1. INTRODUCTION
⌅Concrete is a material that exhibits excellent behavior over time. However, due to the sheer volume of reinforced concrete constructions, any durability issue in the material that requires repair, rehabilitation or replacement intervention of the affected elements entails major economic and social impact. The prevention of deterioration of concrete structures began to be considered in all concrete codes and regulations from the last decades of the 20th century, yet a fully satisfactory solution remains far off.
Most concrete codes establish prescriptive methods for the control of concrete durability (
1
(1)
Aleksander, M.; Thomas, M. (2015). Service life prediction and
performance testing - Current developments and practical applications. Cement and Concrete Research. 78, 155-164.
https://doi.org/10.1016/j.cemconres.2015.05.013
.
),
that regulate the type of components (cement type, aggregates,
additions) that can be used and sets limiting values for concrete mixes
(typically maximum water/cement ratio and minimum cement content).
Although this approach has been successful up to a certain point so far,
the main drawback is that the actual composition of the concrete
delivered on site cannot be determined through testing, meaning the
requirements can be verified in-situ only through documentary control.
To overcome this issue, certain codes have tried to introduce a performance-based specification for concrete durability. This approach is based in the measurement of relevant properties related to the transport properties of the concrete pore structure that can therefore be an indirect index of durability. Durability indicators used in practice might be physical parameters such as permeability to liquids or gases, water absorption, sorptivity, porosity or abrasion resistance, or chemical, physico-chemical and electro-chemical parameters such as diffusivity, conductivity, resistivity or migration coefficients of harmful species like chlorides.
Spanish regulatory codes have set the standard in terms of the treatment of concrete durability, both with prescriptive methods and through the designation of a durability index. The property designated by the Spanish government technical bodies for the indirect measurement of durability is the permeability to water determined by the measurement of the depth of penetration of water under pressure.
The test for determining this property on hardened concrete was standardized by the European Committee for Standardization (
2
(2) UNE-CEN (2020). UNE-EN 12390-8:2020. Testing hardened concrete. Part 8: Depth of penetration of water under pressure. Asociación Española de Normalización (UNE).
)
in 2001. The test is based on applying water at a pressure of 500 kPa
for 72 hours on one of the sides of a specimen of hardened concrete. The
specimen is then divided into two halves and the penetration depth of
the water’s front is measured. The standard specifies that, until the
time of testing, the specimens must be cured while completely submerged
in water.
This method was first used by Soroka at the beginning of the 1970s (
3
(3) Soroka, I. (1974). Permeability of lightweight-aggregate concrete. In Proceedings, Second International CIB/RILEM Symposium on Moisture Problems in Buildings, (11), 10-12, Bouwcentrum, Rotterdam, Netherlands.
)
in order to assess indirectly the durability of different concretes.
Throughout the 1980s, it was adopted by different researchers to
classify the quality of concrete from the point of view of its
durability (
4-6
(4) Diem, P. (1986). Determination of capillary absorptiveness of very dense concrete. Betonwerk und Fertigteil-Technik, 52(11), 719-724.
(5) McCurrich, L.H. (1986). Reduction in permeability and chloride diffusion with superplasticiser. Concrete, Journal of the Concrete Society, 9-10.
(6)
Baluch, M.H.; Al-Nour, L.A.R.; Azad, A.K.; Al-Mandil, M.Y. (1989).
Concrete deterioration due to thermal incompatibility of its components. Journal of Materials in Civil Engineering, 1(3), 105-119. Retrieved from
https://ascelibrary.org/doi/pdf/10.1061/(ASCE)0899-1561(1989)1%3A3(105)?casa_token=3qqtEmOjgSwAAAAA:H_A2EAfpgPaFlo1Em7LBk38H25y3mcgjBsNw2imjfYhE1f1M4AZ_fNqdJnqr4wzATRusJz0KU-0
.
)
due to its apparent simplicity in terms of implementation and
interpretation of results. Although the rationale for the method used by
these authors was the same, there were differences in their procedures
in the curing regime prior to the test (submerged in water, humid
chamber) similarly in the number of pressure steps and the magnitude of
the applied pressure.
With a view to reducing the variability
introduced by the above-mentioned factors, the method was standardized
for the first time in Germany in 1978 (
7
(7) DIN (1978). DIN 1048 Test Methods for Concrete. Deutsches Institut Fur Normung (DIN).
), and later included in an ISO standard (
8
(8) ISO (1983). ISO/DIS 7031. Concrete hardened-determination of the depth of penetration of water under pressure. International Standards Organization (ISO).
