Abstract

1. Introduction

The aim of this research work was to establish the role of cellulose fibers in preserving the impermeable nature of concrete. Permeability is one of the most critical properties that control durability of concrete. It is defined as the movement of fluid through a porous medium under an applied pressure head; it is in turn governed by two primary factors:

(a) Porosity and interconnectivity of pores in the cement paste. These are controlled for most part by w/c ratio, degree of hydration, and the degree of compaction

(b) Micro cracks in concrete especially at the paste –aggregate interface. These micro cracks are innate feature of concrete, they have significant influence on many of its mechanical properties, including deformation, permeability, and strength.

Density and location of interfacial micro-cracks, on the other hand, are determined by the level of applied stress level, external or internal, experienced by the concrete. Internal stress in concrete occurs as a result of shrinkage, thermal gradients, abrupt changes in the hygro-thermal environment and factors causing volumetric instability.

The current work primarily deals with evaluation of influence of externally applied stress on the permeability of the fiber reinforced concrete, which has so far remained an unexplored area. The work done by Hearn [4] on mature concrete revealed no appreciable effect of stress on the water permeability, on other hand Kermani [7] found that permeability increased significantly when the stress level exceeded to 40% of the ultimate strength. An essential difference between these two studies was that Hearn [4] had subjected the specimens to stress prior to carrying out the permeability tests whereas in Kermani’s tests [7], permeability tests were carried out in the presence of an applied stress. However, when Hearn and Lok [5] carried out nitrogen permeability tests while maintaining a stress on concrete, they too found that there is a threshold value of stress beyond which increases in the permeability occurred.

Studies by Banthia et al [3] also revealed that there exists a ‘threshold’ value of stress, below which a decrease in permeability occurs and above which an increase in permeability can be expected. They also found that fiber reinforcement could increase this ‘threshold’ stress value [1].

An even greater increase in permeability is expected if these loads occur at an early age, when concrete has not yet gained much strength. Early load application, which routinely occurs in real life, may thus significantly increase the permeability and produce concrete in service with inadequate durability.

Fiber reinforcement is known to control cracking resulting from externally applied loads as well from environmental effects including thermal and shrinkage strains [1].

The purpose of this research work was to address following issues, role of fibers in preserving durability concrete and acceptable level of stress on early age fiber reinforced concrete. The current research is also expected to generate Life Expectancy Functions (LEFs) for use in Life Cycle Engineering (LCE) Models.

2. Experimental Technique for Evaluation of Permeability

Depending on the transport mechanism, test methods for measuring water permeability can be divided into several categories [6]:

1. Steady State water permeation

2. Non-Steady State water permeation

3. Capillary suction

Method used for current purpose was essentially a steady state method. Time taken to obtain steady –state depends on various factors that includes water to cement ratio, degree of hydration, age, degree of saturation, thickness and size of the sample.

There is no standard test for measuring water permeability of FRC under stress; hence an indigenously built apparatus at UBC was used for measuring steady state water permeability of concrete samples. Due to reasonable thin sample thickness, steady state was attained.

Some of the salient features of this apparatus include:

§ Sample thickness was 25 mm so that time taken for saturation is less

§ Stress could be applied while monitoring permeability of samples

§ Wet area was maximized

§ Constant hydraulic pressure could be applied to the outside walls while keeping the inner wall free under atmospheric pressure

The experimental method employed in this research made use of comparison technique. In each of the conducted experiments permeability of stressed and unstressed fiber reinforced concrete specimen was compared.

3. Permeability Apparatus

A schematic representation of the apparatus for water permeability testing system is shown in Figure 1. The apparatus primarily consists of four main parts permeability cells, water supplying unit, measuring device and universal testing machine for application of load.

Figure 1 : Schematic Representation of Water Permeability Setup

Permeability cell consists of squared top and bottom aluminum plate each of length 1400 mm and 45 mm thick, a 160 mm diameter aluminum tube, a 50 mm diameter piston, and a 100 mm diameter aluminum disk, all the components are labeled in Figure 2. All these components were assembled together by the means of four studs and four nuts. In order to keep the permeability cells water tight O rings were installed in the grooves made in top and bottom plate.

