applied

sciences
Article
Development of High-Efficiency, High-Speed and
High-Pressure Ambient Temperature Filling System
Using Pulse Volume Measurement
SeBu Oh 1 , SeHoon Oh 1, * and ByeongRo Min 2, *
1 Department of Mechanical Engineering, Chungang University, Seoul-si 06974, Korea; ok777@cau.ac.kr
2 Institute of Biotechnology and Bioengineering, Sungkyunkwan University, Suwon-si 16419, Korea
* Correspondence: osh@cau.ac.kr (S.O.); min7887@skku.edu (B.M.);
Tel.: +82-2-820-5314 (S.O.); +82-31-353-6152 (B.M.)




Received: 27 April 2019; Accepted: 17 June 2019; Published: 19 June 2019


Abstract: Adjusting the filling pressure is essential to fit the final gas volume when charging
a carbonated beverage with high pressure. However, in the previous mechanical carbonated
ambient filling system, it was difficult to control and monitor the charging conditions such as
pressure, temperature and flow rate. In this study, we have developed a high efficiency carbonated
ambient filling system capable of high speed and high pressure filling, by using a pulse type electronic
flow-meter. The response speed characteristics of the M(BC) and F(MH) series valves were investigated.
LMS Imagine.Lab Amesim (Siemens PLM Software) was used to calculate the charging and discharging
time of the system under a high CO2 gas pressure condition. The quantitative and precise charging
system was implemented with the change of filling time and monitoring/controlling/correction of flow
rate. Moreover, a dual controller of the high-speed pulse output was established and a high-speed
data processing/flow rate charging algorithm was applied in the system. The filling variation of the
system was in the range of ±3 gram(gr) (standard deviation 0.57). The developed system could be
applied to improve the quality of goods and economic feasibility at various industrial sectors.

Keywords: ambient temperature filler; high pressure filling; high speed filling; flow meter filling;
carbonated beverage filling

1. Introduction
Carbonated beverage fillers can be classified into mechanical and electric fillers. It is difficult to
fill an exact amount with conventional mechanical fillers, which adjust to the filling height due to
the bottles having different levels of expansion when filling under a high pressure (approximately
3.4 kg/cm2 ·G or above), and give a low filling yield value due to excessive filling compared to the
standard amount [1]. Because of this problem, filling machines with an electronic flowmeter that can
fill the correct amount regardless of the expansion level of the filled bottle under a high temperature
and pressure have been developed and are being commercialized [2]. However, the pulse output type
electronic flowmeter has difficulties in calculating a high-speed pulse output during high-speed filling,
so it is difficult to achieve highly accurate filling [3,4]. In this study, a filling system using a ‘pulse
output type flow meter’ was designed. The performance of a mass flowmeter is higher than that of an
electronic flowmeter. However, in our case, more than 100 flowmeters are required per filling machine,
and so the electronic flowmeter was selected as the optimum design considering price competitiveness.
As a result of analyzing the performance of the electronic flowmeter (pulse output, signal distortion
of the output waveform, repetition accuracy, etc.), it was found that the pulse signal output waveform
was distorted at low-speed filling (filling completion time or filling start time). In order to solve this