). In Spain, the adoption of the method is mainly due to the work carried out by the Centro de Estudios y Experimentación de Obras Públicas (CEDEX) in the 1990s. The laboratory work carried out by this organization (
9-11
(9)
Gálligo, J.M.; Rodríguez, F. (1989). El ensayo de penetración de agua
como método para el control de la durabilidad de un hormigón utilizado
en estructuras de puentes de carretera. Hormigón y Acero. 40(171), 143-152. Retrieved from
http://www.hormigonyacero.com/index.php/ache/article/view/825
.
(10) Gálligo, J.M.; Rodríguez, F. (1990). Permeabilidad del hormigón: Influencia de la dosificación y métodos de ensayo. Hormigón y Acero. 41(176), 157-164. Retrieved from
http://www.hormigonyacero.com/index.php/ache/article/view/761
.
(11)
Gálligo, J.M.; Rodríguez, F. (1995). Estudio comparativo sobre las
limitaciones de permeabilidad del hormigón en las normativas española y
europea. Hormigón y Acero. 46(195), 37-47. Retrieved from
http://www.hormigonyacero.com/index.php/ache/article/view/549
.
) resulted in the drafting of a Spanish standard in 1990 (
12
(12)
UNE (1990). UNE 83309:1990 EX. Ensayos de hormigón. Determinación de la
profundidad de penetración de agua bajo presión. Asociación Española de
Normalización (UNE).
). This standard features
significant differences vis-à-vis the current European version of the
test: it was established that, 24 hours before the test, the specimens
had to be dried in an oven at 50±5ºC; during the test, three pressure
steps were applied (100 kPa for 48 hours, 300 kPa and 700 kPa for 24
hours each) instead of a single pressure step; and finally, in addition
to the maximum water penetration depth, the average depth was also
determined.
The Structural Concrete Instruction (EHE) issued in 1998 (
13
(13)
Real Decreto 2661/1998, de 11 de diciembre, por el que se aprueba la
«Instrucción de Hormigón Estructural (EHE)». Boletín Oficial del Estado,
núm. 11, de 13 de enero de 1999, pp. 1525 a 1526. Retrieved from
https://www.boe.es/eli/es/rd/1998/12/11/2661
.
) passed new legislation in Spain that established the determination of the depth of water penetration according to (
12
(12)
UNE (1990). UNE 83309:1990 EX. Ensayos de hormigón. Determinación de la
profundidad de penetración de agua bajo presión. Asociación Española de
Normalización (UNE).
) as an experimental verification
of compliance with the durability requirements of minimum cement
content and maximum water/cement ratio. This verification had to be
carried out for concretes subjected to exposure classes with risk of
chloride corrosion, chemical attack, freeze-thaw or abrasion. The limits
considered sufficient to guarantee the impermeability of concrete were
50 mm for the maximum depth, and 30 mm for the average depth, for all
the mentioned environments and obtained as the average of three
specimens. For individual specimens, the maximum depth allowed is 65 mm
and the average depth is 40 mm).
The EHE was replaced in 2008 by the Structural Concrete Instruction EHE-08 (
14
(14)
Real Decreto 1247/2008, de 18 de julio, por el que se aprueba la
Instrucción de Hormigón Estructural (EHE-08). Boletín Oficial del
Estado, núm. 203, de 22 de agosto de 2008, pp. 35176 a 35178. Retrieved
from
https://www.boe.es/eli/es/rd/2008/07/18/1247
.
), and this in 2021 by the Structural Code (
15
(15)
Real Decreto 470/2021, de 29 de junio, por el que se aprueba el Código
Estructural. Boletín Oficial del Estado, núm. 190, de 10 de agosto de
2021, pp. 97664 a 99452. Retrieved from
https://www.boe.es/eli/es/rd/2021/06/29/470
.
).
The Structural Code maintains the determination of the penetration
depth as proof that the concrete has sufficient impermeability to
guarantee its durability during the structure’s service life. The
verification must be carried out using the single pressure step of the
current European standard, but the EHE-08 introduced the modification,
maintained in the Structural Code that, prior to carrying out the test,
the specimens must be subjected to a drying period of 72 hours in an
oven at 50±5ºC. The limits are the same as in the previous EHE
Instruction, applicable to the same types of environments, though
updated to the exposure classes defined in the European standard EN 206 (
16
(16) UNE-CEN (2018). UNE-EN 206 Hormigón. Especificaciones, prestaciones, producción y conformidad. Asociación Española de Normalización (UNE).
