Лако-красочные материалы — производство

The advantages of externally bonded reinforcement over other methods of strengthening concrete structures include the ability to strengthen part of a structure whilst it is still in use, minimum effect on headroom, low cost and ease of maintenance (see Fig. 6.8, for example). The method has been in use for over 20 years, mainly to enhance flexural capacity, and has been found to produce effective and economical solutions to particular problems.

Historical development. The technique was introduced in France in the late 1960s where the first reported application was a major

bridge strengthening scheme on the Autoroute du Sud. It was also used in several other countries during the early 1970s including Switzerland, South Africa and Japan. For example, up to 1975 well over 200 bridges in Japan had been strengthened to accommodate the large increases in loading that had occurred due to increases in heavy goods traffic since the bridges were designed.

In 1975 the first application to a bridge in the UK occurred when 2 pairs of motorway bridges on the M5 at Quinton were strengthened after cracks which were described as ‘little more than shrinkage cracks’ were observed during a routine bridge inspection. A detailed appraisal of the design showed that the cracking occurred as a result of stopping off too many reinforcing bars in a zone centred at the quarter points of the main span. Loading tests, carried out before and after strengthening, demonstrated the effectiveness of the technique in providing increased flexural stiffness and in reducing crack opening under load(15).

Two bridges on the M20 in Kent were strengthened shortly after construction in 1977. Cracks were discovered in one of the side spans which were found to have insufficient longitudinal steel. Steel plates were subsequently bonded to the side span soffits and the top surface of the deck over the supports (Fig. 6.9).

The floor beams of a North London building were strengthened in 1978 to allow for increased floor loadings following a change in use. The factor of safety provided by the existing structure, in the

event of fire rendering the adhesive bond ineffective, was calculated to be 1.1 and this was deemed to be satisfactory.

Meanwhile the technique was gaining ground overseas. The capacities of solid slab floors in two telephone exchanges in Zurich were raised from 2 to 7 kN/m2 to accommodate new switching equipment. Here, constraints which led to adoption of the technique in preference to other strengthening methods were that there had to be no substantial reduction in room height and the need for minimum delay in reconstruction. Other examples include the strengthening of the roof structure at the Central Railway Station in Warsaw and as part of a structural repair process following blast damage within an apartment of a twenty-six storey building in Brussels. The only known North American application of the technique involves a modern University building in Canada. The floor slabs of a computer room were reinforced in bending to accommodate an increase in loading. Supporting beams were also strengthened using steel strips bonded to the vertical faces to comply with the requirements for shear in the latest code of practice.

A precast, prestressed hollow-box beam skew bridge on the Ml in Yorkshire was required to support a vehicle carrying a generator to Cumbria. The total load of the lorry plus generator was 460 tons. Strengthening the bridge using an additional overlay slab of reinforced concrete was considered but this would have resulted in considerable increases in dead weight and additional road works. Another possible solution considered was to strengthen using longitudinal plates which would have increased the bending strength, but would have done little to enhance the insufficient shear capacity of the bridge. The solution actually adopted was to bond plates to the soffit of the deck in a direction parallel to the abutments which was 45° to the longitudinal axis of the hollow box beams. The strengthening was successful as only small deflections of the deck were observed during the passage of the abnormal load over the bridge.

Since 1983 there have been, on average, one or two applications of bonded plating to buildings per year in the UK. Following a fire in a single-storey school building in Glasgow, prestressed concrete roof beams were redesigned and strengthened using flexural plates, assuming zero residual prestress. Also in Glasgow, concrete roof beams in a distillery were strengthened in flexure and ‘nominally’ in shear following removal of supporting internal walls during reconstruction (Fig. 6.10). The cracking which developed in a West

Midlands police car park soon after construction, as a result of a deficiency in conventional reinforcement, has been well publicised. Here again, additional bonded plate reinforcement was installed in order to control further cracking under live load. An 11 m span, 2 m deep reinforced concrete beam at a shop in Bootle was strengthened to provide a 10% increase in ultimate moment capacity. In so doing it was calculated that the mean horizontal shear stress resisted by the adhesive was 9.84 N/mm2.

In 1985 cracking was noticed in the floor slabs of a multi-storey office building in Leeds. The cracks were adjacent to the external columns and the central lift well and design checks indicated a deficiency in both shear capacity and top flexural reinforcement. A combination of soffit supporting brackets and steel plates bonded to the top surface adjacent to supports was adopted to restore capacity and control cracking (Fig. 6.11). Subsequent load tests revealed that the steel plates were attracting tensile stresses up to 40 N/mm2 at 1.35 times design load.

