repair slab-column connections using cfrp of slab-column connections using cfrp i.n. robertson, g....

Repair of slab-column connections using CFRP

I.N. Robertson, G. Johnson Department of Civil Engineering, University of Hawaii, USA

Abstract

Repair, strengthening and retrofit of reinforced and prestressed concrete members have become increasingly important issues as the World's infra- structure deteriorates with time. In the United States, a tremendous amount of repair work is required on concrete bridges to avoid continued deterioration and the eventual need to replace the structures. Buildings and bridges are often in need of repair or strengthening to accommodate larger live loads as traffic and building occupancies change. In addition, inadequate design and detailing for seismic and other severe natural events has resulted in considerable structural damage and loss of life, particularly in reinforced concrete buildings. Numerous buildings and bridges suffer damage during such events and need to be repaired.

The use o f Carbon Fiber Reinforced Polymer (CFRP) composite fabric bonded to the surface of concrete members is comparatively simple, quick and virtually unnoticeable after installation. In the United States, the use o f composites has become routine for increasing both the flexural and shear capacities of reinforced and prestressed concrete beams. Earthquake retrofit of bridge and building structures has relied heavily on composite wrapping of columns, beams and joints to provide confinement and increase ductility.

This paper presents the results of cyclic testing of three reinforced concrete slab-column connections. Each of the specimens is a half-scale model of an interior slab-column connection common to flat-slab buildings. The specimens are reinforced according to current ACI-318 Building Code requirements and include slab shear reinforcement. While supporting a s lab gravity load equivalent to dead load plus 30 percent of the live load, the specimens were subjected to an increasing cyclic lateral loading protocol. The specimens were then repaired with epoxy crack sealers and CFRP fabric on the slab surface and subjected to the same loading protocol. Repair with epoxy and CFRP on the top surface o f the slab was able to restore both initial stiffness and ultimate lateral load capacity of the original specimen.

Transactions on the Built Environment vol 57, © 2001 WIT Press, www.witpress.com, ISSN 1743-3509

506 Earthquake Resistant Engineering Str'uctwes 111

Introduction

The economy of flat plate buildings has lead to their wide spread use in residential and office buildings throughout the world. However, flat plate buildings have often experienced problems in lateral loading events, particularly seismic events. Flat plate systems generally rely on shear walls or other primary bracing systems to resist lateral loads and control inter-story drift. However, there will still be inter-story drift, which can induce sipficant bendmg and shearing forces at the slab-column connections. In such events the most common mode of failure at the slab-column connections is a punching shear failure. The punchmg failure can occur with little or no warning and can lead to the collapse of a large area of the floor system and possibly a progressive collapse of the entire structure. In the 1985 Mexico City earthquake, 91 flat plate buildings collapsed and 44 were severely damaged due to punchmg failure [l].

Shear reinforcement in the slab adjacent to the column connection is effective in resisting this punchmg shear failure. However, severe crackmg of the slab around the connection may still result during cyclic lateral loadmg. To avoid demolition of the entire structure, these damaged connections must be restored to their o r i p a l condition. A previous study on repair of slab-column connections using epoxy mortar was unable to restore the lateral load capacity of the connection or the initial lateral stiffness [2]. Farhey et al. were able to restore both strength and s w e s s properties by epoxy injection and addition of externally applied steel plates [3]. The cost of epoxy injection and the intrusive nature of the steel plates and bolts on the living space make this an undesirable solution.

This paper reports on the repair of slab-column connections using gravity- feed epoxy crack repair and carbon fiber reinforced polymer fabric ( C m ) epoxied to the slab surface. Three half-scale models of interior slab-column connections were subjected to an increasing cyclic lateral displacement routine while supporting a slab gravity load equivalent to dead load plus 30 percent of the live load. The specimens were then repaired with gravity-feed epoxy crack sealers and CFRP fabric on the slab surface and subjected to the same loading protocol.

