Process analysis of acetaldehyde production from ethanol ...

© The 2020 Pure and Applied Chemistry International Conference (PACCON 2020)

Process analysis of acetaldehyde production from ethanol via non-oxidative

dehydrogenation Thanakarn Suthirojn1, Bunjerd Jongsomjit2*, Pongtorn Charoensuppanimit1*

1Control and Systems Engineering Research Laboratory, Department of Chemical Engineering, Faculty of

Engineering, Chulalongkorn University, Bangkok, 10330, Thailand 2Center of Excellence in Catalysis and Catalytic Reaction Engineering, Department of Chemical Engineering,

Faculty of Engineering, Chulalongkorn University, Bangkok, 10330, Thailand

*E-mail: [email protected], [email protected], [email protected]

In Thailand, the utilization of ethanol is generally for the production of gasohol, a

greener fuel used in automobiles. However, the demand for gasohol may decrease in the

coming future due to the rapid development in energy storage used in the electrical vehicles.

Such advancement may cause the reduction of gasohol usage which could directly affect the

ethanol demand. To tackle this, ethanol may be utilized as a precursor for the synthesis of more

valuable chemicals such as diethyl ether, acetaldehyde, etc. In this work, the synthesis of

acetaldehyde is selected due to acetaldehyde is more costly than ethanol and acetaldehyde has

a wide range of applications. Provided the easy separation products and the attainment of

valuable in-situ hydrogen, the non-oxidative dehydrogenation of ethanol is chosen for process

analysis and economic evaluation in this study. In this research, Aspen Plus will be utilized for

the process simulation of acetaldehyde production from ethanol. A useful sensitivity analysis

of process parameters such as operating conditions, raw material, and production capacities

will be conducted to comprehend the impacts of these effects on the process feasibility. In

addition to the techno-economic results will be employed to justify whether the process is

suitable for ethanol valorization. A profitable process that may be obtained in this work can

generate benefit to the bio-ethanol which is one of the bio-refinery products receiving more

interests in the short run. Lastly, the catalysts activity can case the limitation in purity of the

products and affect the price of acetaldehyde .

1. Introduction

The utilization of ethanol in Thailand

is generally for the gasohol production, a

greener fuel mixed between benzene and

ethanol used in automobiles. However, the

demand for gasohol may decrease in the

coming future due to the rapid development

in energy storage used in the electrical

vehicles (EV).1 Such advancement may

cause the reduction of gasohol usage which

could directly affect the ethanol demand.

These may affect the fuel car ratio in the

future. Due to bioethanol being fuel grade, it

has low cost and value. To tackle this, ethanol

may be utilized as a precursor for the

synthesis of more valuable chemicals such as

diethyl ether, acetaldehyde, etc.

Acetaldehyde is widely used as the

starting material for synthesis of many

industrial chemical products such as acetic

acid, acetate esters, and pentaerythritol.

Acetaldehyde is more costly than ethanol.

Acetaldehyde has a wide range of

applications such as food preservative and as

a precursor to vinylphosphonic acid that is

used as adhesive or ion conductive

membrane. Therefore, in this work, the

synthesis of acetaldehyde is selected due to

previously several decent attributes.

In recent years, ethanol processing

has been more interesting because the

amounts of ethanol have become from

biomass treatment (bio-ethanol). There is

various research that studies ethanol

processing such as partial oxidation and

steam reforming. Therefore, the non-

oxidative dehydrogenation is an essential

first step in ethanol steam reforming and

ethanol partial oxidation reactions. Typically,

there are two methods involved in IE1


acetaldehyde productions - e.g., partial

oxidative dehydrogenation of ethanol and

non-oxidative dehydrogenation of ethanol.

Previous study, they have

investigated non-oxidative dehydrogenation

but there is inert gas in their process.2-5

Although the addition of inert gas to reactions

is more useful in lab-scale reactions, it cannot

be use for scaling up to commercial scale.

When adding inert gas into dehydrogenation

reaction, it can cause problem with separation

process. Provided the easy separation

products and the attainment of valuable in-

situ hydrogen, the non-oxidative

dehydrogenation of ethanol is chosen for

process analysis and economic evaluation in

this study.

