Relationship between Flowability, Entrapped Air Content and Strength of Uhpc Mixtures Containing Different Dosage of Steel Fiber

This study investigated the effects of different fiber contents by volume (0%, 1%, 2%, and 3%) on flowability, entrapped air content, and compressive and flexural strength of ultra high performance concrete (UHPC). Four water-to-binder ratios (0.18, 0.20, 0.22, 0.24) were used and four different dosages of superplasticizer were introduced for preparing each water-to-binder ratio mixture. The experimental results reveal that the increased content of fiber decreases the flowability and entrapped air content of fresh UHPC mixtures. The compressive strength of UHPC increased with a greater addition of steel fiber, from 1% to 3%. The flexural strength was increased slightly by the addition of 1% steel fiber and increased remarkably with the addition of 2% and 3% steel fiber. For every water-to-binder ratio mixture, with and without steel fiber, a good linear relationship can be found between compressive strength and entrapped air content. Therefore, it is suggested that lowering the entrapped air provides a significant contribution to increasing compressive strength of UHPC containing higher steel fibers.


Introduction
As a new cement-based material, ultra high performance (UHPC) has gained interest around the world since it was introduced in the early 1990s.Ultra high performance concrete (UHPC) can show compressive strength from 150 to 810 MPa [1], approximately 3-16 times that of conventional concrete.Due to the improved mechanical strength, impact resistance, fatigue resistance, and its excellent durability [2][3][4], UHPC is also regarded as a construction material for energy saving, material saving, and low carbon emission.Presently, applications of UHPC have been reported in Europe, North America, Australia, Asia and New Zealand [5].
One of the key factors in producing UHPC is to improve the particle packing density of its cementitious matrix.Therefore, many studies have been conducted on the selection of raw materials, such as cement types and fineness, silica fume, fly ash, metakaolin, rice husk ash, nanoparticles and different type of fine aggregate [5][6][7][8][9], optimal mixing proportions, and curing conditions [10].The addition of fiber can improve the flexural properties of UHPC, and also leads to the transformation from brittle failure to ductile failure [11].The fibers used in concrete include carbon, steel, polypropylene and glass fibers, and most researchers have paid attention to the steel fiber.It was reported that the tensile strength of UHPC linearly increased with increased steel fiber volume ratio, from 0% to 5% [12].UHPC mixtures incorporating short steel fibers exhibited enhanced flexural properties compared to those of mixtures with a similar volume of longer steel fibers [13].The shape and distribution of steel Two types of quartz sand produced by Harbin Jinghua Water Treatment Material Company (Harbin, China) were used as aggregates, as shown in Table 2.A mixture of 80% coarse quartz sand and 20% fine quartz sand was used for preparing all the UHPC mixtures.A polycarboxylate-based superplasticizer (SP) from Harbin Qiangshi Company (Harbin, China) was used.It has a water-reducing efficiency higher than 30% and a solid content of about 40%.The steel fibers (as shown in Figure 1) used in this study are copper-coated, smooth fibers with a tensile strength of 2850 MPa.Each fiber is 13 mm long with a diameter of 0.2 mm.

Mixing Proportions of UHPC Mixtures
The ratio of cement:silica fume:quartz sand was kept at the value of 1:0.3:1.1 for all the mixtures, as shown in Table 3.All the mixtures can be divided into four series, with different water-to-binder (cement + silica fume) ratios of 0.18, 0.20, 0.22 and 0.24.In every series, with the same water-to-binder ratio, varying dosages of superplasticizer by mass of cementitious materials were added to obtain different flowabilities for the UHPC mixtures, and different steel fiber contents of 0%, 1%, 2%, and 3% by volume of mixture were used at every superplasticizer dosage level.Therefore, a total of 64 mixtures were manufactured in this study.

