ABECE - Expanding the application range of RC-columns by the … · 2008. 7. 23. · Table 2. Test column detailsVK1 toVK6. Empelmann (2007). Column VK 1 VK 2 VK 3 VK 4 VK 5 VK 6

Tailor Made Concrete Structures – Walraven & Stoelhorst (eds)© 2008 Taylor & Francis Group, London, ISBN 978-0-415-47535-8

Expanding the application range of RC-columns by the use of UHPC

M. Empelmann, M. Teutsch & G. StevenInstitute for Building Materials, Concrete Construction and Fire Protection (iBMB), Technical University ofBraunschweig, Braunschweig, Germany

ABSTRACT: The development of Ultra-High Performance Concrete (UHPC) with compression strengths upto 200 MPa widens the application range for RC-constructions. But UHPC shows, in comparison to normaland high-strength concrete, a brittle material behaviour when the ultimate load-bearing capacity is reached.Experimental research carried out at the iBMB of the Technical University of Braunschweig, Germany, showsthat the load-bearing behaviour of UHPC-columns can be improved considerably by the addition of steel fibres,leading to Ultra High Performance Fibre Reinforced Concrete (UHPFRC).

On this basis, it is now possible to use UHPFRC-columns for constructions, which so far have been reservedfor steel and/or composite constructions. By the use of UHPFRC the load-bearing capacity of RC-columns canbe adjusted in a “tailor-made” way, according to the individual loading situation.

This paper will present results of several numerical and experimental studies, with regard to the use ofUHPFRC-columns in comparison to normal or high-strength concrete columns and composite columns.

1 INTRODUCTION

1.1 General

Today’s engineering structures tend to get taller andmuch more filigree. Consequently, they are highlystressed and a further development of the buildingmaterials used becomes necessary. In this contextreinforced concrete compression members competeagainst steel and steel-composite members, and theload-bearing capacity has to be increased by the useof higher strength concrete or by an increase of thelongitudinal reinforcement.

With regard to nowadays ecological discussions, thesustainability of a construction has to be reflected addi-tionally in order to reduce the energy consumption forerection and operation. Due to the higher expenditureof energy for the steel production in comparison toconcrete, but also due to cost and production reasonsan increase in load-bearing capacity by additional lon-gitudinal reinforcement does not lead to the desiredresults. Therefore, only an increase of the compres-sive concrete strength remains as the most promisingoption.

1.2 Field of application

In these days ultra high strength concretes with com-pression strengths up to 200 MPa can be producedunerringly. This is achieved by a reduction of thewater/binder ratio and an increase of the packing

density by aggregate and admixture optimisation.Additionally the high density of UHPC provides amuch higher resistance against carbonation, chlo-ride penetration, freezing and freeze-thaw loading incomparison to normal concrete.

Thus, the expansion of the application range of rein-forced compression members by the use of UHPC isfocussing on compression members as used in slenderand highly stressed compression members of e.g. high-rise buildings or industrial buildings and on memberswith high durability requirements, such as bridgecolumns or structures in aggressive environmentalconditions.

1.3 Focus of investigation

For the expansion of the application of UHPFRC-columns some remaining questions have to be clarifiedwith regard to the design method and the construc-tive design requirements. In the following the resultsof some specific tests are presented, a proposal forthe current design method is given and the potentialapplication in high-rise buildings is outlined.

2 TESTS ON UHPFRC-COLUMNS

2.1 Experiments with a central normal force

As experimental basis, 12 short columns with a squarecross-section of 20 × 20 cm and a length of 60 cm were

461

Table 1. Test column details S1 to S6. Teutsch (2007)

Column S1 S2 S3 S4 S5 S6

Concrete UHPFRC 155 UHPC 150�vf × lf /df 1.25Vol.-% × 30/0.375 = 1.0 0Asl 4 Ø 28 8 Ø 28 4 Ø 14 4 Ø 28Grade Asl S 670/800 BSt 500 S 670/800 BSt 500ρl 6.16% 12.32% 1.54% 6.16%Stirrups Ø 8/8.4 cm Ø 8/6 cm Ø 8/4.1 cmArrangement A B Afr/fck 0.020 0.029 0.020 0.028 0.042 0.043ρs[Vol.-%] 1.48 2.74 1.48 2.20 3.31

�vf : fibre volumetric ratio, lf /df : fibre length and fibre diameter,fr : mean confining pressure, fck: characteristic concrete compr. strengthρl : longitudinal reinforcement ratio, ρs: lateral reinf. volumetric ratio

Figure 1. Reinforcement arrangementA (left) and B (right).

tested at the iBMB. Some details of the layout of thetest specimens are given in Table 1 and 2.

