Analysis of the variables that affect the strength of concrete with recycled aggregates from prefabricated pipes.

Chapter II – The construction and demolition waste, its recycling and reuse opportunities 478 ANALYSIS OF THE VARIABLES THAT AFFECT THE STRENGTH OF CONCRETE WITH RECYCLED AGGREGATES FROM PREFABRICATED PIPES 1 *Tarela, Ester; 1 Letelier, Viviana; 1 Osses, Rodrigo; 1 Cárdenas, Juan Pablo; 2 Moriconi, Giacomo 1 Departamento de Obras Civiles. Universidad de La Frontera. Av. Francisco Salazar, 01145, Temuco, Chile 2 Department of Scienze e Ingegneria della Materia, dell’Ambiente ed Urbanistica (SIAMU) Università Politecnica delle Marche, Italy e-mail: *[email protected] ABSTRACT The influence of several parameters in the compressive and flexural strength of concrete with recycled aggregates is analyzed. The concretes are dosed for a compressive strength of 30MPa. The recycled aggregates are obtained from debris of prefabricated concrete pipes with a compressive strength of 20MPa. Four variables are considered in this analysis: the percentage of natural course aggregates that are replaced by recycled ones; the amount of mortar adhered to the surface of the aggregates, that will be reduced applying mechanical abrasion processes instead of the chemical traditional ones; the maximum size of the replaced aggregates and the percentage of cement addition specified in the dosing. Taguchi’s statistical method is used considering three levels for each of the four parameters to determine their effects on the material behavior, minimizing the number of the experimental tests required. The tests are performed after curing the samples during 28 and 90 days. The three levels for each parameter are stablished after former studies and/or previous experimental tests. The results are quantified through the analysis of variance methodology (ANOVA), to determine which of the variables has higher effects in the final strengths. The results show that materials with equivalent compressive strengths to those of a control concrete can be obtained restricting the percentage of replaced aggregates to 30% and reducing the amount of adhered mortar through abrasion processes, when the samples have been cured for 28 days. The effect of the size of the recycled aggregates gains significance after 90 days of curing, while the amount of cement addition controls the flexural strength. These results indicate that materials with equivalent mechanical properties to those of a control concrete can be obtained by controlling several parameters, allowing the reuse of debris and reducing the amount of natural course aggregates needed significantly. Keywords: Recycled aggregates; Recycled concrete; Concrete mechanical properties; Compressive strength. Proceedings of the II International and IV National Congress on Sustainable Construction and Eco- Efficient Solutions 479 1.- Introduction Sustainable management is becoming an obligation in construction engineering and in other areas, due to a strong social demand. As a part of this global process, the reuse of concrete obtained from demolished constructions not only concerns to industrialized regions but responds to a worldwide need. Usually, there is almost a 5% rate of loss of product in the precast concrete industry due to defective materials or production faults. This material is rejected and considered a residual in the production plants. Nevertheless, the absence of foreign bodies such as ceramic material, lime, paint traces or rebar makes it a perfect candidate for recycling. Considering its reuse within the same production plant would save costs, not only in raw material but in residual disposure. Previous works on concretes that use recycled aggregates (0,0,0,0,0), show a significant difference in the mechanical behavior of these concretes when compared to conventional ones. One of the main causes of this loss of strength is the old mortar adhered to the surface of the recycled aggregates. Two interfaces have to be considered instead of one when using recycled aggregates; the usual one (old interface), between the old mortar and the aggregates, and a new one (new interface), between the aggregates, with the old adhered mortar, and the new cement matrix (0,0,0). Consequently, the quality and quantity of the adhered mortar will strongly influence the mechanical behavior of the recycled concrete, particularly, its mechanical strength (0,0). Alternatives to control the loss of strength have been analyzed lately. The first and most considered one is to limit the amount of the recycled aggregates. Authors (0,0,0, 0,0,0,0,0,0) prove that concrete properties are not significantly modified when the use of these aggregates is restricted to relatively low percentages. Therefore, nowadays, several European technic regulations allow the use of up to 30% of recycled aggregates in the production of structural concrete. Other ways to obtain a desirable strength have been studied. Adjusting the water/cement rate increasing the amount of cement would reduce the effect of the aggregates and, consequently, limit the loss of compressive strength (0,0,0). The size of the recycled aggregates is also involved in the