Numerical Prediction on the Effects of the Coarse ... - JST

Numerical Prediction on the Effects of the Coarse AggregateSize to the Pipe Flow of Pumped ConcreteMyoungsung Choi

Journal of Advanced Concrete Technology, volume ( ), pp.12 2014 239-249

Ultrasonic Wave Reflection Approach to Evaluation of Fresh Concrete FrictionYannick Vanhove, Chafika Djelal Thierry Chartier,Journal of Advanced Concrete Technology, volume ( ), pp.6 2008 253-260

Relation between the shape of silica fume and the fluidity of cement paste at low water to powder ratiosEtsuo Sakai,Yasuo Kakinuma Kenji Sakamoto, , Masaki DaimonJournal of Advanced Concrete Technology, volume ( ), pp.7 2009 13-20

Effects of Shape and Packing Density of Powder Particles on the Fluidity of Cement Pastes withLimestone PowderEtsuo Sakai, Keisuke Masuda Yasuo Kakinuma, , Yutaka AikawaJournal of Advanced Concrete Technology, volume ( ), pp.7 2009 347-354

https://www.jstage.jst.go.jp/article/jact/6/2/6_2_253/_pdf



http://www.editorialmanager.com/jact/

http://www.j-act.org/index.html

Journal of Advanced Concrete Technology Vol. 12, 239-249, August 2014 / Copyright © 2014 Japan Concrete Institute 239

Scientific paper

Numerical Prediction on the Effects of the Coarse Aggregate Size to the Pipe Flow of Pumped Concrete Myoung Sung Choi1, 2

Received 27 December 2013, accepted 29 July 2014 doi:10.3151/jact.12.239

Abstract The objective of this study is to numerically investigate the effects of coarse aggregate size on the pumpability of con-crete. In actual construction of super-structures, the coarse aggregate size plays a determining factor on concrete pump-ing. Thus, depending on the coarse aggregate sizes, the properties of lubrication layer which is the major role facilitat-ing concrete pumping, were numerically simulated using shear-induced particle migration analysis. Moreover, the parti-cle shape effects according to the coarse aggregate sizes were also considered in numerical simulation to precisely simulate the actual flow conditions. For verification, 1,000m long full-scale tests with three different coarse aggregate sizes (10, 20 and 25mm) were conducted and compared with numerical results. It could be found that when concrete is being pumped, the rheological properties of concrete and lubrication layer highly depend on the coarse aggregate size, which consequently influences the flow condition of pumped concrete, whereas the thickness of lubrication layer re-mained almost constant irrespective of the coarse aggregate sizes.

1. Introduction

Since concrete pumping has been a common technique to transport the fresh concrete for the construction of high-rise buildings or long-span bridge pylons, from a practical point of view, a smaller size coarse aggregate has been empirically used to facilitate the concrete pumping. Even for a structure designed to use a large size coarse aggregate due to structure durability, for example 25 mm coarse aggregate size, based on hands-on experience, a coarse aggregate size is replaced with a smaller size coarse aggregate, for example 10mm or 20mm coarse aggregate size, to secure the pumping of concrete for the high speeds or long distance placement. This implies that the coarse aggregate size is highly related to the pumpability of concrete.

There have been various studies indicating that the rheological properties of concrete are affected by the aggregate characteristics, which may be a function of the particle size, concentration, shape and surface tex-ture (Hu and Wang 2011; Tattersall and Banfill 1983; Tattersall 1991; Ferraris 1999; Struble and Sun 1995; Geiker et al. 2002; Mahaut et al. 2008, Denis et al. 2002; Kurokawa et al. 1996). However, most studies have focused on the effects of the aggregate characteris-tics on the rheological properties of mortar or concrete itself. The study on the effects of the aggregate charac-teristics to the pumpability of concrete has scarcely been

performed thus far due to the difficulty in experimenta-tion on concrete pumping.

Recently, there have been some numerical studies in order to estimate the pumpability of concrete before actual testing (Jo et al. 2012; Choi et al. 2013a). Jo et al. (2012) tried to simulate the concrete flow inside the pipe considering the formation of lubrication layer. Choi et al. (2013a) developed the numerical prediction method based on the consideration of the particle shape effects on pumped concrete flow. However, Jo et al.’s work has some limitations in estimating the real per-formance of concrete pumping due to not considering the effects of the aggregate characteristics and Choi et al.’s work was not considering the effects of coarse ag-gregate size on pumpability of concrete.

Therefore, the objective of this study is to numeri-cally estimate the pumpablity of concrete according to the coarse aggregate size considering the effects of par-ticle shape. Most studies on the concrete pumping (Jo et al. 2012; Choi et al. 2013a; Alekseev 1953; Weber 1968; Morinaga 1973; Sakuta et al. 1989; Jacobsen et al. 2009; Kaplan et al. 2005; Kwon et al. 2013a, Kwon et al. 2013b, Choi et al. 2013b) have indicated that the lubrication layer which is formed at the interface be-tween the concrete and the wall of the pipe plays a cru-cial role in facilitating concrete pumping (Cf. Fig. 1). When concrete is pumped, a redistribution of particles occurs near the wall of the pipe due to the gradient of the shear stress (Phillips et al. 1992; Ingber et al. 2009; Lu et al. 2008). This is a common feature of particle suspensions, and initially well-mixed particles in con-centrated suspension flows undergo migration from re-gions of higher shear stress to those of lower shear stress. The concrete pumping can therefore be consid-ered in most cases as the shearing of an annular layer that is a few millimeters thick and has much lower vis-