Appl. Sci. 2019, 9, 2491; doi:10.3390/app9122491 www.mdpi.com/journal/applsci

Appl. Sci. 2019, 9, x FOR PEER REVIEW 2 of 11

As a result of analyzing the performance of the electronic flowmeter (pulse output, signal
distortion of the output waveform, repetition accuracy, etc.), it was found that the pulse signal output
waveform was distorted at low-speed filling (filling completion time or filling start time). In order to
Appl. Sci. 2019, 9, 2491 2 of 10
solve this problem, ‘Low Flow CUT-OFF’ and ‘Pressure shock’ were set as the function of the
electronic flowmeter to eliminate the instability at low flow rates and to improve flow measurement
accuracy.‘Low
problem, In addition, the accuracy
Flow CUT-OFF’ and of the system
‘Pressure depends
shock’ onas
were set thethe
response
functionspeed
of theofelectronic
the solenoid valve,
flowmeter
diaphragm
to eliminatevalve, the flowmeter
the instability at low and
flowthe control
rates and toalgorithm.
improve flowTherefore, in this study,
measurement accuracy.we In
designed
addition, a
carbonated beverage filling system at ambient temperature that enables high-speed
the accuracy of the system depends on the response speed of the solenoid valve, diaphragm valve, and high-
pressure
the fillingand
flowmeter with
theacontrol
filling algorithm.
algorithm, Therefore,
by pressure in compensation in the high-speed
this study, we designed a carbonatedpulse output
beverage
flow rate
filling measuring
system typetemperature
at ambient [5]. that enables high-speed and high-pressure filling with a filling
algorithm, by pressure compensation in the high-speed pulse output flow rate measuring type [5].
2. Hardware Composition
2. Hardware Composition
2.1. Component of the Flow Rate Filling System
2.1. Component of the Flow Rate Filling System
In this study, the authors designed and built a flow rate filling valve in order to understand the
Incharacteristics
filling this study, theof
authors designed
a high-speed and built soft
carbonated a flow ratefilling
drink fillingsystem
valve inasorder
shownto in
understand the
Figure 1 and
filling
used an characteristics
electronic flowof meter
a high-speed carbonated
to conduct tests to soft drink filling
understand system as shown
the characteristics in Figure
of flow supply1 and
and
used an electronic flow meter to conduct tests to understand the characteristics of
filling. The authors succeeded in high flow rate filling using a pulse output type flow meter. By flow supply and
filling.
addingThe authors succeeded
a dualizing in high
control system, theflow rate was
system filling using a pulse
constructed output
so that type charge
it could flow meter. By adding
at a high speed
a[5].
dualizing control system, the system was constructed so that it could charge at a high speed [5].

Figure 1. Components
Components of a high-speed carbonated soft drink flow rate filling system.

2.2. Design of
2.2. Design of Filling
Filling Valve
Valve
The
The conventional
conventionalfillingfillingsystems
systemsutilized
utilized instruments
instruments based on the
based on thetheory of mechanism
theory of mechanismand fluid
and
dynamics and were composed of the counter pressure process, filling
fluid dynamics and were composed of the counter pressure process, filling valve opening process, valve opening process, filing
process, filing valve
filing process, filingclosing
valve process,
closing and snifting
process, andprocess
snifting(Ohprocess
2016). However,
(Oh 2016). theHowever,
filling method using
the filling
a flow meter measures the filling flow rate of the filling valve and converts
method using a flow meter measures the filling flow rate of the filling valve and converts the electric the electric signal into
pneumatic
signal into pressure,
pneumaticwhich should
pressure, open/close
which shouldthe filling valve
open/close or remove
the filling valvethe orresidual
removepressure, etc.
the residual
Therefore, even if the flow rate is measured precisely, there will still be a difference
pressure, etc. Therefore, even if the flow rate is measured precisely, there will still be a difference between the desired
filling
between amount and thefilling
the desired actual amount
filling amount
and the due to thefilling
actual time delay
amount in the operating
due to the time partsdelay
powered by
in the
a pneumatic
operating system.
parts powered Therefore, it is necessary
by a pneumatic system. toTherefore,
develop aithigh precisiontopneumatic
is necessary develop a highcontrol filling
precision
valve that has a higher response speed. As shown in Figure 1, a filler that
pneumatic control filling valve that has a higher response speed. As shown in Figure 1, a filler that uses a flow meter measures
the
usesfilling
a flowvolume with the flow
meter measures metervolume
the filling and uses thethe
with electric
flow signal
meter from the flow
and uses meter signal
the electric to closefrom
the
filling
the flow valve. Therefore,
meter to close it isthe
notfilling
possible to use
valve. the conventional,
Therefore, it is notmechanical
possible to type.
useForthethis reason, it is
conventional,
necessary to operate each filling mode electrically in accordance with a series
mechanical type. For this reason, it is necessary to operate each filling mode electrically in accordance of pre-determined steps.
In
withtheory,
a seriesif the system is filling steps.
of pre-determined at a rate
In of 150 gr/sec
theory, if the and there
system is a 30 at
is filling msa time
rate of delay in the operation
150 gr/sec and there
time, there will be a 4.5 gr difference in the filling volume. In order to reduce
is a 30 ms time delay in the operation time, there will be a 4.5 gr difference in the filling volume. this difference, the authors
In
added a low-speed filling control valve when it is close to the target filling level, and performed
low-speed filling to reduce the gap.
Appl. Sci. 2019, 9, x FOR PEER REVIEW 3 of 11

order to reduce this difference, the authors added a low-speed filling control valve when it is close
Appl. Sci. 2019, 9, 2491
to
3 of 10
the target filling level, and performed low-speed filling to reduce the gap.

