https://doi.org/10.14311/APP.2022.33.0489 Acta Polytechnica CTU Proceedings 33:489–496, 2022 © 2022 The Author(s). Licensed under a CC-BY 4.0 licence Published by the Czech Technical University in Prague PRESTRESSED ULTRA-HIGH PERFORMANCE CONCRETE (UHPC) BEAMS FOR REUSABLE STRUCTURAL SYSTEMS: DESIGN AND TESTING Dario Redaellia, ∗, Jonathan Moixa, Alex Muresana, b, Jan Brüttingb, Corentin Fivetb a Institute of Construction and Environmental Technologies, University of Applied Sciences Western Switzerland (HES-SO), Fribourg, Switzerland b Structural Xploration Lab, Swiss Federal Institute of Technology (EPFL), Fribourg, Switzerland ∗ corresponding author: dario.redaelli@hefr.ch Abstract. With the aim to reduce the environmental footprint of buildings, this paper presents an original structural system concept for two-way slabs in residential and office buildings. The proposed system extends the best practice in terms of modularity, versatility, demountability, reusability, and durabil- ity. A finite set of elements can be used to form the main load-bearing system of multiple successively constructed buildings having different and unpredicted layouts and static systems. Ultra-High Per- formance Concrete (UHPC) was identified as one of the most promising materials for this application because of its high strength, extreme durability and the opportunities it opens for shape optimization and material consumption reduction. In the first part of this contribution, the main features of the new structural system and preliminary design assumptions for UHPC modules are briefly outlined. Then, the authors present and discuss the results of an experimental campaign on prestressed UHPC beams that were tested in bending and shear, under service and ultimate load conditions. Beams with transversal openings in the web were also tested to assess the influence of these openings on the crack- ing behavior and shear strength of the beams. Experimental results provide useful information for the subsequent modeling of the structural system. Thanks to the presence of fibers, transversal openings in the web only have a limited influence on the response of the beams, thus allowing horizontal technical shafts to pass through the structural thickness of the slab. Keywords: Building, reuse, slab, UHPFRC. 1. Introduction Today, the global resource consumption is 1.7 times higher than what the planet can sustain [1]. The construction industry alone is responsible for approx- imatively one third of material usage [2], energy con- sumption [3] and waste production [4] worldwide, with load-bearing systems (e.g. slabs) contributing the major share of these impacts. Existing strate- gies to face these issues are mainly based on the use of construction materials with a low environmental impact (e.g. natural, locally available, or recycled materials) or on the design of material-efficient sys- tems (e.g. weight minimization). These approaches, however, do not provide a long-term solution since they are not addressing in a decisive way the build- ing’s end-of-life: costs and energy required to demol- ish, transport and transform materials to be recy- cled, and the inevitable degradation (down-cycling) of their properties. An alternative strategy to reduce environmental impacts [5, 6] consists in conceiving building’s com- ponents in order to allow their reuse with minimal interventions in successive buildings and for multi- ple building lifetimes. This strategy, even though not yet widely established from a technological point of view, has the potential to radically transform the con- struction industry towards a more sustainable circu- lar economy. 2. A reusable structural system for two-way slabs The fundamental requirements that a building system must fulfil in order to ensure open-ended reusability are: durability, versatility, modularity, reversibility and adaptability [5]. A new type of structural system for two-way slabs, which complies with all the above requirements, has been conceived within the framework of a collab- orative research project between the Swiss Federal Institute of Technology (EPFL) and the University of Applied Sciences Western Switzerland (HES-SO). The system (Figure 1.a) consists of quadrilateral slab modules that can be interconnected in a plane to cre- ate slabs with different ground plans. Each module (Figure 1, right) is conceived as a grid of orthogo- nal and regularly spaced beams, creating an array of transversal openings that define possible locations both for columns (with the addition of a head capital 489 https://doi.org/10.14311/APP.2022.33.0489 https://creativecommons.org/licenses/by/4.0/ https://www.cvut.cz/en D. Redaelli, J. Moix, A. Muresan et al. Acta Polytechnica