Test Specimens

Each specimen represents a half-scale model of an interior flat-slab-column connection (Figure 1). The gravity load applied during the test is equivalent to a loading on the full-scale structure of the total dead load plus 30 percent of the floor live load. This results in a shear stress on the critical perimeter equal to 25 percent of the direct punching shear capacity given by the ACI-3 18 Building Code [4]. Flexural slab reinforcement consisted of 10 mm diameter deformed bars with a nominal yield strength of 420 MPa. The slab shear reinforcement consisted of smooth 6 mm diameter bars. Three different types of shear reinforcement were used (Figure 2). Reinforcement layout is shown in Figure 3. The location and spacing of stirrup legs was the same for all three specimens.


Enrthqunke Resistant Engineering Structures I11 507

300 mm 250 mm OPEN STIRRLPS WITH

alp C U I S U K

1 I e l

Figure 1. Specimen dimensions Figure 2. Slab shear reinforcement

TOP REINFORCING LAYOUT -1Onn Bars

, - BO'TTOM REINFORCING LAYOUT

Figure 3. Specimen remforcement layout

The specimens were tested as shown in Figure 4. Figure 5 shows a typical specimen in the test frame. The cyclic lateral dtsplacement routine is shown in Figure 6. Because of the stroke limitation of the load actuator, cyclic load reversals were only possible up to 5% lateral drift (Phase I). Beyond this dnft, cycling continued in one direction only (Phase 11). The specimen concrete properties are listed in Table 1. More detad on the test program is available [5].

l' LC -LCADcmL LABORATORY L R =LOADROD

STRONG FLOOR

Figure 4: Test setup for slab-column connection cyclic tests


508 Eurtlzqunke Remtant Enguzeerzng Structures /I!

Figure 5: Specimen 4HSR prior to application of lateral load.

Displacement Steps

Figure 6: Lateral displacement routine used for original and repaired specimens.

Table 1. Specimen material properties,

CFRP Tensile Strength 1 1020 MPa 1 Pull-Off 1 1.37 MPa 1

Specimen Number

4HSR

3 SLR

Shear Reinft.

Headed studs

Single leg

Repair Technique

Epoxy crack repair

Epoxy + top

Compressive Strength (f,)

(MPa) 38.2

43.4

Modulus of Rupture (f,)

(MPa) 3.5

4.3


Earthquak Resisrant Engineering Srrzlctwes 111 509

Original Specimen Tests

During application of the gravity load, slab cracking occurred adjacent to the - ..

column. As increasing lateral -

displacements were applikd, these cracks extended to the edge of the slab while additional flexural and torsional cracks occurred. The final crack pattern of a typical slab after 8% lateral drift is shown in Figure 7. None of the slab-column connections suffered punching shear failure because of the presence of slab shear reinforcement.

All three specimens displayed similar lateral load and drift capacities. A typical lateral load versus drift hysteretic plot is shown in Figure 8 for specimen 4HS. The backbone curve is defmed as the envelope of the peaks of each hysteretic cycle. Figure 8 also shows the backbone curves for the other two original specimen tests.

--------- Gravity Load Cracks

I Positive Displacement Cracks

I . . . . . . . . . . . . . . . . . . . . . . Nsgotivs Displacement Cracks I Figure 7: Cracks in original specimen.

40

- :: 20 .- Y, U

8 0 -1 - I 5 -20 -1

-40

-60

-6 -4 -2 0 2 4 6 8 10 Drift (%)

Figure 8: Specimen 4HS Hysteretic response and Backbone Curve.


5 10 Earthqzlake Resistatlt Engineerit7g Structures III

Repair Techniques

After testing to 8% lateral drift, the three specimens were repaired using epoxy crack sealers and carbon fiber reinforced polymer fabric (CFRP). In all cases, the repair was made to an area extending 600mm from the face of the column. This width was selected based on the standard width of CFRP fabric supplied by Sika Products for the repair. It represents a distance of approximately 5 times the slab thickness fkom the face of the column.