3. Filling Precision Design

3.1. The Control

3.1. The Control Method
Method to
to Increase
Increase the
the Precision
Precision Level
Level of
of Filling
Filling
Figure 2 shows the result of an analysis
analysis on the
the target
target filling
filling rate
rate precision
precision level of ±3
level of ±3 gr for the
filling performance.

Figure 2. Filling error analysis for filling targets.

After excluding
excluding thetheerror
errorininthe
theflow
flowmeter
meteritself,
itself,the filling
the system
filling system hashas
anan
allowance of only
allowance ±2 ±2
of only gr.
If
gr.the filling
If the valve
filling fillsfills
valve at aatrate of 150
a rate gr/sec,
of 150 an allowance
gr/sec, an allowance of 2ofgr2 will give
gr will an an
give allowance window
allowance windowof
operation delay that is 13.3 ms (which is 0.013
of operation delay that is 13.3 ms (which is 0.013 seconds). s).
We selected
selected the XGK model in the XGT series of LS Company, which which has fast data processing
speed, as the PLC of the controller.
controller. At the initial design stage, we could not increase the resolution of
the flowmeter because one PLC CPU could not process a large amount of input/output and control
related programs. Therefore, in order to improve the pulse processing speed by distributing the CPU
load in the controller, the local PLC, controlling only the flow meter and the filling valve control control unit,
and the main PLC, controlling the remaining machine control unit, were
the remaining machine control unit, were constituted. constituted.

3.2. Control Algorithm

3.2. Control Algorithm to
to Increase
Increase the
the Precision
Precision Level
Level of
of Filling
Filling
The
The pulse
pulseflow rate
flow measurement
rate measurement filling system
filling has more
system has I/Os
more(PLC inputs
I/Os (PLCand outputs)
inputs andcompared
outputs)
to the conventional mechanical filling systems. Therefore, it has several times more
compared to the conventional mechanical filling systems. Therefore, it has several times more control steps. This
control
will burden
steps. the PLC
This will CPU
burden theand
PLC affect
CPUthe andresponse rateresponse
affect the of the valve. Therefore,
rate of it Therefore,
the valve. was necessary to
it was
develop a new control algorithm in order to improve this issue.
necessary to develop a new control algorithm in order to improve this issue.
3.2.1. Structured Function Block Program
3.2.1. Structured Function Block Program
A formalized program was developed using a structured function block program. With a module
A formalized program was developed using a structured function block program. With a
program, the number of steps in the program was reduced, and the system was easier to debug, while
module program, the number of steps in the program was reduced, and the system was easier to
having less program errors. By grouping the module program for each valve unit, it was possible
debug, while having less program errors. By grouping the module program for each valve unit, it
to standardize the program and realize the core algorithms in blocks. The structured function block
was possible to standardize the program and realize the core algorithms in blocks. The structured
program could modulate the existing sequence commands program to reduce the number of program
function block program could modulate the existing sequence commands program to reduce the
steps. This resulted in a shorter scan time, improving the filling precision and accuracy in the iteration.
number of program steps. This resulted in a shorter scan time, improving the filling precision and
accuracy
3.2.2. in the
Filling iteration.Algorithm
Processing

3.2.2.The filling
Filling processing
Processing algorithm of the filler using the pulse flow rate method is as shown in the
Algorithm
flow chart in Figure 3. According to the filling process algorithm in Figure 3, the filling valve counter
The filling processing algorithm of the filler using the pulse flow rate method is as shown in the
pressure is a process where the pressure of the filler bowl and the bottle is equalized [5]. The filling valve
flow chart in Figure 3. According to the filling process algorithm in Figure 3, the filling valve counter
and liquid valve is the process where the main valve, which fills up the bottle, is opened. The filling
pressure is a process where the pressure of the filler bowl and the bottle is equalized [5]. The filling
valve slow-fast fill valve is the process where the filling speed is slowed down from the initial high
valve and liquid valve is the process where the main valve, which fills up the bottle, is opened. The
speed to a lower speed when the bottle is almost full (80% full, if the target volume is 500 gr, at 400 gr).
filling valve slow-fast fill valve is the process where the filling speed is slowed down from the initial
The snifting valve is a process where the residual pressure inside the bottle is removed after filling is
over [5]. When filling a bottle, the key element is controlling the level of the bottle and the pressure.
Appl. Sci. 2019, 9, x FOR PEER REVIEW 4 of 11