Figure 2 : Components of Permeability Cell

Water supplying unit consisted of aluminum cylinder with two aluminum plates at its two ends which are fastened together by means of four studs and four nuts. Again, in order to keep the cylinder water tight, O rings were installed, inside the grooves made in top and bottom plate. The driving pressure was regulated by means of air regulator which controlled air pressure applied to the cylinder in order to keep the water at a certain level of pressure head. The pressure head was kept constant through out the progress of test. All the components of water supplying unit are labeled in Figure 3.

Bottom Plate

Aluminum Cylinder

Top Plate

Air Pressure Regulator

Figure 3 : Components of water Supplying Unit

Water measuring device consists of electronic measuring weights connected to the computer. Drained water is continuously monitored and weight is continuously recorded in the computer with an accuracy of 0.01 gram.

The whole setup consist of two identical permeability cells, one of the cell is mounted to universal testing machine [Figure 4] so that a constant level of stress can be applied on one of the samples, where as other cell is kept out side the UTM [Figure 5] so that permeability of the unstressed specimen can be monitored at the same time.

Universal Testing Machine

.

Figure 4 : Stressed Permeability Cell

Figure 5 : Un-Stressed Permeability Cell

4. Experimental Program

Two identical hollow cylindrical concrete specimens were employed to compare the permeability. In order to justify the role of fibers in preserving durability of concrete and acceptable level of stress on early age FRC[1], identical samples with different fiber volume fraction and under different loads were proposed; and tested. A total of eight test pairs were tested at age of 7 days to establish the beneficial role of cellulose fibers in early age concrete. Tables 1 -8 gives the details of test variables.

Table 1: Test Variable

	Series 1(a)		Series 1(b)
Fiber Volume Fraction	0.0%		0.0 %
Applied Stress Level	0	0.3f_u	0	0.5f_u

Table 2: Test Variable

	Series 2(a)		Series 2(b)
Fiber Volume Fraction	0.1%		0.1 %
Applied Stress Level	0	0.3f_u	0	0.5f_u

Table 3: Test Variable

	Series 3(a)		Series 3(b)
Fiber Volume Fraction	0.3 %		0.3 %
Applied Stress Level	0	0.3f_u	0	0.5f_u

Table 4: Test Variable

	Series 4(a)		Series 4(b)
Fiber Volume Fraction	0.5%		0.5 %
Applied Stress Level	0	0.3f_u	0	0.5f_u

5. Test procedure

5.1 Concrete Mix Design

The proportion of different materials used in this experimental work is given in Table 5. CSA Type 10 normal Portland cement (ASTM Type 1), fly ash, coarse aggregate with a maximum size of 9.5 mm, saturated surface –dry (SSD) clean river sand with a fineness modulus of about 2.5 and potable water were used.

Table 5: Mix Design

Materials	Quantity (Kg/m³)
Type 10 Portland Cement	250
Fly Ash	100
Sand	870
Gravel 3/8’’	870
Water	210
W/C	0.6

5.2 Test Sample

Hollow concrete core cylindrical specimens were used for the purpose of testing permeability. Cylindrical concrete specimens with a 100 mm diameter and a height of 200 mm were casted with a 50 mm diameter hollow cylindrical core at the center. Figure 6 shows the hollow core concrete cylindrical specimens.

Figure 6 : Hollow Core Concrete Cylindrical Specimens

In order to evaluate the strength of casted hollow cylindrical samples, three identical regular cylindrical samples were casted at the same time as of hollow cylindrical samples.

To achieve the specified objective three identical cylindrical specimens of 102 mm (4’’) diameter and of height 204 mm (8’’) were used to determine the compressive strength and hollow concrete core cylinder (Figure 6) were used to obtain permeability of concrete.

5.3 Casting

Materials listed in Table 5 were mixed together in a mixer, moisture content of the sand was measured before mixing and the amount of water was adjusted accordingly to keep the w/c ratio at 0.6. Fibers were added after one round of mixing, this composition was allowed to mix for few minutes and then later water was added followed by mixing of few more minutes till we obtained a consistent homogenized mixture of the above ingredients.

Three regular standard cylindrical specimens and two hollow core concrete cylinders were casted in one batch for each of the proposed experiment. Normal PVC molds were used for casting regular cylindrical specimens. However two special molds made at UBC were used for casting hollow core cylindrical specimens. Table 6 gives the details of dimensions of the specimens used in current research work.

