machines

Article
Twin-Shaft Mixers’ Mechanical Behavior Numerical
Simulations of the Mix and Phases
Maria Cristina Valigi 1, * , Silvia Logozzo 1 , Luca Landi 1 , Claudio Braccesi 1 and Luca Galletti 2
1 Department of Engineering, University of Perugia, Via G. Duranti, 1, 06125 Perugia, Italy;
sililog@hotmail.com (S.L.); luca.landi@unipg.it (L.L.); claudio.braccesi@unipg.it (C.B.)
2 Department of Production, Officine Meccaniche Galletti O.M.G. Srl, 06135 Perugia, Italy; luca.galletti@omg.it
* Correspondence: mariacristina.valigi@unipg.it




Received: 27 March 2019; Accepted: 5 June 2019; Published: 7 June 2019


Abstract: In this paper, the mechanical behavior of concrete twin-shaft mixers is analyzed in terms of
power consumption and exchanged forces between the mixture and the mixing organs during the
mixing cycle. The mixing cycle is divided into two macro phases, named transient and regime phase,
where the behavior of the mixture is modeled in two different ways. A force estimation and power
consumption prediction model are presented for both the studied phases and they are validated by
experimental campaigns. From the application of this model to different machines and by varying
different design parameters, the optimization of the power consumption of the concrete twin-shaft
mixers is analyzed. Results of this work can be used to increase productivity and profitability of
concrete mixers and reduce energy waste in the industries which involve mixing processes.

Keywords: Twin-shaft mixer; mechanical optimization; mechanical design; power evaluation

1. Introduction
Nowadays the environmental impact is an important industrial focus, therefore, companies
producing machines must improve product quality and reduce energy waste at the same time. To
remain competitive, this objective must necessarily be achieved using fewer resources, less time, less
money and paying attention to energy consumption.
In this scenario, the development of smart and flexible models able to evaluate the influence of
the project parameters on the consumed energy can be very useful in order to optimize the machine
productivity reducing the absorbed energy.
The mixing process is very important in the alimentary, pharmaceutical, chemical, nuclear and
materials industries which employ different types of mixers. There are many studies in the field of
mixing machines involving computational fluid dynamics (CFD) analyses for many types of mixers,
such as static mixers [1,2] passive mixers [3], dough kneaders [4], stirred tank reactors [5,6], and
continuous mixers [7,8].
The investigation of the mixing in stirred tanks was performed in [9–11], considering a generalized
Newtonian fluid and numerical simulations examined the mixing behavior with Newtonian and
non-Newtonian fluids in simplified geometries of twin-screw mixers [12–14]. Moreover, the influence
of the mixer geometry and operating conditions were studied [15].
This paper focuses on twin-shaft concrete mixers and the concrete mixing machine design and
optimization, which are topics of great significance for achieving high-quality production standards with
reduction of costs and waste. For this reason, scientific research is often focused on the mixing phases
modeling [16–19], in order to study the influence of the various design and production parameters on
the entire mixing process and to find the best way to improve the concrete quality [20–26] and reduce

Machines 2019, 7, 39; doi:10.3390/machines7020039 www.mdpi.com/journal/machines

Machines 2019, 7, 39 2 of 13

phenomena such as power consumption, wear [27–30] and friction among mixing components and
mixtures [31] and to predict the useful life of the mechanical mixing organs [32,33].
Concrete mixers can work by exploiting different geometrical configurations based on the positions
of the mixing shafts and on the machine kinematics [34].
In this paper, the study and optimization of a twin-shaft horizontal mixer are presented and
discussed. The aim of the paper is to present a model to study the mechanical behavior of twin-shaft
mixers, based on the prediction of the exchanged forces between the mixture and the mixing organs and
the power consumption when design parameters vary in a realistic production range. The presented
model is based on the soil mechanics and foundations theory differently from the models proposed in
the state of the art.
The evaluation of power consumption is done in the transient and regime phases of the mixing
procedure, and the model is validated by means of experimental tests in a real production plant. Then
it is applied to different machine models and sizes to define optimal design parameters.
The main studied design parameters influenced by the optimization process are the mixing blades
and arms size and shape, the radius and length of the tank, the shaft spacing, the processed concrete
volume, and the filling level.

2. Material and methods

2.1. The Mixing Cycle

The mixture presents different behaviors for the three general phases of a mixing cycle. The first
stage, called dry phase, comprises three operations: The introduction of aggregates (sand and gravel)
and cement and a dry premixing stage, without any water addition. In this phase, the resistance of the
paste increases until the water is added to the mixture.
The second stage, called the wet phase, comprises two operations: The introduction of water and
additives, which causes a change of the physical properties of the paste. In this phase, after a peak
of resistance of the mixture, it starts to decrease, due to a lubricating action which arises when the
cement paste starts to form. Tests on real mixing machines demonstrated that if water and sand are
introduced contemporarily during the wet phase, the mixing forces are lower than when water and
sand are added separately. This means that the dry material loading in the second case is the worst
case in terms of forces but the mixing operation results to be improved [19].
The third stage is the final mixing phase, which starts after the water introduction and finishes
before the discharge operation. In this phase, the complete homogenization of the paste is reached,
and the resistance tends to have a constant trend (regime phase). The last phase of the mixing cycle is
the discharge operation when the mixer’s tank is emptied, and a new mixing cycle can start.
The changes in the concrete behavior during the entire mixing cycle result in different modeling
criteria both of the paste behavior and of the mechanical interactions between the mixture and the
mixer’s components like blades and arms.
Considering that, in the wet phase the concrete behavior cannot be modeled in a univocal way
and the mixing cycle can be ideally divided into two macro phases: The transient phase, comprising
all the stages until the resistance of the paste tends to have a constant value and the regime phase,
starting when the resistance becomes quite constant and finishing before the discharge operation.
A typical power consumption curve for a generic mixing procedure is represented in Figure 1. In
this picture, label 1 represents the end of the first mixing stage, label 2 the end of the second mixing
phase and 3 the end of the third mixing stage and the beginning of the discharge operation. From
Figure 1 it can be observed that the power consumption is maximum in stage 2. In particular, the peak
power can be found in the time period during the water introduction.
The blue dashed line in Figure 1 divides the two studied macro phases of the mixing cycle,
previously defined.
Machines 2019, 7, 39 3 of 13
Machines
Machines2019,
2019,7,7,39
39 33ofof13
13

Figure1.
Figure
Figure 1.Power
1. Power absorption
Powerabsorption curve
curvefor
absorptioncurve foraageneric
genericmixing
mixingcycle.
mixing cycle.
cycle.