Repair of Specimen 4HS

The surface of the slab was roughened with a needle gun to produce a roughness amplitude of 5mm over the repair area. In addition, badly cracked and weak concrete around the slab-column connection was removed to expose sound concrete. Figure 9 shows a close-up view of a slab top surface prior to repair.

Figure 9: 4HS prior to epoxy repair.

Crack repair was performed using a super low viscosity gravity feed epoxy resin, SikaDur 55 SLV, by ponding on the top surface of the slab in the area to be repaired. In order to prevent leaking through the cracks, the bottom surface of the slab was sealed with SikaDur 3 1, a Hi-Mod Gel. This material was also used to fill any voids in the bottom of the slab adjacent to the column where partial crushing of the slab concrete had occurred during testing. The low viscosity crack sealer was then applied liberally to the top surface of the slab and allowed to seep into all cracks. A light sprinkling of clean silica sand was added to the top surface to provide improved bond with subsequent materials.

Damaged areas on the top surface of the slab were filled using SikaDur 21 Lo-Mod LV, a low modulus, low viscosity epoxy binder mixed with clean silica sand in a ratio of 6 parts sand to 1 part epoxy. After allowing adequate curing time, specimen 4HSR was ready for re-testing.


E~rrhquake Res~stant Engineetwg Structures 111 5 1 1

Repair of Specimen 3SL

The slab cracking was repaired following the same procedure as described above for specimen 4HS. This was followed by application of two layers of SikaWrap Hex 103C, a unidirectional CFRP fabric, to the slab top surface. SikaDur Hex 300 was used both as the CFRP matrix and the bonding agent to attach the CFRP to the concrete surface. The surface of the slab was coated with SikaDur Hex 300 using rollers. The fabric was then applied in 600mm wide by 1500mm long strips, one on either side of the column, in each direction. The first strips were placed in the direction of cyclic testing (Figure 10), with the second layer in the perpendicular direction. The fabric was then rolled with ribbed rollers to ensure full saturation and to remove any air voids below the fabric. After ambient curing for 5 days, the specimen was ready for re-testing.

Figure 10: CFRP application - first layer.

Repair of Specimen 2CS

For specimen 2CS, the same procedure was followed as for specimens 4HS and 3SL. In addition, CFRP layers were also applied to the bottom of the slab surface. This required removal of the SikaDur 31 and roughening of the concrete surface using a needle gun. Spalled concrete around the slab-column connection was repaired using the SikaDur 21 sand mortar described earlier. The CFRP was then applied using SikaDur 330 in place of SikaDur 300 to facilitate better adhesion during application. Properties of the CFRP are listed in Table 1. The pull-off tests were performed after cyclic testing of the specimens.

Repaired specimen tests

The repaired specimens were tested under exactly the same gravity load and lateral displacement routine used for the original specimen tests. Cracking of specimen 4 H S k with epoxy only repair, proceeded similar to the original specimen. However, fewer cracks formed in the epoxy repair area with these cracks opening slightly wider than the original specimen. In specimens 3SLR and 2CSR, cracks outside of the CFRP area, which had not been repaired, re-


5 12 Earthquake Resistant Enpneeving Stmctures 111

opened, but d d not propagate into the CFRP-covered area. Instead, these cracks extended along the edge of the CFRP.

Debonding of the CFRP around the slab-column connection was noted at drift levels in excess of 5 percent. This debonding extended approximately 3 times the slab thtckness from the face of the column, and did not affect the anchorage of the CFRP free edge. In addition, cracking of the CFRP occurred parallel to the unidirectional fiber in the single layers adjacent to the column. In specimen 2CSR a 40mm long rupture occurred across the CFRP fibers adjacent to one corner of the column at the 7.5% dnfl level, indicating that the tensile capacity of the CFRP had been reached at thls location.

Result Comparison

Hysteretic plots for the three repaired specimens are shown in Figures 11, 12 and 13 along with backbone curves. The original specimen backbone curve is also shown for comparison purposes. The epoxy-only repair of specimen 4HSR was able to restore the peak lateral strength of the connection, however, the initial stiffness of the specimen was not recovered by this repair (Figure 11).