1Civil Engineering Research Team, Daewoo Institute of Construction Technology, Suwon, Korea. 2Dept. of Civil and Environmental Engineering, Korea Advanced Institute of Science and Technology, DaejeonKorea. E-mail: [email protected],

M. S. Choi / Journal of Advanced Concrete Technology Vol. 12, 239-249, 2014 240

cosity than the concrete itself. A possible mechanism that contributes to the formation of the lubrication layer is shear-induced particle migration (Jo et al. 2012; Choi et al. 2013a; Phillips et al. 1992). Thus, to estimate the concrete pumpability according to the coarse aggregate size, in this study, the shear-induced particle migration analysis was conducted to analyze the properties of lu-brication layer. Moreover, in order to accurately evalu-ate the effect of coarse aggregate size, along with parti-cle size difference, the particle shape effects according to the coarse aggregate size were also considered. The particle shape effects were determined through the digi-tal image processing techniques (Kwan et al. 1999; Mora et al. 2000; Otsu 1979). For verification, the nu-merically calculated velocity profile and pressure varia-tion were compared with experimental results of 1,000 m long full-scale pumping tests. From the comparison, it could be found that the pumpability of concrete is significantly affected by the coarse aggregate size and the numerical simulation considering the particle shape effects depending on the coarse aggregate size can accu-rately predict the pumpability of concrete.

2. Numerical methodology for the pipe flow of pumped concrete

2.1 General flow equations The concrete flow in a pipe occurs at low particle Rey-nolds numbers (Re<<1) and the laminar flow approxi-mation is applicable (Wang and Lee 1989). Thus, the conservation equations for a steady state, incompressi-ble flow rate are as follows (Lam et al. 2002):

0u∇ ⋅ = (1)

0p τ−∇ +∇ ⋅ = (2)

where u is the velocity vector, p is the mechanical pressure, and τ is the extra stress tensor.

The extra stress tensor (τ ) is simply given by:

( ), , s gη γ φτ γφ= (3)

where the subscripts s and g represent the sand and coarse aggregate, respectively. sφ and gφ are the vol-

ume fraction of sand and coarse aggregate, respectively, and η is the apparent viscosity and γ is the strain rate tensor and the magnitude ( γ ) can be written as fol-lows:

( )0.5 2 2 2 0.52 21 1[ ( : )] [ ]2 2 rr zz rzθθγ γ γγ γγ γ + += = + (4)

The components of strain rate tensor, ( )Tu uγ = ∇ + ∇ in cylindrical coordinate system, are given by:

2 2 2, , ,

r r zrr zz

z rrz zr

u u ur r z

u ur z

θθγ γ γ

γ γ

∂ ∂= = =

∂ ∂∂ ∂

= = +∂ ∂

(5)

For axisymmetric flows and cylindrical coordinates, Eqs. (1) and (2) can be rewritten as:

0r r zu u ur r z

∂ ∂+ + =

∂ ∂ (6)

0rr rr zrpr r r z r

θθττ τ τ∂ ∂∂− + + + − =∂ ∂ ∂

(7)

0rz rz zzpz r r z

τ τ τ∂ ∂∂− + + + =∂ ∂ ∂

(8)

Since the extra stress tensor is symmetric, rz zrτ τ= , for a pressure-driven pipe flow, the following relations are satisfied:

z r∂ ∂∂ ∂

, r zu u , ( ),z zu u r z= , ( )p p z= (9)

These relations, known as the lubrication approxima-tion theory (Wang and Lee 1989), simplify the conser-vations in Eqs. (7) and (8) to give:

1 ( ) 0zupz r r r

γη∂∂ ∂

− =∂ ∂ ∂

(10)

2.2 Shear-induced particle migration A possible mechanism that illustrates the formation of the lubrication layer is shear-induced particle migration (Jo et al. 2012; Choi et al. 2013; Phillips et al. 1992; Leighton et al. 1987a; Leighton et al. 1987b). Leighton et al. (1987a, 1987b) suggested phenomenological mod-els for particle migration in non-homogeneous shear flows that typically result from spatial variation in irre-versible interaction frequency and effective viscosity. Phillips et al. (1992) adapted the scaling arguments of Leighton et al. (1987a, 1987b) and proposed a diffusive flux equation to describe the time evolution of the parti-cle concentration based on a two-body interaction model. In this study, the particle diffusive model pro-posed by Phillips et al. (1992), combined with general flow equations, was extended to solve the flow of con-crete and predict the particle concentration distribution of suspensions in a pressure driven pipe flow.

Fig. 1 Schematic pattern of concrete flow in the pipe (Choi et al. 2013b).


The governing equation of the shear-induced particle migration for the Poiseuille flow was as follows,

{ }2 2 2( )z z z

c

u u ua K K a

t z r rη

φφ ηφ φ φ

η

∂ ∂ ∂∂ ∇+ = ∇ ∇ +

∂ ∂ ∂ ∂

⎛ ⎞⎜ ⎟⎝ ⎠

i (11)

where φ is the particle concentration, t is the time, zu is the velocity component in the flow direction, a

is the particle radius, z is the flow direction, r is the radial direction, η is the apparent viscosity of the con-centrated suspension, and cK and Kη are dimen-sionless phenomenological constants.

In the pipe flow of pumped concrete, the shear stress is the highest at the wall of the pipe and linearly de-creases as the position moves to the center of the pipe. The stress gradient is a driving force to move particles toward the center of the pipe as described in the first term of the right side in Eq. (11). The increase of the particle concentration due to the migration may increase the viscosity and the yield stress, which hinder the addi-tional migration of the particles as described in the sec-ond term of the right side in Eq. (11). As a result, the concentration of the particle inside the pipe is deter-mined by the balance between the two actions, namely, the migration due to the stress gradient and the hin-drance due to the increased rheological properties.