2.2. Twin-Shaft
2.2.Twin-Shaft
2.2. Mixers
Twin-ShaftMixers and
Mixersand Geometrical
andGeometrical Indices
Geometrical Indices for Machine
Indices for
for Machine Comparison
Machine Comparison
Comparison
Concrete twin-shaft
Concretetwin-shaft
Concrete mixersare
twin-shaftmixers horizontalmixers
are horizontal
horizontal mixerswhich
mixers whichwork
which workwith
work withaaarotation
with rotationof
rotation oftwo
of twohorizontal
two horizontal
horizontal
mixing shafts
mixingshafts
mixing (Figure
shafts(Figure 2).
(Figure2). Typical components
2).Typical components
components of of these
of these machines
these machines are
machines are the
are the motors,
the motors, one
motors, one or
one or more
ormore foreach
morefor
for each
each
shaft, the
shaft,the
shaft, gear
thegear boxes,
gearboxes, the
boxes,the shafts,
theshafts, the tank
shafts,the tank and
and the mixing organs:
the mixing
mixing organs: Blades,
organs: Blades,arms,
Blades, arms,and
arms, andconnections.
and connections.
connections.

Figure 2.2. Twin-shaft

Figure Twin-shaft mixer
mixer and
and mixing
mixing components:
components: (a)
(a)The
Themachine,
machine,(b)
(b)the
thetank
tankand
andmixing
mixing
Figure 2. Twin-shaft mixer and mixing components: (a) The machine, (b) the tank and mixing
components: 1: Scraper, 2: Blade,
components: Blade, 3:
3: Arm
Armand
andconnection,
connection,4:4:Shaft.
Shaft.
components: 1: Scraper, 2: Blade, 3: Arm and connection, 4: Shaft.
In
Inthe the studied
studied twin-shaft mixers, the the tank
tank has
hasaatypical
typicalomega
omegashape,
shape,asasrepresented
representedininFigure
Figure3,3,
In the studied twin-shaft mixers, the tank has a typical omega shape, as represented in Figure 3,
where
wheresome someof ofthe
theused
used geometrical
geometrical parameters are shown:
where some of the used geometrical parameters are shown:
h is the filling level (m);
h is the filling level (m);
h Risisthe thefilling
tank level
radius (m);
(m);
R is the tank radius (m);
RD Disisthe thetank
center to center
radius (m); distance between the two circles of the omega (m);
is the center to center distance between the two circles of the omega (m);
k
Dk is is the height of the intersection between thetwo two circlesof(m);
is the
the center
height to of center distance between
the intersection between the the two circles
circles (m); the omega (m);
k zzismax
max the isheight
the filling
of height
the (m);
intersection
is the filling height (m); between the two circles (m);
z(t)isis
zmax
z(t) isthe
thefilling
the fillingheight
filling height(m);
height function(m),
function (m),where
wheret tisisthe
thetime
time(s);(s);
α/2(t)
z(t)
/2(t) is theis the
is the semi-angle
filling representing
height function
semi-angle thefilling
(m), where
representing the filling height
t is height
the timeatat(s);
aacertain
certaintime
timet t(rad);
(rad);
β
/2(t) is the
is angle,
the representing
semi-angle the intersection
representing the height
filling
 is the angle, representing the intersection height (rad). (rad).
height at a certain time t (rad);
 is the angle, representing the intersection height (rad).
Machines 2019, 7, 39 4 of 13
Machines 2019, 7, 39 4 of 13
Machines 2019, 7, 39 4 of 13

Figure 3. Omega
Figure 3. Omega tank
tank and
and geometrical
geometrical parameters.
parameters.
Figure 3. Omega tank and geometrical parameters.
Based
Based on on these
these parameters,
parameters, in in this
this paper
paper somesome additional
additional geometrical
geometrical indices
indices have
have been
been defined
defined
in Based
in order
order to on these
to establish parameters,
establish aa criterion
criterionto in this
tocompare paper
comparedifferent some additional
differentmixing geometrical
mixingmachines:
machines: indices have been defined
in order to establish a criterion to compare different mixing machines:
VVwork C = R𝑅
II== Vwork V CC==D𝐷 𝑅 (1)
(1)
I = Vwork (1)
V 𝐷
I denotes the work index and C represents the compactness of the machine. In the formula of the
I denotes the work
workindex and
andCCvolume
represents the
thecompactness ofofthe machine. In the
theformula of
ofthe
workIindex,
denotes V the
represents index
the total representsfor unit compactness
of length of the the machine.
mixture andIn Vwork formula
is the volume the
work
work index,
index, VVrepresents
represents the
thetotal
totalvolume
volume forfor
unit of length
unit of of the
length of mixture
the and and
mixture VworkV is the volume for
work is the volume
for unit of length processed at the same time by the two shafts of the mixer during the mixing cycle,
unit of length
for unit processed at theatsame time by theby two shafts of the mixer
of the during the mixing cycle, which
which inof length
turn processed
corresponds to the the same
volume time
for unit the two shafts
of length comprised mixer
between during the mixing
the intersection cycle,
of the
in turn
which corresponds
in turn to
corresponds the volume
to the for
volume unit forof length
unit of comprised
length between
comprised the
between intersection
the of theoftwo
intersection the
two circles of the tank and the filling level h (Figure 3).
circles of
two circles the tank
of the and the
tank Iand filling level h (Figure 3).
The work index canthe befilling level h as
considered (Figure 3).
a representation of the machine productivity. The
The
The work
work index
index I Ican
can bebe considered
considered asasa arepresentation
representation ofofthe machine productivity. The
compactness index ranges from 0.5 to 1, as the ratio D/R ranges from 1the machine
to 2. This meansproductivity. The
that the value
compactness
compactness index
index ranges
ranges from
from 0.5
0.5 toto1,1,as
as the
the ratio
ratioD/R
D/R ranges
ranges from
from 11toto2.2.This
This means
means that
that the
thevalue
value
of C increases with the compactness of the machine.
ofofCCincreases
increases with
withthethecompactness
compactness ofofthe
themachine.
machine.
It can be observed that I and C increase accordingly, thus more compact machines can process
ItItcan be observed that I and C increase accordingly, thus thusmore compact machines can process
bigger work volumes. Figure 4 represents the accordingly,
can be observed that I and C increase relationship between more compact
I and C. machines can process
bigger
biggerworkworkvolumes.
volumes.Figure
Figure44represents
representsthe therelationship
relationshipbetween
betweenI Iand andC. C.