high speed to a lower speed when the bottle is almost full (80% full, if the target volume is 500 gr, at
Appl. Sci. 2019, 9, x FOR PEER REVIEW 4 of 11
400 gr). The snifting valve is a process where the residual pressure inside the bottle is removed after
filling is over [5]. When filling a bottle, the key element is controlling the level of the bottle and the
high speed to
Appl. Sci. 2019,
a lower speed when the bottle is almost full (80% full, if the target volume is 500 4gr, at
pressure. In 9,a 2491
situation where the level and pressure should be controlled as in carbonate filling, of 10
a
400 gr). The snifting valve is a process where the residual pressure inside the bottle is removed after
small amount of change in the pressure can significantly affect the filling precision.
filling is over [5]. When filling a bottle, the key element is controlling the level of the bottle and the
In a situation
pressure. In awhere the level
situation whereand
thepressure should
level and be controlled
pressure should beascontrolled
in carbonate
as filling, a smallfilling,
in carbonate amounta
of change in the pressure can significantly affect the filling precision.
small amount of change in the pressure can significantly affect the filling precision.

Figure 3. Filling process algorithm.

Figure 3. Filling process algorithm.

Figure 4 shows the interface, which shows and configures the PID (A piping and
Figure 4 shows
instrumentation the interface,
diagram) of thewhich
level shows and configures
and pressure of thethe PID (A
filling piping
bottle. and instrumentation
Furthermore, for the
diagram) of the level and pressure of the filling bottle. Furthermore, for the controller,
controller, the automatic tuning configuration and programs were implemented in order to optimize the automatic
tuningFigure 4 showsandthe
configuration interface, which
programs shows and configures thethePID
filling(A piping and
the filling bottle pressure and level were
with implemented
each pressureinand order to optimize
level control PID value. bottle pressure
Therefore, it is
instrumentation
and leveltowith diagram) of the level and pressure of the filling bottle. Furthermore, for the
possible findeach pressure and
the optimized PIDlevel
valuecontrol
withPID value.
a single Therefore,
touch. it is possible
The pressure controlto is
find the optimized
implemented as
controller,
PID value the
withautomatic
a single tuning
touch. configuration
The pressure and programs
control is were implemented
implemented as shown inabove.
order toTherefore,
optimize
shown above. Therefore, the system is capable of controlling the pressure up to ±0.1 kg/cm and the 2
the system
the filling bottle pressure
is capable and levelthe
of controlling with each pressure
pressure up to ±0.1andkg/cm
level2 control
and thePID levelvalue.
±3 mm.Therefore, it is
level ±3 mm.
possible to find the optimized PID value with a single touch. The pressure control is implemented as
shown above. Therefore, the system is capable of controlling the pressure up to ±0.1 kg/cm2 and the
level ±3 mm.

Figure
Figure 4.
4. Filler
Filler bowl
bowl pressure
pressure control
control structure.

3.3. Experimental
3.3. Experimental Design
Design of
of Filling
Filling Valve
Valve Response
Response
Figure 4. Filler bowl pressure control structure.
In designing the controlling part, the solenoid valve response rate is considered to be important.
Therefore, the solenoid
3.3. Experimental Designvalve response
of Filling Valvetime that is related to the control part is considered as important.
Response
The response time characteristics of the F company (Germany) and M company (USA) solenoid valve
are shown in Figure 5.
Therefore, the solenoid valve response time that is related to the control part is considered as
important. The response time characteristics of the F company (Germany) and M company (USA)
solenoid valve are shown in Figure 5.
As shown in Figure 5, the response rate of the filling valve was assessed by measuring the time
takenSci.for
Appl. detecting
2019, 9, 2491 the pressure signal at the valve outlet after the solenoid valve is activated
5 ofby
10
giving an operation signal to the said valve.

Appl. Sci. 2019, 9, x FOR PEER REVIEW 5 of 11

In designing the controlling part, the solenoid valve response rate is considered to be important.
Therefore, the solenoid valve response time that is related to the control part is considered as
important. The response time characteristics of the F company (Germany) and M company (USA)
solenoid valve are shown in Figure 5.
As shown in Figure 5, the response rate of the filling valve was assessed by measuring the time
taken for detecting the pressure signal at the valve outlet after the solenoid valve is activated by
giving an operation signal to the said valve.

