Scholarly article on topic 'Protected Paste Volume (PPV) as a parameter linking the air-pore structure in concrete with the frost resistance results'

Protected Paste Volume (PPV) as a parameter linking the air-pore structure in concrete with the frost resistance results Academic research paper on "Civil engineering"

CC BY
0
0
Share paper
Academic journal
Construction and Building Materials
OECD Field of science
Keywords
{Concrete / "Freeze–thaw resistance" / "Air-void structure" / "PPV factor" / "Spacing factor"}

Abstract of research paper on Civil engineering, author of scientific article — Jerzy Wawrzeńczyk, Wioletta Kozak

Abstract A new description is proposed of the air void structure in air-entrained concretes using a PPV (Protected Paste Volume) parameter. The PPV is determined as a ratio of the paste area protected by air voids to the total paste area, obtained from the 2D surface measurements. This approach, although constitutes a development of the Philleo concept, takes into account the number, size and distribution of not only air voids but also aggregate grains, often disregarded by other researchers. The obtained results show that the standard spacing factor L⩽0.20mm fails to guarantee that 100 percent of the paste volume is protected against the harmful effects of frost, which is the basic assumption in the Powers spacing factor model. The preliminary verification was carried out of the relationships between the parameters L and PPV and the results of freeze–thaw experiments. According to the authors, at this stage of the study the PPV provides better correlation with the frost resistance results than the standard spacing factor L. Unlike other indicators representing a statistical estimation of the average spacing of air bubbles, the PPV much better reflects the spatial displacement of the air pores in concrete and can thus be considered as an improved factor of air voids spatial dispersion.

Academic research paper on topic "Protected Paste Volume (PPV) as a parameter linking the air-pore structure in concrete with the frost resistance results"

ELSEVIER

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Protected Paste Volume (PPV) as a parameter linking the air-pore .BicrossMark

structure in concrete with the frost resistance results

Jerzy Wawrzenczyk *, Wioletta Kozak

Kielce University of Technology, al. Tysictclecia Panstwa Polskiego 7, Kielce 25-314, Poland

HIGHLIGHTS

• A novel method to describe an air void structure in A-E concretes is proposed.

• PPV is a ratio of the paste area protected by air voids to the total paste area.

• PPV can be treated as a better factor of air void spatial dispersion.

• Concrete can be frost resistant with the PPV higher than 80%.

• Concrete may be non-frost resistant when PPV is lower than 60%.

ARTICLE INFO

ABSTRACT

Article history:

Received 16 December 2015

Received in revised form 12 February 2016

Accepted 25 February 2016

Available online 4 March 2016

Keywords: Concrete

Freeze-thaw resistance Air-void structure PPV factor Spacing factor

A new description is proposed of the air void structure in air-entrained concretes using a PPV (Protected Paste Volume) parameter. The PPV is determined as a ratio of the paste area protected by air voids to the total paste area, obtained from the 2D surface measurements. This approach, although constitutes a development of the Philleo concept, takes into account the number, size and distribution of not only air voids but also aggregate grains, often disregarded by other researchers. The obtained results show that the standard spacing factor L 6 0.20 mm fails to guarantee that 100 percent of the paste volume is protected against the harmful effects of frost, which is the basic assumption in the Powers spacing factor model. The preliminary verification was carried out of the relationships between the parameters L and PPV and the results of freeze-thaw experiments. According to the authors, at this stage of the study the PPV provides better correlation with the frost resistance results than the standard spacing factor L. Unlike other indicators representing a statistical estimation of the average spacing of air bubbles, the PPV much better reflects the spatial displacement of the air pores in concrete and can thus be considered as an improved factor of air voids spatial dispersion.

© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://

creativecommons.org/licenses/by/4.0/).

1. Introduction

Air entrainment is a basic method of protecting concrete against the destructive action of frost, water and de-icing salts. Although this method has been known for over 70 years, there are still significant problems with obtaining a time-stable system of very small closely spaced air bubbles at the lowest possible total air content [1-8]. Information about the parameters that determine the quality of the air void structure is necessary to maintain its efficiency. The discussions found in the literature concentrate on the selection of adequate parameters, their critical values and correlation with the results of frost resistance tests. Currently, the quality of the air void system in concrete is quantified by the standard

* Corresponding author.

spacing factor L, the content of micropores A300, the total air content A and the specific surface a [16]. The principles of the method used to determine the spacing factor L were formulated by Powers [26]. His model assumes that all bubbles are monosized and uniformly spaced throughout the paste phase. In fact, however, the entrained air voids cover a wide range of sizes, have variable spatial distribution, merge and form groups in the cement paste. The standard test procedure [16] to analyse the cross-sectional area of concrete uses the linear traverse method. Considering the development of automated image analysis, the improvement of methods for describing the air pore structure in concrete seems inevitable and necessary (Fig. 1).