The governing Eqs. (10) and (11) should be supple-mented with appropriate boundary conditions at the pipe wall, which is subjected to the usual no slip condition ( 0u = ) and no particle flux through the wall as ex-pressed below:

2 0z zc

u uK K n

r rηηφ φ φη

⎡ ∂ ∂ ⎤∇⎛ ⎞∇ + ⋅ =⎢ ⎥⎜ ⎟∂ ∂⎝ ⎠⎣ ⎦ (12)

The particle concentration is assumed to be uniform initially at the entrance of the pipe, i.e.

0φ φ= for 0 r R≤ ≤ at 0z = (13)

where 0φ is the initial concentration of particles. The values of cK and Kη were determined to 0.3 and 0.6, respectively (Phillips et al. 1992)

For the shear-induced particle migration according to the coarse aggregate size, in this study, three different coarse aggregate sizes (i.e. 10mm, 20mm and 25mm maximum coarse aggregate sizes) were considered and each particle size is solved using Eq. (11) along with general flow equations. Moreover, to solve the Eq. (11), the apparent viscosity η is the function of particle con-centration and particle shape according to the coarse aggregate size so that it needs to be properly determined to solve the governing equations.

2.3 Effects of particle size on rheological prop-erties Concrete is a composite material consisting of water, cement and aggregates as the main components. As a suspension point of view, concrete is a concentrated suspension of solid particles in a viscous liquid (i.e.

paste or mortar) and aggregate characteristics and con-tent significantly influence the rheological properties of concrete. Although several factors can determine the rheological properties of suspension, in this study, in order to investigate the effects of coarse aggregate size on concrete pumpability, along with coarse aggregate size difference, the particle shape effects according to the coarse aggregate size were considered.

In order to simulate the particle shape effects accord-ing to the coarse aggregate size, the Farris (1968) model which is a modified equation of Krieger and Dougherty (1959) equation was used for the plastic viscosity of concrete, and a similar pattern of equation for the yield stress was used (Chateau et al. 2008; Hafid et al. 2010). It is worth noting that to improve the theoretical validity, it will be necessary to take into account polydisperse system of concrete component (de Larrard 1999; Vu et al. 2010) but these methods is beyond the scope of the present paper.

2.3.1 Yield stress The apparent viscosity η of Eqs. (10) and (11) for the Bingham fluid model can be expressed as follows:

0p

τη η

γ= + i (14)

where pη and 0τ are the plastic viscosity and yield stress of Bingham fluid model, respectively. The yield stress, 0τ , is a function of the particle concentration, and the following existing model for the yield stress was exploited (Chateau et al. 2008; Hafid et al. 2010),

, max ,max0 0 2.5 2.5

, max , max

11( ) (0)

(1 / ) (1 / )s g

gs

s g

φ φ

φφτ φ τ

φ φ φ φ− −

−−=

− − (15)

where sφ and gφ are the volume fraction of sand and coarse aggregate and ,maxsφ and ,maxgφ are the maxi-mum concentration of sand and coarse aggregate, re-spectively.

Through the preliminary investigation on the effects of the rheological properties of the lubrication layer on the concrete pumping (Choi et al. 2013b), the yield stress causes a minor effect on the pipe flow compared to the effects of the viscosity. On the other hand, the plastic viscosity pη

is greatly influencing the flow rates or the pump pressure. Therefore, it needs to be carefully investigated the effect of particle concentration on plastic viscosity. It is noting that in this study, to take into account the particle shape effects on plastic viscos-ity, instead of using estimate of Chateau et al (2008) like yield stress estimate, Farris (1968) theory devel-oped for multimodal suspension was extended.

2.3.2 Plastic viscosity Farris (1968) developed a theory for calculating the vis-cosity of multimodal suspensions. Concrete can be con-sidered as a three-phase material that consists of paste,


sand and coarse aggregate. The viscosity of multimodal suspension can be calculated from the unimodal viscos-ity of each size as long as the relative sizes are sufficient to ensure no interaction between the coarse aggregate and sand particles. The Farris model was used in the following form:

,max,max [ ][ ]

,max ,max

1 1g gs s

sr

gs

g

η φη φφφ

ηφ φ

−−⎛ ⎞⎜ ⎟

⎛ ⎞= − −⎜ ⎟⎜ ⎟⎝ ⎠ ⎜ ⎟

⎝ ⎠ (16)

The Farris model breaks the aggregate parameters into two modes: sand and coarse aggregate. The volume fraction, i.e. sφ and gφ of each mode is determined separately according to the mix proportion and ,maxsφ and ,maxgφ is the maximum solid fraction of sand and coarse aggregate. Through Liu (2000) model, the maximum solid fractions of each mode were determined as 0.675 for the sand and 0.650, 0.645 and 0.638 for the 10 mm, 20 mm and 25 mm coarse aggregates sizes, re-spectively, for this study condition. Where sη and gη are the intrinsic viscosity of the sand and coarse aggre-gate, respectively, and it represents the effects of indi-vidual particle shape on viscosity. It is defined as fol-lows:

