Frost Resistance of Magnesium Oxychloride Cement Mortar Added with Highland Barley Straw Ash

CAO Feng, QIAO Hongxia＊, WANG Penghui, LI Weijia, LI Yuanke

(1.School of Civil Engineering, Lanzhou University of Technology, Lanzhou 730050, China; 2. School of Civil and Transportation Engineering, Qinghai Minzu University, Xining 810000, China; 3. Western Center of Disaste Mitigation in Civil Engineering of Ministry of Education,Lanzhou 730050, China)

Abstract: In order to study the influence of highland barley straw ash (HBSA) prepared under certain conditions on the durability of magnesium oxychloride cement mortar (MOCM) under freeze-thaw damage,rapid freeze-thaw cycle tests were carried out firstly. The relative mass evaluation parameters and the relative compressive strength evaluation parameters, which represent the degradation of freeze-thaw resistance, were used as the indices to study the degradation rule of MOCM. Secondly, nuclear magnetic resonance (NMR)tests were carried out on MOCM under different freeze-thaw cycles to analyze the pore diameter changes in the freeze-thaw process. The microstructure of MOCM was tested by Fourier transform infrared spectroscopy(FTIR), X-ray diffraction (XRD) and scanning electron microscopy (SEM), and then the effect mechanism of HBSA on the anti-freezing performance of MOCM was revealed. Finally, the two-parameter Weibull distribution function was used to analyze the reliability of durability degradation of MOCM added with HBSA under freezethaw cycles. The specific conclusions are as follows: With the increase of HBSA’s addition, the freeze-thaw resistance of MOCM increase firstly and then decrease. When the addition of HBSA is 10%, the decay rate of relative mass evaluation parameters and relative compressive strength evaluation parameters is the slowest, and the frost resistance is the best. The proportion of harmful pores and more harmful pores in MOCM added with 10% HBSA decreases by 25.11% and 21.34%, compared with that without HBSA before and after freeze-thaw cycles. A lot of magnesium silicate hydrate (M-S-H) gels are generated in MOCM with HBSA content of 10%,which fills part of the pores, so that the proportion of harmful pores and more harmful pores is the lowest. The Weibull function can be effectively applied to the reliability analysis of the freeze-thaw cycle of MOCM added with HBSA, and the theoretical results are in good agreement with the experimental results.

Key words: magnesium oxychloride cement; highland barley straw ash; freeze-thaw resistance; pore diameter distribution; microstructure; reliability

1 Introduction

Large areas of salt lakes are distributed in Qinghai Province of China, which have a serious effect on the service life of concrete structures in this area.In addition to the serious corrosion caused by brine,freeze-thaw damage is an important factor leading to concrete damage in this area[1]. A large amount of magnesium chloride is produced as a by-product in the process of extracting potassium from salt lakes. The annual output of magnesium chloride is approximately 60 million tons, but the effective utilization rate is less than 1 million tons. A large amount of this by-product is discharged back into the salt lake as the form of old brine, which results in a waste of magnesium resources and compromises the sustainable development of salt lake resources to a large extent[2,3]. Magnesium oxychloride cement (MOC), also known as Sorrell cement, is a pneumatic cement-based material that Portland cement is not required[4]. Magnesium oxide,magnesium chloride, and water contained in MOC are mixed in accordance with a certain molar ratio, and has the advantages of early strength, fast hardening,bonding property, wear resistance, and brine erosion resistance[5,6]. Two main chemical constituent phases of 5Mg(OH)2·MgCl2·8H2O and 3Mg(OH)2·MgCl2·8H2O in MOC largely determine these prperties[7]. In addition, MOC’s preparation consumes a large amount of magnesium chloride and has the advantage of resisting brine corrosion, and thus, its application in this area has important practical value. Therefore, the demand for the development of enhanced MOC’s freeze-thaw resistance in its popularization and application becomes increasingly urgent.