Table 6: Specimen Details

Specimen Type

Details

· For: Compressive Strength at 7^th Day (as per ASTM C39)

· Three per Mix

· Diameter = 102mm (4’’)

· Height =204mm (8’’)

· For: Permeability Test at 7^th Day

· Two per Mix

· Diameter = 102mm (4’’)

· Inside Diameter=50mm (2’’)

· Height =204mm (8’’)

Figure 7 gives the details of the molds used for casting hollow core concrete cylinders. These molds consist of four main parts which includes upper ring, body, core and base.

Figure 7 : Molds used for casting hollow core concrete cylinders

Upper ring and core of the mold were made of PVC, as it is easy to remove while demolding. It is to be noted that for regular cylindrical molds, oil was used to make demolding easier, but for hollow cylindrical molds, no lubricant was used in order to prevent the effect of oil on concrete permeability.

Concrete consolidation was done using a table vibrator. Casted samples were allowed to achieve there initial set before covering them with plastic sheets

Demolding was done 24 hours after the casting of the samples. Regular Concrete Cylinders were demolded using air pressure; however demolding hollow cylindrical specimens was a dexterous task. Following procedure was adopted to demold these specimens:

§ PVC core was removed first using hydraulic pressure as shown in Figure 8 on the following page.

§ After removal of PVC core, upper ring was removed gently.

§ Body of the mold was removed from the base, and then it was turned upside down and was installed above hollow iron cylinder so that hollow cylindrical specimen could be removed easily by means of gentle pressure.

Figure 8 : Removal of PVC core

5.4 Curing

After demolding of samples, they were kept in saturated lime water bath until the day of testing.

5.5 Sample Preparation

One of the most critical parts of test was preparation of sample, as it essentially decides success of the test. Leakage of water was essentially controlled by the success of sample preparation.

5.5.1 Grinding

To prepare samples for the test, they were removed from saturated lime water tank after 6 days of curing. They were then grinded till two smooth flat parallel surfaces were obtained at both the ends of the samples. These grinded samples were then left for couple of hours for drying.

5.5.2 Sealant

In permeability tests, boundary leakage through sample is a critical issue, hence key requirement is to prevent leakage through sample boundary, which is in this case was done using rubber rings with an outer diameter of 4’’ and an inner diameter of 2’’. These rings were placed on the dried clean surface of samples using a silicone building sealant. DOW CORING 790 was used as a silicone building sealant. Samples were left for 12 hours of drying after installation of rubber rings using silicone building sealant. Upon application of sealant and installation of rubber rings on the surface, these samples were left under small load so that rubbers rings sticks well to the concrete surface as shown in Figure 9. Final prepared samples are shown in Figure 10.


Figure 9 : Samples under miniscule load	Figure 10 : Sample Preparation

5.6 Assembling Set up

After sample preparation, dried sample were placed tightly on permeability cell base, as shown in Figure 11 -14.


Figure 11 : Permeability Cell Base	Figure 12 : Prepared Sample


Figure 13 : Sample Sitting on Cell Base	Figure 14 : Sample top sealed with Aluminum Cap

As shown in Figure 14, the aluminum disk was tightly fitted on sample top to prevent any leakage of water through sample base and top. It is be to be noted that O shaped grooves were made on the cell base and aluminum cap so that when it is pressed hard against the rubber installed on sample surface it makes 0 shaped grooves on rubber surface, to prevent any undue leakage through sample surfaces. Figure 15 shows the formation of 0 rings on rubber surface of prepared samples.

After installation of sample on cell base, main tube made up of aluminum was fitted tightly on cell base, a small amount of oil was applied on tube surface so that it could be removed easily after permeability test. There after top plate was secured tightly on main tube, it was clamped tightly in order to squeeze the upper disk on sample surface, so that O rings as shown in Figure 15 are formed. All the cell components were tightly secured together by four studs and nuts.

Figure 15 : 0 shaped grooves on sample surface to prevent water leakage

The whole assembly was mounted in a testing machine where a certain compressive stress could be applied on the specimen during the test. Two identical cells were used: one under stress in the Universal Testing Machine as shown in Figure 4 and the other outside of the machine with no stress as shown in Figure 5. Finally, the water supply pipes were connected to the cells as shown in Figure 16.