Figure 5. Schematic
Schematic diagram for measuring the response time.

As shown
4. Results in Figure 5, the response rate of the filling valve was assessed by measuring the time
and Discussion
taken for detecting the pressure signal at the valve outlet after the solenoid valve is activated by giving
4.1. Analysis ofsignal
an operation the Response Characteristics
to the said valve. of the Filling Valve
Figures
4. Results and6 and 7 show the results of comparison of the valve response time characteristics of the
Discussion
solenoid valves of M company and F company in designing the filling valves.
4.1. Analysis of the Response Characteristics of the Filling Valve
Figures 6 and 7 show
21 the results of comparison of the valve response time characteristics of the
solenoid valves of M company
Figure 5. and F company
Schematic diagram in for
designing thethe
measuring filling valves.
response time.
The average response rates of the two models of M company ranged from 12 ms to 16ms, while
20
4. Results
the averageand Discussion
response rate of the solenoid valves of F company was 10 ms. In addition, the differences of
the response rates of the solenoid valves were 3 ms for M company and 1 ms for F company, indicating
4.1. the
Analysis of had 19
the Response Characteristics of the Filling Valve
that valves a good response rate and speed difference.
As for the6 response
Figures and 7 showratethe
of results
the filling valve, the maximum
of comparison rateresponse
of the valve differences
timeofcharacteristics
the solenoid valves
of the
msec

18
of M company ranged from 14ms to 16ms, which is significant.
solenoid valves of M company and F company in designing the filling valves. However, the result was 10~11 ms for
F company, indicating that the products of this company performed better in terms of the differences.
17

21

16

20

15
0 1 2 3 4 5 6 7 8 9 10 11
19
Number of filling
msec

18 Figure 6. M Company’s valve response characteristics.

17

16

15
0 1 2 3 4 5 6 7 8 9 10 11

Number of filling

Figure6.6.M
Figure MCompany’s
Company’svalve
valveresponse
responsecharacteristics.
characteristics.
 Appl.Sci.
Appl. Sci.2019,
2019,9,9,2491
x FOR PEER REVIEW 66 of
of 10
11

13

12

11
msec

10

7
0 1 2 3 4 5 6 7 8 9 10 11
Number of filling

Figure7.7.FFCompany’s
Figure Company’svalve
valveresponse
responsecharacteristics.
characteristics.

4.2. Comparison
The average of the response
response characteristics
rates of the twoofmodels
the filling valve
of M company ranged from 12 ms to 16ms, while
the average response rate of the solenoid valves of F company
Figures 8 and 9, Tables 1 and 2 show the filling accuracy of the was‘MHE2
10 ms. model
In addition, the differences
F Company’ using a
of the response rates of the solenoid valves were 3 ms for M company and
pulse output flow meter and the repeat accuracy of the filling valve. The initial test value was1 ms for F disrupted
company,
indicating
by thatfactor,
an external the valves had aingood
resulting someresponse rate and
of the valves speed difference.
exceeding the maximum filling level. However,
after removing the disrupting factors, the levels were all within rate
As for the response rate of the filling valve, the maximum differences of the solenoid valves
the limit.
of MFigures
company 8 and 9 show that the average filling values of the filling valves the
ranged from 14ms to 16ms, which is significant. However, were,result was 10~11
respectively, ms for
501.4 gr
F company, indicating that the products of this company performed better in terms of the
and 501.9 gr, which were within the target value of 500 gr ± 3 gr. In addition, the filling precision and differences.
iteration accuracy were found to be satisfactory under a high pressure and high-speed filling condition
(34.2. Comparison
bar, 350 bpm (bottleof the response characteristics
per minute) or higher).of Therefore,
the filling valve
it is believed that the proposed filling system
can be a good alternative for the conventional mechanical type or the load cell communication 7type
Appl. Sci. 2019, 9, x FOR PEER REVIEW of 11
Table 1. Accuracy measurement value of filling valve.
flow meter.
Standard Dispe Minimu Middle Maximu
Average
510 Deviation Rsion m Value Value m Value
501.9 0.38 0.14 501.3 502.0 502.6

Table 2. Repeatability measurement value of filling valve.