The ultimate load-bearing capacity obtained fromthe tests and the corresponding compressive strainsare given in Table 3, indicating the enormous increaseof the ultimate load by the use of UHPC.

In Figure 2 the force-strain relationships of columnsS1, S4, S5, S6 andVK 1 are shown with a nearly linear-elastic curve up to the peak load, followed by a loaddecrease and a more or less constant deformation curveup to the maximum strain level. For more detailedinformation reference is given to Teutsch (2007) andEmpelmann (2007).

2.2 Post-fracture behaviour of the tested columns

The post-fracture behaviour and the robustness of thetested columns was compared by using the ductilityvalue I10 according to Zaina/Foster (2005). This valueI10 is defined as the ratio between the complete elastic-plastic work and the elastic work as given in Figure 3.Aductility value I10 = 1 characterises a brittle behaviourand I10 = 10 an ideal elastic-plastic material behaviour.Figure 4 shows the derived ductility indices I10 of the12 tested columns.

As a design rule the ductility of UHPFRC com-pression members should reach a robustness, whichis comparable to the robustness of normal reinforcedconcrete columns, for which a long-term international

experience exists. In Zaina/Foster (2005) a ductil-ity index I10 = 6.5 has been determined from testsand calculations for normal strength RC-columns(fck = 40 MPa). With increasing concrete strength, thisvalue decreases to I10 = 4 for columns with C 100 andonly reinforced with the nominal values according tothe design codes.

The value I10 = 6.5 is indicated as a benchmark inFigure 4; the columns S2, S4, S5, VK2 and VK4 toVK6 exceed this value. However, it has to be notedthat with increasing concrete strength this value canonly be achieved with considerable efforts with regardto the steel fibre addition, confinement reinforcementand a minimal longitudinal reinforcement.

Furthermore, it can be seen that in normal strengthcolumns (C 25/30) the addition of steel fibres hasnearly no influence. The supplement of steel fibresbecomes reasonable in columns with high strengthconcrete (C 80/95). Anyhow, when using UHPC withfck ≤ 160 MPa only the above mentioned combina-tion of steel fibres, confinement reinforcement andminimal longitudinal reinforcement led to I10 valueslarger than 5. This can be noted especially for col-umn S3 with ρl = 1.54%, which showed no sufficientrobustness.

Based on these investigations and until furtherknowledge is available, ductility index values I10 ofapproximately 5 to 6.5 are recommended, in order toachieve a robust behaviour of UHPFRC-compressionmembers.

2.3 Numerical investigation of the experiments

In order to further investigate the load-bearing anddeformation behaviour of the tested columns, a3-dimensional FE-model was developed, using thesoftware DIANA. For the computational modelling ofUHPFRC an elastic-plastic constitutive law, alreadyimplemented in DIANA, was used considering a

462

Table 2. Test column details VK1 to VK6. Empelmann (2007).

Column VK 1 VK 2 VK 3 VK 4 VK 5 VK 6

Concrete UHPFRC 160 UHPFRC 155 C 80/95 FRC 70/85 C 25/30 FRC 25/30�vf × lf /df 1.70 1.0 0 0.51 0 0.30Asl 4 Ø 28 4 Ø 26,5 4 Ø 28Grade Asl S 670/800 St 850 BSt 500ρl 6.16% 5.52% 6.16%Stirrups Ø 8/6 cm Ø 8/12 cm Ø 8/17 cmArrangement Afr/fck 0.028 0.028 0.028 0.031 0.062 0.062ρs[Vol.-%] 2.20 1.10 0.78

Table 3. Ultimate load-bearing capacity and the corresponding strain of tested columns S1 to S6 andVK1 to VK6.

Column S1 S2 S3 S4 S5 S6

Nu −6516 kN −7358 kN −5612 kN −6057 kN −6224 kN −6297 kNεu −3.1 ‰ −2.8 ‰ −2.9 ‰ −3.1 ‰ −3.0 ‰ −3.0 ‰

Column VK 1 VK 2 VK 3 VK 4 VK 5 VK 6

Nu −7225 kN −5777 kN −4497 kN −4230 kN −2119 kN −2158 kNεu −3.4 ‰ −3.1 ‰ −3.0 ‰ −3.5 ‰ −3.6 ‰ −3.6 ‰

Figure 2. NR-ε relationship of columns S1, S4 to S6 andVK1.