CTU Proceedings Figure 1. Schematic representation of the structural system (a) and of individual modules realized with glue- laminated timber (b) or UHPC (c). Shear strength Bending strength Stiffness Weight VRd [kN] MRd [kNm] EI [kNm2] [kg/m] 1 layer 11.5 8.8 719 8.7 2 layers 23.1 34.0 5750 17.3 Table 1. Geometric and mechanical properties of glue-laminated timber beams. Figure 2. Glue-laminated timber beams. to support the slab) and for vertical technical shafts. Slab modules can also be stacked vertically to locally increase the static height of the system, thus allow- ing to tailor the strength and stiffness distribution of the slab to meet specific static requirements [7]. Extremely versatile, the proposed structural system allows for multiple re-adaptations in successive build- ings with unpredicted uses, geometries and support positions. For practical applications, the system was designed to resist vertical loads of residential and office build- ings with spans varying between 4 and 12 meters (with multiple layers). A first implementation is pre- sented here as a reference and was designed in GL32h glue-laminated timber. Bolted steel plates fixed to the ends of the beams with epoxy glued steel rods were chosen as connections between modules (Fig- ure 1.b). The geometric and mechanical properties of timber beams are given in Table 1 for one-layer and two-layer configurations (Figure 2). Each beam is made of eight 80 × 30 mm laths. Due to the cross- ing of orthogonal beams, however, only half of the laths are continuous in each direction ((Figure 2, left), which limits the structural performance of this solu- tion. Shear strength, bending strength and bending stiffness given in Table 1 were calculated according to Swiss standards [8] by assuming that only continuous laths contribute, which was confirmed by laboratory experimental tests. Despite the lightness, the positive environmental properties and the good mechanical strength of tim- ber, stronger and stiffer alternative solutions were searched. Ultra-High Performance fiber-reinforced Concrete (UHPC) was identified as a promising alter- native due to its very high strength, extreme durabil- ity and the opportunities it opens for shape optimiza- tion and material consumption reduction. The rest of the paper focuses on design, testing and preliminary modeling of the UHPC implementation. 490 vol. 33/2022 Prestressed UHPC Beams VR MR EI Weight [kN] MRd [kNm] [kNm2] [kg/m] ULS Design values 46.5 36.2 6105 41.8Expected average values 69.7 46.3 Required average values 34.7 43.5 Table 2. Geometric and mechanical properties of UHPC beams. Figure 3. UHPC beams. 3. Conceptual and detailed design of the UHPC implementation With respect to timber, UHPC is heavier, more ex- pensive and its production requires more energy and non-renewable materials. To provide a competitive solution with respect to timber, the following criteria were set as basis of UHPC modules design: • provide with a single UHPC layer (24 cm height) strength and stiffness equal or higher than two lay- ers of glue-laminated timber (48 cm, see Table 1) • optimize the geometry of the cross-section to re- duce self-weight and material consumption • explore the potential to create transversal openings in the webs for horizontal technical shafts, which is not possible for glue-laminated timber beams Detailed design of the beams was carried out ac- cording to Swiss Standard SIA 2052 [9] assuming the mechanical properties of a relatively common tensile- hardening UHPC mixture (class UA, [9]): compres- sive strength fU ck = 150 N/mm2, elastic limit in ten- sion fU tek = 7.5 N/mm2, maximal tensile strength provided by the fibres fU tuk = 9.5 N/mm2 and mod- ulus of elasticity EU = 55′000 N/mm2. Table 2 gives the calculated ULS design values of shear and bend- ing strength, as well as the corresponding average val- ues (that might be expected in case of testing). Also given are the estimated average values of shear and bending strength required in case of testing so that the corresponding design values (ULS) are at least equal to those of two layers of timber (see Table 1 for comparison). UHPC beams are pre-tensioned with 12 mm diame- ter Y1570C pre-stressing wires directly fixed to steel plates at the extremities of the beams (figure 1.c). Since the position of a specific slab module within the structural systems is a priori unknown, each mod- ule must be able to resist both positive and nega- tive moments. Consequently, symmetrically prestress was designed, with one prestressing wire per flange. Flanges were sized to resist the compressive force gen- erated by bending moments at ultimate, while their shape was chosen to allow placing longitudinal and transversal prestressing wires with a minimal cover according to [9]. A web thickness of 30 mm (Fig- ure 3) was chosen to facilitate the casting of UHPC and to obtain a design value of the shear strength (VRd = 46.5 kN, Table 2) significantly higher than the target design value corresponding to two layers of glue-laminated timber beams (VRd = 23.1 kN, Ta- ble 1). Thanks to this shear strength surplus, the creation of transversal openings in the web of UHPC beams can be explored. Despite the presence of few experimental and modeling approaches in the liter- ature [10–13], no recognized design approach exist to the authors’ knowledge for the shear strength of UHPC beams with transversal web openings. 