505
Standard Dispe Minimu Middle Maximu
Average
Deviation Rsion m Value Value m Value
Volume(gr)

501.9 0.38 0.14 501.3 502.0 502.6

500

Figures 8 and 9, Table 1 and 2 show the filling accuracy of the ‘MHE2 model F Company’ using
y = 0.057x + 501.11
a pulse output flow meter and the repeat accuracy of the filling valve. The initial test value was
R² = 0.0823
disrupted by an external
495 factor, resulting in some of the valves exceeding the maximum filling level.
However, after removing the disrupting factors, the levels were all within the limit.

490
0 2 4 6 8 10
Count

Figure8.8.Accuracy
Figure Accuracymeasurement
measurementvalue
valueofoffilling
fillingvalve.
valve.

510
 490
0 2 4 6 8 10
Count

Appl. Sci. 2019, 9, 2491 Figure 8. Accuracy measurement value of filling valve. 7 of 10

510

505
Volume(mg)

500

y = -0.043x + 502.11
R² = 0.106

495

490
0 2 4 6 8 10
Count

Figure9.9.Repeatability
Figure Repeatabilitymeasurement
measurementvalue
valueofoffilling
fillingvalve.
valve.

Table 1. Accuracy measurement value of filling valve.

Figures 8 and 9 show that the average filling values of the filling valves were, respectively, 501.4
gr and 501.9 gr, which Standard
were within the target value
Dispe of 500 gr ± 3 gr.
Minimum In addition,Maximum
Middle the filling precision
Average
and iteration accuracyDeviation
were found toRsion Value
be satisfactory under a highValue
pressure andValuehigh-speed filling
condition 501.9
(3 bar, 350 bpm0.38
(bottle per minute)
0.14 or higher).
501.3 Therefore, it is believed 502.6
502.0 that the proposed
filling system can be a good alternative for the conventional mechanical type or the load cell
communication type flow meter.
Table 2. Repeatability measurement value of filling valve.

4.3. Counter Pressure andStandard

Average Exhaust Time Dispe Minimum
Analysis of the Chamber Middle Maximum
Deviation Rsion Value Value Value
Counter
501.9
pressure and
0.38
exhaust time
0.14
analysis 501.3
of the chamber was analyzed
502.0
using AMESim
502.6
(Advanced Modeling Environment for Simulation of Engineering System) [6].

4.3. Counter Pressure and Exhaust Time Analysis of the Chamber

Counter pressure and exhaust time analysis of the chamber was analyzed using AMESim
(Advanced Modeling Environment for Simulation of Engineering System) [6].
The pressurization time of the filling vessel was determined based on the following assumptions,
taking into account the boundary values of the design variables that have the greatest effect.

- Chamber volume: 523 cc

- Temperature: 293 K
- Supply pressure: 5 bar
- Orifice area: 19.625 mm2
- Chamber initial pressure: 1.013 bar

The pressurization and exhaust time according to the supply pressure condition of the chamber
was analyzed as shown in Table 3 through the analysis of the AMESim model. The model to calculate
the filling/exhaust time of the chamber was configured as shown in Figure 10. Figures 11–13 show
the result of the analysis of the pressurization and exhaust time within the CO2 carbonate gas at
ambient temperature. The pressurization and exhaust time was analyzed to be approximately 0.2 s
and 0.4 s, respectively.
 Appl.Sci.
Appl. Sci.2019,
2019,9,9,x2491
FOR PEER REVIEW 8 8ofof11
10
Appl. Sci. 2019, 9, x FOR PEER REVIEW 8 of 11
Appl. Sci. 2019, 9, x FOR PEER REVIEW 8 of 11
The pressurization time of the filling vessel was determined based on the following
The pressurizationTable 3. Pressure
time of the variation and timewas
filling vessel analysis of the chamber.
determined based on the following
The pressurization
assumptions, time ofthethe
taking into account filling values
boundary vessel ofwas determined
the design based
variables thaton thethefollowing
have greatest
assumptions, taking into account the boundary values of the design variables that have the greatest
assumptions, taking into account the boundary values ofAMESim
effect. the design variables that have the greatest
Analysis
effect. Supply Pressure Condition
effect.
- Chamber volume: 523 cc Counter Pressuring Time Exhaust Time
- Chamber volume: [bar] 523 cc
--Temperature:
Chamber volume:293 K523 cc [s] [s]
- Temperature: 293 K
--Supply
Temperature:
pressure:2935 K
bar
5 0.25 0.5
- Supply pressure: 5 bar 2
--Orifice
Supplyarea:
pressure:
19.6255 mm
bar
4 2 0.23 0.45
- Orifice area: 19.625 mm
--Chamber
Orifice area: 19.625
initial mm2 1.013 bar
pressure:
- Chamber initial pressure:
3 1.013 bar 0.21 0.4
- Chamber initial pressure: 1.013 bar

Figure 10. Modeling of the chamber counter pressure and exhaust time.
Figure 10. Modeling of the chamber counter pressure and exhaust time.
Figure
Figure 10.
10. Modeling
Modeling of
of the
the chamber
chamber counter
counter pressure
pressure and
and exhaust
exhaust time.
time.