In this research, Aspen Plus will be

utilized for the process simulation of

acetaldehyde production from ethanol. A

useful sensitivity analysis of process

parameters such as operating conditions, raw

material and product sale prices as well as

production capacities will be conducted to

comprehend the impacts of these effects on

the process feasibility. Lastly, the techno-

economic results will be employed to justify

whether the process is suitable for ethanol

valorization. A profitable process that may be

obtained in this work can generate benefit to

the bio-ethanol which is one of the bio-

refinery products receiving more interests in

the near future.

2. Materials and Methods

2.1 Acetaldehyde production feedstock

In this research, the feedstock for

acetaldehyde production is 99 %wt ethanol

which is fresh feed and received from the

recycle unit. The amount of ethanol in feed

stream was calculated from the chemical

reaction is as follows:

C2H5OH CH3CHO + H2 (1)

The non-oxidative dehydrogenation

reaction (Eqn. 1) is the main reaction for and

acetaldehyde is the product for process

simulation. Nevertheless, the ethanol

dehydration which is side reactions can be

occurred as well to produce diethyl ether and

ethylene (Eq. 2 and 3) shown as follows:

2C2H5OH C2H5OC2H5 + H2O (2)

C2H5OH C2H4 + H2O (3)

The production rate of 12000 tons of

acetaldehyde per year was assumed,6 the

amount of 99 %wt ethanol in fresh feed

stream was estimated to be 26587.54 tons per

year according to mass balance calculation.

The ethanol conversion is approximately to

80%7; consequently, the amount of ethanol in

the recycle stream was estimated to be 6647

tons per year.

2.2 Product

The product expected to be produced

from this process is laboratory grade

acetaldehyde with a 98 wt% product purity

according to acetaldehyde suppliers.8 There

are two more side reactions from this process,

the ethanol dehydration; furthermore, diethyl

ether and ethylene are produced as by-

products as well.

2.3 Process flow diagram description

2.3.1 Thermodynamics model

In this study, all acetaldehyde

production processes were simulated by

Aspen Plus. The NRTL model was used in

the simulation with the NIST TDE database.

Since the suitable conditions for

acetaldehyde production is 1 atm and

temperature is 200 ํC7, predictive split

models suitable for low pressure reaction,

polar molecules and azeotropic mixtures

(acetaldehyde and diethyl ether, ethanol and

water) - e.g., NRTL, UNIQUAC and Wilson

were selected as a result to Eric Carlson’s

recommendation.9

2.3.2 Design and simulation

In this study, the acetaldehyde plant

was designed with a production rate of 12000

tons of acetaldehyde per year. All processes

were simulated with the RadFrac model for

precision distillation and the Rstoic model for

reactions.

Before designing the reaction and

separation, the azeotropes and boiling points IE2


are investigated by using the NIST TDE

database. The results showed that two

heterogeneous azeotropes between water-

ethanol and diethyl ether-acetaldehyde were

formed; for this reason, it’s more difficult in

the separation step. The conversion of

ethanol in the non-oxidative dehydrogenation

process is 80%. In addition, the selectivity of

acetaldehyde is 48%, resulting in the

selectivity of ethylene and diethyl ether is

51.29% and 0.71% respectively.7

Figure 1 shows the process flow

diagram of the non-oxidative

dehydrogenation of ethanol for acetaldehyde

production. The fresh feed of 99 %wt ethanol

(ETOH-F) is mixed with the ethanol in the

recycling stream (ETOH-RE) in mixer M-1.