[ ] 1limo

r

φ

ηη

φ→

−= (17)

where rη is determined by Eq. (16). 2.4 Determination of particle shape effects ac-cording to the coarse aggregate size When the multimodal suspension viscosity model, i.e. Farris model, was investigated, the same constant value, i.e. 2.5, as an intrinsic viscosity was usually used. In effect, the intrinsic viscosity represents the effects of the individual particle shape on the viscosity. According to Struble and Sun (1995) and Barnes et al. (1989), the intrinsic viscosity is 2.5 for spherical particles, and when the particles deviate from spherical shape, this value should be increased. The well-known Einstein equation is the original form for the dilute suspension of spherical particles. This equation predicts that the intrin-sic viscosity is 2.5 for spherical particles and measured values confirm this prediction. Ball and Richmond (1980) made the assumption that the effect of all the particles in a concentrated suspension is the sum of the effects added sequentially by writing the Einstein equa-tion in different form. The Krieger and Dougherty (1959) expression replaces the 5/2 fraction of the Ball and Richmond (1980) equation fraction with the intrin-sic viscosity. The intrinsic viscosity is the limiting value of the reduced viscosity as the concentration of particles approaches zero (Cf. Eq. 17). By replacing the 5/2 frac-tion with intrinsic viscosity, any particle shapes can be considered with the expression. Study on a range of particulate systems such as submicron spheres, ground particles, and glass rods and fibers, were reviewed by Barnes et at. (1983). Hinch and Leal (1972) showed

how intrinsic viscosity varies with aspect ratio for high and low shear is summarized. Therefore, to simulate the effect of particle shape according to the coarse aggre-gate size on plastic viscosity, it is reasonable to obtain some relationship between the intrinsic viscosity and the shape characterization of the aggregates (Choi et al. 2013a; Barnes et al. 1983; Szecsy 1997).

As concrete constituents deviate from spherical shape, modified intrinsic viscosity values should be used for each coarse aggregate to simulate the exact particle shape. Szecsy (1997) found the relationship between the intrinsic viscosity values and circularity in order to cal-culate the overall viscosity (Cf. Fig. 2). The maximum limit of 10 is set for the intrinsic viscosity as it is the largest reported value for intrinsic viscosity by Barnes et al. (1983). The circularity is calculated as follows:

2

4

Circularity

A

Pπ

π

×

=⎛ ⎞⎜ ⎟⎝ ⎠

(18)

where A is the cross sectional area of particle, and P is the perimeter of particles. In this study, to find out the parameters of circularity according to the coarse aggre-gate size, the digital image processing technique was used and more detailed statements were included in the next chapter. 3. Experimental program

3.1 Concrete mixes Three different maximum coarse aggregate sizes, i.e., 10 mm, 20 mm and 25 mm, were selected and the fine-ness moduli were 4.95, 5.65 and 6.72, respectively, as listed in Table 1. The cement was CEM I 52.5 N with a density of 3150kg/m3. The sand was natural river sand with a density of 2590kg/m3 and a fineness modulus of 2.81. The sand particles size ranged from 0.08 mm to 5 mm with a water absorption capacity of 2.43 %. The coarse aggregate was a limestone aggregate material

Fig. 2 Relationship between intrinsic viscosity values and circularity.


with a water absorption capacity of 0.8 % and a density of 2610kg/m3. The amount of mixing water was cor-rected to take into account the water absorbed by the sand and the coarse aggregates. A polycarboxylate-based high-range water-reducing admixture (HRWRA) was used. As listed in Table 1, its dosage was adjusted so as to obtain the same slump flow, marked as % HRWRA, meaning the percentage of the admixture rela-tive to the binder content (in mass). 3.2 Rheological measurements To calculate the rheological properties of concrete through Eqs (15) and (16), the rheological properties of cement paste were measured by using a commercially available digital Brookfield viscometer (Model DV-II pro) equipped with smooth cylindrical spindle. The de-vice keeps the temperature of the specimen constant during the entire time span of the rheological test through a water circulation system around the sample container. The cement paste was tested at the age of 15 min after water addition time. The tests were executed stepwise at 100, 50, 30, 20, 10 and 5 rpm. At each rota-tional speed, torque and apparent viscosity data were recorded. The recorded results were then converted into viscosity functions such as shear stress (Pa) and shear rate (1/s) using standard procedure (Mitschka 1982). The linear regression analysis was conducted to deter-mine the plastic viscosity and yield stress as slope and intercept of the regression line drawn through the data points in shear stress versus shear rate plot (Lachemi et al. 2004). 3.3 Pumping circuit A horizontal pumping circuit with a length of 1,000 m was installed (Cf. Fig. 3). It consists of nine 180o bends with a diameter of 2.0 m and the pipe diameter was 125 mm and its thickness was 7.7 mm. The concrete pump is a high-pressure piston pump (model BSA14000HP-D, Putzmeister Co., Germany). The specifications of the model are given in Table 2.

The pumping circuit was equipped with 11 pressure gauges (model JUMO dTRANs p20, JUMO GmbH & Co. KG, Germany). The detailed locations of the gauges are shown in Fig. 4.

3.4 Ultrasonic velocity profiler (UVP) In this study, to investigate the thickness variation of the lubrication layer according to the coarse aggregate size, the actual velocity profile near the wall of the pipe was measured with a special sensor known ultrasonic veloc-ity profiler (UVP) (Choi et al. 2013a; Choi et al. 2013b). The detailed specifications of the sensor are given in Table 3. Other information about the sensor, such as its

Table 1 Mix proportions.

Mixes Design strength (MPa)

w/c ratio

Cement(kg/m3)

Sand (kg/m3)

Coarse aggregate (kg/m3)

Coarse aggregate size(mm)

Fineness modulus

Polycarboxylate-based HRWRA

(%)

Slump flow (mm)

C10 10 4.95 0.66 600 C20 20 5.65 0.75 590 C25

50 0.33 500 736 871 25 6.72 0.90 620

Fig. 3 Overview of the 1,000 m full-scale test setup.

