Methods of repairing concrete structures
J Sudhakumar*, Regional Engineering College, India 26th Conference on OUR WORLD IN CONCRETE & STRUCTURES: 27 - 28 August 2001, Singapore

Article Online Id: 100026073 The online version of this article can be found at:


This article is brought to you with the support of  Singapore Concrete Institute    All Rights reserved for CI‐Premier PTE LTD  You are not Allowed to re‐distribute or re‐sale the article in any format without written approval of  CI‐Premier PTE LTD  Visit Our Website for more information  

26 th Conference on Our World in Concrete & Structures: 27 - 28 August 2001, Singapore

Methods of repairing concrete structures
J Sudhakumar*, Regional Engineering College, India


There are three basiG symptoms of distress in a concrete structure: cracking, spalling and disintegration. Each of these basic symptoms in itself is fairly obvious and may be readily detected and differentiated from the others. Cracking is one of the most misunderstood problems of concrete. It is generally regarded as indicative of defective design or materials. Methods of repairing cracks include bonding with epoxies, routing and sealing, stitching, external stressing, grouting, blanketing, use of overlays, etc. Repairing cracks does not usually involve strengthening. In a structure showing spalling and disintegration, it is usual to find that there have been substantial losses of section and/or pronounced corrosion of reinforcement. Hence, the repair involves some requirement for restoration of lost strength. The principal methods used for repair of spalling and disintegration are jacketing, guniting, prepacked concrete, drypack, replacement of the concrete, and the application of overlays of several types. This paper gives a detailed description on all the above methods used for repairing cracks, spalling and disintegration in concrete structures.
Keywords: concrete, cracking, spalling, diSintegration, repair, stitching, grouting, jacketing Introduction Concrete structures will show the degree of deterioration in the form of cracking, spalling and diSintegration. Each one of these is clearly distinguishable. The reasons for their development may be poor materials, poor design poor construction practice, poor supervision or a combination. Crack formation in concrete is most interesting because sometimes the same causes produce a different cracking pattern, and sometimes the same cracking pattern is produced by different causes. Sometimes concrete cracks in a location where no cause can be found out, and in other places it does not crack where there is every reason for cracks to occur. However, fifty percent of the cases are straight forward. Cracks in themselves are seldom indicative of structural danger; accordingly, repair usually does not involve strengthening. So their repairs are basically intended to seal the cracks against an objectionable flow of water or to improve the appearance of the construction. In the repair of a structure showing spalling and diSintegration, disintegration, it is usual to find that there have been substantial losses of section and/or pronounced corrosion of the reinforcement. Both are matters of concern from a structural viewpoint, and repair generally involves some urgency and some requirement for restoration of lost strength [1]. 1.


Repairing cracks In order to determine whether the cracks are active or dormant, periodic observations are done utilizing various types of telltales [Fig.1]. Crack movement can be detected by placing a mark at the end of the crack. Subsequent extension of the crack beyond the mark indicates probable continuance of the activity that produced the defect originally. The deficiency of this technique is that it will not show any tendency for the crack to close or provide any quantitative data on the movement [2]. In another method, a pin or a toothpick is lightly wedged into the crack and it falls out if there is any extension of the defect. The deficiencies of this method, as before, are that there is no indication of closing movement or any quantitative measure of the changes, which occur [2]. A strip of notched tape works similarly. Movement is indicated by tearing of the tape. An advantage in this case is that some indication of closure can be realized by observing any wrinkling of the tape. However, this device is not reliable. The tape is not dimensionally stable under changing conditions of humidity, so that one can never be sure whether the movements are real or are due to shrinkage or swelling of the marker [2]. The device using a typical vernier caliper is the most satisfactory of all. Both extension and compression are indicated and movements of about one-hundredth of an inch can be measured using a vernier caliper of special type. If more accurate readings are desired, extensometers can be used. The reference points must be rigidly constructed and carefully glued to the surface of concrete, using a carborandum stone to prepare the bonding surface before attaching the reference marks. Where extreme accuracy is required resistance strain gauges can be glued across the crack. They are, however, expensive, sensitive to changes in humidity, and easily damaged.