new material strength. Fine recycled aggregates are not recommended because the smaller the size of the recycled aggregates the bigger the surface (old and new interfaces) involved, increasing the amount of old adhered mortar in the cement mix and the concrete production. Its characterization is complex, making it highly unpredictable, producing unexpected modifications in the properties of fresh and hardened concrete 0. The main issue in the use of recycled aggregates relays in the old adhered mortar. Studies have been performed to reduce de amount of mortar on the recycled aggregates. Ismail & Ramli 0 use different acids to minimize o even eliminate the loose mortar particles adhered to the aggregates surface. The physical properties of the concrete significantly improved, a reduction in the mortar improves the quality of the contact surface between the new cement matrix and the aggregates, increasing the mechanical strength of the concrete. Considering these precedents, in this study the optimum limits of the mentioned variables (recycled aggregates percentage, cement addition percentage, size of aggregates and amount of mortar) are evaluated, in aim to obtain similar strengths to the ones of the control concrete. 2.- Problem description 2.1.- Materials The coarse aggregates used in this investigation were obtained from precast concrete debris, specifically, from cylindrical tubes with an average load capacity of Chapter II – The construction and demolition waste, its recycling and reuse opportunities 480 20MPa after 28 days of curing, provided by a local company (Prefabricados Burzios, Temuco, Chile). The debris were grinded in a jaw crusher, obtaining aggregates with a maximum size of 30mm, that were later separated using mechanical sieving to use only the fraction between 6.3 and 19.3mm. Previous works reduce the amount of mortar by mainly applying chemicals dissolving adhered materials. Instead, a mechanical process is used here. A mechanical Los Angeles abrasion machine was used to reduce the adhered mortar through erosion. It will be quantified through the revolutions applied by the abrasion machine to the aggregates. The procedure is illustrated in Fig. 1. Fig. 1 “The debris is grinded in a jaw crusher. Example of the aggregates with no abrasion (0rev), 100rev and 300rev mechanical abrasion processes are applied”. In Table 1 real density and absorption values of these aggregates are shown. The natural aggregates have higher dry density values (Gn) than the recycled ones (Gr), because of the presence of adhered mortar in recycled aggregates. Therefore, density will increase when the amount of mortar is reduced. Sand G N G R G R100R G R300R Dry density (Kg/m 3 ) 2630 2666 2300 2330 2420 Water absorption (%) 1,4 1,7 6,2 5,3 4,0 Table 1 “Aggregates densities and absorption”. Gn Natural gravel; Gr: Recycled gravel without erosion; Gr100r: recycled gravel with erosion applying 100rev; Gr300r: recycled gravel with erosion applying 300rev. Sideways, the water absorption of recycled aggregates is almost 3.6 times bigger than the one of natural aggregates. Nevertheless, when the abrasion level increases, the absorption decreases to values around 3.1 and 2.4 times the one of the natural aggregates when 100rev and 300rev abrasion is applied respectively. The characteristics of pozzolanic cement, that was used to make the concretes, are shown in Table 2. Proceedings of the II International and IV National Congress on Sustainable Construction and Eco- Efficient Solutions 481 Values NCh 148 Of. 68 Requirements Physical and mechanical Characteristics Specific gravity (g/cm3) 2,8 Autoclave Expansion (%) 0,1 1,0 max Initial setting time (hh:mm) 02:40 01:00 min Final setting time (hh:mm) 03:40 12:00 max Compressive strength (kg/cm2) 3 days 280 7 days 320 180 min 28 days 410 250 min Chemical Characteristics Loss on ignition (%) 4,0 5,0 max S03 (%) 3,5 4,0 max Table 2 “Characteristics of pozzolanic cement”. 2.2.- Experimental method Experimental phase consisted in analyzing the differences in the compressive and tensile strength between concrete made with recycled aggregates and a control concrete made with natural aggregates. Several tests are performed to identify the most significant variables involved in the properties of concrete with recycled aggregates from prefabricated debris. According to the results obtained by previous authors and to the ranges stablished by different technic regulations, different values were set to be considered for each variable: - Recycled aggregates percentage: 20%, 30% and 40%. The choice was made considering different technic regulations that nowadays allow the use of recycled aggregates. - Cement mix percentage: 0%, 5% and 10%. Based on previous authors (0,0,0), but using national -Chilean- product. - Recycled aggregates size: 6.3-9.5mm, 9.5-12.5mm and 12.5-19.3mm. - Amount of mortar: 0rev, 100rev, and 300rev abrasion levels. The use of mechanical abrasion to reduce the old adhered mortar is new. The abrasion levels were chosen in base to experimental tests performed prior to the analysis. To analyze all the possible scenarios and figure out the influence of each variable all the combinations