Fig. 4 Schematic ground plan of the 1,000m full-scale pumping circuits and the location of pressure gauges (The values indicate the distance from the beginning of the circuit).

Table 2 Pump specifications.

Item Content Model BSA14000HP-D

Max. Flow rate (m3/h) 82*/54** Max. pressure (bar) 185*/260** Engine house (kW) 350

Stroke/min 26*/17** * rod side, ** piston side

Table 3 Experimental UVP parameters.

Item Content Frequency (MHz) 8 Cycles per pulse 2~32 No. of profiles 1024

Sound velocity (m/s) 2680± 200 Doppler angle (o) 85± 0.5

Spatial resolution (mm) Min. 0.20


principle of measurement, the installation of the ultra-sonic probe and some limitations can be found in earlier works (Choi et al. 2013b; Met-flow 2002). In addition, for application of the UVP using ultrasonic waves, a 1 m transparent engineered plastic pipe with the same diameter as a standard pipe was installed in the last sec-tion, as shown in Fig. 5.

When ultrasonic waves propagate in a medium con-taining coarse particles, such as concrete, the ultrasound pulse hits the particles and part of the ultrasound energy is scattered and lost for the echo measurement. Thus, as the measuring depth increases, the amplitude of echoed ultrasound energy decreases. This is especially true for the high frequency ultrasonic waves used in this work. Thus, above a given depth, the amplitude of echoed ultrasound energy is not sufficiently strong to detect the Doppler shift frequency and give access to the velocity profile. Although this limited measurement range does not allow obtaining the full velocity profile in the cross section, the measured thickness exceeds the expected thickness of the lubrication layer and provides valuable information to examine the variation of thickness de-pending on coarse aggregate size.

4. Numerical simulation for the pumping test

4.1 General The numerical simulations considering pumping tests were conducted using shear-induced particle migration. To consider the effects of particle shape according to the coarse aggregate size, the circularity of particles

were measured through digital image processing and the modified intrinsic viscosity was used to determine the plastic viscosity which is the value for applying the nu-merical simulations. It needs to investigate the detailed calculation results of circularity and the condition of numerical simulation. 4.2 Determination of intrinsic viscosity To determine an optimum intrinsic viscosity considering the particle shape effects, the circularity were quantified using digital image processing technique (Choi et al. 2013a; Kwan et al. 1999; Mora et al. 2000; Otsu 1979). This technique is essentially composed of stepwise tasks (Cf. Fig. 6). First, the image of aggregates from an opti-cal microscope with a Charged Couple Device (CCD) was captured by normal light. The aggregates were then automatically detected in a binary image which was converted from the grayscale image based on a set threshold. Next, each aggregate was labeled and the area and perimeter of aggregates was provided and calcu-lated the circularity value using Eq. (18). Through the relationship between the intrinsic viscosity and the cir-cularity in Fig. 2, the calculated intrinsic viscosities of 10 mm, 20mm and 25 mm coarse aggregates are deter-mined as 5.0, 5.3 and 6.0, respectively, which are the averaged values of representative 50 samples of each aggregate which are selected randomly. The intrinsic viscosity of sand was also determined as 3.5. 4.3 Pipe modeling and analysis conditions Figure 7 shows the two-dimensional axisymmetric modeling of the pipe and part of the mesh. From the previous studies (Jo et al. 2012; Choi et al. 2013a; Kwon et al. 2013b; Choi et al. 2013b), the pressure dis-tribution along the pipe length has a nearly linear rela-tion, which means that the losses of pressure in the el-bows are identical to the losses in the straight pipe. Therefore, for an effective numerical analysis, the 10 m length of the pipe was modeled for the shear-induced particle migration analysis. Moreover, as a macro-scale point of view, concrete could be considered as a homo-geneous material with continuous particle distribution. So, initial concrete is assumed to be homogeneous con-dition at the entrance of the pipe. The commercial CFD code Fluent (2012) software was used to calculate the

Fig. 5 Application point of the UVP and transparent en-gineered plastic.

(a) original image (b) binary image (c) analyzed image

Fig. 6 Quantification process of digital image processing for the calculation of the circularity.


flow conditions. In addition, a subroutine program called the User Defined Scalar (Fluent 2012) was coded to solve Eq. (13) considering the different coarse aggre-gate size and the multimodal suspension viscosity model with particle shape effects. The subroutine pro-gram was interlocked with the general flow analysis. 5. Results and discussion

5.1 Particle shape effects Through the numerical simulation on the pipe flow of pumped concrete considering the shear-induced particle migration, the velocity profile crossing the radius for concrete C10 case was illustrated representatively in Fig. 8. The value is defined as normalized values; the veloc-ity is relative to the maximum velocity and the radius is relative to the unit length. The lubrication layer is a re-gion where the shear rate sharply increases at the wall.