Figure 4. Relationship between work index (I) and compactness of the machine (C).
Figure4.4.Relationship
Figure Relationshipbetween
betweenwork
workindex
index(I)
(I)and
andcompactness
compactnessof
ofthe
themachine
machine(C).
(C).
Machines 2019, 7, 39 5 of 13

3. Power Consumption Prediction Model

The developed power consumption theoretical model is mainly based on the calculation of the
forces exchanged between the arm-blade assemblies and the mixture, during the transient and regime
phases. These forces will depend on the rotating position of the arm-blade assembly and on the material
filling level. From the knowledge of the forces, the power absorption is deduced and compared with
experimental data from test campaigns performed using specific twin mixer models. As the mixture
behavior can be modeled in different ways in the transient and regime phases, the model will be
different for the two considered mixing stages.

3.1. Transient Phase

The power absorption in the transient phase will be calculated considering the model of the dry
phase and predicting values very close to the peak, which occurs in the wet phase, where the mixture
cannot be modeled in a univocal way.
In this stage, the mixture can be modeled as a granular material and treated by employing the
Terzaghi’s mechanics of foundations theory [35,36]. According to this theory, the specific limit load
qd (N/m2 ), reachable by a foundation without soil instabilities can be calculated as the sum of three
contributions represented by forces due to the weight of the soil, forces due to the material cohesion and
forces due to the material internal friction. For this reason, the specific limit load can be expressed by:

1
qd = cNc + γD f ,i Nq + γBNγ (2)
2

where c (N/m2 ) is the material specific cohesion, Nc , Nq , Nγ (dimensionless) are the load capacity
factors as tabulated in [35], Df,i , (m) is the immersed quote of the ith component of the arm-blade
assembly (Figure 5), which can be treated as a foundation, B (m) is a shape factor representing the
width of the considered immersed part, γ (N/m3 ) is the material specific weight expressed as a load on
the arm-blade assembly.
The load capacity factors depend on internal friction. The value of the internal friction angle
considered in this model was 0.61 rad (35 degrees), as it results for the gravel sand in its loose
state [19,35].
Considering Figure 5, the arm-blade assembly can be considered subdivided into three parts: The
blade (a), the connection with the arm (b) and the mixing arm (c).
The immersed quote of the ith component of the arm-blade assemblies depends on time and can
be calculated as:
D f ,i (t) = z(t) − R + rci cos ωt (3)

where ω (rad/s) is the shaft angular speed and rci (m) are the distances of the center of gravity of the
single ith components of the arm-blade assembly from the shaft, as represented in Figure 5 (i = 0 is the
blade a, i = 1 is the connection b, i = 2 is the arm c).
Machines 2019, 7, 39 6 of 13
Machines 2019, 7, 39 6 of 13

Figure5.5.Arm-blade
Figure Arm-bladeassembly
assembly and
and immersed parts.
parts.

From
From the calculation
the calculation ofof
thethe
specific limit
specific load,
limit thethe
load, total force
total acting
force on on
acting thethe
arm-blade assembly
arm-blade can
assembly
becan
calculated as the sum
be calculated as theof all
sum theofthree forces
all the acting
three on acting
forces the centers of gravity
on the centers of
of the threeof
gravity components:
the three
Blade, connection, and arm.
components: Blade, connection, and arm.
The
Theforce
forceononthe
thesingle
singlecomponent
componentcan canbe
beexpressed
expressed by:
by:

𝐹(𝑡) = 𝑐c𝑟𝑒𝑑,𝑖(t(𝑡)𝑞
F(t) = red,i)q L𝑑B𝐿 𝐵
d
(4)(4)

where L(m) and B(m) are the length and width of the considered part respectively. cred,i(t) is a reduction
where L(m) and
coefficient B(m)takes
which are the
intolength
account and width
that the of the considered
immersion partith
of the componentcred,i
respectively. of (t)
theis aarm-blade
reduction
coefficient which takes into
assembly varies over time. account that the immersion of the ith component of the arm-blade assembly
varies over time.
2𝐷𝑓,𝑖 (𝑡) 2𝐷𝑓,𝑖 (𝑡)
𝑐𝑟𝑒𝑑,𝑖 (𝑡) = , 𝑖𝑓 < 1; 𝑐𝑟𝑒𝑑,𝑖 (𝑡) = 1 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 (5)
|𝑅
2D𝑟𝑑f ,i𝑖 (cos
t) 𝜔𝑡| |𝑅 𝑟𝑑𝑖f ,icos
2D (t𝜔𝑡|
)
cred,i (t) = , if < 1; cred,i (t) = 1 otherwise (5)
|R rdi cos ωt| |R rdi cos ωt|

where rdi is the radial length of the i-th component of the arm-blade assembly.
where 𝑟𝑑𝑖 is the radial length of the i-th component of the arm-blade assembly.
The calculation of the peak power absorption Ppeak (t) (W) is done considering all the contributions
The calculation of the peak power absorption Ppeak(t) (W) is done considering all the contributions
of the arm-blade assembly components:
of the arm-blade assembly components:

𝑃𝑝𝑒𝑎𝑘 (𝑡)
Ppeak=(t)(𝐹(𝑡) (t)blade𝑟𝑐0
= (F𝑏𝑙𝑎𝑑𝑒 rc0 +
+ 𝐹(𝑡)
F(t)connection
𝑐𝑜𝑛𝑛𝑒𝑐𝑡𝑖𝑜𝑛 𝑟𝑐1
rc1 + F(+ 𝐹(𝑡)
t)arm )ω 𝑟𝑐2 )𝜔
rc2𝑎𝑟𝑚 (6)(6)