Figure 16 : Test Setup Showing the Permeability Cell mounted on UTM

The final step before starting test was determination of compressive stress to be applied on specimens, as mentioned earlier compressive stress of 30% and 50% of the ultimate strength was applied on the specimens from each series, to study the effect of stress on permeability of early age FRC. Compressive strength of samples was determined by crushing three samples from each batch. Figure 17 shows the compression testing machine with a capacity of 600, 000 lbs (2659 KN) used in the current research work. Compressive strength tests were done in harmony with ASTM C 39 -96.

Sample

Compressive Strength

Figure 17 : Compressive Strength Testing Machine

A test age of 7 days was selected through out, anticipating that at an age of 7 days, in all instances, secondary loads occur on concrete structures, and primary loads occur in most instances. Each of the test pairs was tested at an age of 7 days for about 30 hours after the equilibrium conditions was attained.

The specimens were placed in a specially designed cell as explained above such that water permeated under a pressure of 0.48 MPa through the 25 mm thick outer wall and was collected in the inner hollow core.

The hydraulic pressure was applied to the outside wall of the hollow cylinder. The driving pressure was controlled by a regulator and kept constant throughout the test and a calibrated pressure gauge was used to indicate the inflow pressure. Extreme caution was exercised to detect any leakage in the system.

The collected water in the hollow core was then drained out to a collection reservoir where its mass was measured continuously and accurately using a computer controlled scale as shown in Figure 16. Load relaxation occurred in the machine with time, and the drop in load was corrected by moving the loading arm downward.

Using Darcy’s law, coefficient of water permeability (Kw) of concrete specimens under steady state condition can be determined using following equation:

K_w= (1)

where:

Kw – Coefficient of Water Permeability (m/s)

Q – Rate of water Flow (m³/sec)

L – Thickness of Specimen Wall (m)

A- Permeation Area (m²)

Δh- Pressure head (m)

6. Permeability Test Results

The whole experimental program was divided into four series of tests, based on fiber volume fractions. Each series of test was further subdivided into two sub-series of tests based on percentage of ultimate load to be applied on specimens as shown in Figure 18

Figure 18 : Flowchart showing detailed experimental program

Results for Series 1 (a): Plain concrete specimens were tested for permeability; two identical specimens were tested at an age of 7 Days, one under a stress of 0.3 f_uand another identical specimen under zero stress. Same experiment was done twice to obtain consistent information. Permeability was observed to decrease by a factor of about 0.3 under 30% of ultimate strength as compared to unstressed specimen. Table 7 gives the information of test variable for this series of Test. Figure 19 shows the test results for this series where as Figure 20 shows the averaged results for the same series.

Table 7: Test Details for Series 1(a)

Series 1 (a)

Specimen Age

(Days)

Compressive Strength (f_u) (MPa)

Driving Pressure

(MPa)

Date of Test

Applied Stress Level (MPa)

0.3f_u

S1(a)

17.59

0.48

Sep 18, 05

5.28

S’1(a)

19.53

0.48

Sep 17, 05

5.86

Figure 19 : Results for Series 1(a)

Figure 20 :Averaged results for Series 1(a)

Results for Series 1 (b): In this series plain concrete specimens were tested, stressed specimens carried a stress of 0.5 f_u.At a stress level of 50% of the ultimate strength permeability of the stressed specimen was found to increase by a factor of 1.38. Experiment was repeated twice to obtain consistent information and permeability was found to increase by approximately same factor. Table 8 gives the information of test variable for this series of Test. Figure 21 shows the test results for this series where as Figure 22 shows the averaged results for the same series.

Table 8: Test Details for Series 1(b)

Series 1 (b)

Specimen Age

(Days)

Compressive Strength (f_u) (MPa)

Driving Pressure

(MPa)

Date of Test

Applied Stress Level (MPa)

0.5f_u

S1(b)

17.11

0.48

July 04, 05

8.55

S’1(b)

22.85

0.48

July 08, 05

11.42

Figure 21 : Results for Series 1(b)

Figure 22 : Averaged results for Series 1(b)

Results for Series 2 (a): Permeability of concrete was tested using 0.1% cellulose fiber, two identical specimens were tested at an age of 7 Days, one under stress level of 0.3 f_uand another identical specimen under zero stress. Same experiment was done thrice to obtain consistent information, but lot of variability was observed in the data. Table 9 gives the information of test variables. Figure 23 shows the test results for this series where as Figure 24 shows the averaged results for the same series.