)
(corrosion by marine chlorides in aerial or submerged elements,
corrosion by non-marine chlorides, freeze-thaw, abrasion, and weak and
moderate chemical attack), with the exception of elements situated in
the tidal range, severe chemical attack or pre-stressed elements
subjected to moderate chemical attack, for which the limits are
established at 30 mm for the maximum depth determined as the average of
three specimens and 20 mm for the average depth (40 mm and 27 mm for
individual specimens, respectively).
Despite the method’s relative simplicity, doubts have been cast surrounding it due to the high variability of test results (
17
(17)
The Concrete Society (1988), Permeability Testing of Site Concrete-A
Review of Methods and Experience, Concrete Society Technical Report No.
31, p. 95.
), and the influence of a large number of factors on the results (
18
(18)
Revuelta, D.; Carballosa, P.; García-Calvo, J.L. (2018). Análisis de
factores que afectan al ensayo depenetración de agua bajo presión. Work
presented in V Congreso Nacional de Áridos, Santiago de Compostela, Spain. Retrieved from
https://digital.csic.es/bitstream/10261/214797/1/analispresi.pdf
). In this sense, it seems clear that the previous
treatment can have a significant bearing on the depth of the
penetration fronts for the same type of concrete, since in addition to
the mechanism of transport of water by permeability in a porous medium
due to the pressure gradient, the previous drying established by the
Structural Code can lead to an additional transport mechanism, namely
capillarity, which should lead to greater water penetration and can
contribute to variability. Another issue pointed out since the method
was used for the first time is the difficulty to correctly visualize the
penetration front, especially in specimens cured using total immersion
in water. The foregoing is why it is believed that the previous
treatment was introduced in order to ensure the correct visualization of
the front once the specimen is divided into two halves.
Another durability indicator that could be used to measure the durability performance of concrete is resistivity (
19
(19)
Polder, R., Andrade, C., Elsener, B., Vennesland, O., Gulikers, J.,
Weidert, R., Raupach, M. (2000). Test methods for on site measurement of
resistivity in concrete. Materials and Structures. 33, 603-611.
https://doi.org/10.1007/BF02480599
.
).
Electrical resistivity can be defined as the ratio between an applied
voltage and the resulting electrical current circulating in a unit cell
of a material. The resistivity of concrete mainly depends on its
porosity, water content and the chemical species present in the pore
solution. One of the first authors in developing a testing apparatus for
the measuring of resistivity in solid materials was Wenner (
20
(20) Wenner, F. (1915). A method for measuring earth resistivity. Bulletin of the Bureau of Standards, 12, 469-478.
).
This method uses a probe fitted with four electrodes and has been
successfully applied for the assessment of the corrosion of steel (
21
(21)
Gowers, K. R., Millard, S. G. (1999). Measurement of concrete
resistivity for assessment of corrosion severity of steel using Wenner
technique. ACI Materials Journal, 96(M66), 536-541. Retrieved from
http://www.comsiru.uct.ac.za/sites/default/files/image_tool/images/333/Events/Gowers_and_Millard_1999_-_Measurement_of_Concrete_Resistivity_for_Assessment_of_corrosion_severitty_usin.pdf
.
).
One of the advantages of the four-electrodes probe is its
non-destructiveness, which makes it especially suitable for the
evaluation of existing structures (
22
(22)
Castillo, A., Andrade, C., Martínez, I., Rebolledo, N.,
Fernández-Troyano, L., Ayuso, G., Cuervo, G., Junquera, J., Santana, C.,
Delgado, J. (2011). Evaluación y monitorización de la durabilidad de
las cubiertas del Hipódromo de la Zarzuela de Madrid. Informes De La Construcción. 63(524), 33-41.
https://doi.org/10.3989/ic10.058
.
,
23
(23)
Chinchón-Payá, S.; Torres, J.; Rebolledo, N.; Sánchez, J. (2021).
Evaluación del estado de elementos estructurales del Mercado de Legazpi:
Ataque por sulfatos al hormigón y corrosión de las armaduras. Informes de la Construcción. 73(561), e380.
https://doi.org/10.3989/ic.76737
.
).
A
further method for the determination of resistivity in concrete
consists in two electrode plates directly applied to a concrete specimen
(
24
(24)
Saleem, M.; Shameem, M.; Hussain, S.E.; Maslehuddin, M. (1996). Effect
of moisture, chloride and sulphate contamination on the electrical
resistivity of Portland cement concrete. Construction and Building Materials. 10(3), 209-214.
https://doi.org/10.1016/0950-0618(95)00078-X
.
