Bendt Aarup Manager CRC Technology ApS Østermarken 119, DK-9320 Hjallerup

Transcription

1 1 Slender Columns Bendt Aarup Manager Technology ApS Østermarken 119, DK-930 Hjallerup -mail: Lars Rom Jensen Chie design engineer Hi-Con ApS Gørtlervej 8, DK-930 Hjallerup -mail: Peter llegaard Associate proessor Structural Research Laboratory Department o Building Technology and Structural Design Aalborg University, 9000 Aalborg -mail: pe@bt.aau.dk ABSTRACT is a high-perormance steel ibre-reinorced concrete with a typical average compressive strength in the range o MPa. Design methods or a number o structural elements have been developed since was invented in 1986, but the current project set out to urther investigate the range o columns or which current design guides can be used. The columns tested had a slenderness varying rom 1.11 to 1.76, and a reinorcement ratio (area o reinorcement to area o concrete) ranging rom 0 to 8.8%. A total o 77 tests were carried out 61 columns were tested in ambient conditions and 16 columns were tested in standard ire conditions. The tests showed good correlation between test results and results calculated according to established design guides. The ire tests demonstrate that load capacity o slender columns can be reduced very quickly due to thermal stresses and a reduction o stiness also in cases where temperature at the rebar is still relatively low. However, guidelines or achieving acceptable ire resistance can be determined based on the test results. Key words: columns,, ire resistance, design, tests 1. INTRODUCTION short or Compact Reinorced Composite - is a high-perormance steel ibre-reinorced concrete developed in 1986 [1]. The ibre content is typically -6% by volume and the average

2 compressive strength is in the range MPa. has a very low porosity which means that durability and resistance to corrosion are very good, so that a very small cover to the reinorcement can be used. This is very important because is oten used or slender structures and because a combination o passive reinorcing bars and ibre reinorcement is used in. Over the last 6-7 years, has been used increasingly or a number o small structural applications such as staircases and balcony slabs in Denmark [,3], and there is a growing interest or elements such as beams and columns. has been investigated extensively and part o the development o has been carried out in a number o uropean Research projects. Based on the input rom these projects design guides have been developed [4]. However, the experimental background is relatively limited or columns. Hi-Con, the world's largest producer o elements, who have been producing since 001, wanted to establish a broader base or design o columns. This was done in the current project, sponsored by Mål A uropean Union Regional programme. The project was headed by Hi-Con, with support rom Technology and Carl Bro as. Testing was carried out at Aalborg University (AAU) and the Technical University o Denmark (DTU) in Copenhagen. The project was initiated in September 00 and concluded in September 004. Figure 1 - Cantilevered Hi-Con balcony slabs used in apartments in Aalborg, Denmark.. COLUMNS TSTD UNDR AMBINT CONDITIONS.1 Test programme The programme ocused on centrally loaded columns in ambient conditions where a total o 57 columns were tested. The columns ranged rom 80x80 mm cross-section with a height o 4.

3 3 metres to 00x00 mm cross-section with a height o.7 metres. Other parameters in addition to size and slenderness were shape, reinorcement ratio, size o reinorcement and steel ibre content. The programme also included 4 columns tested with eccentric load with an eccentricity o 5 mm. 6 columns were tested at DTU mostly those with a height diering rom 75 mm, while 51 columns were tested at AAU, including the 16 columns tested in ire conditions and the 4 columns tested with eccentric load. The setups are shown in igure. At AAU the testing was done in a newly built 000 kn press with hinges at the top and the bottom. The centre o rotation was placed so that the physical length o the columns was equal to the theoretical length shown in table. The hinges allow or delections in all directions. Load was introduced in increments and at each load level, 10 measurements o displacements were taken. In each test series, at least one column was loaded to ailure, while or others, the test was stopped ater a load reasonably above the predicted ailure load had been achieved. The testing at DTU was carried out in a 5000 kn press. The columns were simply supported at each end, i.e. such that the ends o the column were ree to rotate in one plane and rotationally restricted perpendicular to this plane. The theoretical column length, which is given in table, was slightly larger than the physical length o the columns as the distance rom the surace o the supports to the centre o rotation was added. The tests at DTU were carried out in displacement control at a constant rate o travel o the crosshead o the testing machine. Figure - Testing setup at DTU (on the let) and AAU.