should be tested. This factorial design would imply developing 3 4 experimental tests. To reduce this number Taguchi 0 experimental method was used. Using a set of so called orthogonal arrays the number of different dosages required can be minimized while giving full information. In this case we consider 4 different independent variables with 3 level values each, so L9 orthogonal array is used, limiting the number of needed experimental tests to 9. The method assumes that the variables do not interact, meaning that the effect of the variation in one variable does not depend on the level settings of the other independent variables. The orthogonal arrays considered here are displayed in Table 3. The results will be analyzed applying ANOVA. Chapter II – The construction and demolition waste, its recycling and reuse opportunities 482 Recycled agg size (mm) Cement mix rate (%) Recyc agg (%) Abrasionlevel (rev) H1 6,3-9,5 0 20 0 H2 6,3-9,5 5 30 100 H3 6,3-9,5 10 40 300 H4 9,5-12,5 0 20 0 H5 9,5-12,5 5 30 100 H6 9,5-12,5 10 40 300 H7 12,5-19,3 0 20 0 H8 12,5-19,3 5 30 100 H9 12,5-19,3 10 40 300 Table 3 “Values of the parameters applied in each test for each of the 4 variables”. Worth to mention that all of the concretes are made using natural sand, only the natural course aggregates will be partially replaced by recycled coarse aggregates. The properties, slump, apparent density and air content of the fresh concretes are measured. Mechanical properties are evaluated through compression and flexural tests. Compression tests are performed following the NCh1037 and the NCh1017 regulations, using three cubic 15x15x15cm test specimens that will be cured during 28 and 90 days. Flexural ones are performed following the NCh1038, using three prismatic 15x15x50cm test specimens cured during 28 days. 3.- Methodology Different mixtures were designed using Faury 0 method calibrated for compressive strength of 30MPa after curing the sample during 28 days. A water/cement ratio of 0.42 was considered for control concrete and concrete without cement aggregates. When concretes with cement addition are considered this ratio will decrease between 5 and 10%. To calculate the final amount of water used in each mixture the recycled aggregates underwent a pre-saturation process reaching 80%. Based on previous works, this pre-saturation process helps sealing the pores in the recycled aggregates, what limits water exchange plus avoids problems on the workability of fresh concrete and the water/cement ratio (0,0). The values used in the dosage of the concretes, considering the pre stablished levels given by Taguchi’s method 0, and the results obtained when applying Abrams cone and air content are shown in Table 4. All the samples were elaborated in laboratory conditions and unmolded after 24 ± 2 hours after the mixture. Then, they were completely immersed in water with lime at 23 ± 3ºC during the curing time required (28 or 90 days). Proceedings of the II International and IV National Congress on Sustainable Construction and Eco- Efficient Solutions 483 Sand (Kg/m3) Gn 2,36 - 19,3 mm Gr Cement (Kg/m3) Water Slump (cm) Air content (%) 6,3 - 19,3 (mm) (Kg/m3) HC N 781,8 1024,2 - 382,3 162,6 2,5 1,1 H1 R 781,8 819,3 204,8 382,3 162,6 1,5 1,5 H2 R 781,8 717,0 307,3 401,4 162,6 1,5 1,3 H3 R 781,8 614,6 409,6 420,5 162,6 1,3 1,3 H4 R 781,8 819,3 204,8 382,3 162,6 4,4 1,7 H5 R 781,8 717,0 307,3 401,4 162,6 3,0 1,6 H6 R 781,8 614,6 409,6 420,5 162,6 3,2 1,1 H7 R 781,8 819,3 204,8 382,3 162,6 3,8 1,0 H8 R 781,8 717,0 307,3 401,4 162,6 5,5 1,4 H9 R 781,8 614,6 409,6 420,5 162,6 3,0 1,8 Table 4 “Tests in control concrete (N) and recycled (R) ones with each dosage”. Compression tests according to Chilean Standards NCh1037.Of2009 were carried out on each of the cubic specimens, which were tested at right angles to the position of casting. Splitting tension tests were carried out on prismatic specimens tested at right angles to the position of casting, in conformance with Chilean Standards NCh1038. Of2009. Three specimens are tested for each case and dosage, considering the mean value as the final result. 4.- Results and discussion The results for compressive and flexural strength obtained for each of the samples are shown in Table 5. Compressive strength (MPa) Flexural strength (MPa) Sample 28 days 90 days 28 days HC 34,8 40,2 5,3 H1 25,5 34,8 4,2 H2 30,5 39,2 5,0 H3 29,6 36,5 5,1 H4 29,7 35,1 4,9 H5 35,9 36,2 4,4 H6 25,0 32,4 4,6 H7 28,7 38,4 4,0 H8 35,7 43,1 4,2 H9 26,1 36,9 5,5 Table 5 “Results obtained for each specimen” 4.1.