction and Building Materials 26 (212) 172 179 179 with the decrease of the hole s diameter. Consequently, increasing the hole s diameter is not an option to improve the connection capacity. Finally, for RC footings, the bond strength does not vary significantly with the embedment length. Acknowledgement The authors acknowledge the financial support of the companies Betão-Liz; DSI; HILTI; SECIL; and SIKA. References [1] Armour T, Groneck P, Keeley J, Sharma S. FHWA-SA-97-7 - Micropile Design and Construction Guidelines Implementation manual: Federal Highway Administration US Department of Transportation; 2. [2] Cyna H, Schlosser F, Frank R, Plumelle C, Estephan R, Altamayer F, et al. FOREVER: Synthesis and recommendations of the French national project on micropiles (1999 23): IREX; 24. [in French]. [3] Rasines JME. Micropiles to footing connection. In: SEMSIG-AETSS, editor. Jornadas técnicas SEMSIG-AETESS 3ª Sesión: Micropilotes Naos Livros; 23. p. 131 41. [in Spanish]. [4] Gómez J, Cadden A, Traylor RP, Bruce DA. Connection capacity between micropiles and existing footings-bond strength to concrete. In: Cadden DABaAW, editor. Geo3 GEO Construction Quality Assurance/Quality Control Conference Proceedings. Dallas/Ft. Worth, TXS25. p. 196 216. [5] Tepfers R. Cracking of concrete cover along anchored deformed reinforcing bars. Mag Concr Res 1979;31(16):3 12. [6] FIB. Bond of reinforcement in concrete. Lausanne, Switzerland: Fédération Internationale du Béton; 2. [7] ACI Committe 48. Bond development of straight reinforcing bars in tension (48R-3). Farmington Hills, Mich.: American Concrete Institute; 23. p. 49. [8] CEB-FIP Model Code 199. CEB/FIP. London: Comité International du Béton (CEB); 1991. [9] Moosavi M, Jafari A, Khosravi A. Bond of cement grouted reinforcing bars under constant radial pressure. Cement Concr Compos 25;27(1):13 9. [1] Hyett AJ, Bawden WF, Macsporran GR, Moosavi M. A constitutive law for bond failure of fully-grouted cable bolts using a modified Hoek cell. Int J Rock Mech Min Sci Geomech Abstracts 1995;32(1):11 36. [11] Hyett AJ, Bawden WF, Reichert RD. The effect of rock mass confinement on the bond strength of fully grouted cable bolts. Int J Rock Mech Min Sci Geomech 1992;29(5):53 24. [12] Yahia A, Khayat K, Benmokrane B. Bond strength of cement grout anchors cast in dry and submerged conditions. Ground anchorages and anchored structures. London, UK: Thomas Telford; 1997. p. 89 99. [13] Malvar LJ. Bond reinforcement under controlled confinement. ACI Mater J 1992;89(6):593 61. [14] Noghabai K. Effect of tension softening on the performance of concrete structures. Experimental, analytical and computational studies. Luleå University of Technology; 1998. [15] ACI Committee 318. Building code requirements for structural concrete and commentary (ACI 318M-5). Farmington Hills, Mich.: American Concrete Institute; 25. p. 438. [16] CEN. Eurocode 2: Design of concrete structures Part 1-1: General rules and rules for buildings (EN 1992-1-1). Brussels: European Committee for Standardization; 24. p. 225. [17] Santos PMD, Júlio ENBS. Development of a laser roughness analyser to predict in situ the bond strength of concrete-to-concrete interfaces. Mag Concr Res 28;6(5):329 37. [18] CEN. EN 196-1 Methods of testing cement Part 1: Determination of strength. European Committee for Standardization; 25. [19] EN 445: Grout for prestressing tendons. Test methods. European Committee for Standarditazion; 2. p. 19. [2] LNEC. Especificação E 397 Static Modulus of Elasticity of Concrete Under Compression. Laboratório Nacional de Engenharia Civil; 1993. p. 2. (in Portuguese). [21] EN 447: Grout for prestressing tendons. Basic requirements. European Committee for Standarditazion; 2. p. 14. [22] CEN. EN 1239-1 Testing hardened concrete Part 1: Shape, dimensions and other requirements for specimens and moulds. European Committee for Standardization; 2. [23] CEN. EN 1239-3 Testing hardened concrete Part 3: Compressive strength of test specimens. European Committee for Standardization (CEN); 23. [24] Timoshenko SP, Goodier JN. Theory of Elasticity. 3rd ed., Rio de Janeiro: Editora Guanabara Dois; 198. (in Portuguese).