A renewed interest in assessing air entrainment quality has been observed in the works of a number of authors. The method described by Philleo [9] allows estimating the volume fraction of potentially protected cement paste within a distance S from the

http://dx.doi.org/10.1016/j.conbuildmat.2016.02.196 0950-0618/© 2016 The Authors. Published by Elsevier Ltd.

This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

3D - volume

• 1 j| # fJE>

''ms raft fir i

2D - oreo

1D - lines

Vh r-T*

-O- J -G- —

OD - points

í ++1

Fig. 1. Methods of measuring air-void characteristics [19].

edge of the air void. Attiogbe [10] proposed a parameter F, which is also the volume of potentially protected paste, but he indicates that the gamma distribution is more suitable. Both approaches are based on the same results of the standard chord length measurements, which are treated as purely statistical, and the difference is only in the method of analysis [16]. Zatocha and Kasperkiewicz [13] proposed the use of a high-resolution flatbed scanner with simultaneous measurement to separate the cement paste and the air voids in the image of a concrete surface. The procedure they presented allows an effective quantitative analysis of the cement paste areas protected against the frost action by the air voids, from which appropriate conclusions can be drawn. Elsen et al. [11] proposed an alternative spacing factor, Lnew, which is the mean 3D distance between air voids, computed using the number of air bubbles per unit volume of concrete, established in an advanced stereological analysis. Mayercsik et al. [12] offered a method that utilizes the linear-path function to ascertain a probability density function for the three-dimensional size distribution of entrained air voids directly from the plane polished section of hardened concrete. The three-dimensional structures are determined using two-dimensional data. Soroushian et al. [14] analyzed microstructural images of concrete captured by environmental scanning electron microscopy and optical fluorescence microscopy, where a 3D structure of concrete is generated from a perpendicular section. In Aligizaki et al. [15], the air content and size distribution of air voids in hardened cement paste were determined from stereoscopic microscope images using the section-analysis method.

A different view was presented by Hasholt [24], who claims that not the spacing of air voids but the total surface area of air voids better describes the relationship between the air pore structure and resistance to salt frost scaling.

The procedure for determining the standard parameters of airpore structure [16] has remained unchanged since the 40 s of the twentieth century. According to the literature, the classification of concrete in terms of its frost resistance can be problematic when based exclusively on the L parameter [17,18,24]. Lazniewska-Pie karczyk [18], while analyzing the literature data came to the conclusion that the spacing factor L depends on a number of factors, including the type of concrete, microsilica content, w/c (or w/b) ratio and the conditions under which the concrete is subjected to freezing. Critical values of the standard spacing factor, Lcryt, may range from 0.25 mm for concretes made with silica fume, frozen in water and with the water to binder ratio w/b = 0.50, to 0.50 mm for high-performance concretes with the w/c ratio in the range from 0.30 to 0.35 and 0.90 mm for concretes containing 6% of silica fume and with the w/b = 0.33.

In this paper, the authors propose the two-dimensional description of the air void structure in air-entrained concretes using a PPV (Protected Paste Volume) parameter. The PPV is a ratio of the paste area located at a distance of 0.20 mm from the edge of the nearest air void (protected paste area) to the total paste area. This

approach takes into account not only the number, size and distribution of air voids but also aggregate grains. Previous studies conducted on a numerical model [20,21] and model concretes [25] showed a significant influence of the air-paste-aggregate system on the air entrainment structure. The paper presents the procedure, results and the comparison of L and PPV factors against the frost resistance results.

2. Materials and methods

The research programme was designed to deal with two important tasks: developing a parameter better describing the air void system in concretes compared with the standard approach and determining the correlation between this parameter and frost resistance results. Air entrainment of concrete is a random process and the design and manufacture of the samples with pre-established parameters of the air pore structure is a difficult task. Considering a relatively large number of pores (850 chords recorded on the traverse with a length of 2400 mm gives L 6 0.20 mm), non-uniformly spaced in the concrete, the biggest differences were expected between the L and PPV descriptions of the air pore structure.