-6 -4 -2 0 2 4 6 8 10 Drift (%)

Figure 1 1 : 4HSR and 4HS response.

Specimen 3SLR with CFRP on the slab top surface displays significantly increased lateral load capacity and recovery of almost all of the initial stiffness (Figure 12). The peak lateral load is 34% greater than in the original specimen, though the drop in load capacity after the peak is more rapid with the repaired specimen. However, the repaired specimen is still able to support the same lateral load as the original specimen after cycling to 8% lateral drift.

Specimen 2CSR with CFRP on top and bottom surfaces of the slab performed almost identically to 3SLR with CFRP on the top only (Figure 13). This confirmed the expectation that reinforcement of the bottom surface would not result in a substantial improvement in performance. The expensive and


Earthquake Resistant Engrneering Structures I11 5 13

difficult task of soffit application of the CFRP is therefore considered unnecessary to restore adequate connection performance.

Drift (%)

Figure 12: 3 SLR and 3 SL response.

-6 -4 -2 0 2 4 6 8 10 Drift (%)

Figure 13: 2CSR and 2CS response.

It must be noted that all three specimens were reinforced with slab shear reinforcement to prevent punchmg shear failure. This shear reinforcement continued to provide shear strength for the repaired specimens, even though larger unbalanced moments were transferred from slab to column at the hgher peak lateral loads. Without the shear reinforcement, the higher lateral loads may have lead to punching shear failure. It is not known from this study whether the CFRP aided in resisting punching shear failure.


5 14 Earthq~lake Resistant Englneerlng Str-uctzlres III

Conclusions

Based on the results of t h ~ s study, the following conclusions were made. 1. Epoxy repair of cracks in a slab-column connection was able to restore the lateral load capacity, but dtd not restore the initial stiffness of the connection. 2. Epoxy repair and CFRP fabric on the top surface of the slab around a damaged slab-column connection was effective at restoring both peak lateral load capacity and initial stiffness of the connection. 3. Application of CFRP fabric to the bottom surface of the slab in the connection region dld not improve the connection performance over repair with epoxy and CFRP on the top surface only. 4. The simplicity of gravity-feed crack repair using super low-viscosity epoxy and the ease of application of CFRP fabric to the top surface of the slab make this a simple and economical repair method for damaged slab-column connections in seismic regions. 5. The tlun profile of the CFRP fabric results in a repair that does not intrude on the useable space in the building. The repair can be performed easily and quickly under existing floor finishes such as carpet and thm-set tile, and requires minimal disruption for the buildmg occupants.

Acknowledgements

The authors wish to thank Andriano Bortolin of Slka Products USA for donation of all epoxy materials used in thls study. The authors also thank Chandler Rowe and Plas-Tech Ltd. for donation of their time, labor and expertise in performing the repair on all three specimens.

References

[ l ] Ghali, A. & Megally, S., Stud Shear Reinforcement for Punching: North American and European Practices, Proceedings of the International Workshop on P u n c h g Shear Capacity of RC Slabs, Kungl Tekniska Hogskolan Institute For Byggkonstruktion, Stockholm, Sweden, June 2000.

[2] Pan, A. D. & Moehle, J. P,, An Experimental Study of Slab-Column Connections, ACI Structural Journal, V. 89, No. 6, Nov. 1992, pp. 626-638.

[3] Farhey, D. N., A d q M. A., & Yankelevsky, D. Z., Repaired RC Flat-Slab- Column Subassemblages under Lateral Loading, Journal of Structural Engineering, November 1995, pp. 1710-1720.

[4] ACI Committee 3 18, Building Code Requirements for Reinforced Concrete (ACI 31 8-99), American Concrete Institute, Detroit, 1999.

[5] Johnson, G, & Robertson, I. N., Seismic Repair and Retro3t of Reinforced Concrete Slab-Column Connections using C m , University of Hawaii Report UHMlCE101-02, May 2001, pp. 134.