Drucker-Prager yield criteria with isotropic hardeningand softening.

The FE-model was validated by a comparisonwith the test results. Figure 5 shows the very goodconformity between the load-compressive strain rela-tionship obtained from the test and the FE-calculationfor column S1.

In order to define minimum requirements tosecure a robust post-fracture behaviour of UHPFRC-compression members, a parametric study with theparameters: reinforced concrete cross-section, lon-gitudinal and lateral reinforcement ratio ρl andρs respectively, steel fibre content vf and steel

Figure 3. Ductility index I10 according to Zaina/Foster(2005).

Figure 4. Ductility index I10 of the 12 tested columns.

fibre aspect ratio lf /df was carried out. The resultsof the numerical investigations are considered inthe constructive design requirements mentioned inparagraph 3.2.

463

Figure 5. Load-compressive strain relationship for columnS1. Comparison between test and FE-calculation.

3 PROPOSAL FOR A DESIGN MODEL OFCENTRICALLY LOADEDUHPFRC-COLUMNS

Based on the test and investigation results, a pro-posal for a design model for centrically loadedUHPFRC-columns loading was developed, which isorientated at the German Standard DIN 1045-1 andthe present design knowledge of UHPC in Germany,documented in DAfStb-Sachstandsbericht UHFB(2007).

3.1 Load bearing capacity

The load-bearing capacity of centrically loadedUHPFRC-columns is the addition of the load carriedby concrete and steel and can be determined accordingto the following relationships:

withAc: concrete gross cross-sectionAs: longitudinal reinforcement cross-section

and

fck: characteristic concrete compression strengthThe factor 0.85 considers the influence of long-term

loading and the difference between in-situ and cylin-der compression strength of UHPFRC according toDAfStb-Sachstandsbericht UHFB (2007).

The safety factor γc is chosen to 1.5 for in-situconcrete members and 1.35 for precast members.

The coefficient γ ′c is used according to the DAfStb-

Sachstandsbericht UHFB (2007) and is given inequation (4):

It is proposed to use the formulation according toDIN 1045-1 for the reinforcement:

fyk: characteristic steel strength of the longitudinalreinforcement at the peak-load concrete compressivestrain εcu

The above formulae represent the current state of theart and are on the conservative side. However, furtherresearch is needed with regard to the value of the safetyfactor, as in comparison to international regulationsand based on test results a lower safety factor appearsto be possible.

3.2 Constructive design requirements

In comparison to conventional concrete compressionmembers a similar robustness of UHPFRC-columnscan be achieved, if the requirements stated in equa-tion (6) to (8) are fulfilled.

a) Steel fibre content:

b) Minimum confining pressure:

c) Minimum longitudinal reinforcement:

Depending on the reinforcement layout the confin-ing pressure of the actual column can be derived fromthe equations in Table 4.

3.3 Further development of the design model

It has to be mentioned here, that the proposed designmodel is based on a test series with a limited amount oftests.Therefore, it has to be verified experimentally forother UHPFRC mixes, different cross-sections (e. g.circular, but especially wall-type columns) and largerdimensions (scale effect). Also altered longitudinalreinforcement layouts and grades should be studiedas, by the use of high strength longitudinal reinforce-ment with a yield strain above the ultimate concretecompressive strain an effective redistribution from the

464

Table 4. Mean confining pressure fr .

fck: characteristic concrete compression strengthAs,lat : stirrup cross-sectionbc: width of the column core measured in outer stirrups axisfyt : stirrup yield strengthsouter : vertical spacing of outer stirrupsinner : vertical spacing of inner stirrup

concrete towards the still unyielded high strength steelcan be achieved.

Furthermore, there are so far no results avail-able concerning the long-term and the load-bearingand deformation behaviour of eccentrically loadedUHPFRC-compression members. This still has to bestudied more in detail.

Finally, it is important for a practical application toalso investigate the behaviour of UHPC at high tem-peratures. It is known that high strength concretes, incontrary to normal concrete, show already at temper-atures of 150–250◦C a significant loss of strength anda considerable, much more unfavourable, explosivespalling behaviour. First tests were carried out at theiBMB with small, unloaded test specimens in order toinvestigate whether the destructive spalling in case ofa fire can be prevented by an addition of PP-micro-fibres. It could be shown that a PP-fibre addition of2.0 kg/m3 is able to reduce the spalling significantly.

However, information concerning the load-bearingand deformation behaviour of UHPC and UHPFRC athigh temperatures are not yet available.