From the perspective of the work objectives, the most interesting concretes are those with a higher potential for problems relating to a uniform distribution of air voids in the cement paste (clustering, segregation of air voids) and/or frost resistance. Contrary to the spacing factor L, the PPV takes into account the real size and placement of air voids and aggregate grains and therefore, in such "special" concretes the difference between the estimations of these parameters is expected to be the largest.

The test samples were chosen from three groups: with very "good", very "bad" and "intermediate" values of air void structure parameters and frost resistance results. The CEM II cement, silica fume (SF) (10% by mass of cement) and basalt aggregate, widely regarded as frost-resistant aggregate were used in all the concretes. Two types of air entrainment were used: traditional air entraining admixture (AEA) or polymer microspheres (MSF) with a nominal diameter of 40 im. The concrete mix compositions are summarized in Table 1.

The examination of the air pore structure was carried out by two methods. The first method was the standard measurement according to PN EN 480-11 [16]. The second, innovative method (PPV) was more complicated and consisted in extracting air pores and cement paste area. The samples, with dimensions 10 x 12 x 3 cm, were cut out from the cubes, ground and polished to provide a smooth surface for microscopic observations. Applying the fluorescent powder as a filling for air voids allowed the use of a UV light for better contrasting and facilitating the separation of the air voids from the background. Owing to proper preparation and special lighting, it was possible to separate the aggregate grains. The images were made using the system composed of a Nikon SMZ 800 microscope, a CCD camera and the Prior measuring table. Two-dimensional analysis was conducted using the ImagePro Plus software [22], which allows computing the ratio of the objects to the whole

Table 1

The components of A+H concretes.

Concrete Cement type w/b ratio Binder [kg/ m3] Air content [%] Volume of paste [%] The AEA or MSF dosage [% mb]

A CEM II/B- 0.42 375 3.0 27.2 -

B M (V-LL) 367 4.1 26.6 AEA 0.20

C 42,5R 356 5.9 25.9 AEA 0.35

D 355 7.0 25.8 AEA 0.50

E 464 - 33.0 MSF 0.05

F 382 - 27.7 MSF 0.15

G CEM II/A- 0.40 404 - 29.5 MSF 0.35

H S 42,5N 0.42 396 - 29.0 MSF 1.11

area of the image. The preparation of a single image across the surface with a resolution of 3.4 im/px was impossible because of the large file size and difficulties in further processing of the image. Therefore three 30 x 80 mm images were taken of each polished section. The images were subjected to correction associated with the preparation of the image for analysis.

To determine the PPV factor, the protected paste area (PpA) and the total surface area of the paste (PA) have to be established.

In stage I, the total area of the paste (PA) (Fig. 2c) is determined by subtracting air void phase (Fig. 2a) and aggregate grain phase (Fig. 2b) from the area of the image.

In stage II, the protected paste area (PpA) is determined. For each air void, the center of mass coordinates (x, y) is found, along with equivalent diameter fixed as the diameter of the circle with an area equal to the cross-sectional area of the measured object. The protected paste shell with a thickness of 0.20 mm is drawn (Fig. 3b) around each air void (Fig. 3a). The areas occupied by the aggregate grains (Fig. 3c) are subtracted from the protected paste image (Fig. 3b). The resultant image (Fig. 3d) reflects the actual area of the protected paste (PpA). Finally, the PPV is calculated as a PpA/PA ratio.

The frost resistance test was conducted according to the modified ASTM C 666A method [23] using the beam specimens with dimensions 8 x 8 x 35 cm, frozen and thawed in water. Generally it is regarded that the method of ASTM C 666 A procedure is very severe, because tests are carried out on samples at the extremely high degree of saturation (the samples are subjected to freeze-thaw cycles at 14 days of water curing). The modification allowing the samples to dry before freeze-thaw cycles consisted of curing the samples in water for 7 days, then under laboratory conditions for 21 days, and after another 7 days of soaking in water they were subjected to freeze-thaw cycles. A similar way of sample preparation for frost resistance testing is used in other European standards [27,28]. The process of the concrete destruction due to freeze-thaw cycles was assessed by measuring the changes in mass and linear dimensions of the samples after 300 cycles. Concrete was considered as non frost resistant when samples expansion at any time of the test exceeded 1 mm/m.

3. Results and analysis

The air void structure parameters and the internal frost-induced cracking resistance results are shown in Table 2. The frost resistance test results, expressed as changes in mass, are presented in Fig. 4 and as the linear deformation in Fig. 5.