Calculation of costs for seismic rehabilitation of historical buildings

M. Bostenaru Dan Institute for Construction Management and Machinery, University of Karlsruhe, Germany

Abstract:

The years 191 8-1 940 in Bucharest saw the construction of the buildings which gave the central part its present day look. Research about inter-bellum Bucharest and its buildings is widespread and information about them is both available and reliable. Disaster prevention includes the reduction of seismic risk through retrofitting existing buildings in order to meet seismic safety requirements. The planning of alterations to existing buildings differs from new planning through an important condition: the existing construction must be taken as the basis for all planning and building actions. This paper presents the differences in the methods of approaching and calculating the costs for the rehabilitation of old buildings when seismic retrofit becomes a special issue.

1 Introduction

Prevention of disasters caused by earthquake has become more and more important in recent years. The target of economic efficiency of retrofit measures on existing buildings has a special importance, besides their structural, functional and aesthetic aspects. The mounting of retrofitting systems is, regarding the estimation and calculation of costs, a special form of rehabilitation of endangered buildings. Planning of further measures is based on information about the building and its location. While in some western countries there are elaborated methods for estimating costs, in the countries like Romania, this is not the case.


5 16 Earthquake Resrstant Engineering Structures l!!

2 Requirements on this methodology

A major goal is, to find a method which doesn't require location specific values. The only criteria should be the information available on each particular location, such as knowledge about the specific building practice in that country, the characteristics of the existing building and the possible retrofit measures.

The method should be flexible. It should be possible to apply it in different regions, to adapt it to different requirements and to extend and develop it according to the growth of knowledge in this field. It should be computer based as computer based tools prove to have many advantages such as; proper interface for the many actions required in the complex process of seismic retrofit; possibility of the use of a large database; adaptability of the database to different local prices and the prevention of man-made errors.

3 General methods of costs planing

Planing the costs for buildings includes the method to be employed, an adequate structural analysis of the building, scheduling of the structural elements and their quantities, the calculation of the costs itself, the documentation of the results accompanied by an analysis and for the optirnising of alternatives.

There is wide range of cost elements, which can be used in different stages of planning. However, the range of methods is more influenced by the availability of comparison data, according to the level of precision required. The kind of data available can be very different from one location to another and there is a strong relationship between the data collected and the specific costs estimation methods of that country. For example, in the USA surveys were undertaken more than ten years ago. At the beginning of the last decade, on behalf of the Federal Management Agency, these were converted into a database, which allows the estimation of costs of retrofit for the described buildings [l]. The Baukosteninformationszentrum in Stuttgart, Germany, has collected more detailed information for building costs [ 2 ] . As seismic risk is not a major problem there, this information can only be used for new buildings or for general renewal of the old ones, according to the cost calculation and estimation practice. In Romania there was no such data collected, but the more detailed a method is, the easier it is to find location specific catalogue data.

3.1 New buildings

There are two main German indexes concerned with the structure of a building with regard to costs: one of them giving a better understanding of constructing a building regarding the estimation of costs, while the other one is better suited for a calculation [3].

According to the first standardised index, a building will be divided into spaces with different functions. Using a database with buildings having similar construction types and functions the costs for a new building can be estimated


Enrthqunke Resistant Engineering Structures 111 5 17

through the comparison of some key types of areas such as gross floor area or main function area. This estimation is possible in incipient stages of planning.

The other norm can be used only in later stages of planing but is more precise. It relies on classifying building elements after their execution type. Its structure has different depth levels. On the first level a building is divided into real estate, furbishing up and making accessible, building - building engineering, building - technical assets, external arrangements, equipment and works of art, building extra expenses. This division is called "vertical division". On this level the building costs can be estimated. An example of horizontal division of the third costs group (building engineering) differentiates building cavity, foundation, external walls, internal walls, floors, roofs, building structural mountings and other measures for building engineering. On this level the building costs can be calculated. On a third level for example the external walls are divided into bearing external walls, not bearing internal walls, external columns, external doors and windows, exterior finishes of exterior walls, interior finishes of exterior walls, modulated external walls, anti-glare shield, external walls, others. On this level the building costs can be plotted. Further levels are allowed, according to the specific requirements of each project.