" ..g. 1 Tell lule$

2.1 Types of cracks Cracks can be divided into active cracks and dormant cracks. It should be considered that, in a sense, every crack is an active one. For example, if there is any change in the load supported by a member, whether an increase or a decrease, the member must deform to accommodate the change. In this case, deformation is bound to occur and be concentrated at the pOints of weakness, ie. at the cracks. Thus, the proper differentiation between active and dormant cracks is one of magnitude of movement, and the telltales are a measure of the difference. If the magnitude of the movement, measured over a reasonable period of time (say 6 months or 1 year), is sufficient to displace or show significantly on the telltales, we can treat the crack as an active one. If the movements are smaller, the crack may be considered as dormant. Cracks can also be divided into solitary or isolated cracks and pattern cracks. Generally, a solitary crack is due to a positive overstressing of the concrete either due to load or shrinkage. The cause for this will become apparent when the line of the crack is compared with the layout of the portion of the concrete, its reinforcement and the known stresses in it. Overload cracks are fairly easily identified because they follow the lines demonstrated in laboratory load tests. A crack due to setting and hardening shrinkage is formed in the first week of the life of the concrete. If the length of concrete under inspection is more than about 9 m, it is not likely that there is a solitary crack; usually there will be another of a similar type. The analYSis of this second crack confirms the findings from the first. In a long retaining wall or long channel, the regular formation of cracks indicates faults in the design rather than the construction, but an irregular distribution of SOlitary cracks may indicate poor construction as well as poor design [1,2].


Regular patterns of cracks may occur in the surfacing of concrete and in thin slabs. These are called pattern cracks. The term pattern cracking is used to indicate that all the cracks visible have occurred more or less at the same time.

2.2 Methods of repairing cracks Methods of repairing cracks include bonding with epoxies, routing and sealing, stitching, external stressing, grouting, blanketing, use of overlays, etc. Various factors are to be considered in deciding the type of repair to be carried out. The details of these methods are described in the next sections. 2.3 Bonding with epoxies Cracks in concrete may be bonded by the injection of epoxy bonding compounds under pressure. Usual practice is to drill into the crack from the face of the concrete at several locations; inject water or a solvent to flush out the defect; allow the surface to dry (using a hot air jet, if needed); surface-seal the cracks between the injection points; and inject the epoxy until it flows out of the adjacent sections of the crack or begins to bulge out the surface seals, just as in pressure grouting. Usually the epoxy is injected through holes of about % inch in diameter and % inch deep at 6 to 12 inches centers. Smaller spacing is used for finer cracks. The limitation of this method is that unless the crack is dormant or the cause of cracking is removed and thereby the crack is made dormant, it will probably recur, possibly somewhere else in the structure. Also, this technique is not applicable if the defects are actively leaking to the extent that they cannot be dried out, or where the cracks are numerous [2]. 2.4 Routing and sealing This method involves enlarging the crack along its exposed face and filling and sealing it with a suitable material [Fig.2]. This is the most simple and common technique for sealing cracks and is applicable for sealing both fine pattern cracks and larger isolated cracks. The cracks should be dormant. This technique is not applicable for sealing cracks subject to a pronounced hydrostatic pressure. The routing operation consists of following along the crack with a concrete saw or with hand or pneumatic tools, opening the crack sufficiently to receive the sealant. A minimum surface width of % inch is desirable. Smaller openings are difficult to work with. The surfaces of the routed joint should be rinsed clean and permitted to dry before placing the sealant. The method used for placing the sealant depends on the material to be used and follows standard techniques. Routing and sealing of leaking cracks should preferably done on the pressure face so that the water or other aggressive agents cannot penetrate the interior of the concrete and cause side effects such as swelling, chemical attack, or corrosion of the reinforcement. The sealant may be any of the several materials, depending on how tight or permanent seal is desired. This is a method where thorough water tightness of the jOint is not required and where appearance is not important [1,2].

-····1/-/ ,





.. ~r~--· - .