Mixing Procedure and Specimen Preparation
Compared with normal strength concrete, the UHPC mixture in its fresh state is more sensitive to mixing procedures and environment.To minimize the influence of the operation process, all the mixtures were prepared by using a planetary mixer (Wuxi Jianyi Instrument & Machinery Co., LTD, Wuxi, China) (as shown in Figure 2) using the same procedure.Dry ingredients, including cement, silica fume, and quartz sand, were first mixed for 3 min at a low speed of about 140 ± 5 rpm.Then, water and superplasticizer were added and mixed for approximately 3 min at low speed.After, steel fibers (for mixes containing steel fibers) were uniformly added in the mixing pot and mixed for another 3 min at a high speed of about 285 ± 10 rpm until homogeneous mixtures were obtained.
When the UHPC mixtures were ready, prism specimens with a size of 40 × 40 × 160 mm 3 were prepared by using a plastic mold (as shown in Figure 3).The mixture was cast in molds in two layers and vibrated for 90 s to consolidate the mixtures.The specimens in the molds were covered with

Mixing Proportions of UHPC Mixtures
The ratio of cement:silica fume:quartz sand was kept at the value of 1:0.3:1.1 for all the mixtures, as shown in Table 3.All the mixtures can be divided into four series, with different water-to-binder (cement + silica fume) ratios of 0.18, 0.20, 0.22 and 0.24.In every series, with the same water-to-binder ratio, varying dosages of superplasticizer by mass of cementitious materials were added to obtain different flowabilities for the UHPC mixtures, and different steel fiber contents of 0%, 1%, 2%, and 3% by volume of mixture were used at every superplasticizer dosage level.Therefore, a total of 64 mixtures were manufactured in this study.

Mixing Procedure and Specimen Preparation
Compared with normal strength concrete, the UHPC mixture in its fresh state is more sensitive to mixing procedures and environment.To minimize the influence of the operation process, all the mixtures were prepared by using a planetary mixer (Wuxi Jianyi Instrument & Machinery Co., LTD, Wuxi, China) (as shown in Figure 2) using the same procedure.Dry ingredients, including cement, silica fume, and quartz sand, were first mixed for 3 min at a low speed of about 140 ˘5 rpm.Then, water and superplasticizer were added and mixed for approximately 3 min at low speed.After, steel fibers (for mixes containing steel fibers) were uniformly added in the mixing pot and mixed for another 3 min at a high speed of about 285 ˘10 rpm until homogeneous mixtures were obtained.plastic sheet and placed under room temperature (22 ± 3 °C) for 24 h to minimize loss of water from the surface.Afterwards, the specimens were taken out from their molds and exposed to steam curing at 90 °C for 72 h.Finally, the specimens were stored in a standard curing room (22 ± 3 °C, RH 90%) until the time of mechanical measurements.

Flowability and Air Content Test
The flowability of all the fresh mixtures was carried out by using the test method for fluidity of cement mortar, in accordance with the Chinese National Standard GB/T 2419-2005 [19].A mini cone mold was placed at the center of a jolting table, as shown in Figure 4.As soon as the mixing procedure was completed, the UHPC mixture was cast into the cone mold in two layers, and each layer is tamped 15 times with a tamping rod to fill the mixture uniformly in the mold.The mold was lifted vertically and the sample on the jolting table was dropped 25 times.Then, two diameters, perpendicular to each other, were determined and the mean value was recorded to evaluate the flowability.
The air content in fresh UHPC mixture was measured using the water column method, according to British standard EN 12350-7 [20], and the equipment (FORM + TEST Seidner & Co. GmbH, Riedlingen, Germany) used is shown in Figure 5.The fresh UHPC mixture was poured into the cylinder container in two layers.Each was consolidated by a 15-fold drop on the table from a height of 30 mm.After dropping, the upper part of the apparatus was applied to the lower container and measurement in a pressure apparatus was conducted.For every mixture, the air content was tested twice.If the difference between the two measured values was less than 0.2%, the average value was determined as the test result.Conversely, if the difference between the two measured values was more than 0.2%, the test result was invalid.When the UHPC mixtures were ready, prism specimens with a size of 40 ˆ40 ˆ160 mm 3 were prepared by using a plastic mold (as shown in Figure 3).The mixture was cast in molds in two layers and vibrated for 90 s to consolidate the mixtures.The specimens in the molds were covered with plastic sheet and placed under room temperature (22 ˘3 ˝C) for 24 h to minimize loss of water from the surface.Afterwards, the specimens were taken out from their molds and exposed to steam curing at 90 ˝C for 72 h.Finally, the specimens were stored in a standard curing room (22 ˘3 ˝C, RH 90%) until the time of mechanical measurements.
Appl.Sci.2016, 6, 216 4 of 13 plastic sheet and placed under room temperature (22 ± 3 °C) for 24 h to minimize loss of water from the surface.Afterwards, the specimens were taken out from their molds and exposed to steam curing at 90 °C for 72 h.Finally, the specimens were stored in a standard curing room (22 ± 3 °C, RH 90%) until the time of mechanical measurements.