,
25
(25)
Woelfl, W.R.L.; Spangler, M.B. (1980). The electrical resistivity of
concrete with the emphasis on the use of electrical resistance for
measuring moisture content. Cement, Concrete and Aggregates. 1(2), 64-67.
https://doi.org/10.1520/CCA10405J
.
).
This test, known as the direct method, determines the bulk electrical
resistivity of concrete, yielding fewer variable results than the
four-electrodes test, thus meaning it is considered as the reference
testing method for resistivity in concrete.
Standards for both test methods for the determination of resistivity in concrete have been issued in Spain (
26
(26)
UNE (2008). UNE 83988-1:2008. Durabilidad del hormigón. Métodos de
ensayo. Determinación de la resistividad eléctrica. Parte 1: Método
directo (método de referencia). Asociación Española de Normalización
(UNE).
,
27
(27)
UNE (2014). UNE 83988-2:2014. Durabilidad del hormigón. Métodos de
ensayo. Determinación de la resistividad eléctrica. Parte 2: Método de
las cuatro puntas o de Wenner. Asociación Española de Normalización
(UNE).
), and at the time of writing this article (end
of 2021), a work item is underway in CEN for the publication of a
European standard collecting the two methods. However, although several
publications try to promote the use of electrical resistivity as a
quality control tool (
28
(28)
Torres-Acosta, A.A.; Presuel-Moreno, F.; Andrade, C. (2019). Electrical
resistivity as durability index for concrete structures. ACI Materials Journal, 116(6), 245-253.
https://doi.org/10.14359/51718057
.
),
this technique is usually not included in the current concrete codes
and regulations like the recently published Spanish Structural Code.
As stated before, resistivity is mainly influenced by the moisture content of concrete. As such, in order to successfully characterize the concrete pore network connectivity, resistivity measurement should be performed on completely water-saturated specimens. For this reason, both testing methods establish that resistivity must be measured on concrete specimens cured in water
This work aims to evaluate the influence of three factors, the nature of the aggregate, the water/cement ratio and the previous treatment (curing submerged in water or previous drying in an oven as per the Structural Code procedure), in the results of penetration of water under pressure obtained on concretes manufactured with aggregates of a granitic, limestone and siliceous nature, common materials for the manufacture of concrete in large areas of the Iberian Peninsula. This interest is due to the difficulties highlighted by manufacturers of concrete prepared in order to meet the requirement established in the previous Instruction EHE-08, requirements that have been transposed to the new Structural Code.
Alongside the above, to aid in the search for a possible alternative index that could be used with guarantee to control the durability requirements, the work tries to evaluate the influence of two factors, aggregate and water/cement ratio, on the measurement of resistivity on saturated specimens, since it makes little sense to determine this property on dried specimens.
2. METHODOLOGY
⌅2.1. Materials
⌅For
the manufacture of the concrete used in this study, a cement resistant
to sulfates with the addition of ground granulated blast-furnace slag
was used, type III/A 42.5 N/SRC according to the Spanish standard UNE
80303-1 (
29
(29) UNE (2017). UNE 80303-1. Cementos con características adicionales. Parte 1: Cementos resistentes a los sulfatos. Asociación Española de Normalización (UNE).
), from the company Cementos Tudela Veguín, S.A.
The
aggregates were granitic, limestone and siliceous rocks, the first one
being supplied by the company General de Hormigones, S.A., and the last
two by Hanson Hispania. S.A.
Table 1
lists the main physical characteristics of the aggregates, density and
water absorption, determined following the UNE-EN 1097-6 standard (
30
(30) UNE-CEN (2014). UNE-EN
1097-6. Tests for mechanical and physical properties of aggregates -
Part 6: Determination of particle density and water absorption. Asociación Española de Normalización (UNE).
).
| Nature | Size | Density (g/cm3) | Water absorption (%) |
|---|---|---|---|
| Granitic | 0/4 | 2.65 | 0.09 |
| 6/12 | 2.56 | 1.16 | |
| 12/20 | 2.58 | 1.20 | |
| Limestone | 0/4 | 2.69 | 0.43 |
| 6/11 | 2.69 | 0.38 | |
| 11/22 | 2.69 | 0.38 | |
| Siliceous | 0/6 | 2.66 | 0.30 |
| 6/20 | 2.64 | 0.57 |
In
order to achieve a similar workability for each of the concrete batches
(between 40 and 100 mm of slump measured with the Abram’s cone
according to the standard UNE-EN 12350-2 (
31
(31) UNE-CEN (2020). UNE-EN 12350-2. Testing fresh concrete - Part 2: Slump test. Asociación Española de Normalización (UNE).