4. Experimental campaign on pre-stressed UHPC beams with or without web openings To acquire a better knowledge on the influence of web openings on the behavior of the studied system and to provide basic information for further analyt- ical and numerical modeling, an experimental cam- paign on prestressed UHPC beams with and without web openings was carried out. 4.1. Description of the test series Four beams were fabricated for testing (figure 2). Beams S1 and B1 have no web openings, while beams S2 and S3 have openings with different geometries. Beams S1, S2 and S3 were designed to fail in shear and were tested in three-point bending with 1.5 m span. These beams were fabricated with a length of 1.5 times the test span, two test regions in the outer thirds and a strong region in the central part. By changing the position of load and supports (black and red configurations, figure 2), it was possible to carry out two tests per beam, named "side a" and "side b" hereafter. To avoid any risk of bending failure, the 491 D. Redaelli, J. Moix, A. Muresan et al. Acta Polytechnica CTU Proceedings Figure 4. Geometry of the four tested beams. Beam / Batch Compressive Strength fU cm* [N/mm2] Elastic modulus EU cm* [N/mm2] Elastic limit in tension fU tem** [N/mm2] Tensile Strength fU tum** [N/mm2] Hardening strain in tension εU tum** [!] B1 177.2 (3) 60’100 (3) 8.25 (6) 8.90 (4) 1.87 (4) S1 167.2 (3) 55’300 (3) 6.97 (3) 9.82 (2) 5.21 (2) S2 169.1 (3) 56’567 (3) 6.17 (6) 13.53 (6) 3.08 (6) S3 177.1 (3) 58’900 (3) 6.48 (6) 10.83 (6) 2.30 (6) * Measured on 70 × 140 mm cylinders. ** Calculated by inverse analysis of bending tests on 30 mm thick plates according to [9]. Note: (2), (3), (4), (6) indicate the number of tests used to calculate the average value. Table 3. Measures mechanical properties for UHPC at 28 days central part of these beams was reinforced with an additional " 20 B500B steel rebar. Beam B1 was de- signed to fail in bending and was tested in four-point bending with 3.0 m span. Each beam was fabricated with a separate UHPC batch. Prestress forces of 115 kN per wire were freed three days after casting. Instantaneous and delayed shortening of the beams were monitored until the day of testing for a precise evaluation of prestress losses. For "S" elements, longitudinal strains on "sides b" were monitored while testing "sides a": strains very close to zero were detected, showing that the prestressing level on "sides b" was not influenced by cracking and failure that occurred on sides "a". Tests on "side a" were stopped relatively early after the peak force, while tests on "side b" were pursued much further in the post-peak to verify the ductility of the beams. 4.2. Materials The commercial product Ductal FM was used as UHPC mixture for the test series, with self- compacting consistency and 270 kg/m3 (3.4% in vol.) high-strength straight steel fibers (length lf = 14 mm, diameter df = 0.2 mm). Table 3 summarizes the average 28 days mechanical properties for each batch. Measured values are comparable but slightly higher than those assumed for preliminary design. However, relatively low values of the elastic limit in tension, fUte, were measured for some specimen. This could be partly explained by the surface grinding procedure which was applied to 30 mm thick plates according to [9], which may have damaged these specimens. 