4 FIELD OF APPLICATION FORUHPFRC-COLUMNS

4.1 Example: High-rise buildings

With regard to the possible application rangeof UHPFRC-columns, highly stressed prefabricated

Figure 6. View and foyer columns of Post Tower duringerection in 2002 [right: www.spannverbund.de]

compression members for high-rise buildings areinteresting elements. Due to architectural reasons it isvery often the designer’s task and target not to changethe geometry and construction method of the columnswithin the structure, if possible. As it is shown in thefollowing, the range of the achievable load-bearingcapacity for a given concrete column cross-section canbe widened by the use of UHPC considerably.

As an example for this application range someselected columns of the office building of the Ger-man Mail – called “Post Tower” – in Bonn/Germany(Figure 6) will be looked at in more detail.

The high-rise building consists of 42 floors witha total height of 162 m above ground. The groundfloor columns have to carry design loads NEd up to30,000 kN, with a column diameter of 762 mm and abuckling length of 15.65 m, resulting in a slendernessrate of λ = 82 (see Figure 6, right). The upper floorcolumns have a maximum outer column diameter of508 mm and a buckling length of 3.55 m, carrying adesign load NEd of 20,000 kN.

With the currently available and used concretegrades, i.e. C 100/115, and allowable longitudinal rein-forcement ratios ρl ≤ 9%, the necessary load-bearingcapacities for the given cross-sections, especially inthe lower floors, could not be achieved in an econom-ical way. That is why, all of the approximately 1,600columns in the 42 upper floors were carried out as steelcomposite columns.

4.2 Basic input data for the investigations

The following study reflects on the possible fieldof application of UHPFRC for highly stressed com-pressive members and shows the possibility of aconstruction as a reinforced concrete column withC 100/115, as an UHPFRC-compression member andas a composite column. Figure 7 shows the structural

465

system, the loading and the important geometricalvalues of the investigated columns.

This study is based on the design model pro-posed in Section 3, whereas it has to be rememberedthat this model is still under development (see para-graph 3.3). Conventional reinforced columns with aconcrete compression strength of C 100/115 were anal-ysed according to DIN 1045-1. The structural designcalculation for steel columns was carried out accordingto DIN 18800, and those for the composite columnsaccording to EC 4.

4.3 Foyer column

For the highest stressed interior column in the foyer ofthe “Post Tower” (NEd = 30,000 kN) with a bucklinglength sk = 15.65 m, the investigation results are givenin Figure 8.

The calculation with a concrete grade C 100/115would have resulted in a reinforcement ratio ofρl = 17% , which is practically impossible and wellabove the allowable value according to DIN 1045-1.That is why a mixed reinforcement with an additionalinner steel core was chosen, as shown in Figure 8. A

Figure 7. System, column loading and geometry.

Figure 8. Comparison foyer column, column withC100/115, UHPFRC 160 and composite column (left toright).

Table 5. Unit price and CED.

Price CED Price CED

Reinforcement 1200 €/t 13700 MJ/t UHPFRC 160 1200 €/m3 6900 MJ/m3

(BSt 500 S)Sectional Steel S 355 1400 €/t 13700 MJ/t C 100/115 450 €/m3 2600 MJ/m3

Formwork 80 €/m2 50 MJ/m2 C 35/45 130 €/m3 1200 MJ/m3

significantly lower reinforcement ratio of ρl = 9% isobtained for a column constructed with UHPFRC. Thecomparable composite column consists of a steel tubesteel S 355 with a wall thickness of 10 mm, an innerquadratic steel core 38/38 cm of S 355 and reinforcedconcrete C 35 with a reinforcement ratio of ρl = 4%.

It can be stated here that by the use of UHPFRC itwould have been possible to erect this highly, loadedslender column as a reinforced column.

In addition to the comparison of the design possibil-ities, the sustainability and the costs were examined.Table 5 shows the assumed unit prices and the cumu-lated energy demand (CED) of the used materials takenfor this comparison. The CED-values were consideredaccording to Schießl (2007).

Figure 9 shows the comparison of the total priceand the energy demand (CED) for the foyer columnincluding concrete, reinforcement steel, steel sectionand formwork, if necessary. It can be stated here thatthe UHPFRC variant is considerably more economicalas the composite column (44%) and that the energyconsumption is also significantly lower (58%).