),
which corresponds to the interval defined as ‘soft’ consistency in the
Structural Code, including tolerances), the polyfunctional admixture ME
3850 was used, provided by BASF.
2.2. Experimental design
⌅To appraise the factors influencing the depth of water penetration under pressure, an experimental design featuring three factors (aggregate, water/cement ratio and treatment prior to the test) was performed, with three repetitions per combination of factors, resulting in a total of N = 36 results of testing.
For the water/cement ratio (w/c), two levels have been chosen, w/c = 0.5 and w/c = 0.7. The level w/c = 0.5 corresponds to a concrete that meets the requirement of maximum allowed w/c, established for one of the most unfavorable classes of exposure defined in the Structural Code that require the verification of the penetration depth. Hence, it was assumed that they would yield results within the limits of the Structural Code for both cases involving prior treatment, since the limits have been stablished for the case of previous drying. This case should be more unfavorable and therefore should yield higher penetration depth values. The second level chosen, w/c = 0.7, would correspond to concrete that does not meet the requirements, and therefore in theory should give results outside the limits established by the Code, at least for the treatment by drying during the previous 72 hours. The w/c ratio declared in Table 2 is the effective w/c ratio, i.e., not considering the water absorbed by the aggregates. The Spanish Structural Code establishes the w/c ratio limits for the total water, including absorbed water. Thus, the use of the limits with the effective w/c ratio is more unfavorable, so the results of the water penetration depth in case that the total w/c ratio should have been used would be lower.
| Component | w/c = 0.5 | w/c = 0.7 | w/c = 0.5 | w/c = 0.7 | w/c = 0.5 | w/c = 0.7 |
|---|---|---|---|---|---|---|
| kg/m3 | ||||||
| CEM III/A 42,5 N/SR | 300 | 300 | 300 | 300 | 300 | 300 |
| Water | 150 | 210 | 150 | 210 | 150 | 210 |
| Granitic sand 0/4 (49%) | 930 | 854 | -- | -- | -- | -- |
| Granitic aggregate 6/12 (15%) | 280 | 257 | -- | -- | -- | -- |
| Granitic aggregate 12/20 (36%) | 680 | 624 | -- | -- | -- | -- |
| Limestone sand 0/4 (50%) | -- | -- | 965 | 887 | -- | -- |
| Limestone aggregate 6/11 (15%) | -- | -- | 290 | 267 | -- | -- |
| Limestone aggregate 11/22 (35%) | -- | -- | 685 | 630 | -- | -- |
| Siliceous sand 0/6 (51%) | -- | -- | -- | -- | 990 | 911 |
| Siliceous aggregate 11/22 (49%) | -- | -- | -- | -- | 960 | 884 |
Therefore, two types of concrete were manufactured for each aggregate, with the same cement content of 300 kg/m3, and two different w/c ratios. The choice of cement content was made to be consistent with the durability requirement of minimum content in the Structural Code for the most unfavorable environments, and which require the verification by the water penetration test, considering that the lower cement content would be the most unfavorable case, since there are ambient classes that require higher cement contents (325-350 kg/m3). To maintain the cement contents in each of the formulas, in order to isolate this factor, the volume of the aggregates was adjusted so that the same proportions were mutually maintained in volume. The six nominal dosages used are listed in Table 2 .
Regarding the treatment prior to the test, two levels have been compared: on the one hand, the one given in the UNE-EN 12390-8 standard, conservation of the test specimens under water at 20±2°C until the moment of the test; and on the other hand, after the curing in water for 28 days, drying prior to the test for 72 hours in a forced air ventilation oven at 50 ± 5ºC.
For the assessment of resistivity, only two factors, aggregate and w/c, are analyzed. Only limestone and siliceous aggregates were considered, since the resistivity measurement equipment wasn’t available when the granitic batches were manufactured. Since one reading over a concrete specimen is considered as an individual result, a total of N = 42 results of testing were obtained.
To comply with
the number of tests provided for in the experimental design, a total of
18 batches were manufactured with the components described previously, 3
batches for each of the mixtures in
Table 2
.