4.3. Test results All load tests were carried out at an age of UHPC comprised between 28 and 34 days. Table 4 summa- 492 vol. 33/2022 Prestressed UHPC Beams Beam/ Batch Web opening Maximal measured force Numerically predicted max. force Maximal measured shear Maximal measured moment Expected strength (see tab. 2) Required strength (see tab. 2) Shear/moment at 1st detectable* crack B1 No 86.4 kN - - 43.2 kNm 46.3 kNm 43.5 kNm 39.0 kNm S1-a No 146.3 kN 158.8 kN 73.2 kN - 69.7 kN 34.7 kN 54.9 kNS1-b 174.3 kN 87.2 kN - 69.7 kN 34.7 kN 75.2 kN S2-a "96 149.6 kN 151.0 kN 74.8 kN - - 34.7 kN 35.0 kNS2-b 151.3 kN 75.7 kN - - 34.7 kN 44.9 kN S3-a 96 × 150 139.7 kN 127.8 kN 69.9 kN - - 34.7 kN 40.1 kNS3-b 135.4 kN 67.7 kN - - 34.7 kN 45.2 kN * Cracks detected either by visual inspection or by digital image correlation. Table 4. Summary of test results. Figure 5. Measured force-displacement curve and crack pattern at failure for beam B1. rizes the maximal applied force and the correspond- ing maximal shear force and bending moment for each test. These values are compared to expected and re- quired average strength values according to prelimi- nary design (see Table 2) and to strength predictions obtained with non-linear numerical modeling (see sec- tion 5). Figures 3, 5, 6 and 7 present force versus mid-span displacement curves (at two different dis- placements scales) and pictures of the crack patterns at failure for each beam. Based on observations and measured values data, the main experimental findings are summarized: • all beams showed the expected failure mode • for all beams, including beams with web openings, the measured strength is equal or higher than the required values (see tables 1 and 2). For beams "S", the location of the critical shear crack was strongly influenced by the position where the additional fă20 rebar ends • test results of sides "a" and "b" are practically iden- tical for beams S2 and S3, but show aăsignificant difference for beam S1 (figure 5). This difference could be explained by the accidental presence of a failure plane with particularly poor number or ori- entation of fibers on side "a". More detailed analy- ses are however required to explain this behavior • by taking as a reference the average strength of beams S1-a and S1-b, the reduction of shear strength due to presence of transversal web open- ings is in any case less than 20% • for all beams, ultimate behavior is extremely ductile, with the progressive formation of a col- lapse mechanism showing very significant post- peak residual strength • cracks were only detected (tableă4, note 1) at rel- atively high load levels, indicating that UHPC beams will be uncracked or micro-cracked in ser- vice. 5. Non-linear finite element modeling To gain a deeper understanding on test results and to carry out further parametric analyses, the struc- tural behavior of type "S" beams was modeled with a non-linear finite element software [14]. The software solution implements a coupled tensile fracture, com- pressive plasticity model and allows to define strain hardening/strain softening constitutive laws within a smeared crack modeling approach. For the sake of simplicity, a unique strain-hardening tensile law was used for all the beams, with fU te, fU tm and εU tu cal- culated as the average of all the bending tests carried out for batches S1, S2, S3 and B1 (figure 4, black). For the softening part, a stress-crack opening soft- ening law according to [15] was assumed, with zero stress for a crack opening equal to lf /2. To consider 493 D. Redaelli, J. Moix, A. Muresan et al. Acta Polytechnica CTU Proceedings Figure 6. Tensile stress-strain and stress-crack opening laws for modeling. Figure 7. Measured vs simulated force-displacement curve and crack pattern at failure for beam S1. 494 vol. 33/2022 Prestressed UHPC Beams Figure 8. Measured vs simulated force-displacement curve and crack pattern at failure for beam S2. Figure 9. Measured vs simulated force-displacement curve and crack pattern at failure for beam S3. 495 D. Redaelli, J. Moix, A. Muresan et al. Acta Polytechnica CTU Proceedings the less favorable alignment of fibers in beams than in thin bending plates, the post-cracking part of the tensile law was scaled on the stress axis by aăfactor 0.8 (figure 4, red). Young’s modulus and compressive strength were set according to values given in Table 3, while other parameters of the three-dimensional me- chanical model were set according to state-of-the-art knowledge and are not detailed here. Numerical solutions are compared to experimental ones in figures 5 to 7: despite some numerical insta- bilities in the post-peak and the need to validate the sensitivity of the numerical solution to variations in the numerous mechanical parameters, the experimen- tal response is quite well reproduced both in terms of crack pattern and forces vs mid-span displacements. 