When examining the velocity profile, a significant change in the slope is shown within a limited zone, which represents the lubrication layer and shear rates which is the slope of the velocity profiles concentrated in this layer. Thus, the multimodal suspension approxi-mations can be effectively used to simulate the lubrica-tion layer. However, although the multimodal suspen-sion viscosity model, i.e. Farris Model, can simulate the formation of the lubrication layer, there is some dis-crepancy between the numerically calculated velocity profile and experimentally measured velocity profile using UVP as shown in Fig. 8. The experimentally measured velocity profile had a more abrupt change in its slope near the wall, which results in higher flow rate than that of numerically calculated. The discrepancy was caused by the neglect of particle shape effects when conducting shear-induced particle migration analysis. To take into account particle shape effects, the modified intrinsic viscosity values, i.e. 3.5 for the sand and 5.0, 5.3 and 6.0 for the 10 mm, 20 mm and 25 mm coarse aggregate sizes, respectively, were used in the shear-induced particle migration analysis. After the particle shape effects were applied with increased intrinsic vis-cosity, the calculated velocity profile patterns become almost close to the measured velocity profile, as shown in Fig. 9. Through this comparison, it could be found that the increased intrinsic viscosity reflecting the parti-cle shape effects should be carefully used to simulate the real concrete flow in the pipe. 5.2 Calculation of rheological properties Figure 10 showed the numerically calculated results for the plastic viscosity and yield stress crossing the radius in the pipe. The values are relative to the results of con-crete C10. When examining the calculated results, the plastic viscosity and yield stress varied greatly depend-ing on the coarse aggregate size. As the coarse aggre-gate size becomes larger, the rheological properties be-come increased. This stems from the fact that a concrete

Fig. 8 Comparison of the normalized velocity profile crossing the radius between numerical results using the multimodal suspension viscosity model (i.e. Farris model) and experimental results using UVP for concrete C10 (The UVP measured only certain amount of depth due to insufficient echoed ultrasound energy).

(a) Modeling of the pipe

(b) Mesh section (above dashed quadrangular)

Fig. 7 Pipe modeling and mesh configuration.


mix having larger size coarse aggregate demands a higher level of consistency for a homogeneous matrix, which results in higher rheological properties for the

concrete and the lubrication layer. Thus, through nu-merical analysis, it could be found that the rheological properties can be varied depending on the coarse aggre-gate size, which results in a governing reason for deter-mining the flow condition in the pipe. 5.3 Thickness of lubrication layer In order to investigate the thickness variation of lubrica-tion layer depending on the coarse aggregate size, nor-malized velocity profiles which were calculated numeri-cally are compared in Fig. 11. As shown in the normal-ized velocity profiles, the thickness of the region which underwent a considerable change, namely the lubrica-tion layer, is nearly identical regardless of coarse aggre-gate size, which means that the thickness of the lubrica-tion layer has nearly constant values regardless of the coarse aggregate size. This result could be explained by the fact that the variation of the thickness of the lubrica-tion layer is offset by the effect of the rheological prop-erties and the stress gradient of the coarse aggregate. As the coarse aggregate size become larger, the rheological properties and the stress gradient become increased si-multaneously, so that the increased rheological proper-ties hinder the additional migration of the particle while the larger stress gradient which is a driving force to move particles toward the center of the pipe is created due to the larger coarse aggregate size. Thus, the thick-ness of lubrication layer was constantly maintained by the balance between the hindrance due to increased rheological properties and the migration due to the in-creased stress gradient.

To verify the numerical results, the axial velocities experimentally measured by means of UVP are com-pared in Fig. 12 depending on the coarse aggregate size. The concentration thickness, i.e., the thickness of the lubrication layer, has a nearly identical profile in all cases, which showed a good agreement with the nu-merical results. Thus, it could be confirmed that the thickness of lubrication layer does not vary according to the coarse aggregate size.

Fig. 9 Comparison of the normalized velocity profile crossing the radius between the numerical results con-sidering modified intrinsic viscosities and experimental results using UVP for concrete C10.

Fig. 11 Numerical results on the normalized velocity pro-files crossing the radius depending on the coarse aggre-gate size.

(a) Plastic viscosity

(b) Yield stress

Fig. 10 Numerical results on the rheological properties profiles crossing the radius depending on the coarse aggregate size. (The values are relative to the results of concrete C10).


5.4 Concrete pumpability depending on the coarse aggregate size To take a look at the effect of the coarse aggregate size on the flow rate, the relative velocity profile was calcu-lated in Fig. 13. The values are relative to the results of concrete C25. Under same pressure inlet boundary con-dition, as the coarse aggregate size becomes smaller, the relative velocity profile becomes increased. On the con-trary, concrete having a larger coarse aggregate size demands higher rheological properties to maintain the consistency of the concrete matrix, which result in rela-tively small velocity profile.

Moreover, the pressure distributions along the pump-ing circuit depending on the coarse aggregate size are investigated. The values are relative to the results of concrete C10. As shown in Fig. 14, under same flow rate condition, the relative magnitudes of the pressure show significantly different along the pipe depending on the coarse aggregate size. To verify the numerical re-sults of pressure distribution, the experimentally meas-ured pressures along the pipeline using designated pres-sure gauges are compared in Fig. 14 and show a good agreement with the numerical results.

With these comparison results, for a better under-standing, the required pressures per unit length are cal-

culated and compared under the same flow rate condi-tion, in this case 30 m3/h, depending on the coarse ag-gregate size. The results are illustrated in Fig. 15 and also obtain a good agreement with experimental results. According to previous studies (Choi et al. 2013b; Kwon et al. 2014b; Kaplan et al. 2005; Feys et al. 2005) the obtained values of pressure per unit length are reason-able. When examining the required pressures according to the coarse aggregate size, the concrete using 25 mm coarse aggregate size requires nearly twofold the pres-sure compared to the 10 mm coarse aggregate size, which implies that by changing the coarse aggregate size in the concrete, the required pressure could be sig-nificantly reduced, making it possible to securely pump concrete at low pressures or long distances placement in actual sites.