The mixed stream (stream 3) is delivered to

furnace (H-1) to heat the stream to 400 ํC, a

proper temperature for acetaldehyde

production. The stream discharged from H-1

(stream 4) is fed to reactor R-1,and reaction

takes place in R-1. All vapor products from

reactor R-1 (stream 5) are sent into the

compressor (C-1) to raise the pressure and

heat exchanger (E-2, E-3 and E-4 ) to cooled

down temperature using boiler feed water,

cooling water and glycol solution

respectively. The process stream 5 was

cooled to 10 ํC and then stream 10 was sent

to knock-out drum (D-1) where the

separation between gas and liquid phases

occurred. The high pressure liquid in stream

11 was sent to valve to reduce the pressure to

about 2.3 atm which was proper for

acetaldehyde separation by using distillation

column, S-1. The distillation column was

used to distilled acetaldehyde from other

components. There were three product

streams from this column including

ACETALD, 13 and 14 streams. The

RADFRAC model was used to perform this

process. The column contained 10 stages in

the rectifying section and 7 stages in the

stripping section with a total stage excluding

the condenser and the reboiler. By utilizing

the Design Specification feature, the optimal

reflux ratio was computed to be about 1.56

which corresponded to the purity of 98.0 wt%

in the product stream. According to water and

ethanol is an azeotropic mixture, the

extractive distillation column (S-2) was used

to extracted ethanol from water by using

ethylene glycol. The ethanol from S-2 was

used as the recycled stream (ETOH-RE) for

mixing with ethanol fresh feed (ETOH-F)

and then fed to reactor repeatedly. The

distillation column (S-3) was used to distilled

water from ethylene glycol, and ethylene

glycol in the recycled stream (EG) is used for

S-2 repletely.

2.3.3 Economic analysis

In this work, the ethanol’s cost was

assumed to be 4.36 US$ per liter10 for raw

material cost. The purity of ethanol was

assumed to be 99%. The acetaldehyde’s price

which is produced from this process is 532.9

and 469.28 US$ per liter8 was used as a basis

for economic analysis. The costs of utility

were obtained from the Analysis, Synthesis

and Design of Chemical Processes11 and

given in the Table 1.

In the final, unit operation sizing

including cost estimations were performed by

Aspen Economic Evaluation to obtain

Internal Rate of Return (IRR), Profitability

Index (PI) and Payout Period (POP).

Table 1. Summary of utilities price.11 Utility Price Unit

Electricity 0.06 US$/kWh

Cooling water 0.067 US$/ton

Chilled water 0.185 US$/ton

Boiler feed water 2.45 US$/ton

Low pressure steam 12.68 US$/ton

Medium pressure steam 13.71 US$/ton

High pressure steam 16.64 US$/ton

3. Results and discussion

3.1 Acetaldehyde production performance


of ethanol was reacted in gas phase reaction.

The production rate of acetaldehyde was

assumed to be 12000 tons/year in the outlet

stream from the reactor (stream 5) with 80%

conversion of ethanol. On the contrary, the

capacity of purified acetaldehyde from the IE3


separation unit was obtained to be 10315.1

tons/year with acetaldehyde purity of 98 %wt

in the product stream (ACETALD

stream). According to various by-products

that occurred during the reaction such as

water, ethylene diethyl ether and hydrogen,

some amount of acetaldehyde may be lost

during the phase separation process (D-1).

Figure 1. Process flow diagram of the non-oxidative dehydrogenation of ethanol for

acetaldehyde production.

Table 2. Summary of thermal and electrical duties in each unit of the acetaldehyde production

process.

Table 3. Summary of Internal Rate of Return (IRR), Profitability Index (PI) and Payout

Period (POP) at different acetaldehyde prices.

Unit code Unit name Thermal duty (MW) Electrical duty (MW)

H-1 Fire heater 1.52 -

R-1 RSTOIC reactor - -

C-1 Compressor - 0.29

E-1 Heat exchanger 0.65 -




S-1 Distillation column 1.15 -

S-2 Extractive distillation column 0.75 -

S-3 Distillation column 1.45 -

Total 6.81 0.29

Acetaldehyde price

(US$ per liter)

%IRR PI POP (Year)

120 Not return profit 0.988 Not return profit

125 24.5 1.01 8.3

140 49.8 1.08 4.1

400 2869.2 1.74 1.04

445 78162.8 1.79 1.001

469.28 Not return profit 1.82 Not return profit

532.9 Not return profit 1.88 Not return profit

IE4


3.2 Effect of diethyl ether to

acetaldehyde purity

As mentioned in section 3.1)

Acetaldehyde production performance,

several by-products obtained during the

reaction. The by-products that almost effect

the purity of acetaldehyde is diethyl ether.

Although the selectivity of diethyl ether in

the side reaction is only 0.71%, but this

diethyl ether could form the azeotrope

mixture with acetaldehyde and the near

boiling point of acetaldehyde and diethyl

ether is 20.2 ํC and 34.6 ํC respectively. As

a result, the maximum purity of

acetaldehyde is only possibly to 98%wt.