2.5 Stitching The tensile strength of a cracked concrete section can be restored by stitching in a manner analogous to sewing cloth. Concrete can be stitched by iron or steel dogs in the same way as timber. A series of stitches of different lengths should be used so that they do not all exert their load in one line at the side of the crack. The best method of stitching is to bend bars into the shape of a broad flat bottomed letter U between 1 foot and 3 feet long and with ends about 6 inches long, and to insert them in holes drilled to match in the concrete on either side of the crack. The bars are then grouted up, some grout being placed in the holes in advance of the bars. Fig.1 shows the details of stitching [1,2,3].
Usually cracks start at one end and run away from the starting place quicker on one side of the concrete than on the other. The stitching should be on the side, which is opening up first. The following points should be observed in general. Any desired degree of strengthening can be accomplished, but it


must be considered that the strengthening also tends to stiffen the structure locally. This may give more force to the restraints causing the cracking and reactivate the condition. Stitching the crack will tend to cause the problem to migrate elsewhere in the structure. If it is decided to stitch, investigate and, if necessary, strengthen adjacent areas of the construction to take the additional stress. In particular, the stitching dogs should be of variable length and/or orientation and so located that the tension transmitted across the crack does not devolve on a single plane of the section, but is spread out over an area. Where there is a water problem, the crack should be sealed as well as stitched so that the stitches are not corroded, and because the stitching itself will not seal the crack. Sealing should be completed before stitching is commenced, to avoid the corrosion and also because the presence of dogs tends to make it difficult to apply the sealant. Where possible, stitch both sides of the concrete section so that further movements of the structure will not exert any bending action on the dogs. In bending members, it is possible to stitch one side of the crack only, but this should be the tension side of the section, where movement is originating. If the member is in the state of axial tension, then a symmetrical placement of the dogs is a must, even if excavation or demolition is required to gain access to opposing sides of the section. In order to resist shear along the crack, it is necessary to use diagonal stitching. One set of dogs can be placed on each side of the concrete, if necessary. The lengths of dogs are random so that the anchor points do not form a plane of weakness. The dogs must be grouted with a non-shrink or expanding mortar, so that they have a tight fit, and movement of the crack will cause simultaneous stressing of both old and new sections. If this is not possible, proportion the stitching to take the entire load without participation by the existing materials [Fig.3]. The dogs are relatively thin and long and so cannot take much in the way of compressive force. Accordingly, if there is a tendency for the crack to close as well as to open, the dogs must be stiffened and strengthened by encasement in an overlay or by some similar means.

2.6 External stressing

Development of cracking in concrete is due to tensile stress and can be arrested by removing these stresses. Further, the cracks can be closed by inducing a compressive force, sufficient to overcome the tension and to provide a residual compression. This compressive force is applied by use of the usual prestressing wires or rods. The principle is very similar to stitching, except that the stitches are tensioned; rather than plain bar dogs which apply no closing force to the crack and which may in fact have to permit the crack to open up a bit before they begin to take the load. All the points noted regarding stitching must be considered [2]. An additional problem is that of providing an anchorage for the prestressing wires or rods. Some form of abutment is needed for this purpose. The effect of the tensioning force on the stress conditions in the structure should be analyzed [FigA].