A recent example of a bridge strengthening scheme in this country is the upgrading of a pedestrian bridge at a service station on the М2 in Kent. After design checks revealed that there was insufficient steel in the deck and a bridge inspection showed that there were transverse cracks in the soffit, longitudinal plates were applied to

the soffit at each end of the bridge to provide the extra strength required. In 1987 an access ramp and bridge were strengthened at Felixstowe docks to accommodate the increased axle loading now required by EEC regulations.

These examples illustrate the range and type of applications which are likely to benefit from steel plate bonding. All have been carried out successfully and the required extra strength has been achieved.

Structural design. For applications involving the addition of steel plate reinforcement to enhance the flexural capacity of an existing reinforced concrete structure, design calculations may be based upon normal reinforced concrete theory. One of the advantages of the technique over other forms of strengthening is the ability to carry out the work while the structure is still, at least partially, in use. Thus, any temporary propping of the structure to relieve dead load stresses is unlikely to be attractive. The calculation process under design working loads is best carried out in stages:

(1) Calculation of existing stresses based upon the dead and permanent loads acting on the original section.

(2) Calculation of additional stresses based upon the imposed or live loads acting on the strengthened section.

(3) The stresses are added algebraically and a check made to ensure that those in the concrete and both sets of reinforcement are acceptable.

(4) Calculation of the interfacial shear stress in any adhesive layer and a check made to ensure that this is acceptable.

For the purposes of these calculations steel areas may be transformed into equivalent concrete areas by using a modular ratio adjusted for permanent loads where appropriate. It should be noted that, although the effect of plating is to change the position of the neutral axis and hence to increase the area of concrete resisting flexural compression, the need to limit the stresses within the existing sections may govern the design. A final check should be made to ensure that the moment of resistance of the section under ultimate conditions is adequate and that a ductile failure mode would occur.

A well prepared adhesive joint using a cold cure epoxy resin will have an intrinsic shear bond strength of between 10 and 20 N/mm2 depending on the adhesive. This is well above the shear strength of the concrete to which it is bonded which will probably be not more than about 4 N/mm2. It is important that under ultimate conditions the local shear stresses in the adhesive do not anywhere exceed the shear strength of the concrete. For steel-to-steel joints, such as for multilayer plates or cover plates at butt-joints, a higher value of bond shear strength may be taken. Typical values of allowable mean shear stress in the adhesive layer for design under service loading are 1.2 N/mm2 for a steel-to-concrete joint and 3.0 N/mm2 for a steel-to-steel joint.

The plate thickness must be selected to allow reasonable flexibility for conforming with concrete surface irregularities. However, 3 mm is likely to be the minimum practical thickness to avoid distortion during gritblasting. A wider thinner plate gives better results because of reduced shear and normal stresses resulting from the bigger contact area but if the plate is too thin there is a risk of out of plane transverse bending which may cause the edges of the plate to lift from the surface. The most satisfactory results may be expected from plates having a width/thickness (bit) ratio of about 60. This will ensure failure by yielding of the plate at ultimate load with little or no risk of sudden springing off of the bonded plate, while at the same time maintaining a significant stiffening effect and increase in failure load. Ratios of less than 40 should be avoided, particularly in continuous beams at regions of hogging bending moment where shear and bond stresses in the concrete are likely to be relatively high.

Wherever possible plates should extend over the complete length of the region in tension but in simply supported spans requiring flexural strengthening this is unlikely to be practical. In such cases the plates should extend over at least 80% of the span. Elsewhere, a minimum transmission length of 200 mm is a reasonable assump­tion. A contribution to plate anchorage will also arise if bolts have been used to support the weight of the plate during cure of the adhesive. If the above recommendations on plate geometry have been followed, the use of bolts is not strictly necessary but their use, particularly at plate ends, is a wise precaution against unexpected defects or poor workmanship. The contribution of bolting to plate anchorage is difficult to assess without a full finite element analysis since the distribution of both shear and normal stresses varies rapidly and will depend on the particular configuration of the plate relative to the structure. Typical stress distributions within the adhesive layer are illustrated in Fig. 6.12. It is generally safer to design an adequate anchorage based on adhesive bonding alone and to provide secondary anti-peel bolts on an empirical basis. It is recommended that such bolts be designed to resist a shear stress of three times the mean value over the effective anchorage area. The effective anchorage area is obtained from the product of the effective anchorage length (/a) and plate width (b). The effective anchorage length increases with decreasing bit ratio and may be obtained from Fig. 6.13. Such bolts will also be capable of resisting any normal stresses at the plate ends.

More than one layer of plates may be needed to transmit the required load and yet meet the width/thickness ratio requirements for individual plates. Alternatively, it may be necessary to form a butt-joint between plate ends in which case the load must be transferred through a bonded cover plate. Such joints should be kept to a minimum since they result in a change of section stiffness. They should also be avoided at locations of high deformation, e. g. structural plastic hinges, and in the region of concrete construction joints. Cover plates should have the same dimensions as the main plates and overlay lengths must be sufficient to minimise the force which the concrete is called upon to transmit. Research(16) suggests that the total overlap length for plates having the recommended bit ratio of 60 should be at least 400 mm.