For each batch, 7 cylindrical specimens measuring Ø15x30 cm and 2
cylindrical specimens measuring Ø10x 20 cm, in line with the UNE-EN
12390-2 standard (
32
(32) UNE-CEN (2020). UNE-EN 12390-2. Testing hardened concrete - Part 2: Making and curing specimens for strength tests. Asociación Española de Normalización (UNE).
),
were manufactured. All the specimens were kept for at least 28 days
immersed in water at 20±2ºC. After the curing period, resistivity was
measured according to the standard UNE 83988-1 on the seven Ø15x30 cm
specimens per batch at the age of 28 days in saturated conditions.
Three Ø15x30 cm specimens from each batch were directly subjected to the test of water penetration under pressure according to the standard UNE-EN 12390-8. Three other Ø15x30 cm specimens were subjected to the drying stipulated in the Structural Code (72 hours in an oven at 50 ± 5ºC). One Ø15x30 cm was stored as reserve. For each test, the maximum and average depth of penetration of water under pressure were determined as the average of the three specimens.
The two Ø10x20 cm specimens were tested under compression according to the standard UNE-EN 12390-3 (
33
(33) UNE-CEN (2020). UNE-EN 12390-3. Testing hardened concrete - Part 3: Compressive strength of test specimens. Asociación Española de Normalización (UNE).
), at 28 days of age.
3. RESULTS AND DISCUSSION
⌅3.1. Compressive strength
⌅Table 3 displays the compressive strength results at the age of 28 days obtained on the two Ø10x20 cm cylindrical test specimens, for each of the 6 mixes by type of aggregate. The results are consistent with what expected forecasts, that is, an appreciable decrease in compressive strength when increasing the water/cement ratio from 0.5 to 0.7 (34%, 43% and 37% drop for granitic, limestone and siliceous aggregates respectively).
| Water/cement ratio | Granite | Limestone | Siliceous |
|---|---|---|---|
| 0.5 | 46.4 | 48.5 | 43.0 (*) |
| 46.9 | 55.5 | 40.7 | |
| 42.3 | 48.9 | 36.9 | |
| Mean | 45.2 | 51.0 | 40.2 |
| Std. Deviation | 2.5 | 3.9 | 3.1 |
| 0.7 | 26,9 | 31,3 | 26,9 |
| 33,8 | 27,9 | 29,5 | |
| 28,7 | 28,7 | 19,0 | |
| Mean | 29.8 | 29.3 | 25.1 |
| Std. Deviation | 3.6 | 1.8 | 5.5 |
(*) Result of a single specimen due to an unsatisfactory failure
3.2. Water penetration depth
⌅Figure 1 shows the box-and-whisker plots of the results obtained for the maximum and mean depth of water penetration. It is striking that the penetration, obtained as the average of three specimens for each of the combinations of factors, is in all cases below the limits established in the Spanish Structural Code for mass or reinforced concrete elements. This fact, which would be expected in concrete with a w/c ratio = 0.5, is nevertheless contradictory with what is expected for concrete with a w/c ratio = 0.7, since the purpose of the test is precisely to detect concretes that do not comply with the composition limits established in the Structural Code for durability reasons. Regarding the individual specimens, only one of them (out of a total of 54 individual specimens with w/c = 0.7) resulted in a value outside the limits (granite aggregate, w/c = 0.7, oven-dry, with a maximum penetration depth of 75 mm). Another observation that emerges from Figure 1 is the overly low average penetration values obtained in concretes made with limestone and siliceous aggregates, for both w/c ratios.
As the literature (
17
(17)
The Concrete Society (1988), Permeability Testing of Site Concrete-A
Review of Methods and Experience, Concrete Society Technical Report No.
31, p. 95.
) has pointed out, a large dispersion of the
results is observed, mainly in relation to the maximum depth of
penetration. This high dispersion is highly influenced by the presence
of outlier values in individual specimens. Although they can be detected
by any statistical analysis of outliers, it was decided by the authors
to keep them in the calculation of the results, since no experimental
reason was found to consider its rejection beyond the variability of the
test itself. In general, the previous oven drying treatment increases
the dispersion compared to the specimens that were stored submerged in
water until the time of the test, although it is true that the drying
treatment gives, as expected, values above the specimens only cured in
water.