Flowability and Air Content Test
The flowability of all the fresh mixtures was carried out by using the test method for fluidity of cement mortar, in accordance with the Chinese National Standard GB/T 2419-2005 [19].A mini cone mold was placed at the center of a jolting table, as shown in Figure 4.As soon as the mixing procedure was completed, the UHPC mixture was cast into the cone mold in two layers, and each layer is tamped 15 times with a tamping rod to fill the mixture uniformly in the mold.The mold was lifted vertically and the sample on the jolting table was dropped 25 times.Then, two diameters, perpendicular to each other, were determined and the mean value was recorded to evaluate the flowability.
The air content in fresh UHPC mixture was measured using the water column method, according to British standard EN 12350-7 [20], and the equipment (FORM + TEST Seidner & Co. GmbH, Riedlingen, Germany) used is shown in Figure 5.The fresh UHPC mixture was poured into the cylinder container in two layers.Each was consolidated by a 15-fold drop on the table from a height of 30 mm.After dropping, the upper part of the apparatus was applied to the lower container and measurement in a pressure apparatus was conducted.For every mixture, the air content was tested twice.If the difference between the two measured values was less than 0.2%, the average value

Flowability and Air Content Test
The flowability of all the fresh mixtures was carried out by using the test method for fluidity of cement mortar, in accordance with the Chinese National Standard GB/T 2419-2005 [19].A mini cone mold was placed at the center of a jolting table, as shown in Figure 4.As soon as the mixing procedure was completed, the UHPC mixture was cast into the cone mold in two layers, and each layer is tamped 15 times with a tamping rod to fill the mixture uniformly in the mold.The mold was lifted vertically and the sample on the jolting table was dropped 25 times.Then, two diameters, perpendicular to each other, were determined and the mean value was recorded to evaluate the flowability.

Strength Measurement
The compressive strength and flexural strength for every mixture were measured on the prepared prism specimens at 28 days of age, according to GB/T 17671-1999 [21].A three-point flexural test was conducted, and averages of the three specimens for each batch were reported as tested results of flexural strength.Six broken samples, after the flexural test, were used to measure compressive strength with a loading rate of 1.5 kN/s.The average of six samples for every mixture was reported as the tested compressive strengths.

The Flowability of Fresh Mixture
Figure 6 presents the effects of steel fiber content, water-binder ratio and superplasticizer dosage on flowability of fresh UHPC mixtures.For all four different water-to-binder ratios, the flowability decreased with the increase of steel fiber content when the superplasticizer dosage was kept the same.The addition of 1% steel fiber led to a notable reduction of flowability when compared with the blank mixture without fiber, and a relatively slight decrease of flowability occurred when the addition of fiber increased from 1% to 3% for most of mixtures.For instance, the flowability of mixtures with 1%, 2%, and 3% steel fibers were reduced by 22.3%, 25.8%, and 28.8% compared to that of the blank mixture with a water-to-binder ratio of 0.20, containing 1.4% superplasticizer.This flowability reduction effect might be attributed to the increased specific surface area of steel fiber [16], leading to an increased demand of cement paste for covering the surface of fibers and aggregates.On the other hand, the steel fibers were randomly distributed in the mortar matrix and acted as a skeleton, and eventually prevented the flow of fresh UHPC mixture [14].
When the same water-to-binder ratio and steel fiber content was used, the more dosage of superplasticizer increased the fluidity as expected.To increase the flowability, a higher dosage of The air content in fresh UHPC mixture was measured using the water column method, according to British standard EN 12350-7 [20], and the equipment (FORM + TEST Seidner & Co. GmbH, Riedlingen, Germany) used is shown in Figure 5.The fresh UHPC mixture was poured into the cylinder container in two layers.Each was consolidated by a 15-fold drop on the table from a height of 30 mm.After dropping, the upper part of the apparatus was applied to the lower container and measurement in a pressure apparatus was conducted.For every mixture, the air content was tested twice.If the difference between the two measured values was less than 0.2%, the average value was determined as the test result.Conversely, if the difference between the two measured values was more than 0.2%, the test result was invalid.