In order to improve MOC’s freeze-thaw resistance, its water resistance and mechanical properties might be improved firstly, and adding the active blending materials is an effective method. Previous studies have found that fly ash[8], rice husk ash[9], wheat straw ash[10]and other active admixture could improve cement-based materials’ performance significantly. Adding fly ash to MOC could optimize its pore structure to obtain a more compact structure, and the M-S-H produced could improve its water resistance[11,12]. The addition of rice husk ash to magnesium oxychloride fiber cement could improve its microstructure, mechanical properties, and durability effectively[13]. Wheat straw ash is a feasible pozzolanic ash material that could replace a certain proportion of cement, and has no adverse effect on its performance[14]. Rape straw ash could improve the concrete’s compactness, the stability of passivation film on the reinforcement’s surface, and its corrosion resistance[15]. Durability of wheat straw ash concrete exposed to freeze-thaw damage increases as the wheat straw ash replacement level increased from 5% to 15%[16]. The artificial volcanic ash materials above have a significant effect on cement-based materials, and the common reason is that they all contain a certain amount of active silica. Reaction of activated silica with hydration products of cement-based materials occurs to form new products which are more favorable to their properties.

Highland barley is a type of cereal crop suitable for growing in the cold weather of Qinghai-Tibet Plateau, and has strong cold resistance, a short growing period, early maturity, and high yield[17]. At present, the area planted with highland barley in Qinghai Province reaches 1 million mu, and accounts for approximately 33% of the entire province’s grain crop planting area[18].However, after the highland barley harvest, a large number of abandoned crop straw has brought some difficulties to the local people’s production. Burning straw in farmland is forbidden on the Qinghai-Tibet Plateau where the temperature is low with the low rainfall, and the treatment of straw returning appears to have little effect. Local farmers use the highland barley straw primarily as a kind of fuel and livestock feed. However,the ash amount is high, particularly the amount of silica, so it is not suitable for the use of domestic fuel and livestock feed[19]. Therefore, the higher amount of silica in highland barley straw ash (HBSA) makes it possible to be used as a kind of active blending materials.Reactive silica reacts with hydration products in MOC to generate M-S-H, which improves the MOC’s pore structure[20], mechanical properties and durability[21,22].As for the preparation conditions of HBSA, our team conducted a lot of research work in the early stage, and conducted an in-depth study on the activity of HBSA from the aspects of microscopic test and mechanical properties. Finally, it is found that HBSA obtained by calcination at 600 ℃ for 2 h and grinding for 2 h have the highest content of active silica, the largest specific surface area, the smallest average particle size and the best activity. Therefore, it is of great significance to study the effect on freeze-thaw resistance of adding HBSA prepared under certain conditions to MOC mortar (MOCM).

However, the effects of HBSA on MOCM’s freeze-thaw resistance as an active blending material have not been reported. In this work, HBSA prepared under certain conditions was added into MOCM in different proportion to study HBSA’s effects on MOCM’s freeze-thaw resistance. The relative mass evaluation parameters and relative compressive strength evaluation parameters were used to study the anti-freezing performance degradation of MOCM added with HBSA.The optimal proportion of HBSA added into MOCM was determined, and the MOCM’s pore structure and microstructure were analyzed. The Weibull distribution function was used to analyze the reliability of the freeze-thaw resistance of MOCM, and further verify the accuracy of the test results.

2 Experimental

2.1 Raw materials

The raw materials of MOCM are composed largely of light burning magnesium oxide (MgO), industrial magnesium chloride (MgCl2·6H2O), sand, water, a water reducing agent, and water reducer. Light burning magnesium oxide and industrial magnesium chloride are produced by the Magnesium Chloride Plant of Qarhan Salt Lake in Golmud City, Qinghai Province. The amount of MgO is accounted for 98% of light burning magnesium oxide and the proportion of active MgO is 62.4%. The proportion of MgCl2·6H2O is accounted for 96% of industrial magnesium chloride. Sand with a particle size less than 4.75 mm are obtained fromGuide River in Qinghai Province, which are graded well. Tap water is used to mix the admixture, and meets the requirements of the concrete mixing water standard.The water reducer adopts a polycarboxylate-type system high efficiency water reducer, with an efficiency of 21%. The amount of H3PO4is not less than 85%, and the chromaticity Hc unit is not more than 25. The highland barley straw was obtained from the Nanmenxia area of Huzhu County, Qinghai Province. After weeds and other impurities were removed, it was burned to ash in an outdoor natural environment. The burning point is approximately 300 ℃, and the burning time is approximately 3 h. After soil, sand, gravel, and other impurities were removed from the ash, the integrated SX2-12-10A intelligent box muffle furnace was used for secondary calcination under laboratory conditions.The ash after secondary calcination is milled mechanically by roller ball mill.