Standard measurements indicated that the air content was A = 2.38^6.05%, with the micropore content A300 = 0.29^2.14%, while the spacing factor was in the range L = 0.06^0.74 mm. The L factor was less than or equal to 0.20 mm only for two concretes (G, H).

The 2D measurements showed the air content A2D = 2.19^6.98% with the A3Do = 0.21^1.74%. The PPV varied from 13.0 to 89.9%. The highest PPV values were obtained for concretes D and H. Contrary to the assumption made in the Powers' model, these results indicate that the spacing factor L 6 0.20 mm does not ensure that 100% of the cement paste volume is protected from the harmful effects of frost.

Differences are observed in estimating the total air content (A and A2D). Probably they are due to the methodological assumptions of both methods and because of certain errors associated with measurements.

Two concretes (D and H) demonstrated very good frost resistance after 300 freeze-thaw cycles. The remaining concrete proved to be non-frost resistant - significant changes in mass and linear deformation increments were recorded. The changes in mass indicate poor scaling resistance of concrete surfaces.

Only one of the two concretes with L 6 0.20 mm (H) and the two concretes with the PPV higher than 80% (D, H) were found to be frost-resistant. These results indicate that the PPV provides a better correlation with the frost resistance results.

The comparison of the results for two pairs of concrete, D, F and D, H are quite interesting (Fig. 10). The standard spacing factors in the D and F concretes are similar, whereas the values of PPV vary. In the case of the D and H concretes, spacing factor values vary, while the PPVs are similar. Analysis of the quantity and distribution of the chords (1D) and the pore diameters (2D), presented in Figs. 6-9, explains these dependencies.

For concretes D and F, this is due to a similar number of chords up to 300 im (nD00 and n300) (Fig. 6) and varied numbers of pores up to 300 im in diameter (nD00 and np00) (Fig. 7). The difference in the numbers of pores affects not only the PPV but also the frost resistance results. The values of L indicate that both concretes may be considered as non-frost resistant. However, the D concrete is frost resistant in spite of the spacing factor greater than 0.20 mm.

In D and H concretes the values of standard spacing factor L are quite different. This is due to the fact that the number of chords registered for concrete H (nH00) is 3.5 times higher than that for concrete D (nD00) (Fig. 8). The values of PPV are similar despite the fact that the number of pores recorded for H concrete (nH00) is 2 times higher than that for the D concrete (nD00) (Fig. 9). Recording more than 15,000 micropores within an area of 72 cm2 constitutes a very good result, confirmed by the frost resistance results for both concretes. Only 10% difference between the PPV values with nearly 15,000 additional air pores in the H concrete indicates that these pores form groups in the cement paste, the shells of the protected paste overlap and the presence of this additional air bubbles is ineffective in terms of frost resistance.

According to the authors, the PPV can be treated as an assessment factor of the actual spatial dispersion of air pores. The parameters proposed by Attiogbe, Philleo or Powers can only be treated as a statistical evaluation of the distance between air voids, not their real spatial dispersion.

Fig. 10 shows the relationship between the standard spacing factor L, the PPV and frost resistance. The tests results suggest that the concrete with the standard spacing factor L greater than 0.20 mm may be frost resistant. The authors observed that the PPV factor provides a better correlation with the frost resistance results than the standard spacing factor. At this stage of the study, authors suggests that the concrete can be frost resistant with the PPV higher than 80% and non-frost resistant when the PPV is lower than 60%. The literature does not provide any data to verify this

Fig. 2. Scheme of determining PA: (a) air voids; (b) aggregate grains; (c) total paste area PA (yellow). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

U • . . • 1 . 1 • • 1 Kvf&pi^j

(5 • 0

Fig. 3. Scheme of determining PpA: (a) air voids; (b) shells of protected paste around air voids; (c) aggregate grains; (d) protected paste area PpA (blue). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 2

The air void structure parameters and freeze-thaw resistance results.