Figure 11. Counter pressure and exhaust time of the chamber (5 bar).
Figure11.
Figure Counterpressure
11.Counter pressureand
andexhaust
exhausttime
timeofofthe
thechamber
chamber(5(5bar).
bar).
Figure 11. Counter pressure and exhaust time of the chamber (5 bar).

Figure12.
Figure Counterpressure
12.Counter pressureand
andexhaust
exhausttime
timeofofthe
thechamber
chamber(4(4bar).
bar).
Figure 12. Counter pressure and exhaust time of the chamber (4 bar).
Figure 12. Counter pressure and exhaust time of the chamber (4 bar).
 Appl. Sci. 2019, 9, 2491 9 of 10
Appl. Sci. 2019, 9, x FOR PEER REVIEW 9 of 11

Figure13.
Figure Counterpressure
13.Counter pressureand
andexhaust
exhausttime
timeof
ofthe
thechamber
chamber(3(3bar).
bar).

According
The to the and
pressurization result of thetime
exhaust experiment,
accordingthe filling
to the timepressure
supply of the chamber
conditionwas induced
of the chamber as
shown below:
was analyzed as shown in Table 3 through the analysis of the AMESim model. The model to calculate
−5
t = (0.92was
the filling/exhaust time of the chamber − 1/Pcs ) × 7.04 ×
configured asV/S × 10in
shown Figure 10. Figures 11–13 show (1)
the result of the(1)analysis
Equation of the pressurization
can be deduced and exhaust
from the experiment timeand
formula within the CO2 carbonate
is a combination gasand
of linear at
ambient temperature. The pressurization and exhaust time was analyzed to be approximately
exponential modelling. The pressurization time in the chamber in the actual design was approximately 0.2 sec
and
0.260.4 sec,0.5
s and respectively.
s was spent for the exhaust time, confirming that the pressurization time and the exhaust
time were similar to the design time and the result of the AMESim model [6].
Table 3. Pressure variation and time analysis of the chamber.
4.4. Performance Evaluation of the Developed Valve AMESim analysis
Supply pressure
Counter pressuring
It was difficult to either control or monitor each of the filling condition variables (pressure,
condition Exhaust time
temperature, time, flow rate, etc.) in the conventional time mechanical carbonated beverage fillers. In this
[bar] [sec]
study, the variation of filling time with the temperature [sec] and the pressure change, and the filling flow
rate with the electronic flow 5 meter, were monitored 0.25 and controlled, and 0.5the filling deviation was
measured by individual correction4 0.23
of the filling process. The filling quantity0.45 accuracy was measured
by measuring 10 different3 valves. The filling quantity 0.21 repeatability was measured 0.4 10 times with the
same valve, and the filling deviation was within ± 3 gr (standard deviation 0.57).
According to the result of
During a high-pressure the experiment,
carbonated beverage thefilling
fillingfor
time of the chamber
ambient temperature wasfilling,
induced as shown
filling occurs
below:
while turbulence increases, but it was difficult to control individually in the mechanical type. Therefore,
in this study, we achieved filling optimization by adjusting the back pressure −5 and the abovementioned
t = (0.92 − 1 / P cs ) × 7.04 × V / S × 10
calibration of the filling pressure and time among the filling condition variables. Through this study, (1)
we determined
Equation (1) that
can when the surface
be deduced areathe
from is constant
experiment in the orifice and
formula flow isofaa combination
filling valve, we can predict
of linear and
faulty filling and filling processes by predicting the flow speed when
exponential modelling. The pressurization time in the chamber in the actual design was opening a filling valve, monitoring
and controlling0.26
approximately filling
sec pressure
and 0.5 sec changes (controlling
was spent the supply
for the exhaust pressure
time, confirmingand exhaust pressure).
that the pressurization
In this study, we were able to control filling just by flow rate control
time and the exhaust time were similar to the design time and the result of the AMESim model while solving the inconstancy
[6].
in filling pressure by adjusting the back pressure during filling. In mechanical filling, we were not able
to control
4.4. Performancevarious filling condition
Evaluation variables,
of the Developed Valvesuch as temperature, pressure, time, etc., individually,
but as we switched to the pulse flowmeter type electronic filling valve, we removed all uncertainties
It was difficult to either control or monitor each of the filling condition variables (pressure,
of the inconstancy in the electronic flow rate pressure through compensation, thereby implementing
temperature, time, flow rate, etc.) in the conventional mechanical carbonated beverage fillers. In this
optimization. When filling with pressurized fluid, there are valves that match the filling pressure
study, the variation of filling time with the temperature and the pressure change, and the filling flow
to match the pressure of the filler bowl and the pressure of the filling bottle, and the filling valve
rate with the electronic flow meter, were monitored and controlled, and the filling deviation was
separately, so we are able to match the same pressure, decreasing faulty filling and increasing filling
measured by individual correction of the filling process. The filling quantity accuracy was measured
efficiency. This increases the filling time as more back pressure is applied if there is differential pressure.
by measuring 10 different valves. The filling quantity repeatability was measured 10 times with the
At this time, we are able to monitor pressure in the case of filling valves and decreasing faulty filling
same valve, and the filling deviation was within ± 3 gr (standard deviation 0.57).
and increasing filling efficiency by gas volume when filling carbonate by controlling the differential
During a high-pressure carbonated beverage filling for ambient temperature filling, filling
pressure and calibrate the pressure deviation by adjusting the differential pressure. Thus, we were able
occurs while turbulence increases, but it was difficult to control individually in the mechanical type.
to complete a filler with high versatility and good efficiency.
Therefore, in this study, we achieved filling optimization by adjusting the back pressure and the
abovementioned calibration of the filling pressure and time among the filling condition variables.
Appl. Sci. 2019, 9, 2491 10 of 10