- Compressive strength In Fig. 2 the results for compressive strength after curing the samples for 28 days are displayed. Only two of the concrete samples with recycled aggregates, H5 and H8, reaching values of 35.9 and 35.7MPa respectively, equal or exceed the compressive strength of the control sample (34.8MPa). Both of them use 30% of recycled aggregates that suffered 100rev abrasion, and 5% cement addition. Despite all the concrete samples were designed to reach a minimum resistance of 30MPa, only 3 of Chapter II – The construction and demolition waste, its recycling and reuse opportunities 484 them, H2, H5 and H8, did exceed this value, the three of them containing 30% of recycled aggregates. Fig. 2 “Compressive strength after 28 days”. The minimum values of compressive strength were obtained for samples H1, H6 and H9, presenting a decay of between 25% and 28% of the value obtained for the control concrete. For these three samples a 40% of non-abrasion recycled aggregates was used with no cement addition. The other three samples (H3, H4 and H7) presented a decay of compressive strength of between 15% and 17% of the value obtained for the control sample. Fig. 3 shows the results for compressive strength after curing the samples for 90 days. After a long curing time all of the samples satisfy the design strength. Either ways, only H8 sample equals the compressive strength obtained by the control sample. H6 presents the minimum value, 19% lower than the control, and all of the others are between 5% and 13% lower than the control sample. Fig. 3 “Compressive strength after 90 days” These results reveal that concretes with recycled aggregates have a higher gaining rate after 90 days rather than after 28 (Fig. 4). This conclusion agrees with previous analysis (0,0) where the compressive strength was attributed to a stronger union in the interface between the cement and the coarse aggregates. Yu & Shui 0 postulates that during the cement hydration process some of the particles won’t react. When using recycled aggregates these particles are present in the surface adhered mortar, Proceedings of the II International and IV National Congress on Sustainable Construction and Eco- Efficient Solutions 485 so, when a new hydration process is performed the particles may be able to reactivate, producing a higher gaining compressive strength rate in the long term. Furthermore, considering that the hydrating and hardening processes are significantly slower when recycled aggregates are used (0,0,0) stronger bonds may be created between the cement and the coarse recycled aggregates, what may be able to compensate, up to certain degree, the negative effects of weaker aggregates when a long term curing is considered. Fig. 4 “Relative compressive strength after 28 days/ after 90 days”. 4.2.- Influence of studied parameters In this section the results obtained using Taguchi optimization method 0 are discussed to understand the influence of each parameter in the compressive strength after 28 and 90 days. Fig. 5 shows the influence of two parameters, the abrasion level and the percentage of cement added, in the final compressive strength. Both values evolve similarly, increasing their levels leads to bigger compressive strength values. The effect is more significant when the sample is cured during 28 days. An increase of cement addition from 0% to 10% results in an 8% more resistant sample, and an increase of abrasion from 0rev to 300rev results in a 14% more resistant material. After 90 days the cement addition has almost no more influence (about 1% increased compressive strength), what could be due to the original mechanical properties of the concrete controlling the maximum values that can be achieved in a new one with recycled aggregates 0. After 90 days the compressive strength increases a 6% when 300rev are applied. This behavior agrees with observations from Matias et al 0, where the use of angular plane particles is studied. They consider that this kind of particles may present layer stratification, what would limit the durability and compressive strength of the concrete. These effects may be controlled using rounder aggregates. When recycled aggregates are used, abrasion provokes a change in their shape, softening corners and ending with more rounded material. Furthermore, Ismael and Ramli [4] conclude that one of the main factors in low compressive strength in concrete with recycled aggregates is the mortar adhered to the aggregates; and Gonzalez & Etxeberria 0 point out that the use of recycled aggregates with higher absorption level affects the final properties of the concrete. The abrasion process, as shown in Table 1, leads to a decrease in these absorption levels because of the decrease of the adhered mortar to the recycled aggregates. Chapter II – The construction and demolition waste, its recycling and reuse opportunities 486 Fig. 5 “Compressive strength variation with abrasion level and cement addition after 28 and 90 days”. The effect of the amount of recycled aggregates and their minimum size is analyzed in Fig. 6. When the maximum size of the aggregates grows from 9.5mm to 19.3mm the compressive strength increases 8% when cured during 28 days and 7% for 90 days. The use of bigger recycled aggregates reduces the amount of old mortar in the mixture (interfaces), what reduces the weakened zones 0. When the amount of recycled aggregates is 40%, the compressive strength decays a 14% and a 12% of the values obtained for 30% of aggregates, after 28 and 90 days respectively. Kwan et al. 0 and Gonzalez & Etxeberria 0 conclude that this decay obeys to the low quality of the adhered mortar. Fig. 6 “Compressive strength variation with amount and size of recycled aggregates after 28 and 90 days”. Table 6 shows the results of the ANOVA obtained using Taguchi method.