Concrete 1D measurements 2D measurements dL*

A [%] A300 [%] L [mm] N [psc] A2D [%] A2D0[%] PPV [%] Np [psc] mm/m

A 3.42 0.29 0.74 170 2.67 0.21 13.0 1065 2.39

B 2.38 0.46 0.39 262 2.94 0.50 21.7 2649 3.22

C 5.32 0.95 0.31 502 4.83 0.76 38.2 5874 1.59

D 6.05 1.26 0.24 684 6.98 1.74 80.7 15,998 0.03

E 2.67 0.48 0.39 307 3.15 0.31 37.3 6881 4.59

F 2.63 0.54 0.23 479 2.52 0.33 41.1 9507 3.75

G 3.31 0.58 0.20 632 3.49 0.64 55.7 8382 4.43

H 3.36 2.14 0.06 2022 2.19 1.26 89.9 29,121 0.00

Where:

N - number of chords per traverse length, Np - number of air voids per surface area,

A2D - the ratio of the surface area of air voids to the total surface area of the sample,

A3oo - the ratio of the surface area of air voids with a diameter of not more than 300 im to the total surface area of the sample, dL - expansion of samples.

relationship, thus further testing with the use of various methods is necessary.

4. Conclusions

In the paper authors present an innovative method of performing the examination and description of an air void structure in air-entrained concrete using a PPV (Protected Paste Volume) factor. This parameter is determined as a ratio of the protected paste area (located at a distance of 0.20 mm from the edge of the nearest air void) to the total paste area. The PPV factor accounts for the number, size and distribution of air voids, along with the surface measurements. Unlike other approaches that represent a statistical estimation of the average spacing of air voids, the PPV seems to better reflect the air-paste-aggregate system in concrete and thus

can be considered as an improved factor of air voids spatial dispersion, although it applies only to surface measurements.

In order to calculate the value of the PPV factor, first the paste area and the protected paste area need to be determined. Therefore, microscopic examination has to be performed to extract three phases: air pores, aggregate and the shells of the protected paste, which is a difficult and time consuming procedure. Considerable work needs to be undertaken to further develop the existing methodology.

The results presented in this paper, as well as those described in the previous analyses of the numerical model and model concretes, indicate that the spacing factor L 6 0.20 mm fails to guarantee that 100% of the cement paste is protected against deleterious frost effects. This suggests that the Powers' assumption is not true. The standard method of determining the spacing factor L takes into

2D MEASUREMENTS

Number of cycles

Fig. 4. Changes in mass of A+H samples after freeze-thaw cycles in water.

Number of cycles

Fig. 5. Changes in the linear deformation of A+H samples after freeze-thaw cycles in water.

1D STANDARD MEASUREMENTS

160 -|

rds 100 -

h o 80 -

nD300=571 nF300=431

0 50 100 150 200 250 300 Chord [^m]

Fig. 6. Measurements of chords (1D) up to 300 im for concrete D and F.

account the number of pores (chords) and their surface, regardless of whether they are grouped, close to one another or uniformly distributed in the cement paste. Determination of the PPV value also takes into account how air voids are placed, whether they are they grouped or whether the shells overlap. This directly affects the value of PPV and illustrates the advantage of the PPV over other parameters.

I 2000 -■3

100 150 200 Diameter [^m]

Fig. 7. 2D measurements of diameters up to 300 im for concrete D and F.

1D STANDARD MEASUREMENTS

100 200 Chords [^m]

Fig. 8. Measurements of chords (1D) up to 300 im for concrete D and H.

2D MEASUREMENTS

100 200 Diameter [^m]

Fig. 9. 2D measurements of diameters up to 300 im for concrete D and H.

Second very important problem, difficult to solve, is a comparison of the determined air pore structure parameters (L and PPV) to the frost resistance test results. The differences in the values of PPV and L are not large in the cases of frost resistant concretes with

£ 40 Ph

frost resistance concrete non frost resistance concrete

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 Spacing factor L [mm]

Fig. 10. Relationship between the standard spacing factor L, PPV and frost resistance results.

very "good" air entrainment quality or non frost resistant concretes with very "bad" air entrainment system. The biggest differences should be expected in concretes, where difficulties in a uniform air voids distribution in the cement paste are due to clustering, segregation (aggregate grains, mortar and air voids) and/or problems with frost resistance assessment. At this stage of the study, the authors suggest that the PPV provides better correlation with the frost resistance results than the standard spacing factor L and concretes with the value of PPV of at least 80% can be considered as frost resistant and at PPV below 60% as non-frost-resistant. Because of lack of suitable data in the literature further studies are needed to verify this relationship.

References

[1] J. Wawrzenczyk, A. Molendowska, Air void structure in relation to the frost resistance of air-entrained concrete by with microspheres, Cement Wapno Beton 5 (2011) 278-287.

[2] J. Wawrzenczyk, A. Molendowska, Use of microspheres as an alternative method of concrete air-entrainment, Budownictwo Technologie Architektura 4 (2011) 51-55.