Permeability was observed to increase in initial few hours followed by sudden decrease in the permeability value relative to the permeability of unstressed specimen.

Table 9: Test Details for Series 2(a)

Series 2 (a)	Specimen Age (Days)	Compressive Strength (f_u) (MPa)	Driving Pressure (MPa)	Date of Test	Applied Stress Level (MPa) 0.3f_u
S2(a)	7	21.04	0.48	Sep 11, 05	6.31
S’2(a)	7	19.42	0.48	Sep12, 05	5.82
S’’2(a)	7	17.20	0.48	Sep 15, 05	5.16

Figure 23 : Results for Series 2(a)

Figure 24 : Averaged results for Series 2(a)

Results for Series 2 (b): In this case also permeability of concrete was tested using 0.1% cellulose fiber; two identical specimens were tested at an age of 7 Days, one under stress level of 0.5 f_uand another identical specimen under zero stress. Same experiment was done twice to obtain consistent information. Permeability of stressed specimen decreased relative to permeability of unstressed specimen by a factor of 0.75. Table 10 gives the information of test variables. Figure 25 shows the test results for this series where as Figure 26 shows the averaged results for the same series.

Table 10: Test Details for Series 2(b)

Series 2 (a)

Specimen Age

(Days)

Compressive Strength (f_u) (MPa)

Driving Pressure

(MPa)

Date of Test

Applied Stress Level (MPa)

0.5f_u

S2(a)

19..99

0.48

Sep 05, 05

9.99

S’’2(a)

17.81

0.48

Sep 12, 05

8.90

Figure 25 : Results for Series 2(b)

Figure 26 : Averaged results for Series 2(b)

Results for Series 3 (a): Fiber reinforced concrete specimens with fiber volume fraction of 0.3% were tested for permeability; two identical specimens were tested at an age of 7 days, one under a stress of 0.3f_uand another identical specimen under zero stress. Same experiment was done twice to obtain consistent information. Permeability was observed to decrease by a factor of about 0.55 under 30% of ultimate strength as compared to permeability of unstressed specimen. Table 11 gives the information of test variable for this series of Test. Figure 27 shows the test results for this series where as Figure 28 shows the averaged results for the same series.

Table 11: Test Details for Series 3(a)

Series 3 (a)

Specimen Age

(Days)

Compressive Strength (f_u) (MPa)

Driving Pressure

(MPa)

Date of Test

Applied Stress Level (MPa)

0.3f_u

S3(a)

20.32

0.48

Aug 09, 05

6.09

S’3(a)

20.06

0.48

Aug 11, 05

6.02

Figure 27 : Results for Series 3(a)

Figure 28 : Averaged results for Series 3(a)

Results for Series 3 (b): In this series concrete specimen with a fiber volume fraction of 0.3% was tested, stressed specimens carried a stress of 0.5 f_u.At a stress level of 50% of the ultimate strength permeability of the stressed specimen was found to decrease by a factor of 0.76. Experiment was repeated twice to obtain consistent information; permeability was found to decrease by approximately same factor. Table 12 gives the information of test variable for this series of Test. Figure 29 shows the test results for this series where as Figure 30 shows the averaged results for the same series.

Table 12: Test Details for Series 3(b)

Series 3 (b)

Specimen Age

(Days)

Compressive Strength (f_u) (MPa)

Driving Pressure

(MPa)

Date of Test

Applied Stress Level (MPa)

0.5f_u

S3(b)

17.75

0.48

July 28, 05

8.87

S’3(b)

19.18

0.48

July 29, 05

9.59

Figure 29 : Results for Series 3(b)

Figure 30 : Averaged results for Series 3(b)

Results for Series 4 (a): Fiber reinforced concrete specimens with fiber volume fraction of 0.5% were tested for permeability; two identical specimens were tested at an age of 7 days, one under a stress of 0.3 f_uand another identical specimen under zero stress. Same experiment was done twice to obtain consistent information. Permeability was observed to decrease by a factor of about 0.44 under 30% of ultimate strength as compared to permeability of unstressed specimen. Table 13 gives the information of test variable for this series of Test. Figure 31 shows the test results for this series where as Figure 32 shows the averaged results for the same series.