3.3 Energy requirement for acetaldehyde

production

From Table 2, showed a summary

of energy requirements for acetaldehyde

production. It can be seen that the total

energy of thermal duty used is 6.81 MW.

Due to the process reaction temperature is

high 400 ํC, which means the heat

exchanger using in this process must be

cooled down the temperature from very

high to low temperature before delivered to

phase separation process. Not only high

reaction temperature, but also installation

of a compressor before cooling down the

process stream. The compressor (C-1)

using in this process is to compress the gas

stream before entered the phase separation

unit for preventing acetaldehyde flow out to

gas stream (GAS).Thus, the total electrical

duty was from the compressor unit using in

this process is 0.29 MW.

3.4 Economic and investment analysis

The acetaldehyde’s price as shown in

Table 3 was obtained from Aspen

Economic Evaluation. Table 3 summarizes

the IRR, PI and POP, showing that the

acetaldehyde’s price should be 125-445

US$ per liter that could return a profit to the

production plant from PI and POP had

value more than 1. The price 120 US$ of

acetaldehyde per liter could not return a

profit to the production from PI had value

less than 1 as shown in the table. For the

maximum price evaluation, the price of

acetaldehyde 469.28 and 532.9 US$ per

liter which obtained from the supplier could

not return a profit due to the price being

exaggerated.

According to the price from Table

3, the minimum price of acetaldehyde that

could return a profit to the production is 125

US$ per liter. The result shows that the

process would make a profit in 8.3 years

with the %IRR of 24.5 and the possible

maximum price of acetaldehyde is 445 US$

per liter. The price of 445 US$ per liter

could return a profit to the production in

1.001 years with the IRR of 78162.8%.

4. Conclusion


of ethanal to acetaldehyde. Possible purity

of acetaldehyde from the process is only 98

%wt due to an azeotropic mixture forming

between acetaldehyde and diethyl ether

which affect the distillation process.

According to simulation and economic

analysis, the possible minimum and

maximum price which could return a profit

is 125 and 445 US$ per liter, respectively,

evaluated by Internal Rate of Return,

Profitability Index and Payout Period.

In the final, the recommendation

from authors, the activity of catalysts used

in the non-oxidative dehydrogenation had

the selectivity of diethyl ether 0.71% which

caused the maximum purity of

acetaldehyde is 98 %wt. To recommend

this, for the improvement of catalysts

should not have side reactions which may

produce diethyl ether.

Acknowledgements

The authors would like to gratefully

acknowledge the financial support of the

Chemical Engineering department, Faculty

of Engineering, Chulalongkorn University.

References 1. Tunpaiboon, N. Business and Industrial

trends in 2018-2020: Ethanol Industrial. https://www.krungsri.com/bank/getme

IE5


IE6

dia/dcdd704f-25c0-4493-a4b3-

a2eab039cdc4/IO_Ethanol_2018_TH.a

spx (accessed June 17, 2019).

2. Shan J.; Janvelyan N.; Li H.; Liu J.;

Egle T. M.; Ye J.; Biener M. M.; Biener

J.; Friend M.; Flytzani-Stephanopoulos

M. Applied Catalysis B:

Environmental. 2017, 205, 541-550.

3. Neramittagapong A.; Attaphaiboon W.;

Neramittagapong S. Chiang Mai J. Sci.

2018, 35(1), 171-177.

4. Giannakakis G.; Trmpalis A.; Shan J.;

Qi Z.; Cao S.; Liu J.; Ye J.; Biener J.;

Flytzani-Stephanopoulos M. Catalysis.

2018, 61, 475-486.

5. Shan J.; Liu J.; Li M.; Lustig S.; Lee S.;

Flytzani-Stephanopoulos M. Appl.

Catal. B Environ. 2018, 226, 534-543.

6. Methanol Chemicals Company.

Historical Summary of Projects

Completed/Plants in Operation

https://chemanol.com/en/Default.aspx?

pageid=477 (accessed Aug 25, 2019)

7. Autthanit C.; Praserthdam P.;

Jongsomjit B. J. Environ. Chemical

Engineering. 2018, 6, 6516-6529.