Table 1 Chemical composition of HBSA/%

Fig.1 Macroscopic and microscopic image of HBSA

Fig.2 FTIR and XRD spectrum of HBSA

The highest activity of HBSA was obtained by secondary calcination at 600 ℃ for 2 h and ground for 2 h, and the HBSA added into MOCM were prepared under this condition in this work. The macroscopic and microscopic morphology of HBSA are shown in Fig.1.FTIR and XRD spectra of HBSA are shown in Fig.2.The infrared spectrum of HBSA includes five distinct absorption bands. Among them, the absorption bands at 464 cm-1are caused by the symmetric variable angle vibration of Si-O-Si, the absorption bands at 800 and 1 048 cm-1are caused by the symmetric and antisymmetric stretching vibration of Si-O-Si respectively,and the absorption bands at 1 609 and 3 435 cm-1are caused by the variable angle and stretching vibration of crystal water. In the XRD spectrum of HBSA, the main diffraction peak was SiO2.

The chemical composition of HBSA prepared under this condition is shown in Table 1. Silica is the main component, and the content reached to 61.751%.The burning loss measured by the test is 4.55%,which meets the standard requirement of no more than 10%[23]. The content of SO3is 1.751%, which meets the standard requirement of no more than 3.5%[23]. The 28 d compressive strength ratio reaches 1.03, meeting the standard requirement of not less than 0.65[23]. The particle size distribution of HBSA is shown in Fig.3.The average particle size of HBSA was 8.6 μm, and the specific surface area was 2 088 m2/kg. The 45 μm square hole sieve allowance is 0.95%, which meets the standard requirement of less than 20%[24].

2.2 Specimen preparation

The mixture ratio of MOCM from our previous research results is shown in Table 2[25]. HBSA was added into the MOCM mixture in the proportions of 0, 5%, 10%, 15%, and 20% of the mass of magnesium oxide. These specimens were prepared for 40 mm×40 mm×160 mm prism test blocks to test the freeze-thaw resistance, in which each addition had 10 test blocks and each group included 3 pieces, in a total of 50 groups and 150 pieces. The specimen was demolded under the indoor natural condition curing for 24 h, and then cured to 26 d. It was immersed in water for another 2 d, a total of 28 d. A 40 mm×40 mm×40 mm cube specimen was cut off from the prism specimen with a cutting machine for NMR test.

Table 2 Mix ratio of MOCM/(kg/m3)

Fig.3 Particle size distribution of HBSA

2.3 Test method

2.3.1 Freeze-thaw resistance

Before the rapid freeze-thaw test, one group of test blocks was removed from the water, and the compressive strength was tested. After the other 9 groups of test blocks were removed from the water, the surface moisture was wiped. The rapid freeze-thaw test was carried out immediately, in which each freezethaw cycle was 4 h. After each 5 freeze-thaw cycles,the specimens marked with corresponding freeze-thaw cycles were removed. Their surface was dried, the mass was weighed, and the compressive strength was tested.When the freeze-thaw specimen’s compressive strength loss rate is not more than 25% and the mass loss rate is not more than 5%, the antifreeze performance meets the requirements (JGJ/T 70)[26]. The relative mass evaluation parameter,ω1, and the relative compressive strength evaluation parameter,ω2, are calculated according to the following formula:

where,mris the ratio of the mass of the specimen afternfreeze-thaw cycles to that before the freeze-thaw cycles, andfris the ratio of the specimen’s compressive strength afternfreeze-thaw cycles to before the freezethaw cycles. When 0≤ω1,ω2≤1, freeze-thaw resistance is satisfied, whileω1＜0 andω1＜0, freeze-thaw resistance is not satisfied.