3.2 Existing buildings

Methods for evaluating costs of upgrading new buildings were only developed beginning with the last decade. The cost estimation and calculation methodologies can be classified after the criteria of the system used, eg. system of space usage, system of building surfaces, system of building elements and system of execution steps. The future use of the building is generally the only required knowledge for a cost estimation making use of a database of already constructed buildings. For a costs calculation there are more details about the use of the spaces in the building needed. The way of execution and the building substance is generally taken in account only for the costs plot.

Rolf Neddermann [2] developed a method for estimation and calculation renewal costs of old building on the basis of the method used for new ones. He goes further as the last described level in a system of a new defmed type of elements: the "old building elements". The advantage of using ready-made service packs during the planning within the element methodology influenced his decision: Each restoration pack has a description, a current number and descriptions and numbers for the individual building services which have to be executed to build that element. While the building elements were taken into account before only as a collection of single execution steps, Neddermann tries to define real constructional elements. According to his approach, a building can be divided into elements, which consist of their own constructionable parts.


5 18 Earthquake Resistant Engineering Strzictures 111

4 Special conditions for retrofitting existing buildings

The methods based on surface and volume costs are not appropriate in this case, as the measures taken to retrofit does not affect in the same way the whole building substance but mainly the bearing elements. Retrofitting measures are not oriented towards the new use of the building.

4.1 Performance level

The seismic performance of a building can achieve one of the following three objectives [4]: immediate occupancy, damage control or life safety. The fust criteria means normal use. The building can be used immediately after the earthquake, although some installations and non structural elements might be damaged. The second one means damage limitation. Non structural elements may be stronger damaged, structural elements may be damaged in isolated places. The last criteria is defined as collapse prevention. The building may suffer non repairable damage, but it should at least not collapse.

For this proposed methodology, the rehabilitation expense degree used for the estimation of upgrading existing buildings in existing methods corresponds to the three performance objectives in seismic design. The present danger degree corresponds to the damage degree in case of normal upgrading.

4.2 Decision support

The decision of adopting local measures is similarly to usual renewal of old buildings based on state of building substance (Figure 1). In case of global measures, new elements have to be built in. Detailed calculations cannot be found at the time of the estimation of costs. As static calculations take time in the planing process, the methodology proposed here creates a framework, in which these values can be introduced.

Figure 1 : Decision support between local and global measures


Earthquake Rcslstrrrit Englneerlng Strzrctures 111 5 19

4.3 Types of buildings

Town analysis differentiates between three types of historical buildings. Buildings with cultural value belonging to national heritage. Buildings with architectural value which are representative for a certain building style. The knowledge of that style can help in considering retrofit methods for that building restricted through the demand to maintain the characteristics of that style. Buildings with environmental value are existing buildings with a positive influence on the townscape. In this case, the building record is the basis for town image analysis. For all of these different levels of the depth of the costs, also influenced from the rehabilitation necessity and the balance of the life expectancy of the building.

4.3.1 Inter-bellum Bucharest buildings New building technologies such as the wide-scale use of reinforced concrete reached Romania in the years between 1920-1945 [5]. The most recent buildings were residential ones: villas, apartment blocks or social houses. The seven to eleven floor high apartment blocks, with a unreinforced masonry infill concrete frame structure, designed to resist vertical loads only are exposed to high seismic risk.

The inter-bellum architecture in Bucharest has a non contextual approach, concentrated to set new landmarks in the build context (Figure 2). The buildings lie stylistically very close to turn-of-century European styles. This means a lack of ornaments and gaining aesthetics through simplicity, like the preservation of the right angle and the use of a pure circle. An important feature of the architectural stile of this period is the emphasis of the horizontal lines through window bands and balconies. Balconies and bow windows are bounded to one continuous facade area. The transition element between bow window and balcony is in most cases a corner window stressing the non bearing character of these facade elements. Vertical accents through non-bearing elements are exceptions. Another feature of these buildings are the recesses in the facade. As the modem reinforced concrete buildings were much higher then the existing ones in the neighbourhood, there was a rule that each floor over the cornice line should have about a one meter return to the one bellow. Through this, the shadow image of a slope roof was generated through an interesting flat roof play.