Thedependence
The dependenceononthe
thetime
timeisisnecessary
necessary as
as the
the angular
angular position,
position, and
and the
the immersion
immersionquote
quoteofof
each arm-blade assembly are different at each instant.
each arm-blade assembly are different at each instant.
The no-load power consumption was calculated by taking into account the inertia of the arm-
The no-load power consumption was calculated by taking into account the inertia of the arm-blade
blade assemblies and their number. Both the calculated peak power and no-load power were used to
assemblies and their number. Both the calculated peak power and no-load power were used to validate
validate the model by comparison with the experimental measures.
the model by comparison with the experimental measures.
The total peak absorbed power is:
The total peak absorbed power is:
𝑃𝑝𝑒𝑎𝑘 𝑇𝑂𝑇 = 𝑃𝑝𝑒𝑎𝑘 (𝑡) + 𝑃𝑛𝑜−𝑙𝑜𝑎𝑑 (7)
Ppeak TOT = Ppeak (t) + Pno−load (7)

3.2.Regime
3.2. RegimePhase
Phase
During the regime phase, the paste behaves as a viscoplastic material called Bingham fluid. This
During the regime phase, the paste behaves as a viscoplastic material called Bingham fluid. This
means that the mixture can be assimilated to a rigid body when subjected to low stresses and to a
means that the mixture can be assimilated to a rigid body when subjected to low stresses and to a fluid
fluid when subjected to high stresses. Thus, the behavior of the paste can be described when the shear
when subjected to high stresses. Thus, the behavior of the paste can be described when the shear yield
yield stress τ0 (Pa) and the plastic viscosity µ (Pa s/m) are known.
stress τ0 (Pa) and the plastic viscosity µ (Pa s/m) are known.
The law describing the stress τ (Pa) as a function of the velocity v (m/s) for a Bingham fluid is
The law describing the stress τ (Pa) as a function of the velocity v (m/s) for a Bingham fluid is the
the following:
following:
dv𝑑𝑣 (8)
𝜏τ =
= τ𝜏00++µ𝜇dt (8)
𝑑𝑡
Machines 2019, 7, 39 7 of 13

Machines 2019, 7, 39 7 of 13
During this mixing phase, the force Fτ exchanged between the paste and the blades can be modeled
as the During
force necessary to detach
this mixing phase,thethe
mixture
force Ffrom the tank’s wall. For this purpose, the area A (m2 ) of
τ exchanged between the paste and the blades Scan be
the wet surface
modeled as theofforce
the tank mustto
necessary bedetach
calculated as:
the mixture from the tank’s wall. For this purpose, the area
AS (m ) of the wet surface of the tank must be calculatedβ !as:
2
α
AS = 2R + Lm (9)
𝛼
2 2𝛽 (9)
𝐴𝑆 = 2R ( + ) 𝐿𝑚
β
2 2
𝛼α 𝛽
whereLm
where Lm(m)
(m)isisthe
thelength
lengthof
ofthe
the mixer
mixer and + 2)isisthe
and (( 2 + theangle
anglerepresenting
representingthe
thearea
areaofofthe
thetank’s
tank’swall
wall
2 2
covered by the mixture (Figure 3).
covered by the mixture (Figure 3).
𝛼 α𝛽 β 𝑅R−− 𝑧(𝑡)
z(t) 𝑅 k− 𝑘
R−
+2 + = arccos
= arccos +
+ arccos
arccos (10)(10)
2 2 2 𝑅R R 𝑅
Thus,
Thus,
Fτ = τ0 As (11)
𝐹𝜏 = 𝜏0 𝐴 𝑠 (11)
The
Theregime
regimepower absorptionPPregime
powerabsorption (t) (W) results:
regime (t) (W) results:

𝑃P𝑟𝑒𝑔𝑖𝑚𝑒 =F𝐹
regime = τ𝜏ω𝜔RR (12)(12)

The calculated regime power was used to validate the model by comparison with the
The calculated regime power was used to validate the model by comparison with the
experimental measures.
experimental measures.
The total regime absorbed power is:
The total regime absorbed power is:
𝑃𝑟𝑒𝑔𝑖𝑚𝑒 𝑇𝑂𝑇 = 𝑃𝑟𝑒𝑔𝑖𝑚𝑒 + 𝑃𝑛𝑜−𝑙𝑜𝑎𝑑 (13)
Pregime TOT = Pregime + Pno−load (13)

4.4.Experimental
ExperimentalCampaigns
Campaigns and
and Model
Model Validation
Validation

4.1.Model
4.1. ModelValidation
Validation
Thecomparison
The comparisonbetween
betweenthe
the theoretical
theoretical and
and the
the experimental
experimentalpower
powerabsorption,
absorption,isispresented
presentedinin
Figure 6, where the only significant difference between the theoretical model and the experimentisis
Figure 6, where the only significant difference between the theoretical model and the experiment
noticeableininthe
noticeable theearly
earlystages
stagesof
ofmaterial
material loading,
loading, as
as visible
visible in
in the
the zoomed
zoomedred
redbox
boxofofFigure
Figure6)6)[19].
[19].

Figure6.
Figure 6. Comparison between
between theoretical
theoreticaland
andmeasured
measuredpower.
power.
Machines 2019, 7, 39 8 of 13
Machines 2019, 7, 39 8 of 13