2.3.2 Pore diameter test

NMR technology could measure the true pore diameter distribution in the sample more accurately because it does not destroy the pores during the testing process. Hence, the NMR test is used to test the changes in the pore structure of MOCM with different proportion of HBSA in different freeze-thaw periods,and the test range of pore diameter is 2 nm - 1 mm.Before the NMR test, the following treatments were carried out: Firstly, after the dust on the surface of the intercepted cube sample block was removed, the block was placed in water for 10 h at a vacuum pressure of 0.1 MPa. Secondly, the sample block was removed after it was saturated completely, the water on the surface was wiped off, and the NMR test was conducted immediately.

NMR technology can be used to calculate the volume of pores in porous media by measuring the mass and density of water in the sample to obtain such parameters as porosity and pore size distribution. By testing the signal of an H atom in the sample, a greater number of H atoms indicates more water contained,i e,the higher proportion of the pore where water is under this signal[27]. The pore diameter distribution could be calculated by the transverse relaxation timeT2distribution as follows[28]:

whereT2is the medium’s transverse relaxation time,ms;Sis the pore’s surface area, μm2;Vis the pore’s volume, μm3,ρ2isT2’s surface relaxation ratio, μm/ms.ris the pore radius,Fsis the shape factor, 3 for spherical pores and 2 for columnar pores.

3 Results and discussion

3.1 Evaluation of freeze-thaw resistance

3.1.1 Relative quality evaluation parameters

The relation between the relative quality evaluation parameter,ω1, and the number of freeze-thaw cycles,N, is shown in Fig.4. As the number of freezethaw cycles increases, the MOCM’s relative mass evaluation parameters decrease to varying degrees with different content of HBSA. In the early freezethaw stage, theω1of MOCM added with 5% HBSA is slightly higher than that of those with 10% HBSA added. When MOCM specimens are added with 0, 15%,and 20% HBSA, the relative quality evaluation parameter,ω1, decreases to 0 with approximately 43, 45, and 40 freeze-thaw cycles, respectively. The freeze-thaw resistance of specimens with 0, 15%, and 20% HBSA do not meet the requirements with 45 freeze-thaw cycles. After 45 freeze-thaw cycles,ω1is still above 0.5 for specimens with 5% and 10% HBSA. MOCM added with 10% HBSA has the slowest decay inω1, followed by 5%, and the fastest decay when the HBSA content is 20%. It can be seen that the addition of 5% and 10%HBSA could improve MOCM’s freeze-thaw resistance significantly, and 10% HBSA results in the best effect.

Fig.4 The relation between relative quality evaluation parameters and number of freeze-thaw cycles

3.1.2 Relative compressive strength evaluation parameters

Fig.5 The relation between relative compressive strength evaluation parameters and the number of freeze-thaw cycles

The relation between the relative compressive strength evaluation parameter,ω1, and the number of freeze-thaw cycles,N, is shown in Fig.5. Theω2of specimens added with 10% HBSA attenuates is the slowest, followed by those with 5% HBSA, while those with 20% HBSA attenuates is the fastest. In the early freeze-thaw stage, theω2of MOCM specimens with 5%HBSA is slightly higher than that of those with 10%.With the increase in freeze-thaw cycles, the specimens with 10% HBSA content show excellent freeze-thaw resistance. When the specimens with 0, 15%, and 20%HBSA are freeze-thawed approximately 41, 40, and 36 times, respectively, theω2is less than 0, and the freezethaw resistance is unqualified. Added with 5% and 10%HBSA, theω2is still greater than 0 with 45 freeze-thaw cycles, and the freeze-thaw resistance is good. It can be seen that when the HBSA amount is 5% and 10%,MOCM’s freeze-thaw resistance could be improved significantly. The freeze-thaw resistance of MOCM is the best with 10% HBSA, and theω2is still above 0.3 with 45 freeze-thaw cycles. Thus, clearly, adding a certain amount of HBSA could improve MOCM’s freezethaw resistance significantly. This is because adding HBSA causes MOCM’s secondary hydration, and the hydration products improve its microstructure, and thus improve its freeze-thaw resistance.