8. Acetaldehyde.

https://us.vwr.com/store/

product/16902475/acetaldehyde-98

(accessed Jan 18, 2020).

9. Carlson, E.C. Don't gamble with

physical properties for simulations.

Chem. Eng. Prog. 1996, 35-46.

10. Ethyl Alcohol (Industrial Grade) 99%.

https://www.chemipan.com/a/th-

th/244-สิ น ค้า / 328-เ ค มี ทั่ว ไ ป / 16552-เ อ ทิ ล -

แอลกอฮอล์-เกรดอุตสาหกรรม-99-ethyl-alcohol-

industrial-grade-99-thai-250ml-m.html


11. Turton, R.; Bailie, R. C.; Whiting, W.

B.; Shaeiwitz, J. A. Analysis, Synthesis,

and Design of Chemical Processes,

fourth ed. Prentice Hall, 1998.


Techno-economic analysis of acetaldehyde production via partial oxidative

dehydrogenation of ethanol Sudarat Sompong1, Bunjerd Jongsomjit2*, Pongtorn Charoensuppanimit1*

1Control and Systems Engineering Research Laboratory, Department of Chemical Engineering, Faculty of

Engineering, Chulalongkorn University, Bangkok 10330, Thailand 2Center of Excellence in Catalysis and Catalytic Reaction Engineering, Department of Chemical Engineering,

Faculty of Engineering, Chulalongkorn University, Bangkok 10330, Thailand

*E-mail: [email protected], [email protected], [email protected]

In the near future, the demand for ethanol may decrease due to the technological

advancement in the electrical vehicles that could bring about the reduction of gasohol. To

remedy this; therefore, the ethanol may be utilized as a precursor for the synthesis of more

valuable chemicals such as diethyl ether, acetaldehyde and etc. In this work, the synthesis of

acetaldehyde is selected due to several advantages. The techno-economic evaluation is also

performed in this work.

In this research, Aspen Plus was utilized for the process simulation of acetaldehyde

production via partial oxidative dehydrogenation of ethanol under different catalyst, i.e.,

Ag/HAp and V-Zr-La/SBA-15. Process parameters such as product sale prices, production

capacity and energy requirement were investigated to study the impacts of these effects on the

process profitability. According to techno-economic results, it was found that the vapor phase

oxidation of ethanol using V-Zr-La/SBA-15 as a catalyst was appropriate for the production of

acetaldehyde because 1) the process consumed less energy, 2) the amount of CO2 was produced

relatively low, and 3) the process had a high %IRR.

1. Introduction Over the years, fossil fuels are

generally used to produce energy. However,

the fossil fuels are depletable since they are

non-renewable that can’t not be replenished

to match the energy consumption of human.

Therefore, the renewable energy is

considered as the alternative to fossil fuels

since it is generated from natural resources

such as biomass, wind, sunlight or water.1

Bioethanol is one of the biomass

products that can be produced by the

fermentation of starch or sugar crops. It can

used as a mixture in fuel for vehicles or as a

starting material for the chemical industries

such as acetaldehyde, diethyl ether, ethyl

acetate, and others.2 In Thailand, ethanol is

mainly used as a fuel additive for gasohol, a

greener fuel used in vehicles.3 However,

electrical vehicles (EV) with the

advancement in battery technology have

gained increased attentions due to its

environmentally friendliness and its saving

on fuel and maintenance costs compared to

the conventional internal combustion engine

cars (ICE).4 As such, the ethanol demand for

gasohol may decrease in the future. Thus, the

use of ethanol as a precursor to produce

higher valued chemical such as acetaldehyde

is a potential way for the bioethanol

utilization due to its several advantages

including; 1) Acetaldehyde is more

expensive than ethanol; 2) Acetaldehyde has

a wide range of applications such as a food

preservative, as a flavoring agent, and as an

intermediate for producing acetic acid, acetic

anhydride, pyridine, pentaerythritol,

peracetic acid and many other chemicals.5

Typically, the main chemical reactions

for acetaldehyde productions from ethanol

include dehydrogenation and partial

oxidative dehydrogenation reactions.6

Provided the lower operating temperature

and less coke formation on the catalyst

compared to another method, the partial

oxidative dehydrogenation reaction was

chosen for process analysis and economic

evaluation in this work. IE7

mailto:[email protected]



Thus, the aim of this work is to simulate

processes for acetaldehyde production using

various catalysts. Techno-economic analysis

of the simulated processes using Aspen Plus

are to be determined. Techno-economic

results provided herein can prove the

feasibility of this ethanol valorization process

via acetaldehyde production which may

suggest an alternative way of utilizing

ethanol in the future.