Strength Measurement
The compressive strength and flexural strength for every mixture were measured on the prepared prism specimens at 28 days of age, according to GB/T 17671-1999 [21].A three-point flexural test was conducted, and averages of the three specimens for each batch were reported as tested results of flexural strength.Six broken samples, after the flexural test, were used to measure compressive strength with a loading rate of 1.5 kN/s.The average of six samples for every mixture was reported as the tested compressive strengths.

The Flowability of Fresh Mixture
Figure 6 presents the effects of steel fiber content, water-binder ratio and superplasticizer dosage on flowability of fresh UHPC mixtures.For all four different water-to-binder ratios, the flowability decreased with the increase of steel fiber content when the superplasticizer dosage was kept the same.The addition of 1% steel fiber led to a notable reduction of flowability when compared with the blank mixture without fiber, and a relatively slight decrease of flowability occurred when the addition of fiber increased from 1% to 3% for most of mixtures.For instance, the flowability of mixtures with 1%,

Strength Measurement
The compressive strength and flexural strength for every mixture were measured on the prepared prism specimens at 28 days of age, according to GB/T 17671-1999 [21].A three-point flexural test was conducted, and averages of the three specimens for each batch were reported as tested results of flexural strength.Six broken samples, after the flexural test, were used to measure compressive strength with a loading rate of 1.5 kN/s.The average of six samples for every mixture was reported as the tested compressive strengths.

The Flowability of Fresh Mixture
Figure 6 presents the effects of steel fiber content, water-binder ratio and superplasticizer dosage on flowability of fresh UHPC mixtures.For all four different water-to-binder ratios, the flowability decreased with the increase of steel fiber content when the superplasticizer dosage was kept the same.The addition of 1% steel fiber led to a notable reduction of flowability when compared with the blank mixture without fiber, and a relatively slight decrease of flowability occurred when the addition of fiber increased from 1% to 3% for most of mixtures.For instance, the flowability of mixtures with 1%, 2%, and 3% steel fibers were reduced by 22.3%, 25.8%, and 28.8% compared to that of the blank mixture with a water-to-binder ratio of 0.20, containing 1.4% superplasticizer.This flowability reduction effect might be attributed to the increased specific surface area of steel fiber [16], leading to an increased demand of cement paste for covering the surface of fibers and aggregates.On the other hand, the steel fibers were randomly distributed in the mortar matrix and acted as a skeleton, and eventually prevented the flow of fresh UHPC mixture [14].
Appl.Sci.2016, 6, 216 6 of 13 superplasticizer was needed for UHPC mixtures with a lower water-to-binder ratio, due to a smaller amount of existing free water.For water-to-binder ratios of 0.24 and 0.22, the flowability of the mixtures reached as high as 200 mm by using more superplasticizer, which is favorable for casting.However, for water-to-binder ratios of 0.18 and 0.20, the increased dosage of superplasticizer led to a limited increase of flowability.Therefore, the fluidity was more sensitive to superplasticizer dosage in higher w/b than in lower w/b mixtures [22].Such results reveal that a very low water-to-binder ratio may result in increased difficulty in compacting UHPC mixtures.

The Entrapped Air in Fresh Mixture
The test results of air content for all 64 UHPC mixtures are shown in Figure 7.It can be seen that most of the UHPC mixtures had 4%-6% air content, remarkably higher than that of normal concrete (1%-2%), when no air entraining agent was used.The higher amount of entrapped air has a detrimental effect on the mechanical properties of UHPC, and special mixing technology to lower the air content has been developed [23].Obviously, air content of a UHPC mixture decreased with the increase of steel fiber content, and the addition of 3% steel fiber led to about 0.9%-2.4% reduction of entrapped air content for different water-to-binder ratios.Such decreasing effects of steel fiber on air content were also found in the experiments of normal strength concretes [24].The increasing air content in normal concrete was also reported by other researchers, being attributed to the difficulties When the same water-to-binder ratio and steel fiber content was used, the more dosage of superplasticizer increased the fluidity as expected.To increase the flowability, a higher dosage of superplasticizer was needed for UHPC mixtures with a lower water-to-binder ratio, due to a smaller amount of existing free water.For water-to-binder ratios of 0.24 and 0.22, the flowability of the mixtures reached as high as 200 mm by using more superplasticizer, which is favorable for casting.However, for water-to-binder ratios of 0.18 and 0.20, the increased dosage of superplasticizer led to a limited increase of flowability.Therefore, the fluidity was more sensitive to superplasticizer dosage in higher w/b than in lower w/b mixtures [22].Such results reveal that a very low water-to-binder ratio may result in increased difficulty in compacting UHPC mixtures.