3.2 Pore structure analysis

Fig.6 T2 spectrum of MOCM without HBSA in different numbers of freeze-thaw cycles

TheT2spectrum distribution of MOCM specimens without HBSA in 0, 15, 30, and 45 freeze-thaw cycles are shown in Fig.6. The area of the peak in the spectrum represents the number of pores, and the position of the peak represents their size. The first wave mainly represents harmless pores and less harmful pores, and the last two waves mainly represents harmful pores and more harmful pores. Before the freezethaw cycle, the spectrum’s range is more concentrated and the wave peak is higher. It indicates that the pores’distribution range is relatively concentrated and the proportion of micro pores is high. With 15, 30, and 45 freeze-thaw cycles, theT2spectrum widens, indicating that the pores’ distribution range has become wider and more new pores were produced. It can be seen from the first wave that as the number of freeze-thaw cycles increases, the proportion of the original pore size decreases gradually, while the range of the pores increase gradually. It demonstrates that the original pores develop into new pores under the freeze-thaw action. It can be seen from the second and third wave stages that as the number of freeze-thaw cycles increases, the range of the last two wave stages becomes larger and the wave peak value increases. This is because the original relatively high proportion of tiny pores develop gradually into new, larger pores under the action of freezethaw cycles.

Fig.7 The pore diameter distribution of MOCM with different numbers of freeze-thaw cycles

Table 3 The pore diameter distribution statistics of MOCM with 0 and 45 freeze-thaw cycles

The pore diameter distribution of MOCM for 0 and 45 freeze-thaw cycles are shown in Fig.7. The pore whose diameter smaller than 0.02 μm is harmless pore, between 0.02 and 0.05 μm is less harmful pore,between 0.05 and 0.2 μm is harmful pore, and larger than 0.2 μm is more harmful pore[29]. After 45 freezethaw cycles, the greatest number of more harmful pores larger than 0.2 μm is in the specimens without HBSA,while it is the least with 10% HBSA. It indicates that more harmful pores form in MOCM without HBSA under freeze-thaw action, while the formation of harmful pores is avoided effectively when 10% HBSA is added.The freeze-thaw effect changes the proportion of the original pore size, and forms more new pores. When the addition of HBSA is 10%, the proportion of harmless pores and less harmful pores are higher, while the proportion of harmful pores and more harmful pores are lower. However, the pore diameter distribution of MOCM with 0 and 5% HBSA is the opposite, and the MOCM without HBSA have the highest proportion of harmful pores and more harmful pores. The result indicates that adding HBSA could improve MOCM’s pore structure significantly, and thus improve its frost resistance.

Table 3 shows the statistics for pore diameter distribution of MOCM added with 0, 5%, and 10% HBSA in 0 and 45 freeze-thaw times. Before the freezethaw cycle, the proportion of harmful pores and more harmful pores is the highest in MOCM without HBSA.When the addition of HBSA is 10%, the harmful pores and more harmful pores decrease by 25.11% compared with that without HBSA. After 45 freeze-thaw cycles,the proportion of more harmful pores in MOCM with 0, 5%, and 10% HBSA increases by 2.92, 2.04, and 1.93 times, respectively. The proportion of harmful and more harmful pores, and harmless and less harmful pores in MOCM with HBSA content of 5% are basically the same as that with HBSA content of 0%. While the proportion of harmful pores and more harmful pores in MOCM with 10% HBSA content decrease by 21.34% compared with that without HBSA. It can be seen that the addition of HBSA reduces the proportion of harmful pores and more harmful pores significantly,while it increases the proportion of harmless pores and less harmful pores. Thus, the structure’s compactness increases and the freeze-thaw resistance improves effectively.