2.1 Simulation data

In this work, two processes of

acetaldehyde productions that used different

catalysts were investigated.

2.1.1 Silver nanoparticles supported on a

hydroxyapatite (Ag/HAp)7

The chemical reaction to produce

acetaldehyde was expressed in eq. (1).

C2H5OH + 0.5O2 CH3CHO + H2O (1)

The Ag/Hap showed high catalytic

activity, giving 18% conversion of ethanol

and acetaldehyde selectivity of 100% at the

reaction temperature of 277 °C and

atmospheric pressure.

2.1.2 Vanadium oxide doped in SBA-15

supported catalyst with Zr and La

modification (V-Zr-La/SBA-15)8

At 300 °C and atmospheric pressure,

about 98% ethanol conversion and 41%

acetaldehyde selectivity could be achieved

via partial oxidative dehydrogenation of

ethanol over V-Zr-La/SBA-15 catalyst.

The main reactions that takes place

were provided in the following equation.

C2H5OH + 0.5O2 CH3CHO + H2O (2)

The four main by-products of

acetaldehyde production were also occurred

including carbon dioxide, carbon monoxide,

ethylene and diethyl ether (DEE) as follows.

C2H5OH + 3O2 2CO2 + 3H2O (3)

C2H5OH + 2O2 2CO + 3H2O (4)

C2H5OH C2H4 + H2O (5)

2C2H5OH C2H5OC2H5 + H2O (6)

The simulation results of each process

were compared to select a suitable catalyst

for producing acetaldehyde from ethanol.

2.2 Feedstock estimation

In order to determine the total amount

of ethanol and air that must feed into the

system, acetaldehyde production capacity of

12,000 tons/year at the reactor outlet stream

was fixed and used to determine the size of

both processes. Table 1 shows the calculation

results based on stoichiometric ratio of each

process. Please note that the oxygen was

obtained from the atmosphere consisting of

approximately 21 mol% oxygen.

Table 1. The raw material quantities for

acetaldehyde production.

According to the table, since two

processes gave different ethanol conversion

and acetaldehyde selectivity, the total flow

rate of ethanol and air is therefore different at

fixed acetaldehyde production at reactor

outlet stream.

2.3 Process flow diagram and unit

operations

2.3.1 Process simulation and

thermodynamic model

As described in Section 2.1, the

operating conditions of these two processes for

acetaldehyde production were occurred at low

pressure. Also, the process consisted of polar

molecules and binary azeotropes (i.e., ethanol-

water and DEE-acetaldehyde mixture).

Therefore, according to Eric Carlson’s

guideline9, the activity coefficient model

(NRTL) was selected to describe the

acetaldehyde production process using Aspen

Plus.

2.3.2 Flowsheet description

Catalyst Mass flow rate (tons/year)

Ethanol Air

Ag/HAp 69,696.94 18,728.05

V-Zr-La/SBA-15 24,822.39 67,832.09

IE8


Figure 1. Process flow diagram of acetaldehyde production using Ag/HAp catalyst.

2.3.2.1 Acetaldehyde production using

Ag/HAp catalyst

As depicted in Figure 1, the process

started with mixing the fresh ethanol (F-

ETOH) and the ethanol recycle stream (R-

ETOH) in the mixer, M101. Then, the mixed

stream (M-ETOH) was delivered and mixed

with air in the mixer, M102 before entering

the fired heater, FH101 that preheated the

mixed stream to desired temperature of 277

°C, a suitable temperature for acetaldehyde

production. The discharge from FH101 was

fed to the reactor, R101 where the chemical

reactions took place. Then, the reactor

effluent stream was passed through to the

shell and tube heat exchangers, H101, H102

and H103, respectively, to cool down the

product stream from the temperature of 277

to 10 °C. The cooled stream was fed into a

gas-liquid separator, PS101. The liquid

stream from PS101 was sent to a pump where

the stream pressure was raised to about 2 bar

before entering the distillation column, D101.