Figure 2: Street front with an inter-bellum building


520 Earthquake Resistant Engrneerlng Strwtures III

Figure 3: Marcel Iancu building (about 1935), 3 Pictor St. Luchian Street [6]

4.3.2 Retrofit considerations about historical buildings One of the advantages of this methodology is the break-up with statistics, thus permitting a customised approach in applying the methodology to different kinds of buildings. It is possible to develop guidelines about the characteristics to be maintained in the case of a building which belongs to a certain stile (Figure 3) and to provide a catalogue with the adequate building elements in this case. However, this shouldn't replace consulting an architect. For example, retrofit measures based on the extension of the active building's element section are causing the greatest problems for architecturally or environmentally valorous buildings. The steps of this methodology are reaching an adequate depth to evaluate the suitable methods for this type of historical buildings and could therefore, be thus used.

5 Costs calculation methodology

Structural studies to design the retrofit measures are taking a long time and as there are no tools for calculating costs this happens usually after generating unnecessary time wasting. Through this methodology after the building survey acquisition of data for costs determination will take place in almost the same time when engineering measures are in project such the estimation of costs can take place as soon as the retrofit measures are designed.

5.1 The building record

The way the building record is made has to focus on the elements present on the last division level. The building record has to focus on recognising the structural elements and their problems. This means not only the damages but also clear differentiation between bearing and non bearing parts, identifying which bearing


parts are working properly together and also details like reinforced and not reinforced walls. A rapid visual screening has to be done before, in order to specify where to search after structural details. By digitising the building record the bearing parts should be clearly differentiated from the other but divided accordingly to the structure, which allows the costs calculation.

The building record has different detailing ranges. Damages and construction type and materials are recorded also in not so detailed ones. The most deep details are made when working with monuments. Building record plans in Bucharest have additionally different exactness levels. Sometimes there are no records available, sometimes they contain no hints to the structural system. Due to the needs of detailed static calculation the building record has to be a detailed one. It is important to have control about the exactness achieved on the site and possible post editing in the office. Professional equipment is expensive and sometimes not available, thus, the record methodology must allow computer assistance but be easy to make with low-end equipment.

The classical building record results in the plans of different building floors, one after the other and without 3D reference. One possible approach is 3D laser scanning of the existing buildings with marking of the elements, which are defining the GIS modules. If possible interior scanning and correlation of the 2D plans obtained from the vectoring is very usehl. Otherwise, the exterior scanning and use of plan sketches of the building can help.

A further step is the detailed damage record. This means its position and dimension and information about its causes, risk potential and redemption possibilities. This step is in case of retrofitting even more important as the simple building record, but cannot be made without it.

The data from the building record have to be stored in a room book. This consists of a database table where all rooms of the building and their subordinate elements, with their characteristics, are stored. The building elements which belong to a room are the elements surfaces and only if they are exterior parts, the whole elements. This means that the static measures should be divided into under elements, if needed. After developing the static measures, these will be recorded in a second book, the measures book.

5.2 Retrofit elements

The decision support algorithm described in this paper helps choosing retrofit elements. The retrofit measures provided result in some building elements. Further division levels are here like the interior and exterior wall in their bearing parts and a subdivision of these into linear, surface and node elements (Figure 5 and 5). A retrofit element consists of all works which have to be done in order to strengthen, repair, rebuild or even build a constructive component. In case of building retrofit we cannot speak only of "new" building elements but also not only about the " o l d ones. Thus, the elements taken in account are named "retrofit elements" and are from four different types: old elements, which are simply strengthened existing building elements; new elements, which are completely new elements, added in the old substance and by which it should be


522 Earthquake Resistant Engineeying Str.rlctures IIl

taken care of the connection newlold; retrofitted elements, which are old elements with new extensions and fmally replaced elements, which are new build elements instead of the deteriorated old ones.