Table 13: Test Details for Series 4(a)

Series 4 (a)	Specimen Age (Days)	Compressive Strength (f_u) (MPa)	Driving Pressure (MPa)	Date of Test	Applied Stress Level (MPa) 0.3f_u
S4(a)	7	17.67	0.48	Sep 05, 05	5.30
S’4(a)	7	17.52	0.48	Sep 08, 05	5.28
S’’4(a)	7	20.12	0.48	Sep 12, 05	6.03

Figure 31 : Results for Series 4(a)

Figure 32 : Averaged results for Series 4(a)

Results for Series 4(b): In this series concrete specimen using a fiber volume fraction of 0.5% was tested, stressed specimens carried a stress of 0.5 f_u.At a stress level of 50% of the ultimate strength, permeability of the stressed specimen was found to decrease by a factor of 0.62. Experiment was repeated twice to obtain consistent information and permeability was found to decrease by approximately same factor. Table 14 gives the information of test variable for this series of Test. Figure 33 shows the test results for this series where as Figure 34 shows the averaged results for same the series.

Table 14: Test Details for Series 4(b)

Series 4 (b)

Specimen Age

(Days)

Compressive Strength (f_u) (MPa)

Driving Pressure

(MPa)

Date of Test

Applied Stress Level (MPa)

0.5f_u

S4(b)

17.06

0.48

June 30, 05

8.53

S’4(b)

17.33

0.48

July 02, 05

8.66

Figure 33 : Results for Series 4(b)

Figure 34 : Averaged results for Series 4 (b)

7. Conclusions

Cellulose fiber reinforcement proved to be beneficial in preserving permeability of stressed concrete. At a stress level of 50% f_upermeability was found to decrease at all level of fiber volume fractions that is at (0.1%, 0.3% and 0.5%). Though, 0.1 % and 0.3 % fiber volume fraction had same effect at a stress level of 50% of the ultimate strength of the tested sample, 0.5% fiber had more pronounced effect in preserving permeability of concrete.

At a stress level of 30% f_ueffect of fibers was ambiguous, which indicates that at this stress level fibers do not play significant role in reducing water permeability of the concrete. Table 15 and Figure 35 show the effect of fibers at varying stress level.

Table 15: Effect of Fiber Reinforcement on Permeability of Concrete

Stress Level	Fiber Volume Fraction
Stress Level	0%	0.1%	0.3%	0.5%
0%	1	1	1	1
30%	0.29	0.68	0.55	0.44
50%	1.38	0.75	0.76	0.62

Figure 35 : Effect of Fiber Reinforcement

8. References

[1.] Banthia, N., Biparva, A., and Mindess S., Permeability of Stressed Concrete, Cement and Concrete Research, in press, 2005.

[2.] Banthia, N. and Mindess, S., Water Permeability of Cement Paste, Cement & Concrete Research (USA), 19(5), 1989, pp. 727-736.

[3.] Banthia, N., Pigeon, M., Marchand, J. and Boisvert, J., Permeability of Roller Compacted Concrete, ASCE J. of Materials in Civil Eng. , 4(1), 1992, pp. 27-40.

[4.] Hearn, N., Effect of Shrinkage and Load Induced Cracking on Water Permeability of Concrete, ACI Materials J., 96(6), 1999, pp. 234-241.

[5.] Hearn, N. and Lok, B., Measurement of Permeability Under Uniaxial Compression—A Test Method, ACI Materials J. 95(6), 1998, pp. 691-694.

[6.] J. Kropp and H. K. Hilsdorf, “Performance criteria for concrete durability”, Rilem report 12, (1992).

[7.] Kermani, A., Permeability of Stressed Concrete, Building Research and Information, 19(6),1991, pp. 360-366

[8.] Mehta, P. Kumar and Monteiro, Paulo J.M., Concrete Structure, Properties, and Materials, Prentice Hall, Second Edition, 1993.