4.4 Interior column Ø 508 mm

For the interior columns (NEd = 20,000 kN) theobtained cross-sections are given in Figure 10. For thecolumn alternative with C 100/115 an unacceptablehigh longitudinal reinforcement ratio of ρl = 16.5%became necessary, leading – as for the foyer column –to an impractical solution with reinforcement steelonly, and required also a mixed reinforcement layoutwith an additional inner steel core.

The total price and the total energy consumption forthe fabrication of these columns are given in Figure 11.The UHPFRC-compression member costs only 59%

Figure 9. Unit price and CED per m foyer column.

466

of the composite column and the energy consumptionwould have been 59% lower.

4.5 Load-bearing capacity for a givencross-section

In order to show the possible range of the load-bearingcapacity for a compression member with a squarecross-section of 30 × 30 cm and a buckling lengthsk = 3.55 m, the calculated resisting normal forces NRdare given in Figure 12. It can be seen that due tothe application of HPC and UHPC a wide range offorce levels can be considered in the construction ofconcrete columns; in particular between 1,387 and9,300 kN, which means that the design load of a nor-mal “C 30/37-Column” can almost be tripled by a“C 160-Column”.

Figure 13 shows the comparison of the neces-sary cross-sections for the three columns (UHPFRC-compression member, composite column and steelsection with a fire protection casing) with the samedesign resisting force of NRd = 9,300 kN.

Figure 10. Comparison interior column, column withC100/115, UHPFRC 160 and composite column (left toright).

Figure 11. Costs and CED per m interior column.

Figure 12. Calculated NRd of the square RC-column.

Figure 14 shows the comparison of total price andCED of the columns shown in Figure 13, indicatingonce again the advantages of the UHPFRC-columnwith regard to costs and energy demand, compared tothe steel or the composite column.

5 CONCLUSION

Based on experimental and numerical investigationscarried out at the iBMB a proposal for a design modeland constructive design requirements for the construc-tion of UHPFRC-columns could be given. Due to thelimited number of tests on the one hand and the numer-ous input parameters on the other hand, this designmodel is still under development.

However, the presently available results can besummarised as follows:

• For UHPFRC-columns a design model based onthe basic design assumptions of normal and high-strength RC-columns is possible. The differentsafety parameters are adjusted accordingly.

• A robustness similar to conventional RC-columnscan be achieved for UHPFRC-compression mem-bers, if a combination of steel fibre addition, con-finement reinforcement (stirrups) and longitudinalreinforcement (preferably high strength steel) isused. The necessary steel fibre contents, and therequired reinforcement ratios are given in this paper.

• Spalling of the concrete surface of the usedUHPFRC at high temperatures can be avoided bythe addition of 2.0 kg/m3 PP-micro fibres.

The investigations on the possible application rangeof UHPC showed that load-carrying capacities can beachieved, which so far have been reserved for steeland/or composite constructions. Based on an exam-ple of a high-rise building, it was shown that with the

Figure 13. Comparison between an UHPFRC 160 column,a composite column and a steel section column (left to right).

Figure 14. Costs and CED per m square column.

467

development of UHPFRC-compression members allcolumns of this building could have been erected byusing reinforced concrete.

The use of UHPFRC would widen the applicationrange of reinforced compression members. Further-more, UHPFRC used for slender and highly stressedcolumns is much more favourable with regard to erec-tion costs and demands less energy compared to steelor composite columns.

REFERENCES

DAfStb-Sachstandsbericht UHFB. 2007. (Not published).DIN 1045-1. 2001.Tragwerke aus B., Stahl- und Spannbeton.DIN 18800. 1990. Stahlbauten. Bemessung und Konstruktion.

Empelmann, M. & Teutsch, M. & Steven, G. 2007. Trag-und Verformungsverhalten von Stb.-Stützen aus normal-hoch- und ultrahochfestem Beton. Braunschweig (Notpublished).

Empelmann, M. & Teutsch, M. & Steven, G. 2008. Load-Bearing Behaviour of Centrically Loaded UHPFRC-Columns, Proceedings of 2nd Int. Symp. on UHPC,Kassel.

Eurocode 4 (EC4). 1994. Design of composite structures.Schießl, P. & Stengel,T. 2007. Der kumulierte Energieaufwand

(KEA) ausgewählter Baustoffe. www.cbm.bv.tum.de.Teutsch, M. & Steven, G. 2007. DFG-SPP 1182-

Arbeitsbericht UHPFRC-Druckglieder. Braunschweig(Not published).

Zaina, M. & Foster, S.J. 2005. Modelling of fibre-reinforcedHSC columns, UNICIV Report R 439.

468