The Entrapped Air in Fresh Mixture
The test results of air content for all 64 UHPC mixtures are shown in Figure 7.It can be seen that most of the UHPC mixtures had 4%-6% air content, remarkably higher than that of normal concrete (1%-2%), when no air entraining agent was used.The higher amount of entrapped air has a detrimental effect on the mechanical properties of UHPC, and special mixing technology to lower the air content has been developed [23].Obviously, air content of a UHPC mixture decreased with the increase of steel fiber content, and the addition of 3% steel fiber led to about 0.9%-2.4% reduction of entrapped air content for different water-to-binder ratios.Such decreasing effects of steel fiber on air content were also found in the experiments of normal strength concretes [24].The increasing air content in normal concrete was also reported by other researchers, being attributed to the difficulties of orientation and distribution of long fibers [25].Therefore, the reduction effect of entrapped air content in UHPC mixtures can be explained by the good dispersion of short fibers when a high dosage of superplasticizer was used.Other possible reasons should be studied in future.to-binder ratio.The influence of the water-binder ratio on the air content is not significant when a similar fluidity was obtained.

The Relationship between Flowability and Entrapped Air Content
The relationship of flowability and entrapped air content of UHPC mixtures for every steel fiber addition is plotted in Figure 8.It can be seen that the air content generally decreased with the increase of flowability.It was believed that the nature of changes in air content for non-air-entrained concrete mixtures mainly depend on rheological parameters.Previous experiments showed that when yield stress and plastic viscosity decrease, the diameter of slump flow increases and time of slump flow decreases, and the air content in the SCC (Self Compacting Concrete) mixture also decreases [26].This can be used to explain the tendency of change in entrapped air content with flowability for the UHPC mixtures in this study.The influence of superplasticizer dosage on air content was related to water-to-binder ratio and the addition of steel fiber.The fluctuation range of air content decreased with lowering of the water-to-binder ratio.The influence of the water-binder ratio on the air content is not significant when a similar fluidity was obtained.

The Relationship between Flowability and Entrapped Air Content
The relationship of flowability and entrapped air content of UHPC mixtures for every steel fiber addition is plotted in Figure 8.It can be seen that the air content generally decreased with the increase of flowability.It was believed that the nature of changes in air content for non-air-entrained concrete mixtures mainly depend on rheological parameters.Previous experiments showed that when yield stress and plastic viscosity decrease, the diameter of slump flow increases and time of slump flow decreases, and the air content in the SCC (Self Compacting Concrete) mixture also decreases [26].This can be used to explain the tendency of change in entrapped air content with flowability for the UHPC mixtures in this study.

Compressive Strength
Figure 9 shows the change in compressive strength of UHPC with different water-to-binder ratio, superplasticizer dosage, and steel fiber content.It can be seen that the compressive strength increased distinctly with higher fiber content for UHPC mixtures with the same water-to-binder ratio.When the fiber content increased from 0% to 3%, the average increase of compressive strength for 0.18, 0.20, 0.22, and 0.24 water-to-binder ratio mixtures were 59.1%, 53.9%, 49.8%, and 38.9%, respectively.As we know, increased steel fiber content could decrease the average space between fibers, which made more fibers sustain the load and leads to multi-cracks.Moreover, the addition of steel fiber delays the formation and propagation of cracks, and, thus, resulted in an increase of strength [14].On the other hand, the increased strength might be directly related to the reduction of entrapped air content [23].
For the low water-to-binder ratio of 0.18, a greater addition of superplasticizer reduced the compressive strength.For other water-to-binder ratios, the increase of superplasticizer dosage either increases or decreases the effect on compressive strength.For most of mixtures, the lower compressive strength could be attributed to the increase of entrapped air content in the fresh mixture.
The addition of steel fiber has a great influence on the failure mode of a UHPC specimen, as shown in Figure 10.The UHPC specimens without steel fiber have the typical characteristic of brittle failure under uniaxial compression.A loud noise occurred upon failure, and the specimen was There was nearly a linear relationship between flowability and entrapped air content when no fiber was added.Additionally, this relationship became worse when a higher dosage of fiber was used.This is related to the influence of fiber on air content, and other possible reasons should be further investigated.