3.3 Microstructure

3.3.1 FTIR

MOCM specimens with different HBSA content were tested for infrared spectrum, as shown in Fig.8.The weak absorption peak at 3 693 cm-1is caused by the stretching vibration of Mg-OH. The strong absorption peak at 1 609 cm-1is caused by the angular vibration of crystal water, indicating that a large number of 5-phase crystals are produced in the hydration products of MOCM with three kinds of HBSA content. The absorption band at 1 008 cm-1is caused by the antisymmetric stretching vibration of Si-O, and the absorption band at 781 cm-1is caused by the symmetric stretching vibration of Si-O. These two absorption bands are the marks of M-S-H formation[34]. The absorption band of MOCM with 5% and 10% HBSA content at 1 008 and 781 cm-1are significantly stronger than those without HBSA. The absorption band at 3 610 cm-1is caused by the stretching vibration of crystal water, which again indicates the existence of hydration product magnesium silicate hydrate in MOCM added with 5% and 10% HBSA. In addition, MOCM with 5% and 10%HBSA content produces significantly different peaks at 527 cm-1compared with 0% HBSA. The absorption band is caused by the asymmetrical stretching vibration of SiO32-, which indicates that M-S-H does exist in MOCM with HBSA content of 5% and 10%.

Fig.8 FTIR spectra of MOCM with different HBSA content

3.3.2 XRD

The test results of XRD for MOCM with different HBSA content are shown in Fig.9. It could be seen that when the HBSA content was 5%, the diffraction peak of 5-phase crystal is stronger than that of 0% and 10%,so it has higher early mechanical properties. This also explains why, in the early stage of freeze-thaw action,the evaluation parameter of freeze-thaw resistance when HBSA content is 5% is slightly higher than that of 10% content. In addition, the diffraction peak of MgO with 5% content is stronger than that with 0%and 10% HBSA content. The results indicates that there was more MgO and less active SiO2in MOCM with 5% HBSA content, and all active SiO2participate in the hydration reaction to form M-S-H. The MgO content is the lowest when the HBSA content is 0%, but the diffraction peaks of hydration products Mg(OH)2and MgCO3are stronger. When the HBSA content is 10%,the MgO content is lower than that with 5% HBSA content, which shows that more MgO reacts with active SiO2to produce more M-S-H. The results show that the incorporation of HBSA could transform the unfavorable hydration product of Mg(OH)2into the favorable product of M-S-H.

Fig.9 XRD spectra of MOCM with different HBSA content

3.3.3 SEM

SEM images of MOCM with different HBSA content are shown in Fig.14. When the HBSA content is 0%, the micromorphology is mainly composed of needle-like 5-phase crystal and a small amount of M-S-H gel. The 5-phase crystals crisscross each other,providing mechanical properties for MOCM. When the HBSA content is 5% and 10%, the micromorphology shows that a large number of M-S-H gels are formed.Especially, M-S-H gels in MOCM specimen with 10% HBSA content uniformly adhere to the surface of 5-phase crystal and solid particles. M-S-H gels fill a large number of 5-phase crystal gaps and solid particle gaps to form a dense structure. In the MOCM specimen with 5% HBSA content, although some M-S-H gels are formed, the gels are not enough to completely adhere to the surface of 5-phase crystal and solid particles, and there are still some 5-phase crystal gaps and solid particle gaps that are not effectively filled. However, when HBSA content is 10%, sufficient M-S-H gels are generated in MOCM, which play a positive and effective role in improving the pore structure.

Fig.10 SEM images of MOCM with different HBSA content

Fig.11 Weibull distribution probability graph

The mechanism that the addition of a certain amount of HBSA in MOCM can effectively improve its freeze-thaw resistance is mainly as follows: The active SiO2in HBSA could react with the hydration product Mg(OH)2of MOCM to form M-S-H gels, which could transform the product which is unfavorable to its performance into the favorable one. By adding a certain amount of HBSA, a large number of M-S-H gels could be produced. The gels adhere to the surface of 5-phase crystal and solid particles and fill with a large number of harmful and multi-harmful pores. The structure is more compact and the freeze-thaw resistance is significantly improved.