4.2. Experimental
4.2. Experimental Campaigns
Campaigns
Four experimental
Four experimentalcampaigns
campaigns were carried
were out to
carried validate
out the model
to validate the on the basis
model on of thebasis
the comparison
of the
between calculated and measured values of the peak, regime, and no-load
comparison between calculated and measured values of the peak, regime, and no-load powers.powers.
An experimental
An experimentalcampaign
campaignwas wascarried out out
carried on a on
mixer model model
a mixer named named
“Model “Model1”, which1”,processes
which
4.5 m3 of mixture
processes 4.5 m3 ofand returns
mixture m3 of vibrated
and3 returns 3 m3 of concrete. “Model 1”
vibrated concrete. has two
“Model 1” motors
has twowith a nominal
motors with a
installedinstalled
nominal power ofpower
55 kW of for eachformotor.
55 kW The geometrical
each motor. parameters
The geometrical of the
parameters ofmixer usedused
the mixer for the
for
experimental
the campaigns
experimental campaigns are are
shown
shownin Table 1. 1.
in Table
The experimental
The experimental campaigns
campaigns were were performed
performed by by using
using aa concrete
concrete composition
composition with with the
the same
same
characteristics of the recipe used for the calculations, with cohesion of 1300 N/m 2 , a shear yield stress
characteristics 2 , a shear yield stress
of 600
of 600 Pa
Pa and
and a degree of humidity equal to 5%, which is coherent with the cohesion value.
The power
The power consumption
consumption curve curve forfor aa single
single motor
motor of of “Model
“Model 1” 1” mixer,
mixer, resulting
resulting from
from the
the
experimental tests, is
experimental is represented
representedin inFigure
Figure7,7,where
where the mixing
the mixingtime
timewas extended
was extended with respect
with to the
respect to
typical
the working
typical timetime
working of the machine
of the machineandand
the the
installed nominal
installed power
nominal (55 kW)
power is highlighted.
(55 kW) is highlighted.

Table 1.
Table Main geometrical
1. Main geometrical parameters
parameters of
of the
the experimental
experimental mixer.
mixer.

Parameters
Parameters Model 1 Model 1
R (m)R (m) 0.695 0.695
h (m)h (m) 0.253 0.253
Lm (m)
Lm (m) 2.276 2.276
Processed Volume3 (m3 ) 4.500
Processed Volume (m ) 4.500

The
The value
value of
of regime
regime power
power for
for the
the model
model validation
validation was
was calculated
calculated as
as the
the average
average value
value of
of the
the
measured powers in the area where the curve becomes flat before the discharge phase.
measured powers in the area where the curve becomes flat before the discharge phase. The value The value of
the peak power was calculated as the average value of the measured powers during
of the peak power was calculated as the average value of the measured powers during the water the water
introduction
introduction phase.
phase. InIn particular,
particular, one
one can
can observe
observe the
the established
established bounds
bounds for
for the
the evaluation
evaluation of
of the
the
power
power consumption in the different phases by means of the gray vertical lines showed in Figuresin7
consumption in the different phases by means of the gray vertical lines showed
Figures
and 8. 7 and 8.

Figure 7.
Figure Experimental power
7. Experimental power consumption
consumption curve
curvefor
forthe
thesingle
singlemotor
motorofof“Model
“Model1”1”
with indication
with of
indication
power levels.
of power levels.

The average
The averagevalues
valuesofofthe
thetotal
totalpeak
peak power
power (49.5
(49.5 kW),
kW), thethe total
total regime
regime power
power (10.5(10.5
kW),kW), the
the no-
no-load power (4.5 kW) are highlighted according to the bounds indicated in the Figures
load power (4.5 kW) are highlighted according to the bounds indicated in the Figures 7 and 8. Zone 7 and 8. Zone
“A”
“A” represents
representsthetheno-load
no-loadpower,
power,zonezone“B”
“B”corresponds
corresponds to to
thethe
water introduction
water introductionperiod
periodused for for
used the
evaluation
the of theofpeak
evaluation the power, zone “C”
peak power, represents
zone the regimethe
“C” represents condition,
regime when the power
condition, whenconsumption
the power
Machines 2019, 7, 39 9 of 13
Machines 2019, 7, 39 9 of 13

consumption
reaches a quitereaches a quite
constant constant
value. value.
The values ofThe
the values
power of the power consumption
consumption obtained fromobtained from the
the application of
application of the power consumption
the power consumption prediction
prediction model, model,inare
are reported reported
Table 2. in Table 2.
As
Asititcan
canbebeobserved,
observed,the
theexperimental
experimentalandandcalculated
calculatedpowers
powersare
arecompletely
completelysuperimposable,
superimposable,
and the model results to be validated.
and the model results to be validated.

Table 2.
Table Analyticalresults
2. Analytical resultsof
of“Model
“Model1”
1”from
fromthe
theprediction
predictionmodel,
model,considering
consideringboth
boththe
themotors
motorsof
of
themachine.
the machine.

Parameters
Parameters Model 1 Model 1
RegimeRegime
PowerPower
(kW) (kW) 20.69 20.69
Peak Power (kW) (kW)
Peak Power 98.92 98.92
No-load
No-load PowerPower
(kW) (kW) 9 9
Nominal Power (kW) 110
Nominal Power (kW) 110

Other
Otherexperimental
experimentalcampaigns
campaignswerewerecarried
carriedout
outwith
withthe
thesame
samemixer
mixerand
andmixture
mixturecomposition,
composition,
but
but with different filling levels h, so that the processed mixing volume changed.In
with different filling levels h, so that the processed mixing volume changed. Inparticular,
particular,three
three
different filling levels and mixing volumes were analyzed, as reported in Table 3, up to the working
different filling levels and mixing volumes were analyzed, as reported in Table 3, up to the working
limit of the machine. Table 3 reports all the test and prediction results, which represent a further
limit of the machine. Table 3 reports all the test and prediction results, which represent a further
validation of the model.
validation of the model.
Figure 8 shows the experimental power consumption curve from the first test, where the mixing
Figure 8 shows the experimental power consumption curve from the first test, where the mixing
time was extended with respect to the typical working time of the machine.
time was extended with respect to the typical working time of the machine.

Figure 8. Experimental power consumption curve for h = 0.344 m and one single motor.
Figure 8. Experimental power consumption curve for h = 0.344 m and one single motor.