Compressive Strength
Figure 9 shows the change in compressive strength of UHPC with different water-to-binder ratio, superplasticizer dosage, and steel fiber content.It can be seen that the compressive strength increased distinctly with higher fiber content for UHPC mixtures with the same water-to-binder ratio.When the fiber content increased from 0% to 3%, the average increase of compressive strength for 0.18, 0.20, 0.22, and 0.24 water-to-binder ratio mixtures were 59.1%, 53.9%, 49.8%, and 38.9%, respectively.As we know, increased steel fiber content could decrease the average space between fibers, which made more fibers sustain the load and leads to multi-cracks.Moreover, the addition of steel fiber delays the formation and propagation of cracks, and, thus, resulted in an increase of strength [14].On the other hand, the increased strength might be directly related to the reduction of entrapped air content [23].

Flexural Strength
The flexural strength results for different UHPC mixtures are presented in Figure 11.When compared with the compressive strength, the addition of 1% steel fiber had a limited increase in flexural strength.In addition, the addition of 2% and 3% steel fiber provided a remarkable increase in flexural strength of about 24.9%-58.2%and 48.8%-88.7%,respectively.This is attributed to the mechanical interlock and friction at the fiber-matrix interface, which suppressed the propagation and development of cracks.Furthermore, the higher addition of fiber provides more crack-bridging effects, thus, increasing the flexural load carrying capacity [24].The variation in water-to-binder ratio and superplasticizer dosage had similar influences on flexural strength as that on compressive For the low water-to-binder ratio of 0.18, a greater addition of superplasticizer reduced the compressive strength.For other water-to-binder ratios, the increase of superplasticizer dosage either increases or decreases the effect on compressive strength.For most of mixtures, the lower compressive strength could be attributed to the increase of entrapped air content in the fresh mixture.
The addition of steel fiber has a great influence on the failure mode of a UHPC specimen, as shown in Figure 10.The UHPC specimens without steel fiber have the typical characteristic of brittle failure under uniaxial compression.A loud noise occurred upon failure, and the specimen was completely crushed (small particle spalled).The addition of steel fiber increased the ductility of UHPC to some extent [3].The specimen was almost intact after the compressive strength test, with only minor cracks on the surface and damaged particles connected by steel fibers.

Flexural Strength
The flexural strength results for different UHPC mixtures are presented in Figure 11.When compared with the compressive strength, the addition of 1% steel fiber had a limited increase in flexural strength.In addition, the addition of 2% and 3% steel fiber provided a remarkable increase in flexural strength of about 24.9%-58.2%and 48.8%-88.7%,respectively.This is attributed to the mechanical interlock and friction at the fiber-matrix interface, which suppressed the propagation and development of cracks.Furthermore, the higher addition of fiber provides more crack-bridging effects, thus, increasing the flexural load carrying capacity [24].The variation in water-to-binder ratio and superplasticizer dosage had similar influences on flexural strength as that on compressive

Flexural Strength
The flexural strength results for different UHPC mixtures are presented in Figure 11.When compared with the compressive strength, the addition of 1% steel fiber had a limited increase in flexural strength.In addition, the addition of 2% and 3% steel fiber provided a remarkable increase in flexural strength of about 24.9%-58.2%and 48.8%-88.7%,respectively.This is attributed to the mechanical interlock and friction at the fiber-matrix interface, which suppressed the propagation and development of cracks.Furthermore, the higher addition of fiber provides more crack-bridging effects, thus, increasing the flexural load carrying capacity [24].The variation in water-to-binder ratio and superplasticizer dosage had similar influences on flexural strength as that on compressive strength.For most of mixtures, less entrapped air content in the fresh state leads to an increased flexural strength of UHPC specimens, as expected.

The Relationship between Compressive Strength and Entrapped Air Content
At the same water-to-binder ratio, the relationship between compressive strength and air content of UHPC specimens, containing different dosages of steel fiber and superplasticizer, are plotted in Figure 12.It can be found that the compressive strength of UHPC increased linearly with the decrease of entrapped air content, regardless of the variable fiber content, ranging from 0% to 3%.