2.7 Grouting Grouting of cracks can be performed in the same manner as the injection of an epoxy, and this technique has the same areas of applications and limitations. However, the use of an epoxy is the better solution except where the considerations of fire resistance or cold weather prevent such use, in which case grouting is the comparable alternative. The procedure is similar to other grouting methods and consists of cleaning the concrete along the crack; installing built-up seats at intervals along the crack, sealing the crack between the seats with a cement paint or grout, flushing the crack to clean it and test the seal; and then grouting the whole. The grout itself is high early strength Portland cement. An alternative and better method, where it can be performed, is to drill down the length of the crack and grout it so as to form a key. The grout key functions to prevent relative transverse movements of the sections of concrete adjacent to the crack. However, this technique is applicable only where the cracks run in a reasonably straight line and are accessible at one end. The drilled hole should preferably be 2 or 3 inches in diameter and flushed to clean out the crack and permit better penetration of the grout [1,2,3,4]. 2.8 Blanketing Blanketing is similar to routing and sealing, but is used on a larger scale and is applicable for sealing active as well as dormant cracks. Preparing the chase is the first step in blanketing. The chase usually is cut square [Fig.5]. The bottom should be chipped as smooth, as level, and as clean as possible to facilitate breaking the bond between sealant and concrete. The sides of the chase should be prepared to provide a good bond with the sealant material. To this end, they should be dry. All loose, disintegrated, or otherwise unsatisfactory concrete adjacent to the crack should be removed, even if the resulting shape is irregular and the volume to be removed is large. If the concrete adjacent to the chase is porous, a bond coat of waterproof mortar, laid on the sides and bottom of the chase is required. The bond coat must be properly dried and cured before proceeding to place the sealant. The first consideration in the selection of sealant materials is the amount of movement anticipated and the extremes of temperature at which such movements will occur. It must be capable of deforming the required amount under the applicable conditions of temperature. The material should be able to take traffic, be resistant to chemical spillage, and be capable of being pigmented, if desired. The first type of blanket joints use sealant materials, which are known as elastic sealants. They return to their original shape, when not under an externally induced stress, ie. acts elastically. A bond breaker should be used at the bottom of the chase, so that the sealant is free to deform. Otherwise, the stress in the top fibres would be almost doubled, when compared with what it would be with the bond breaker, and the bottom fibres will tend to tear. Good bond breakers include polyethylene, waxed paper and foil [Fig.6]. The second type of blanket joints use sealant materials that are known as mastic sealants. Their details are similar to that of an elastic sealant, except that the bond breaker is omitted and the sealant is bonded to the bottom as well as to the sides of the chase. The sealant is a mastic rather than a compound having elastic properties. This type of joint is for use where the anticipated movements are small and where trafficability or appearance are not considerations. The advantage is that the mastic is less costly than the elastic type of sealant material. The third type of blanket jOints are called mortar-plugged jOints. They are used in cases where the joint must resist a pressure head acting on the face of the plug. The mortar plug provides the strength of the joint, and resists the pressure on the jOint by arching the load to the sides of the chase. Theoretically, the edges of the chase should be undercut so that the mortar plug, which dries more quicker and more thoroughly at the surface, does not shrink differentially and so pull away from the sides and bottom of the chase. Whatever be the type of detail used, when cutting the chase, it is probable that the reinforcement will be exposed. If this is the case, cut deep enough so that the sealant will be behind the reinforcing, clean the bars, and paint them with bitumen as a protection against moisture penetration and corrosion. If the crack is an active one with substantial movements, cut the bars so that they do not impede the movement of the joint. In any event, the reinforcement should not be left in a position where it is partly embedded in two different matrices. For example, it should not be partly in the mortar plug and partly in the sealant, or partly in the joint and partly in the original concrete.


2.9 Use of overlays Overlays may be used to seal the cracks and they are very useful and desirable where there are large numbers of cracks and treatment of each individual defect would be too expensive. Sealing of an active crack by use of an overlay requires that the overlay be extensible and not flexible alone. The occurrence of prolongation of a crack automatically means that there has been an elongation of the surface fibers of the concrete. Accordingly, an overlay which is flexible but not extensible, ie. can be bent but cannot be stretched, will not seal a crack that is active. A two- or three- ply membrane of roofing felt laid in a mop coat of tar, with tar between the plies; the whole covered with a protective course of gravel, concrete or brick, functions very well for this purpose. The type of protective course depends on the use to which it will be subjected. Gravel is typically used for roofs, concrete or brick are used where fill is to be placed against the overlay. An asphalt block pavement also works well where the area is subjected to heavy traffic. 3. Repairing spalling and disintegration Cracks in themselves are seldom indicative of structural danger; accordingly, repair usually does not involve strengthening. So their repairs are basically intended to seal the cracks against an objectionable flow of water or to improve the appearance of the construction. In the repair of a structure showing spalling and disintegration, it is usual to find that there have been substantial losses of section and/or pronounced corrosion of the reinforcement. Both are matters of concern from a structural viewpoint, and repair generally involves some urgency and some requirement for restoration of lost strength [5]. The principal methods used for repair of spalling and disintegration are jacketing, guniting, prepacked concrete, drypack, replacement of concrete, use of various types of overlays etc. 3.1 Jacketing The use of jacketing is primarily applicable to the repair of deteriorated columns, piers and piles. It is especially useful where all or a portion of the section to be repaired is under water. Jacketing consists of restoring or increasing the section of an existing member, principally a compression member, by encasement in new concrete. This method is applicable for protecting a section against further deterioration as well as strengthening. The form for the jacket should be provided with spacers to assure clearance between it and the existing concrete surface. The form may be temporary or permanent and may consist of timber, wrought iron, precast concrete or gauge metal, depending on the purpose and exposure. For marine environments or elsewhere where it is desired to protect concrete from chemical reaction with its environment or from weathering, the use of permanent timber forms is recommended; provided the appearance of the form is not objectionable and does not constitute a fire hazard. Wrought iron makes a very satisfactory, permanent form, but is expensive and so is limited to installations where the additional cost is justified by the increased life of the repair or where it is desired to protect the concrete from severe abrasion, such as due to heavy masses of moving ice. Gauge metal and other temporary forms can also be used under certain conditions.