Tables 4 and 5 show the analysis of factorial variance (ANOVA) performed on the results obtained for penetration of water under pressure. It follows that the three factors composing the analysis, aggregate, water/cement ratio and previous preparation, are statistically significant, since the three factors present p-values lower than 0.05 (column “Sig.”) for both measurements, maximum and mean penetration depth. This observation implies that the separate influence of the three factors is appreciable from a statistical perspective, which means that changes in the w/c ratio influence the results, regardless of the type of aggregate and the treatment. Figure 2 represents the mean values obtained for each factor, independently of the rest.
| Test of Between-Subjects Effects. Dependent Variable: Maximum penetration depth | ||||||
|---|---|---|---|---|---|---|
| Source | Type III sum of squares | Degrees of freedom | Mean square | F | Sig. | Observed powerb |
| Corrected Model | 4141.333a | 9 | 460.148 | 7.085 | .000 | 1.000 |
| Intercept | 11664.000 | 1 | 11664.000 | 179.588 | .000 | 1.000 |
| Aggregate | 1322.000 | 2 | 661.000 | 10.177 | .001 | .975 |
| W/c ratio | 529.000 | 1 | 529.000 | 8.145 | .008 | .784 |
| Preparation | 1965.444 | 1 | 1965.444 | 30.261 | .000 | 1.000 |
| W/c ratio*Preparation | 16.000 | 1 | 16.000 | .246 | .624 | .077 |
| Aggregate*Preparation | 108.222 | 2 | 54.111 | .833 | .446 | .177 |
| Aggregate*W/c ratio | 200.667 | 2 | 100.333 | 1.545 | .232 | .298 |
| Error | 1688.667 | 26 | 64.949 | |||
| Total | 17494.000 | 36 | ||||
| Corrected total | 5830.000 | 35 | ||||
| a. R squared = 0.710 (Adjusted R squared = 0.610) b. Computed using α = 0.05 |
||||||
| Test of Between-Subjects Effects. Dependent Variable: Mean penetration depth | ||||||
|---|---|---|---|---|---|---|
| Source | Type III sum of squares | Degrees of freedom | Mean square | F | Sig. | Observed powerb |
| Corrected Model | 1507.139a | 9 | 167.460 | 19.746 | .000 | 1.000 |
| Intercept | 667.361 | 1 | 667.361 | 78.691 | .000 | 1.000 |
| Aggregate | 932.056 | 2 | 466.028 | 54.951 | .000 | 1.000 |
| W/c ratio | 173.361 | 1 | 173.361 | 20.442 | .000 | .992 |
| Preparation | 78.028 | 1 | 78.028 | 9.201 | .005 | .831 |
| W/c ratio*Preparation | .250 | 1 | .250 | .029 | .865 | .053 |
| Aggregate*Preparation | 62.722 | 2 | 31.361 | 3.698 | .039 | .627 |
| Aggregate*W/c ratio | 260.722 | 2 | 130.361 | 15.371 | .000 | .998 |
| Error | 220.500 | 26 | 8.481 | |||
| Total | 2395.000 | 36 | ||||
| Corrected total | 1727.639 | 35 | ||||
| a, R squared = 0,872 (Adjusted R squared = 0.828) b. Computed using α = 0.05 |
||||||
Regarding the type of aggregate, it is observed that the difference is mainly attributed to the use of granitic aggregates, since the values of maximum penetration are similar in concretes manufactured with siliceous and limestone aggregates. The average penetration value in these two cases is practically negligible, and only granite displays a considerable value.
An increase in both maximum and mean water penetration depth values is observed leading to an augmentation in the w/c ratio, which confirms the hypothesis by which this test was established as an instrument for controlling variations in the composition of concrete in terms of its water content established since the Instruction EHE; and vice versa: regardless of the aggregate and the w/c ratio, the previous drying produces variation in the results: in this case, the introduction of drying also increases the water penetration depth value. The observed statistical power for the three factors considered independently is greater than 0.80, representing a 95% confidence level.
With regard to the interaction between factors, the ANOVA indicates that there would not be a combined effect between paired factors (w/c ratio*preparation, aggregate*preparation or aggregate*w/c ratio) statistically significant for the case of maximum penetration depth, since F-values are reduced in all cases and the p-values are above 0.05. However, this result must be interpreted cautiously, since the observed power for the interaction term is low. For the mean penetration depth, there would be an interaction when introducing the aggregate factor. Again, this interpretation must be viewed with caution, due in this case to the reduced value that this measure has for the case of limestone and siliceous aggregates, which introduces statistical significance, yet without a direct practical meaning.
3.3. Resistivity
⌅Figure 3 shows the box-and-whisker plot of the results obtained for the resistivity at the age of 28 days. As expected, an increase in the w/c ratio decreases resistivity values, since a larger pore network entails a larger content of water in the concrete mass and therefore, since water is the main conductive medium, ions will move more easily.