The Relationship between Compressive Strength and Entrapped Air Content
At the same water-to-binder ratio, the relationship between compressive strength and air content of UHPC specimens, containing different dosages of steel fiber and superplasticizer, are plotted in Figure 12.It can be found that the compressive strength of UHPC increased linearly with the decrease of entrapped air content, regardless of the variable fiber content, ranging from 0% to 3%.The regression equations are given in Figure 12a-d, where Fc is the compressive strength (MPa), and Ac is the air content (%).

Conclusions
(1) The addition of steel fiber decreased the flowability and entrapped air content of fresh UHPC mixtures.The fluidity was less sensitive for lower w/b mixtures to superplasticizer dosage than that for higher w/b mixtures.The influence of superplasticizer dosage on entrapped air content was related to the water-to-binder ratio and the addition of steel fiber.
(2) More addition of steel fiber (from 1% to 3%) led to an increasing compressive strength of UHPC specimens with different water-to-binder ratios and superplasticizer dosages.However, the addition of 1% steel fiber causes little increase in flexural strength, and the addition of 2% and 3% steel fiber provided a remarkable increase in flexural strength.
(3) To prepare flowable UHPC mixtures, a very high dosage of superplasticizer was added for lower water-to-binder ratios, which may have an adverse effect on strength gain.Additionally, the more addition of superplasticizer increased the strength of the UPHC mixtures with a higher waterto-binder ratio, due to the lower air content caused by the improved flowability.
(4) For the same water-to-binder ratio, there is a good linear relation between entrapped air content and compressive strength of UHPC specimens containing different dosages of steel fiber.Therefore, the decreasing effect on entrapped air content by the addition of steel fiber has a significant contribution to the improvement of the compressive strength of UPHC.For normal-strength concrete, every 1% decrease in air content increases compressive strength of the concrete by 4%-6% [27], which is much lower than that found in UHPC specimens in this study.It was also mentioned that the air-void characteristics, such as size distribution, spacing (dispersion), and shape of air voids, have substantial effects on the mechanical properties of concrete [28].Unfortunately, there is little information and experimental results available in the literature about this topic.Therefore, the variation in entrapped air structure may be another aspect for influencing the compressive strength of UHPC containing different dosages of steel fiber.Furthermore, the fact that the entrapped air content has a much higher influence on the compressive strength of UHPC than observed for normal concrete should be further investigated.

Conclusions
(1) The addition of steel fiber decreased the flowability and entrapped air content of fresh UHPC mixtures.The fluidity was less sensitive for lower w/b mixtures to superplasticizer dosage than that for higher w/b mixtures.The influence of superplasticizer dosage on entrapped air content was related to the water-to-binder ratio and the addition of steel fiber.
(2) More addition of steel fiber (from 1% to 3%) led to an increasing compressive strength of UHPC specimens with different water-to-binder ratios and superplasticizer dosages.However, the addition of 1% steel fiber causes little increase in flexural strength, and the addition of 2% and 3% steel fiber provided a remarkable increase in flexural strength.
(3) To prepare flowable UHPC mixtures, a very high dosage of superplasticizer was added for lower water-to-binder ratios, which may have an adverse effect on strength gain.Additionally, the more addition of superplasticizer increased the strength of the UPHC mixtures with a higher water-to-binder ratio, due to the lower air content caused by the improved flowability.
(4) For the same water-to-binder ratio, there is a good linear relation between entrapped air content and compressive strength of UHPC specimens containing different dosages of steel fiber.Therefore, the decreasing effect on entrapped air content by the addition of steel fiber has a significant contribution to the improvement of the compressive strength of UPHC.

Figure 1 .
Figure 1.Example of the steel fibers used in this study.

Figure 1 .
Figure 1.Example of the steel fibers used in this study.

Figure 5 .
Figure 5. Measuring instrument for air content.

Figure 5 .
Figure 5. Measuring instrument for air content.

Figure 5 .
Figure 5. Measuring instrument for air content.

Table 1 .
Chemical composition of cementitious materials.

Table 2 .
Physical and mechanical performances of quartz sand.

Table 3 .
Mix proportion of ultra high performance concrete (UHPC).