Table 3.
Table Powerresults
3. Power resultsfrom
fromthe
theexperimental
experimentaltest
testand
andfor
forthe
theprediction
predictionmodel,
model,for
fordifferent
differentfilling
filling
levels h, considering both the motors of the machine.
levels h, considering both the motors of the machine.
Prediction
Prediction Prediction
Prediction Prediction
Prediction
Test
Test 1 1 Test Test
2 2 TestTest
3 3
Model
Model Model
Model Model
Model
Processed Volume
Processed (m33))
Volume(m 5 5 5 5 5.1 5.1 5.1 5.1 5.6 5.6 5.6 5.6
hh (m)
(m) 0.344
0.344 0.344
0.344 0.3600.360 0.3600.360 0.450
0.450 0.4500.450
Total Regime Power (kW) 22 21.9 22.1 22.0 22.2 22.2
Total Regime Power
Total Peak Power (kW) 22102.6 21.9102.5 22.1104.7 22.0 104.5 22.2110.5 22.2110.9
(kW)
No-load Power (kW) 9 9 9 9 9 9
Total Peak Power(kW)
Nominal Power (kW) 102.6110 102.5110 104.7 110 104.5 110 110.5110 110.9 110
No-load Power (kW) 9 9 9 9 9 9
Nominal
5. Results Power
and (kW)
Discussion 110 110 110 110 110 110

Results of Table 3 show that using the same mixer and changing the filling level and the processed
volume up to the technical limitations of the machine, the total regime power remains quite constant,
Machines 2019, 7, 39 10 of 13

while the total peak power presents a slight increase with h. Another aim of the paper was to study
how the power consumption changed for different machine models and if the geometrical design
parameters of a certain machine were changed of a certain amount. For this purpose, after the model
validation, the prediction model was applied to another twin mixer model, named “Model 2”, which
processes 3 m3 of mixture and returns 2 m3 of vibrated concrete. “Model 2” has two motors with a
nominal installed power of 37 kW for each motor.
The two mixing machines “Model 1” and “Model 2”, were compared, and Table 4 reports all
the calculated data of power consumption for the two studied machines and mixing phases. The
power consumption was evaluated also considering some geometrical changes of the mixers, in order
to assess what design parameters could be tuned to reduce the power consumption, especially in
the transient phase. Small changes of radius and length of the machines (in the order of 1%) were
considered to avoid changes in the number of arm-blade assemblies mounted on the mixer’s shafts
and thus to perform a consistent comparison. In the comparison, both the shaft spacing distance D and
the processed volumes were considered fixed.

Table 4. Results of the prediction model with different geometrical parameters.

Parameters Model 2 Model 1

R (m) 0.620 0.695
h (m) 0.325 0.253
Lm (m) 1.706 2.276
Processed Volume (m3 ) 3 4.5
Total Regime Power (kW) 13.31 20.69
Total Peak Power (kW) 62.57 98.92
No-load Power (kW) 8.23 9
Nominal Installed Power (kW) 74 110
Regime Power (kW) for R increased by 1% 13.37 20.79
Peak Power (kW) for R increased by 1% 54.93 98.11
Regime Power (kW) for Lm increased by 1% 13.36 20.76
Peak Power (kW) for Lm increased by 1% 61.91 98.32

From these results it is clear that in the regime phase the power absorption increases both if the
machine is lengthened and widened, leaving the shaft spacing distance as constant. This result is in
agreement with the model formulas, as the regime power depends on the wet surface extent. One can
observe that for both the machines and for a certain percentage of variation of R and Lm the regime
power consumption is higher if the variation occurs for the radius. Using the compactness index
already described, the results demonstrate that the absorbed power during the regime phase increases
with compactness C, as it can be observed in Figure 9.
In the transient phase, both the increase of radius and length cause a reduction of the filling level
h. If the filling level decreases, also the immersion quotes of all the components of the arm-blade
assemblies decrease and this results in a reduction of the total peak power. Furthermore, as a result, an
increase of radius is better than an increase of length, as it causes a higher decrease of the peak power.
This result was obtained without any measurement of the mixture homogeneity, which is neglected in
this study.
Nevertheless, the results from Table 4 show that geometrical design parameters of the mixers
need to be optimized with respect to the aim of the project. For instance, if the aim is an increase of
processed volume and the transient phase is particularly critical for the power consumption, these
results suggest the widening of the machine, according to the technical construction limitations.
Machines 2019, 7, 39 11 of 13
Machines 2019, 7, 39 11 of 13
Machines 2019, 7, 39 11 of 13

Figure 9. Regime power vs. shaft spacing distance (D) and index of compactness (C).
Figure9.9.Regime
Figure Regimepower
powervs.
vs.shaft
shaftspacing
spacingdistance
distance(D)
(D)and
andindex
indexofofcompactness
compactness(C).
(C).

Another
Another applicationof theof the prediction
modelmodel was performed to investigate how power
Anotherapplication
application of prediction
the prediction was
modelperformed to investigate
was performed to how power consumption
investigate how power
consumption
isconsumption is
affected by the affected
blades’ by the
shape. blades’
Two shape.
kinds of Two kinds
blades, of blades,
normally normally
used in the used
studied in the studied
twin-shaft twin-
mixers,
is affected by the blades’ shape. Two kinds of blades, normally used in the studied twin-
shaft
were mixers, were considered: One with a “T” shape and one with an “L” shape, as represented in
shaftconsidered:
mixers, wereOne with a “T”
considered: Oneshape
with and
a “T”one withand
shape an one
“L” with
shape,an as
“L”represented in Figure 10.
shape, as represented in
Figure
Results 10. Results
show that show
the thatpeak
total the total peak
power is power
9% is 9%for
higher higher
the Tfor the Twith
shape shape with respect
respect the Ltotothe L
Figure 10. Results show that the total peak power is 9% higher for the T shape with torespect shape.
the L
shape. Furthermore,
Furthermore, the total the total regime
regime power
3.8%ishigher
3.8% higher
thefor the T shape with respect toshape.
the L shape.
shape. Furthermore, the total power
regimeispower is 3.8%for
higher Tfor
shape
the Twith
shaperespect to the L
with respect Thus,
to the L shape.
Thus, reductions
reductions of the of
powerthe power
absorption absorption
can be can
obtainedbe obtained
by using by
the using
blades the
with blades
an L with
shape an L shape
designed by
Thus, reductions of the power absorption can be obtained by using the blades with an L shape
designed
the by the
manufacturer manufacturer
of the studiedof of the studied
mixers. mixers.
designed by the manufacturer the studied mixers.