Filling up the forms can be done by pumping the grout, by using pre packed concrete, by using a tremie, or, for subaqueous works, by dewatering the form and placing the concrete in the dry. Filling the form by pumping the grout, offers an advantage compared to others, because reliable results can be obtained with less dependence upon the skill of workmen. The use of a grout having a cementsand ratio by volume, between 1:2 and 1:3 , is recommended. The richer grout is preferred for thinner sections and the leaner mixture for heavier sections. The grout should be placed as soon as possible after the rinsing operation and under an air pressure, sufficient to assure a smooth and continuous flow. The forms should be filled to overflowing, the grout allowed to settle for about 20 minutes, and the forms refilled to overflowing. The outside of the forms should be vibrated during placing of the grout. The top of the jacket should be finished with a collar of concrete in such a manner that a smooth transition between repaired and existing work will result.

3.2 Guniting
Gunite is also known as shotcrete or pneumatically applied mortar. Gunite is used for the restoration of concrete surfaces where the deterioration is relatively shallow. It can be used on vertical and overhead, as well as on horizontal surfaces and is particularly useful for restoring surfaces spa lied due to corrosion of reinforcement. Gunite is a mixture of Portland cement, sand and water, shot into the place by compressed air. In structural applications, the sand and cement are mixed dry in a mixing chamber, and the dry mixture is then transferred by air pressure along a pipe or hose to a nozzle, where it is forcibly projected on to the surface to be coated. Water is added to the mixture by passing it through a spray injected at the nozzle. The flow of water at the nozzle can be controlled to give a mix of desired stiffness, which will adhere to the surface against which it is projected. The existing surface must be made rough to afford a good keying effect. Anchor bolts tying the new work to the old concrete are essential. A layer of galvanized welded wire mesh can also be provided. The mesh is connected to the hook bolts and the existing reinforcing bars, with galvanized tie wire. Incompatibility of gunite with the old concrete is a major problem, and the keying effects of the rough surface and the doweling effects of the anchor bolts are necessary to assure interaction between the two materials. Sand for gunite should be uniformly graded, as for conventional concrete. Hard particles are desirable, since there is a tendency to grind and crumble the grains as they pass through the discharge hose. The sand should contain 3 to 5 percent moisture for efficient operation of the equipment. Because of impact, a certain amount of the material being projected against the surface to be coated will bounce off. This material is known as rebound. It consists primarily of the coarse sand particles and has a much smaller cement content than the mix, as projected from the nozzle. The occurrence of some rebound is unavoidable, and it amounts to 20 to 30 pecent. Because of the rebound, the cement content of the mortar, in place, will be substantially greater than that of the materials as fed to the mixer. Thus, if a 1:3 mix is desired in place, a 1:4 mix fed into the mixer might be adequately rich [1,2,3]. It is difficult to mention mix proportions for gunite because of uncertainties and variations in the amount of rebound, because of difficulties in making test cylinders, and because there is little control over the water-cement ratio of the material in place. In practice, it is usual to use as much water as possible without causing fallouts, as this minimizes the rebound. Roughly, this might mean a watercement ratio of 0.5 to 0.6. Starting with this assumption and assuming a 20 to 30 percent rebound, the proportions required for the mix, as fed to the mixer, can be approximated for any desired strength. In practice, a 1:3 or 1:3.5 mix, by volume, is fed into the mixer. The sand and cement shall be thoroughly mixed in dry state. The time of mixing shall not be less than 1.5 minutes. The nozzle shall be held between 2 ft and 4 ft from the surface to be coated, and held in such a position that the flow of material will strike it as near to a right angle as possible. In shooting vertical or sloped surfaces, the placing shall be started at the bottom and carried up. In such cases, the mortar shall be placed in layers of such thickness that the weight of the plastic mass does not cause it to sag. When more than one layer is to be used to complete the final thickness of the work, the delay between application of successive layers shall be ample to prevent sagging or fallout of the mass, but not so long that the underlying layer has completely set and developed a glaze coating. It is observed that 30 minutes to 1 hour is usually a proper interval. Before any gunite is applied, care shall be taken to remove any sand or rebound clinging to the surfaces. No gunite shall be applied to a surface on which there is running or free water. When shooting around reinforcing bars or anchor bolts, the nozzle shall be moved from side to side and angled to place the gunite back of the rod. At the end of the day's work, or at similar stopping periods, the gunite shall be tapered to a thin edge. Before shooting the adjacent section, this tapered portion shall be thoroughly cleaned and wetted. Operations shall be suspended when wind velocity is such