6. Conclusions and future work This paper presented the design, testing and prelim- inary modeling of UHPC beams that were conceived as a possible technical implementation of a new mod- ular, reusable and extremely versatile structural sys- tem for two-way slabs of residential and office build- ings. Experimental and modeling results demonstrate that UHPC beams can be designed to meet the strength and serviceability requirement of the new modular structural system with the imposed limi- tations of height and with a relatively light cross- section. Thanks to the tensile contribution of the fibers, the presence of web openings of significant size only has a limited influence on the mechanical performances of the beams, thus allowing horizontal technical shafts to pass through the structural thickness of the slab. In the next steps of the research program, the eco- nomic and environmental impact of the system will be evaluated and compared to other possible imple- mentations of the new modular system (e.g. timber, steel or mixed-construction) and to traditional, non- reusable constructive systems for two-way slabs. The sensitivity of finite element modeling with re- spect to variation of the input mechanical parameters will be checked, then the model will be used as a tool for parametric analysis and for further optimization of the proposed concept. Acknowledgements The authors are very grateful to the Whitbybird Foun- dation for funding J. A.’s initial work on the origi- nal paper [7]; and to the Open-Oxford-Cambridge Arts and Humanities Research Council (AHRC grant number AH/R012709/1) Doctoral Training Partnership for fund- ing the continued project. Funding for A. M. came from the School of Engineering and Innovation at the Open University References [1] European Environmental Agency. National Footprint Accounts, 2018. [2] Danish Environmental Protection Agency. Building a Circular Future, 3rd edition, 2018. https://gxn.3xn.com/wp-content/uploads/sites/4/ 2018/09/Building-a-Circular-Future_3rd-Edition _Compressed_V2-1.pdf. [3] L. Pérez-Lombard, J. Ortiz, C. Pout. A review on buildings energy consumption information. Energy and Buildings 40(3):394-8, 2008. https://doi.org/10.1016/j.enbuild.2007.03.007. [4] J. Layke, E. Mackres, S. Liu, et al. Accelerating Building Efficiency, Washington: World Resources Institute, pp. 156, 2016. https://www.wri.org/resear ch/accelerating-building-efficiency. [5] FC. Fivet. Design of Load-Bearing Systems for Open-Ended Downstream Reuse. IOP Conference Series: Earth and Environmental Science 225, 2019. https://doi.org/10.1088/1755-1315/225/1/012031. [6] C. Fiver, J. Brütting, Nothing is lost, nothing is created, everything is reused: structural design for a circular economy, The Structural Engineer 98(1):74-81, 2020. https://www.researchgate.net/profile/Jan-B ruetting-2/publication/338342629_Nothing_is_los t_nothing_is_created_everything_is_reused_stru ctural_design_for_a_circular_economy/links/5e 4581da299bf1cdb9283f37/Nothing-is-lost-nothing -is-created-everything-is-reused-structural-d esign-for-a-circular-economy.pdf. [7] A. Muresan, J. Brütting, J. Canada J, et al. Proc. of the IASS Symp. Creativity in Structural Design ed C Mueller and S Adriaenssens, Boston: MIT, 2018. [8] Swiss Society of Engineers and Architects. SIA 265 - Timber construction, 2012. [9] Swiss Society of Engineers and Architects. SIA 2052 - Ultra-High Performance Concrete (UHPC) - Materials, design and construction, 2016. [10] J. Hegger, G. Bertram. Tailor Made Concrete Structures: New Solutions for our Society, CRC Press, p. 96, 2008. [11] R. Zagon, S. Matthys, Z. Kiss. Shear behaviour of SFR-UHPC I-shaped beams. Construction and Building Materials 124:258-68, 2016. https: //doi.org/10.1016/j.conbuildmat.2016.07.075. [12] L. Moreillon, J. Nseir, J. Tomas, et al. RILEM-fib-AFGC Int. Sym. on Ultra-High Performance Fibre-Reinforced Concrete, pp 549-556, 2013. https://www.afgc.asso.fr/app/uploads/2014/ 01/pp-549-556-Moreillon-1.pdf. [13] Redaelli D and Nseir J 2020 Fibre Reinforced Concrete. From Design to Structural Applications, fib Bulletin No. 95 pp 474-483. [14] V. Červenka, L. Jendele, J. Červenka. ATENA Program Documentation - Theory, 2020. https://www.cervenka.cz/assets/files/atena-pdf /ATENA_Theory.pdf. [15] D. Redaelli. Comportement et modélisation des éléments de structures en Béton Fibré à Ultra-Hautes Performances avec armatures passives. École Polytechnique Fédérale de Lausanne, Switzerland, Lausanne, Ph.D. Thesis, p. 285, 2009. 496 https://gxn.3xn.com/wp-content/uploads/sites/4/2018/09/Building-a-Circular-Future_3rd-Edition_Compressed_V2-1.pdf https://doi.org/10.1016/j.enbuild.2007.03.007 https://www.wri.org/research/accelerating-building-efficiency https://doi.org/10.1088/1755-1315/225/1/012031 https://www.researchgate.net/profile/Jan-Bruetting-2/publication/338342629_Nothing_is_lost_nothing_is_created_everything_is_reused_structural_design_for_a_circular_economy/links/5e4581da299bf1cdb9283f37/Nothing-is-lost-nothing-is-created-everything-is-reused-structural-design-for-a-circular-economy.pdf https://doi.org/10.1016/j.conbuildmat.2016.07.075 https://www.afgc.asso.fr/app/uploads/2014/01/pp-549-556-Moreillon-1.pdf https://www.cervenka.cz/assets/files/atena-pdf/ATENA_Theory.pdf