Figure 10.
Figure Analysis of
10. Analysis of different
different blades’
blades’ shape: (a) L
shape: (a) L shape,
shape, (b)
(b) T
T shape.
shape.
Figure 10. Analysis of different blades’ shape: (a) L shape, (b) T shape.
6. Conclusions
6. Conclusions
6. Conclusions
The paper focuses on the concrete twin-shaft mixers and on a proposed mathematical model able
The paper focuses on the concrete twin-shaft mixers and on a proposed mathematical model
to evaluate the power
The paper consumption
focuses as a function
on the concrete of the
twin-shaft designand
mixers parameters of the mixers.
on a proposed In this paper,
mathematical model
able to evaluate the power consumption as a function of the design parameters of the mixers. In this
the transient
able andthe
to evaluate regime
powerphases of the mixing
consumption cycle were
as a function studied
of the designtoparameters
calculate theof forces exchanged
the mixers. In this
paper, the transient and regime phases of the mixing cycle were studied to calculate the forces
between
paper, thethe concrete
transientmixture and the
and regime mixing
phases oforgans and the
the mixing power
cycle consumption
were studied toincalculate
twin-shaftthemixers.
forces
exchanged between the concrete mixture and the mixing organs and the power consumption in twin-
For this purpose,
exchanged betweena power absorption
the concrete mixtureprediction model organs
and the mixing was presented, validated,
and the power and applied
consumption to
in twin-
shaft mixers. For this purpose, a power absorption prediction model was presented, validated, and
different mixing
shaft mixers. machine
For modelsato
this purpose, investigate
power how to
absorption reduce the
prediction power
model consumption.
was Variationsand
presented, validated, of
applied to different mixing machine models to investigate how to reduce the power consumption.
design
applied parameters such
to different as the
mixing tank’s radius
machine modelsandto length, the filling
investigate how tolevel, the the
reduce processed volume, and
power consumption.
Variations of design parameters such as the tank’s radius and length, the filling level, the processed
the shape ofofthe
Variations mixing
design blades were
parameters suchanalyzed. Results
as the tank’s show
radius andthat increasing
length, the filling
the filling level, level and the
the processed
volume, and the shape of the mixing blades were analyzed. Results show that increasing the filling
processed
volume, and volume for a fixed
the shape of thetank’s
mixinggeometry and number
blades were analyzed.of arm-blade
Results showassemblies, the total
that increasing theregime
filling
level and the processed volume for a fixed tank’s geometry and number of arm-blade assemblies, the
power remains
level and quite constant,
the processed volume while
for athe total
fixed peakgeometry
tank’s power presents a slight
and number ofincrease
arm-blade with h.
assemblies, the
total regime power remains quite constant, while the total peak power presents a slight increase with
total regime power remains quite constant, while the total peak power presents a slight increase with
h.
h.
Machines 2019, 7, 39 12 of 13

If the processed volume is kept fixed, in the regime phase the absorbed power increases with
the index of compactness C and the work index I, defined in this paper and both with the tank’s
radius and length. The increase with radius is higher than the increase with length. In the same
conditions, the peak power registered in the transient phase decreases both when radius and length
increase, but its reduction is higher when the radius is increased by the same percentage. This result
was obtained without any measurement of the mixture homogeneity, which is neglected in this study.
Results demonstrate also that if the mixer mounts the typical T blades, realized by the manufacturer,
the total peak power is 9% higher than the case when the mixer mounts the L blades. In the regime
phase, the total regime power with the T blades is 3.8% higher than the case with the L shape. Mixer’s
manufacturers can benefit the results of this research for designing and producing machines which
allow energy saving and enhanced production rates. In fact, the proposed model allows to study
the machines with the aim of obtaining a greater production volume of concrete with the same
installed power.

Author Contributions: Conceptualization, M.C.V., S.L., C.B., L.L.; data curation, M.C.V., S.L., L.G.; formal
analysis, M.C.V., S.L.; funding acquisition, M.C.V.; methodology, M.C.V., S.L., C.B., L.L.; project administration,
M.C.V., C.B., L.L.; software, M.C.V., L.L., S.L., C.B.; supervision, M.C.V., L.G.; validation, M.C.V., S.L., C.B., L.L.;
writing—original draft, S.L., M.C.V.
Funding: This research received funding from Officine Meccaniche Galletti O.M.G. Srl, Perugia, Italy.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Zalc, J.M.; Szalai, E.S.; Muzzio, F.J. Mixing dynamics in the SMX static mixer as a function of injection location
and flow ratio. Polym. Eng. Sci. 2003, 43, 875–890. [CrossRef]
2. Heniche, M.; Reeder, M.; Tanguy, P.; Fassano, J. Numerical investigation of blade shape in static mixing.
AIChE J. 2005, 51, 44–58. [CrossRef]
3. Tseng, L.; Yang, A.; Lee, C.; Hsieh, C. CFD-Based optimization of diamond-obstacles inserted micromixer
with boundary protrusions. Eng. Appl. Comput. Fluid Mech. 2011, 5, 210–222. [CrossRef]
4. Jongen, T. Characterization of batch mixers using numerical flow simulations. AIChE J. 2000, 46, 2140–2150.
[CrossRef]
5. Dhanasekharan, K.M.; Kokini, J.L. Design and scaling of wheat dough extrusion by numerical simulation of
flow and heat transfer. J. Food Eng. 2003, 60, 421–430. [CrossRef]
6. Wang, W.; Manas-Zloczower, I. Temporal distributions: The basis for the development of mixing indexes for
scale-up of polymer processing equipment. Polym. Eng. Sci. 2001, 41, 1068–1077. [CrossRef]
7. Lamberto, D.J.; Alvarez, M.; Muzzio, F.J. Experimental and computational investigation of the laminar flow
structure in a stirred tank. Chem. Eng. Sci. 1999, 54, 919–942. [CrossRef]
8. Alvarez, M.M.; Zalc, J.M.; Shinbrot, T.; Arratia, P.E.; Muzzio, F.J. Mechanisms of mixing and creation of
structure in laminar stirred tanks. AIChE J. 2002, 8, 2135–2148. [CrossRef]
9. Zalc, J.M.; Alvarez, M.M.; Muzzio, F.J.; Arik, B.E. Extensive validation of laminar flow in a stirred tank with
three Rushton turbines. AIChE J. 2001, 47, 2144–2154. [CrossRef]
10. Lamberto, D.J.; Alvarez, M.M.; Muzzio, F.J. Computational analysis of regular and chaotic mixing in a stirred
tank reactor. Chem. Eng. Sci. 2001, 56, 4887–4899. [CrossRef]
11. Connelly, R.K.; Kokini, J.L. Examination of the mixing ability of single and twin screw mixers using 2D finite
element method simulation with particle tracking. J. Food Eng. 2007, 79, 956–969. [CrossRef]
12. Connelly, R.K.; Kokini, J.L. 2-D numerical simulation of differential viscoelastic fluids in a single-screw
continuous mixer: Application of viscoelastic FEM methods. Adv. Polym. Technol. 2003, 22, 22–41. [CrossRef]
13. Kozic, M.S.; Ristic, S.S.; Linic, S.L.; Stetić-Kozic, S. Numerical analysis of rotational speed impact on mixing
process in a horizontal twin-shaft paddle batch mixer with non-newtonian fluid. FME Trans. 2016, 44,
115–124. [CrossRef]
14. Rivera, C.; Foucault, S.; Heniche, M.; Espinosa-Solares, T.; Tanguy, P.A. Mixing analysis in a coaxial mixer.
Chem. Eng. Sci. 2006, 61, 2895–2907. [CrossRef]
Machines 2019, 7, 39 13 of 13