An environment which combines high moisture levels and de-

icing salts with elevated temperatures is likely to be particularly deleterious to the long-term durability of the adhesive bond. Research(17) suggests that many of the two component, cold-cure epoxy resin adhesives which have been used for bridge strengthening purposes are likely to suffer a significant reduction in elastic properties and bond strength above the temperature range 40-50 °С or after long periods of immersion in water. However, mechanical properties are largely recoverable if the rise in temperature has been caused by an isolated extreme in ambient conditions.

Requirements of the adhesive. A full compliance spectrum for steel/concrete bonding has been published by the authors(18) and is reproduced as an Appendix at the end of the book. The purpose of the adhesive is to produce a continuous bond between steel and concrete to ensure that full composite action is developed by the transfer of shear stress across the thickness of the adhesive layer. Experience has shown that the best chance of success is likely to be achieved by using cold-cure epoxy based adhesives which have been specially developed for use in the construction industry. Provided that the surfaces have been prepared properly, these bond well to both steel and concrete and do not suffer shrinkage and cracking problems such as may occur with other systems like polyesters. For these purposes a cold-cure adhesive is defined as one which is capable of curing to the required strength between the

temperatures of 10 °С and 30 °С. The resin component will normally be based upon diglycidyl ether of ‘bisphenol A’ or ‘bisphenol F or a blend of the two. The hardener, or curing agent, will normally be from the poly amine group, since these tend to produce adhesives with better resistance to moisture than do the polyamides and are less likely to give concern over creep performance under sustained load than with polysulphides. Other additives such as diluents, flexibilisers, plasticisers, toughening agents and inert fillers may also be incorporated into the formulation to improve the application or performance characteristics of the adhesive. In the case of inert fillers these may alternatively be supplied as a third component for inclusion at the time of mixing. It is important that the filler be of a non-conductive material, be highly moisture resistant, be capable of withstanding temperatures well in excess of maximum service temperature and typically shall have a maximum particle size of 0.1 mm. This latter requirement is to minimise the possibility of moisture penetration around the surface of the particles. The toxicity of the chemicals used in the adhesive and any associated primer for use on the steel surface must be low enough to enable safe use on a construction site.

To ensure thorough mixing it is helpful if the resin and hardener are of dissimilar colour. The adhesive should also mix to a smooth paste-like consistency suitable for spreading on both vertical and horizontal surfaces of either concrete or treated steel, in layers from 1 to 10 mm thick to allow for concrete surface irregularities. The pot-life of the mixed adhesive, which determines the time after mixing within which it must be used before it starts to harden, generally needs to be at least 40 minutes at 20 °С. A similar joint open time, which represents the time limit during which the joint has to be closed after the adhesive has been applied to the surfaces to be joined, is also necessary. If the joint is closed after this time the strength of the bond may be dramatically reduced because the exposed surface of the hardener reacts with moisture and carbon dioxide in the atmosphere in a way which impairs adhesion. This effect can be minimised by roughening the surface layer of the wet adhesive just before closing the joint. Since both pot-life and open time are temperature and humidity dependent it is evident that the sequence of operations must be planned carefully to ensure that the adhesive can be applied and the joint completed within the allowable times. For applications involving repair or strengthening the adhesive usually needs to be capable of curing sufficiently to give the required mechanical properties within a period of 3 days at 20 °С in relative humidities up to 95%.

The necessary mechanical properties can be sub-divided into those required of the hardened adhesive itself and those representing bond efficiency with appropriate adherends. In the former case, limiting properties for moisture and temperature resistance together with minimum values of flexural modulus and shear strength are suggested in Table 6.2(18). The limit on water uptake by the adhesive is to assist durability of the steel/adhesive or concrete/adhesive interface even if moisture uptake is not deleterious to the adhesive itself. The lower limit for heat deflection temperature is to ensure that maximum likely ambient temperatures in the UK do not affect the efficiency of the bond. The minimum value for flexural modulus is to guard against problems due to creep of the adhesive under sustained loads, whereby the stiffening efficiency of the additional steel might be impaired. The upper limit on flexural modulus is to limit stress concentrations arising from strain incompatibilities at changes of section. The minimum value of bulk shear strength ensures that the adhesive is at least as strong as the concrete to which it is to be bonded with an appropriate factor of safety. It also assists in minimising the creep by keeping the working shear stress in the adhesive layer, typically 0.5-1.0 N/mm2, at a relatively low pro­portion of its ultimate strength.