I

structural connected parts

I l

I parts included In /

ctrlunlru -1 m "*] o t h e r a s ~ b l i a ,

L-' I- L

I I

I I ---l*-.- concrete slot extentlon , remforcement , -- - i

I mantehng , l

Figure 4: Retrofit elements and retrofitted elements, the example of linear ones

"surfare" elements (also parts

I I 1

i j partial

I ~ - - 1 - - - - - - , /_ - - m

unm of the seuctural p m interior null - 1 rttenor T-- wall --..l

-+ -."--I__-

, braclng ! manteling

manteling * plates I concrete

I---l --A

- ,

Figure 5: Retrofit elements and retrofitted elements, the example of flat ones


Eauthquake Resrstant Etzg~neet-ing Stmctures 111 523

The use of this methodology requires a clear identification of the building elements and their marking in the 3D digital GIS model. This has also the advantage, that the quantities can be calculated through the CAD software. Thus, the building elements can be edited as service specification for costs calculation in consecrated Tendering-Allocation-Billing modules. Therefore, the elements should consist only of given service specification positions given in an already defined database or, when not, the costs for the new services should be provided separately. The costs for this elements du not need to be calculated fictitious as average from the ones of existing buildings but can be really leaded back to the execution steps which caused them.

5.3 Spatial relationship

Tracing of costs gets a spatial relation. The building will be divided in simple elements as nodes (points), beams or columns (lines), floors or walls (areas). The executed works are related to this elements and static programs can also communicate with them.

This elements will be assigned as in a GI-System, using the concepts of edge, surface and node. The advantages of using a G1 type system lie in the wide calculation possibilities, the highlighting of the elements, which are causing the most costs and not at least in the possible compatibility with static programmes. This compatibility derives from the trimmed data storage in CAD and database.

5.4 Database

The transition to GIS is made through bounding the data to spatial elements. The content of the database (Figure 6) tables is a detail which will not be described in this paper.

Figure 6: Entity-relationship diagram for the database


524 Earthquake Resistant Engineering Structures III

The costs are resumed to rooms, which enable further use of the database for building management, but also during the time of execution, for example to control the "Out of use" times and find better strategies. A second resume to the whole element would be interesting for static purposes.

6 Conclusions

As the method wasn't applied until now, statements about the certainty degree of the calculated costs cannot be made. Certainty degrees can only be estimated when another methods or databases are in use in the same region, and this is not the case for Romania. An advantage of the methodology is, that from the same database a service directory can be hold out.

The modular structure of the database allows starting generalisation of the methodology. As soon as the method has been used for some projects, the values can be also stored according to a standardised format in order to build a database of retrofitted objects. This will help estimating costs for similar future projects.

References

[ l ] Hart, Gary; Srinivasan, Mukund, Typical Costs for Seismic Rehabilitation of Buildings, F E M A , 1994.

[2] Neddermann, Rolf, Kostenermittlung in der Altbauerneuerung, Wemer Verlag: Dusseldorf, 2000.

[3] Baukostenberatungsdienst der Architektenkammer Baden-Wiirttemberg, Baukostengliederung, Architektenkammer Baden-Wiirttemberg: Stuttgart, 1988

[4] VSP Associates, A Benefit-Cost Model for the Seismic Rehabilitation of Buildings, FEMA, 199 1.

[5] Union of Romanian Architects, CATALOGUE: "Bucharest in the 1920- 1940s: Between Avant-Garde and Modernism ", Simetria Publishing House, Union of Romanian Architects: Bucharest, 1994.

[6] Simetria Publishing House, The Union of Romanian Architects, (eds.). Marcel Janco in interwar Romania. Architect, Jine artist, theorist, Simetria Publishing House, The Union of Romanian Architects, Meridiane Publishing House: Bucharest, 1996.


repair slab-column connections using cfrp of slab-column connections using cfrp i.n. robertson, g....

Documents