that it blows away the spray from the nozzle and prevents proper control of the consistency. As soon as dry patches begin to appear on the surface of the newly placed mortar, curing by use of a water spray or application of two coats of an approved sealing compound shall be commenced. Minimum curing period shall be 7 days.

3.3 Prepacked concrete
This method is particularly useful for carrying out the repair under water and elsewhere where accessibility is a problem. Prepacked concrete is made by filling forms with coarse aggregate and then filling the voids of the aggregate by pumping in a sand-cement grout. As the grout is pumped into the forms, it will fill the voids, displacing any water in them, and form a concrete mass. Prepacked concrete is used for refacing of structures, jacketing, filling of cavities in and under structures, and underpinning and enlarging piers, abutments, retaining walls and footings. The coarse aggregate used should be carefully selected and must be clean. The void content of the aggregate mass should be minimized to reduce the required volume of the more costly grout. In normal practice, the voids ratio ranges from 35 to 45 percent. The intruded mortar contains fine sand, Portland cement, a pozzolanic material of low mixing-water requirement, an agent designed to increase fluidity and to inhibit early stiffening of the grout, and mixing water. The forms must constitute a closed, watertight system, vented at the top only. If not watertight, the travel of grout cannot be controlled. If not effectively vented, a back-pressure will be created, with the result that the concrete fill may contain voids. Pumping of mortar should commence at the lowest point and proceed upward. The grout pipes should not be more than 5 ft on centers, and the grout level in the mass should be brought up uniformly, as determined by observations of grout levels in the grout pipes. Placing of grout should be a smooth, uninterrupted operation.

3.4 Drypack
Drypacking is the hand placement of a very dry mortar and the subsequent tamping of the mortar into place, producing an intimate contact between the new and existing works. Because of the low water-cement ratio of the material, there is little shrinkage, and the patch remains tight. So it will be of good quality with respect to durability, strength and water tightness. Drypacking is used for filling small, relatively deep holes, such as those resulting from the removal of form ties, and narrow slots cut for repair of cracks. The usual mortar mix is 1:2.5 to 1:3.

3.5 Replacement of concrete
This method consists of replacing the defective concrete with new concrete of conventional proportions, placed in a conventional manner. This method is a satisfactory and economical solution where the repair occurs in depth (at least beyond the reinforcement), and where the area to be repaired is accessible. This method is particularly indicated where a water-tight construction is required and where the deterioration extends completely through the original concrete section.

3.6 Overlays
In addition to seal cracks, an overlay may also be used to restore a spalled or diSintegrated surface. Overlays used include mortar, bituminous compounds, and epoxies. They should be bonded to the existing concrete surface.



When repairing cracks, do not fill the crack with new concrete or mortar. A brittle overlay should not be used to seal an active crack. The restraints causing the cracks should be relieved, or otherwise the repair must be capable of accommodating future movements. Cracks should not be surface-sealed over corroded reinforcement, without encaSing the bars. The methods adopted for repairing spalling and disintegration must be capable of restoring the lost strength.

References :
[1] [2] Champion, S. Failure and Repair of Concrete Structures. John Wiley & Sons Inc. New York, 1961. Sidney M. Johnson. Deterioration, Maintenance and Repair of Structures. Mc Graw-Hill Book Company. New York, 1965. [3] Lee How Son and George C.S. Yuen. Building Maintenance Technology. Macmillan Distribution Ltd. England. 1993. [4] Thomas H. McKaig. Building Failures. Mc Graw-Hill Book Company. New York, 1962. [5] Jagadish, R. Structural Failures - Case Histories. Oxford & IBH Publishing Co. Pvt. Ltd. New Delhi.1995.