In the case of bond strength, steel-to-steel joints are recommended in Table 6.2 for determining fracture toughness and static and fatigue strengths. This is because steel-to-concrete joints are usually dependent on the concrete and as such provide no measure of the bonding efficiency with the external steel plate. The environmental conditions during operation and the required length of service for civil engineering structures need to be taken carefully into consideration when selecting an adhesive. A market evaluation sponsored by the Scottish Development Agency(19) suggests that the minimum required life for this strengthening technique applied to concrete bridges is 30 years. Clearly any accelerated laboratory tests selected, employing for example the wedge cleavage test, should demonstrate, as far as is practically possible that joints made with the chosen combination of surface pretreatment, primer and adhesive can survive the wide range of temperatures and condensing humidities likely at bridge sites whilst also subject to spray from de-icing salts, or, in the case of maritime structures, from the sea.

Surface preparation and curing. The concrete surface to be bonded must be sound, uncontaminated and free from chlorides. Before preparation any cracks wider than 0.3 mm and liable to leakage should be filled by injection of a suitable low viscosity resin. The existing surface must be roughened, using grit — or sandblasting, scabbling or a needle gun, to remove any weak material, surface laitance or contaminated concrete. Prior to applying the adhesive the prepared surface must be dry and free from dust. The surface of the steel to be bonded must also be free of contaminants including mill-scale, rust and most importantly, grease or oil. The degreasing and roughening procedure outlined in Table 6.3 is suitable(19),(20). A final solvent degrease after gritblasting is sometimes recommended but in the authors’ experience this can redistribute contamination and do more harm than good (e. g. Fig. 3.9); certainly great care must be taken to ensure that there is an adequate time interval for solvent products to evaporate prior to adhesive spreading. Although the application of an epoxy based anti-corrosion primer to the prepared steel surface is often considered prudent to minimise any risk of subsequent corrosion, advice must be obtained to ensure that it is compatible with the adhesive system chosen or there may be a risk of interfacial bond breakdown. If there is likely to be any significant lapse of time, say greater than 24 hours, between the application of the primer and the adhesive, a primed steel surface should be given a final degrease before application of the adhesive itself.

Thorough mixing of the packaged and measured adhesive com — Table 6.3. Steel surface preparation procedure

(1) If necessary remove heavy layers of rust by hand or mechanical abrasion with emery cloth or by wire brushing to give rust grades A or В as defined in Swedish Standard SIS 055900 (20).

(2) Degrease by brushing with a suitable solvent, e. g. acetone or 1.1.1 trichloroethylene, and allow to evaporate.

(3) Grit blast to grade 2| of Swedish Standard SIS 055900 (20) to achieve a maximum peak-to-valley depth of at least 50 p, m, using a hard angular clean metal grit which is free of any grease contamination.

(4) Remove surface dust by brushing, vacuuming or blowing with a clean uncontaminated air supply.

Note: Step (4) should be followed as soon as possible by application of the adhesive.

ponents is carried out either by hand or with a slow speed mechanical mixer to avoid trapping air bubbles. The adhesive is then spread by hand in a thin layer to both steel and concrete surfaces, before the two parts are brought together and wedged or bolted into position. The final adhesive thickness should not be less than 1 mm although in practice the actual value will vary in excess of this depending on the flatness of the concrete surface to which the steel is bonded. During the curing period the ambient temperature needs to be maintained at a level of at least 10 °С for 24 hours with most formulations. Thus, it may be necessary to provide some form of external heating for winter repairs in the UK. Any temporary supports should be left in place until tests on control samples stored under the same conditions as the actual joint indicate that the required strength has been attained. In general there appears to be no need to stop normal traffic from using strengthened bridges while the bonding operations take place or during the curing period. Finally, the exposed steel surfaces should be painted with an anti­corrosion paint. This is ideally lapped over the exposed edge of the hardened adhesive and onto the concrete surface to minimise the possibility of moisture ingress along the steel/adhesive interface.

Adhesive quality control. Routine quality control tests on site are likely to be less comprehensive than the full scale materials testing programme which will be necessary for initial adhesive selection. In the former case check tests on pot-life and open time should be carried out on each batch of adhesive to ensure consistent results. Other routine tests suggested include four-point bending on samples of hardened adhesive (Fig. 6.14), joint shear strength using steel double lap specimens (Fig. 6.15) and pull-off adhesion on the concrete surface of the structure. It is also advisable to manufacture additional specimens for storage on site and testing at a later date

in order to assess long-term durability.

Tests to measure the bond which can be obtained with the concrete of the structure to be strengthened are best carried out on the structure itself. A possibility is to utilise a pull-off test as developed for the non-destructive testing of concrete(21). A circular steel probe is bonded to the concrete surface and specially designed portable apparatus is then used to pull off the probe, along with a bonded mass of concrete, by applying a direct tensile force. Any defects in bond would be revealed by the occurrence of failures at the adhesive-concrete interface.