15. Rathod, M.L.; Kokini, J.L. Effect of mixer geometry and operating conditions on mixing efficiency of a
non-Newtonian fluid in a twin screw mixer. J. Food Eng. 2013, 118, 256–265. [CrossRef]
16. Cazacliu, B. In-mixer measurements for describing mixture evolution during concrete mixing. Chem. Eng.
Res. Des. 2008, 86, 1423–1433. [CrossRef]
17. Cazacliu, B.; Legrand, J. Characterization of the granular-to-fluid state process during mixing by power
evolution in a planetary concrete mixer. Chem. Eng. Sci. 2008, 63, 4617–4630. [CrossRef]
18. Cazacliu, B.; Roquet, N. Concrete mixing kinetics by means of power measurement. Cem. Concr. Res. 2009,
39, 182–194. [CrossRef]
19. Braccesi, C.; Landi, L. An analytical model for force estimation on arms of concrete mixers. In Proceedings of
the ASME Design Engineering Technical Conference, San Diego, CA, USA, 30 August–2 September 2009.
20. Yao, Y.; Feng, Z.; Chen, S. Strength of concrete reinforced using double-blade mixer. Mag. Concr. Res. 2013,
65, 787–792. [CrossRef]
21. Yao, Y.; Feng, Z.; Chen, S.; Li, B.Q.; Zhao, L.; Zhao, W. A Double-Blade Mixer for Concrete with Improved
Mixing Quality. Arab. J. Sci. Eng. 2016, 41, 4809–4816. [CrossRef]
22. Hu, C.; De Larrard, F. The rheology of fresh high-performance concrete. Cem. Concr. Res. 1996, 2, 283–294.
[CrossRef]
23. De Larrad, F.; Ferraris, C.F.; Sedran, T. Fresh concrete: A Herschel-Bulkley material. Mater. Struct. 1998, 31,
494–498. [CrossRef]
24. Gram, A.; Silfwerbrand, J.; Lagerblad, B. Obtaining rheological parameters from flow test—Analytical,
computational and lab test approach. Cem. Concr. Res. 2014, 63, 29–34. [CrossRef]
25. Ferraris, C.F. Measurement of the rheological properties of high performance concrete: State of the art report.
J. Res. Natl. Inst. Stand. Technol. 1999, 104, 461–478. [CrossRef]
26. Marsili, R.; Moretti, M.; Rossi, G.; Speranzini, E. Image Analysis Technique for Material Behavior Evaluation
in Civil Structures. Materials 2017, 10, 770.
27. Valigi, M.C.; Logozzo, S.; Rinchi, M. Wear resistance of blades in planetary concrete mixers. Design of a new
improved blade shape and 2D validation. Tribol. Int. 2016, 96, 191–201. [CrossRef]
28. Valigi, M.C.; Logozzo, S.; Rinchi, M. Wear resistance of blades in planetary concrete mixers. Part II: 3D
validation of a new mixing blade design and efficiency evaluation. Tribol. Int. 2016, 103, 37–44. [CrossRef]
29. Valigi, M.C.; Fabi, L.; Gasperini, I. Wear resistance of new blade for planetary concrete mixer. In Proceedings
of the 5th World Tribology Congress, WTC 2013, Torino, Italy, 8–13 September 2013.
30. Valigi, M.C.; Logozzo, S.; Gasperini, I. Study of wear of planetary concrete mixer blades using a 3D optical
scanner. In Proceedings of the International Mechanical Engineering Congress & Exposition, IMECE2015,
Houston, TX, USA, 13–19 November 2015.
31. Valigi, M.C.; Rinchi, M.; Caputo, S. Behavior Simulation of a Concrete Mix in a Turbine Pan Mixer. Available
online: https://www.bft-international.com/en/artikel/bft_Behavior_simulation_of_a_concrete_mix_in_a_
turbine_pan_mixer__2332434.html (accessed on 7 June 2019).
32. Valigi, M.C.; Gasperini, I. Planetary vertical concrete mixers: Simulation and predicting useful life in steady
states and in perturbed conditions. Simul. Model. Pract. Theory 2007, 15, 1211–1223. [CrossRef]
33. Valigi, M.C.; Gasperini, I. Model-based method predicting useful life of concrete mixers [Modellgestütztes
Verfahren zur Vorhersage der Nutzungsdauer von Betonmischern]. Betonw. Und Fert. -Tech./Concr. Plant
Precast Technol. 2005, 71, 38–42.
34. Ferraris, C.F. Concrete mixing methods and concrete mixers: State of art. J. Res. Natl. Inst. Stand. Technol.
2001, 106, 391–399. [CrossRef]
35. Terzaghi, K.; Peck, R.B. Soil Mechanics in Engineering Practice, 3rd ed.; John Wiley and Sons Inc.: New York,
NY, USA, 1996.
36. Lancellotta, R.; Calavera, J. Fondazioni; Mc Graw Hill Education: New York, NY, USA, 1999; pp. 269–273.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).