Through these numerical and experimental results, it can therefore be concluded that the pumpability of con-crete is significantly affected by the coarse aggregate size so that this parameter is a dominant factor for con-crete pumping. As the coarse aggregate size becomes larger, higher pressure is needed to overcome the in-creased rheological properties while maintaining a con-stant lubrication layer thickness. This result was noted in all tested cases in this study.

Fig. 12 Normalized velocities by means of UVP depend-ing on the coarse aggregate size.

Fig. 13 Numerical results on the relative velocity profile crossing the radius depending on the coarse aggregate size.

Fig. 14 Comparison of relative pressure between nu-merically calculated results and experimentally meas-ured data depending on the coarse aggregate size.

Fig. 15 Comparison of the required pump pressure per unit length to obtain a 30 m3/h flow rate according to the coarse aggregate size.


6. Conclusion

The numerical simulations for the pipe flows of pumped concrete depending on the coarse aggregate size using the shear-induced particle migration were conducted and compared with 1,000 m long full-scale pumping test and the following conclusions were obtained. 1. The shear-induced particle migration using multi-

modal suspension approximations can be effec-tively used to simulate the formation of lubrication layer and to investigate the flow conditions in the pipe.

2. The increased intrinsic viscosity reflecting the par-ticle shape effects according to the coarse aggre-gate size should be used to simulate the real con-crete flow in the pipe.

3. Through the numerical simulation, the rheological properties can be varied so much depending on the coarse aggregate size, which plays an important role in determining the flow condition in the pipe.

4. However, the thickness of lubrication layer was constantly maintained by the balance between the hindrance due to increased rheological properties and the migration due to the increased stress gradi-ent.

5. The velocity profile crossing the radius and pres-sure distribution along the pipeline are varying so much depending on the coarse aggregate size so that by changing the coarse aggregate size in con-crete it makes possible to securely pump concrete with low pressures or at long distances placement in actual sites.

6. Through these numerical and experimental results, the pumpability of concrete is significantly af-fected by the coarse aggregate size, which could be one of the dominant factors to determine the con-crete pumping.

Acknowledgement This research was supported by a grant from the Con-struction Technology Innovation Program (08CTIPE01-Super Long Span Bridge R&D Center) funded by Min-istry of Land, Transportation and Maritime Affairs (MLTM) of Korean government. References Alekseev, S. N., (1952). “On the calculation of

resistance in pipe of concrete pumps.” Mekhanizatia Storitel’stva, 9(1), 8-13.

Barnes, H. A., Hutton, J. F. and Walters, K., (1989). “An introduction to rheology.” Elsevier Science.

Ball, R. and Richmond, P., (1980). “Dynamics of colloidal suspensions.” Journal of Physical Chemistry and Liquids, 9, 99-116.

Choi, M. S., Kim, Y. J. and Kwon, S. H., (2013a). “Prediction on pipe flow of pumped concrete base on shear-induced particle migration.” Cement and Concrete Research, 52, 216-224.

Choi, M. S., Roussel, N., Kim, Y. J. and Kim, J. K., (2013b). “Lubrication layer properties during con-crete pumping.” Cement and Concrete Research, 45(3), 69-78.

Chateau, X. Ovarlez, G. and Trung, K. L., (2008). “Homogenization approach to the behavior of suspension of non-colloidal particles in yield stress fluids.” Journal of Rheology, 52, 489-506.

De Larrard, F., (1999). “Concrete mixture proportioning: A scientific approach.” E&FN Spon, London.

Denis, A., Attar, A., Breysse, D. and Chauvin, J. J., (2002). “Effect of coarse aggregate on the workability of sandcrete.” Cement and Concrete Research, 32(5), 701-706.

Ferraris, C. F., (1999). “Measurement of the rheological properties of high performance concrete: state of the art report.” J. Res. Natl. Inst. Stan., 104(5), 461-477.

Farris, R. J., (1968). “Prediction of the viscosity of multimodal suspensions from unimodal viscosity data.” Trans. Soc. Rheol., 12, 281-301.

Feys, D., Verhoeven, R. and De Schutter, G., (2005). “Pipe flow velocity profiles of complex suspensions, like concrete.” Gent University.

Fluent Inc., (2012). “User’s Guide FLUENT 13.0.” Fluent, Pittsburgh.

Geiker, M. R., Brandl, M., Thrane, L. N. and Nielsen, L. F., (2002). “On the effect of coarse aggregate fraction and shape on the rheological properties of self-compacting concrete.” Cement, Concrete and Aggregates, 24(1), 3-6.

Hu, J. and Wang, K., (2011). “Effect of coarse aggregate characteristics on concrete rheology.” Construction and Building Materials, 25, 1196-1204.

Hafid, H., Ovarlez, G., Toussaint, F., Jezequel, P. H. and Roussel, N., (2010). “Estimating measurement artifacts in concrete rheometers from MRI measurements on model materials.” Proceedings of SCC2010, Montreal, Canada, 127-137.

Hinch, E. J. and Leal, L. G., (1972). “The effect of Brownian motion on the rheological properties of a suspension of non-spherical particles.” J. Fluid Mech., 52, 683-712.

Ingber, M. S., Graham, A. L., Mondy, L. A. and Fang, Z., (2009). “An improved constitutive model for concentrated suspensions accounting for shear-induced particle migration rate dependence on particle radius.” International Journal of Multiphase Flow, 35, 270-276.

Jo, S. D., Park, C. K., Jeong, J. H., Lee, S. H. and Kwon, S. H., (2012). “A computational approach to estimating a lubricating layer in concrete pumping,” Computers Materials and Continua, 27 (3), 189-210.