Author Contributions: S.-B.O. conceived and designed the analysis, collected and analyze the data, funding
acquisition; S.-H.O. supervisor, validation, funding acquisition; B.-R.M. supervisor, validation.
Funding: Some results of this paper are the result of the SME Innovation Technology Development Project
implemented by the Small and Medium Business Administration (SA113921). This research was supported by the
Chung-Ang University research grant in 2014.
Conflicts of Interest: The authors declare no conflict of interest.

Nomenclature
Pc Gas pressure in the pilot cavity
Po Pressure in the orifice inlet
Vc Cavity volume when initial stroke is 0
Tc Gas temperature in the pilot cavity
.
mc Flow to pilot cavity
.
mo Flow through the orifice in the pilot cavity
fo Surface area of the orifice
µcv Flow coefficient
Ti Gas inlet temperature
R Gas constant
x Travel distance of the bellows
Pcs Chamber supply absolute pressure (Mpa)
V Chamber capacity (cm2 )
S Orifice effective section area (cm2 )

References
1. Thriveni, T.; Ahn, Y.J.; Choon, H.; Chilakala, R.; Ahn, J.W. A Strategy of Precipitated Calcium Carbonate
(CaCO3 ) Fillers for Enhancing the Mechanical Properties of Polypropylene Polymers. Korean J. Chem. Eng.
2015, 32, 206–215.
2. Eblinger, H.M. Handbook of Brewing: Processes, Technology, Markets; Wiley-VCH: Weinheim, Germany, 2009;
pp. 292–293.
3. Kuang, S.; Nie, Y. Preliminary Study on Processing Oil-Based Calcium Carbonate Mud (OBCCM) into Rubber
Fillers. Environ. Eng. 2006, 24, 45–47.
4. Middendorf, P.J.; Maclntosh, D.L.; Tow, L.V.; Williams, P.L. Performance of Electronic Flow Rate Meters Used
for Calibration of Air Sampling Pumps. Am. Ind. Hyg. Assoc. 2010, 40, 472–476.
5. Oh, S.B. Rotating-Type High-Speed Fluid Filling System Using Pulse-Type Electronic Flow Meter; WO2016/148325A1;
WIPO: Geneva, Switzerland, 2016; pp. 1–23.
6. Jang, J.S.; Kwon, O.S.; Lee, K.W.; Cho, I.H. Analysis of Operating Characteristics and Design Review of
Oxidizer Fill-Drain Valve. Korea Aerosp. Res. Inst. 2011, 10, 79–88.

© 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/).