3.4 Reliability analysis

Weibull function is a common distribution function in engineering reliability, which is widely used in reliability analysis because it could get relatively accurate prediction with small sample data[30]. Weibull distribution model has two parameters, namely shape parameter and scale parameter. The shape parameter mainly affects the shape of the distribution curve, and the scale parameter mainly affects the dispersion degree of the curve on the coordinate axis[31]. Based on the two-parameter Weibull distribution function in this work, the degradation of relative compressive strength evaluation parameter for MOCM with different HBSA content under the action of freeze-thaw cycles was numerically simulated.

3.4.1 Distribution test of Weibull function

The two-parameter Weibull function is expressed as follows:

The sample data of the evaluation parameterω2for the test results are tested by hypothesis distribution test. When the HBSA content is 0% and 10%, the hypothesis distribution test results are shown in Fig.11 (a)and (b), respectively. All the data points of the samples are within the 95% confidence band, indicating thatω2could well obey the Weibull function distribution when the HBSA content is 0% and 10%.

3.4.2 Parameter estimation

There are several methods for parameter estimation of Weibull distribution. Least square method,moment estimation method and maximum likelihood method could be used for parameter estimation. The least square method is used for parameter estimation under small sample data with high accuracy. Therefore,the least square method is used for parameter estimation of Weibull distribution function in this work. Take the logarithm of both sides of Eq.(6), and simplify it to obtain the reliability function as follows:

According to the fitting resultskandbof the least square method for the sample data when the HBSA content is 0% and 10% respectively, the corresponding scale parametersαand shape parametersβcould be obtained. When the HBSA content is 0%,α=26.667 4,β=1.930 9; When the HBSA content is 10%,α=45.638 3,β=1.897 6.

Fig.12 Parameter fitting results of least square method

3.4.3 Reliability analysis

According to Eq.(6) and the value of parameter estimationαandβcalculated above the reliability function distribution curve of MOCM under the action of freeze-thaw cycle can be obtained as shown in Fig.13.When the freeze-thaw cycle reaches 25 times, the corresponding reliability is 0.42 and 0.73 with the HBSA content 0% and 10%, respectively. According to the experiment results, when the freeze-thaw cycle reaches 25 times, the corresponding reliability is 0.44 and 0.74 when the HBSA content is 0% and 10%, respectively.The goodness of fit between Weibull reliability model and the experiment results are 95.5% and 98.6%, respectively. Therefore, the two-parameter Weibull distribution function could be effectively used to analyze the reliability of MOCM added with HBSA under the action of freeze-thaw cycles.

Fig.13 The distribution curve of the reliability function for MOCM

The reliabilityR(t) of MOCM specimen with HBSA content of 0% decreases to 0 when freeze-thaw cycles reach to 65 times. For MOCM specimen with 10% HBSA content, the reliabilityR(t) decreases to 0 when freeze-thaw cycles reach about 115 times. Compared with the MOCM samples without HBSA, the number of freezing and thawing resistance cycles of MOCM specimens with 10% HBSA increase by about 1.77 times. It can be seen that the addition of 10%HBSA into MOCM could effectively improve the resistance of freeze-thaw cycles for MOCM.

4 Conclusions

a) The freeze-thaw resistance of MOCM is different with the different content of HBSA. When the HBSA content is less than 10%, the frost resistance of MOCM increases with the increase of the content of HBSA, and it was better than that without HBSA.When the HBSA content is 10%, the freeze-thaw resistance evaluation parameters attenuate most slowly and the freeze-thaw resistance is the best.

b) The addition of an appropriate amount of HBSA in MOCM could improve the pore structure effectively, and reduce the proportion of harmful pores and more harmful pores significantly. When the HBSA content is 10%, a lot of M-S-H gels are generated in the MOCM, and the 5-phase crystal and solid particle gaps are filled. The proportion of harmful pores and more harmful pores in MOCM is the least, and the pore structure is the best.

c) The Weibull function could be well used for the reliability analysis of MOCM added with HBSA under the action of freeze-thaw cycles. According to the reliability analysis results, when the HBSA content is 10%, the freeze-thaw resistance of MOCM is about 1.77 times higher than that without HBSA.