With regard to water penetration, a widespread dispersion of results is also observed, especially for the case of siliceous aggregate. However, this variability does not confound the differences introduced by changes in the w/c ratio, as can also be verified through the results of the average values grouped by individual factors ( Figure 4 ). Table 6 displays the ANOVA performed on the results obtained for resistivity at 28 days. In this case, only the w/c ratio is statistically significant (Sig.< 0.05). This means that the test is able to detect differences between concretes with w/c ratios of 0.5 and 0.7. The aggregate factor is not significant, which can be due to the great variability showed, especially for the siliceous aggregates (hence the low observed statistical power for this factor). Yet, even with this level of variability, the resistivity seems to be able to detect the differences in the water content, which would make it adequate as an alternative test for the determination of a durability index for its control.
| Test of Between-Subjects Effects. Dependent Variable: Resistivity at 28 days | ||||||
|---|---|---|---|---|---|---|
| Source | Type III sum of squares | Degrees of freedom | Mean square | F | Sig. | Observed powerb |
| Corrected Model | 239551.960a | 3 | 79850.653 | 198.106 | .000 | 1.000 |
| Intercept | 10885312.600 | 1 | 10885312.600 | 27005.922 | .000 | 1.000 |
| W/c ratio | 198090.102 | 1 | 198090.102 | 491.452 | .000 | 1.000 |
| Aggregate | 871.183 | 1 | 871.183 | 2.161 | .146 | 0.306 |
| W/c ratio*Aggregate | 40857.038 | 1 | 40857.038 | 101.364 | .000 | 1.000 |
| Error | 31439.563 | 78 | 403.071 | |||
| Total | 11091228.250 | 82 | ||||
| Corrected total | 270991.523 | 81 | ||||
| a. R squared = 0.884 (Adjusted R squared = 0.880) b. Computed using α = 0.05 |
||||||
Regarding the interaction between factors, the ANOVA indicates that there exists interaction between the w/c ratio and the aggregate (F = 101.364; p-value = 0.00) and with a good statistical power (β = 1.00). This fact is corroborated through Figure 5 , where it can be noticed that the lines do not follow a parallel path. The drop in resistivity is steeper in the case of the siliceous aggregate, which can be attributed to temperature variations or the different electrical properties of the aggregates.
4. CONCLUSIONS
⌅The water penetration test under pressure carried out on concretes manufactured from granitic, limestone and siliceous aggregates proved to be sensitive to changes both in the water/cement ratio and to the treatment by prior oven-drying established in the Spanish concrete codes. The results obtained on limestone and siliceous aggregates are similar, though not the values measured on granite aggregates, which present differences with respect to the other two types. The results showed a great dispersion, especially in the cases in which the specimens were previously dried according to the procedure stipulated in the Structural Code. Surprisingly, the test did not result in water penetration depth values outside the limits established in the Code, even when testing concretes with a water/cement ratio higher than the maximum water content limits established in the Code and when they were subjected to pre-drying.
The power observed in the test campaign does not allow to draw sufficiently conclusive statistically significant inferences regarding the effect of the interactions between various factors. Likewise, the difference in behavior observed between concretes with different types of aggregate (especially granitic) or the obtaining of penetration values below those expected in the regulation, suggests it is advisable to carry out future studies to bolster the results, increasing for example the power of the campaign by increasing the number of samples, extending it to a greater number of levels per factor or studying the influence of other factors such as the content of fines or the possible presence of entrapped air, among others.
Regarding resistivity, the results carried out on concretes manufactured from limestone and siliceous aggregates show that this test is also sensitive to changes in water/cement ratio. Results were similar for both types of aggregates. The variability of the results is also large, especially for siliceous aggregates, yet this variability does not confound the changes in water/cement ratio.
This study similarly stresses the
danger of using indirect relationships between various properties of
concrete such as compressive strength and indirect indexes of
durability, correlated in occasions but without causality (
34
(34)
Lee, J.; Harada, K. (2021). A simple method for estimation of
permeability of concrete from the compressive strength and pore size
distribution based on literature survey. Journal of Asian Architecture and Building Engineering.
https://doi.org/10.1080/13467581.2021.2008943
.
).
This type of relationship can lead to incorrect deductions implying,
for example, that concretes with higher compression strength are less
permeable. In this work, granitic concrete showed higher compressive
strength than siliceous concrete, while its water permeability was
clearly higher.