4 4 All columns were produced at Hi-Con as part o their normal production with the precision which is normal or the industry regarding placing o reinorcement, preparation o ends and initial curvature. Square columns were cast on the side while round columns were cast standing up. Composition or 1 m 3 was: binder Sand 0- mm Sand -4 mm Water 940 kg 664 kg 661 kg 154 kg binder is a mix consisting o cement, micro silica and dry super plasticizer. The steel ibre content was 160, 30 or 480 kg depending on whether a, 4 or 6% mix was used. The steel ibres were straight, smooth and had a length o 1.5 mm and a diameter o 0.4 mm. Generally, cover to the reinorcement was 15 mm except in the case o the columns with cross-sections o 00x00 mm, which had a nominal cover o 5 mm... Results or central loads The properties used or calculations are shown in table 1. The table shows 4 sets o values, all based on results or 100x00 mm cylinder tests, a sample size which is standard or : xpected mean values (conservative estimate) based on other tests with [4] Characteristic the 5% ractile value o expected values Design design value or modulus is the same as the characteristic value, while the design value or compressive strength is obtained by dividing by a material actor o 1.65 Test results ound in testing at AAU or this speciic project on production batches The test values or the mix with 4% o ibres were expected to all between the values achieved with and 6% o ibres, but the values are relatively low. This could perhaps be attributed to dierences in exact water content and compaction. Mixes with 4 and 6% o ibres were produced in smaller batches than the mixes with % o ibres, as the % mixes are part o the normal production at Hi-Con. Fibres are added manually. For the 4% mixes it was observed, that there was little variation in the properties measured or test specimens rom one batch, while there was a relatively large dierence rom one batch to another. The standard deviation was generally larger than what is observed in the normal quality control at Hi-Con. Table 1 - Properties used or calculations. Fibre content expected charact. design Test, mean test stan.dev. expected charact. design test, mean test, stan. dev. Vol.% GPa GPa GPa GPa GPa MPa MPa MPa MPa MPa Some o the main results o the column tests are shown in table. There was considerable variation in the test loads that were carried, but in general the carrying capacity was larger than

5 5 expected. The estimated capacity shown in table was calculated using the properties measured in the project and marked Test, while the design capacity was calculated based on design properties. Also shown in table are two ratios. Ratio 1 is the maximum test load divided by the estimated capacity, while ratio is the maximum test load divided by the design capacity. In a number o cases the columns were not loaded to ailure as testing was stopped ater the estimated capacity had been achieved. This is indicated with * and in these cases the maximum test load carried corresponds to the minimum carrying capacity or the column. The ormulas used or calculating slenderness index α, and capacity N,CR are shown below. They have been derived rom tests carried out in the URKA project Compresit [5] and the Brite/uRam project HITCO [6], where short columns were tested and the Brite/uRam project MINISTRUCT [7], where also slender columns were tested. The ormulas dier only slightly rom the conventional calculation methods, but they predict a slightly higher load capacity than conventional methods. As the increase in strength or compared to conventional concrete is much higher than the increase in Young s modulus the slenderness index or will oten be relatively high. With other types o aggregate the ratio between stiness and strength would be dierent, i.e. with calcined bauxite as aggregate compressive strength would typically be 00 MPa while Young s modulus would be 75 GPa. N,CR where: is the lower value o: cr β min β ( ( cr s s s s 1 = 1+ α λ = π ) ) (1) () α 0 (3) l c r g s A A s A s l c λ = (4) r g : ree column length : radius o gyration : uni-axial compressive strength o matrix : strength o reinorcement : compressive stress in matrix : cross-sectional area : cross-sectional area o reinorcement : cross-sectional area o matrix : modulus o elasticity o matrix : modulus o elasticity o reinorcement A s β = (0.95 ) i α < 1.5 A β = 0.95 i α>= 1.5

6 6 Table - Results at ambient conditions with central loading. * shows that the column was not tested to ailure. Crosssection Length Slenderness Reinorcement Fibre stimated Design Maximum Ratio1 Ratio index content capacity capacity test load mm mm Vol.% kn kn kn x ø * 1.56*.39* x ø ø6 10X none x ø x ø x ø x ø x ø x ø x ø x ø x ø x ø x ø x ø Ø ø Ø ø Ø ø x ø5+ 4 ø x ø * 1.01* 1.87* 81* 1.01* 1.87* 1087* 1.13* 1.90* * 1.54*.57* * 1.34*.16* * 1.61*.6* * 1.39*.13* 1770* 1.45*.* * 1.41*.8* 1484* 1.31*.13* * 1.16*.08* * 0.79* 1.9* 198* 1.08* 1.75*