[3] R. Springenschmidt, R. Breitenbucher, M.J. Setzer, Air-Entrained Concrete— Recent Investigations on the Fine Sand Composition. Waiting Time before Compaction and Redosing of Air-Entraining Agents, Betonwerk+Fertigteil-Technik, 11/1987, pp. 742-748.

[4] K.A. Snyder, A numerical test of air void spacing equations, Adv. Cem. Based Mater. 8 (1998) 28-44.

M. Pigeon, P. Plante, M. Plante, Air-void stability, part I: influence of silica fume and other parameters, Am. Concr. Inst. Mater. J. 86 (5) (Sept.-Oct. 1989) 482490.

P. Plante, M. Pigeon, F. Saucier, Air-void stability, part II: influence of superplasticizers and cement, Am. Concr. Inst. Mater. J. 86 (6) (November-December 1989) 581-589.

F. Saucier, M. Pigeon, P. Plante, Air-void stability, part III: fields tests of superplasticized concretes, Am. Concr. Inst. Mater. J. 87 (1) (1990) 3-11. B. lazniewska-Piekarczyk, The influence of admixtures type on the air-voids parameters of non-air-entrained and air-entrained high performance SCC, Constr. Build. Mater. 41 (2013) 109-124.

R.E. Philleo, A method for analyzing void distribution in air-entrained concrete, Cem. Concr. Aggr. (2) (1983) 128-130.

E. Attiogbe, Volume fraction of protected paste and mean spacing of air voids, Am. Concr. Inst. Mater. J. 94 (M66) (1997) 588-591.

J. Elsen, N. Lens, J. Vyncke, T. Aarre, D. Quenard, V. Smolej, Quality assurance and quality control of air entrained concrete, Cem. Concr. Res. 24 (7) (1994) 1267-1276.

N.P. Mayercsik, R. Felice, M. Tyler Ley, K.E. Kurtis, A probabilistic technique for entrained air void analysis in hardened concrete, Cem. Concr. Res. 59 (2014) 16-23.

D. Zalocha, J. Kasperkiewicz, Estimation of the air entrained concrete using a flatbed scanner, Cem. Concr. Res. 35 (2005) 2041-2046.

P. Soroushian, M. Elzafraney, Morphological operations, planar mathematical formulations, and stereological interpretations for automated image analysis of concrete microstructure, Cem. Concr. Compos. 27 (2005) 823-833. K.K. Aligizaki, P.D. Cady, Air content and size distribution of air voids in hardened cement pastes using the section-analysis method, Cem. Concr. Res. 29 (1999) 273-280.

Polish Standards Committee, PN-EN 480-11 Determination of the Characteristics of Air Pores in Hardened Concrete, Warszawa, Polish Standards Committee, 2008 (in Polish).

E. Attiogbe, Mean spacing of air voids in hardened concrete, Am. Concr. Inst. Mater. J. 90 (M19) (1993) 174-181.

B. lazniewska-Piekarczyk, The frost resistance versus air voids parameters of high performance self compacting concrete modified by non-air-entrained admixtures, Constr. Build. Mater. 48 (2013) 1209-1220. M. Pigeon, R. Pleau, Durability of Concrete in Cold Climates, E&FN SPON, London, 1995.

J. Wawrzenczyk, W. Kozak, A method of assessing parameters of air-void structure in air-entrained concretes, J. Civil Eng. Archit. 9 (7) (2015) 798-806. J. Wawrzenczyk, W. Kozak, A method of analyzing porosity structure in air entrained concrete, Archit. Civil Eng. Environ. (2) (2015) 77-82. Image-Pro Plus, Media Cybernetics.

ASTM C 666 A, Standard Test Method for Resistance of Concrete to Rapid Freezing and Thawing.

M.T. Hasholt, Air void structure and frost resistance: a challenge to powers' spacing facing, Mater. Struct. 47 (5) (2013) 911-923.

J. Wawrzenczyk, W. Kozak, A method of analyzing the porous microstructure in air-entrained concrete on the basis on 2D image analysis, Procedia Eng. 108 (2015) 102-107.

T.C. Powers, The requirement of frost resistance concrete, Proc. Highway Res. Board 29 (1949) 194-211.

PKN-CEN/TS 12390-9, Testing Hardened Concrete. Part 9: Freeze-Thaw Resistance. Scaling, 2007.

CEN/TR 15177, Testing the Freeze-Thaw Resistance of Concrete - Internal Structural Damage Technical Report, April 2006.