Jacobsen, S., Haugan, L., Hammer, T. A. and Kalogiannidis, E., (2009). “Flow conditions of fresh mortar and concrete in different pipes.” Cement and Concrete Research, 39 (11), 997-1006.

Kurokawa, Y., Tanigawa, Y., Mori, H. and Nishinosono,


K., (1996). “Analytical study on effect of volume fraction of coarse aggregate on Bingham’s constants of fresh concrete.” Trans. Japan. Concr. Inst., 18, 37-44. (in Japanese)

Kaplan, D., de Larard, F. and Sedran, T., (2005). “Design of concrete pumping circuit.” ACI Materials Journal, 102(2), 110-117.

Kwon, S. H., Jo, S. D., Park, C. K., Jeong, J. H., Lee, S. H., (2013a). “Prediction of concrete pumping: Part I. Development of New Tribometer for Analysis of Lubricating Layer.” ACI Materials Journal, 110(6), 647-656.

Kwon, S. H., Jo, S. D., Park, C. K., Jeong, J. H. and Lee, S. H., (2013b). “Prediction of concrete pumping: Part II. Analytical prediction and experimental verification.” ACI Materials Journal, 110(6), 657-668.

Krieger, I. M. and Dougherty, T. J., (1959). “A mechanism for non-Newtonian flow in suspensions of rigid spheres.” Transactions of The Society of Rheology, 3, 137-152.

Kwan, A. K. H., Mora, C. F. and Chan, H. C., (1999). “Particle shape analysis of coarse aggregate using digital image processing.” Cement and Concrete Research, 29, 1403-1410.

Lu, G., Wang, K. and Rudolphi, T. J., (2008). “Modeling rheological behavior of highly flowable mortar using concepts of particle ad fluid mechanics.” Cement and Concrete Composites, 30, 1-12.

Lam, Y. C., Chen, X., Tan, K. W., Ma, J., Tam, K. C. and Yu, S. C. M., (2002). “Experimental measurements and simulation of particle migration in a pressure-driven tube flow: applications to flow in runner.” Journal of Injection Molding Technol., 6, 45-57.

Leighton, D. and Acrivos, A., (1987). “The shear-induced self-diffusion in concentrated suspensions.” Journal of Fluid Mech., 181, 415-439.

Leighton, D. and Acrivos, A., (1987). “Measurement of shear-induced self-diffusion in concentrated suspensions of Spheres.” Journal of Fluid Mech., 177, 109-131.

Liu, D. M., (2000). “Particle packing and rheological property of highly-concentrated ceramic suspensions:

maxφ determination and viscosity prediction.” Journal of Materials Science, 35, 5503-5507.

Lachemi, P. M., Hossain, K. M. A., Lambros, V., Nkinamubanzi, P. C. and Bouzoubaa, N., (2004). “Performance of new viscosity modifying admixtures in enhancing the rheological properties of cement

paste.” Cement and Concrete Research, 34, 185-193. Mahaut, F., Mokeddem, S., Chateau, X., Roussel, N.

and, Ovarlez, G., (2008). “Effect of coarse particle volume fraction on the yield stress and thixotropy of cementitious materials.” Cement and Concrete Research, 38, 1276-1285.

Morinaga, S., (1973). “Pumpability of concrete and pumping pressure in pipelines.” Proceedings of Rilem Seminar, Leeds, 3, 1-39.

Mora, C. F. and Kwan, A. K. H., (2000). “Sphericity, shape factor, and convexity measurement of coarse aggregate for concrete using digital image processing.” Cement and Concrete Research, 30(3), 351-358.

Mitschka, P., (1982). “Simple conversion of Brookfield R.V.T readings into viscosity functions.” Rheol. Acta., 21, 207-209.

Met-flow, S. A., (2002). “Model UVP-duo with software version 3 user’s guide.” Met-flow Co, Switzerland.

Otsu, N. A., (1979). “Threshold selection method from gray level histogram.” IEEE Trans. Syst. Man. Cybernet, 9(1), 62-69.

Phillips, R. J., Armstrong, R. C. and Brown, R. A., (1992). “A constitutive equation for concentrated suspensions that accounts for shear-induced particle migration.” Physics of Fluids, 4, 30-40.

Struble, L., Sun, G. K., (1995). “Viscosity of Portland cement paste as a function of concentration.” Advanced Cement Based Materials, 2(2), 62-69.

Sakuta, M., Kasanu, I., Yamane, S. and Sakamoto, A., (1989). “Pumpability of fresh concrete.” Takenaka Technical Research Laboratory, Tokyo, 125-133.

Szecsy, R. S., (1997). “Concrete rheology.” Ph.D thesis in Urbana, Illinois.

Tattersall, G. H., Banfill, P. F. G., (1983). “The rheology of fresh concrete.” Pitman, Boston.

Tattersall, G. H., (1991). “Workability and quality control of concrete.” E & FN Spon, New York.

Vu, T. S., Ovarlez, G. and Chateau, X., (2010), “Macroscopic behavior of bidisperse suspensions of noncolloidal particles in yield stress fluids.” J. Rheol., 54(4), 815-833.

Weber, R., (1968). “The transport of concrete by pipeline.” Cement and Concrete Association, London.

Wang, H. P. and Lee, H. S., (1989). “Numerical techniques for free and moving boundary problems.” In: Tucker, C. L. ed., Computer Modeling for Polymer Processing, Oxford University Press.

Documents

Numerical Prediction on the Effects of the Coarse ... - JST