7 7.3 Results or eccentric load The ormulas used or calculating load capacity and displacements under eccentric loads are equivalent to the methods used in the Danish standard DS411 and are given in the ollowing: The modulus o elasticity is determined as [4]: c c,0 k is set to 0.14 rom limit values. c,max c,min = 1 k( ) (1 k)( ) (5) c The ultimate capacity or the column is determined the traditional way as shown in DS411 - and includes the second order moments rom the deormations. The sectional orces are given by the axial orce N s and the moment M = M 0 + (e 1 + e )N s, where M 0 is the moment rom transverse loading, e 1 is the eccentricity or the axial orce and e is the deormation at the middle o the column. ls e is determined by the curvature o the column um = κ m. 10 c, max c,min κ m = where c, max and c, min are respectively the largest and smallest concrete c h compressive stress in the cross section and h is the distance between the points in the cross section with stresses c, max and c, min. The stresses are given by: N s N s M M c,max = + and c,min = (6) A W A W where A is cross-section area and W is the rotational section modulus. The ultimate bearing capacity o an eccentric loaded column is determined as the load N cr where the cross-section ails due to a combination o N cr and M. The results are shown in tables 3 and 4. The tables show loads and displacements in ultimate limit state as well as the expected service loads and displacements at that level. Ultimate capacity is calculated based on test -properties, while design capacity is calculated based on design - properties. The service loads were determined rom the design capacity by assuming that 60% o the load on the column is dead load, while 40% is live load with a saety actor o 1.3. The columns were not actually loaded to ailure as this could have caused damage to the displacement transducers, but testing was stopped shortly ater the loads had exceeded the ultimate load capacity. The initial eccentricity e 1 in the tests was 5 mm. Table 3 - Results rom column testing at ambient conditions with eccentric load comparisons between calculated capacity and test loads. Crosssection index content capacity capacity load Length Slenderness Reinorcement Fibre Ultimate Design Service Test load mm mm Vol.% kn kn kn kn 10x ø x ø x ø x ø c

8 8 Table 4 - Results at eccentric load comparisons between calculated displacements and results in tests. Crosssectioment deorm. load deorm. load deorm. Reinorce- Meas. Ultimate xp. Charact. xp. Test load Meas. deorm. mm kn mm kn mm kn mm mm 10x10 4 ø x10 4 ø x10 4 ø x10 4 ø Discussion As shown in table, the test loads are always higher than the design capacity, and in most cases test loads are also higher than the ultimate capacity calculated with properties obtained in the material testing. This is in part due to the steel ibres, which provide the matrix with a tensile strength higher than 7 MPa [4]. The real variations in the results are lower than they appear at least or the tests carried out at AAU as only some o the columns were actually loaded to ailure, as described earlier. In some cases the columns were slightly curved prior to testing, which led to eccentric loading, early deormations and thus a lower carrying capacity in the test. The dierence or similar columns is shown in igures 3 and 4, a case which was probably the most extreme. The graphs show loading o the columns along with displacements in the centre and at the quarter points. The column shown in igure 3 had a slight curvature prior to testing and, as indicated on the graph, the column started to delect at a relatively low load and actually ailed in bending, while the column shown in igure 4 showed only small delections. In igures 5 and 6 the columns are shown ater the test. The column shown in igure 5 had a displacement o 30 mm at maximum load, and the ailure was very ductile, while the column shown in igure 6 had a brittle type o ailure where displacement at maximum load was only mm Load [kn] 750 Upper 1/4-point Midpoint Lower 1/4-point Delection/displacement [mm] Figure 3 - Load-displacement curve or DTU test on column with 10x10 mm cross-section, % ibres, length 3898 mm, reinorcement 4Y16, maximum test load 570 kn.

9 Load [kn] 750 Upper 1/4-point Midpoint Lower 1/4-point Delection/displacement [mm] Figure 4 - Load-displacement curve or DTU test on column with 10x10 mm cross-section, % ibres, length 3898 mm, reinorcement 4Y16, maximum test load 1430 kn. Figure 5 - Column tested at DTU with 10x10 mm cross-section, % ibres, length 3898 mm, reinorcement 4Y16, maximum test load 570 kn.