The partial condenser was used to condense

the overhead product (98 wt% acetaldehyde)

as well as released any remaining gas. While,

the BOTTOM stream was fed into the

extractive column, D102 where the ethanol-

water azeotrope was eliminated using

ethylene glycol as a solvent, obtaining a high

purity of ethanol in the overhead stream. For

the bottom stream from D102, ethylene

glycol-water mixture was fed into a

distillation column, D103 where the solvent

was recovered. Pure ethanol and ethylene

glycol separated in the distillation unit were

fed back to M101 and D102, respectively.

2.3.2.2. Acetaldehyde production using V-

Zr-La/SBA-15 catalyst

According to Figure 2, ethanol

solution (F-ETOH) and air were mixed in the

mixer M2-101. Then, the mixed stream (M-

ETOH) was fed into the fired heater, FH201

where the mixed stream temperature was

raised to 300 °C, the operating condition of

the acetaldehyde synthesis. The mixed

stream (M-ET-A-G) was fed to a reactor,

R201. After the reaction, the product stream

was sent to the shell and tube heat exchanger,

H201. The product stream was cooled and

exited the H201 at 90 °C before entering a

compressor, C201 which had a discharge

pressure of 10 bar. Next, the product stream

was fed into a series of the shell and tube heat

exchangers H202, H203 and H204 for

reducing the temperature from 428 to 10 °C.

The discharge from H204 was fed into a gas-

liquid separator, PS201. The liquid stream

from PS201 was sent to a valve, which

lowered its pressure to 2.34 bar. Then, the

stream (DE-P) was fed into a distillation

column D2-101, to obtain the acetaldehyde

with a purity of 98 wt% appeared in the ACT

stream.

2.4 Economic analysis In this work, the techno-economic

analysis of acetaldehyde production process

was conducted using the Economics

Evaluator in Aspen Plus. The ethanol cost

was estimated to be about 4.36 US$ per

liter,10 the raw material used in this research.

A list of utility prices was provided in Table

2. IE9


Figure 2. Process flow diagram of acetaldehyde production using V-Zr-La/SBA-15 catalyst.

Table 2. Summary of utility prices11

3. Results & Discussion

3.1 Effect of DEE by-product on

acetaldehyde purity By applications, acetaldehyde with

the purity of 99 wt% and 99.5 wt% had been

extensively used. However, according to

Section 2.1.2, the partial oxidative

dehydrogenation of ethanol on V-Zr-

La/SBA-15 catalyst produced the diethyl

ether selectivity of about 1%. As a result,

the acetaldehyde purity could not fall into

the 99wt% range. This was because the

boiling temperature of diethyl ether

(34.6°C) closed to the boiling temperature

of acetaldehyde (20.2°C) that forms an

azeotropic mixture. From the simulation

results of this process, the maximum

possible purity of acetaldehyde was 98

wt%. Therefore, 98 wt% was selected for

the production of acetaldehyde which was

used in the techno-economic analysis in this

work.

3.2 Acetaldehyde production capacity

In the vapor phase oxidation of

ethanol, both Ag/HAp and V-Zr-La/SBA-

15 catalysts produced the equal amounts of

acetaldehyde productions of about 12,000

tons/year at the reactor outlet stream

(PRODUCTS). Nevertheless, after the

purification unit, acetaldehyde production

capacity was obtained to be about 9,935

tons/yr for Ag/HAp and about 7,901 tons/yr

for V-Zr-La/SBA-15 with a product purity

of 98 wt% in product stream (ACT). It

could be observed that the two processes

showed the different amount of

acetaldehyde production. Since there were

several by-products occurring as mentioned

in Section 2.1.2, some acetaldehyde content

was lost during the separation of these by-

products compared to a chemical reaction

using Ag/HAp catalyst that had only one

by-product, e.g. water.

3.3 Energy consumption

As seen in Table 3, for acetaldehyde

production using V-Zr-La/SBA-15 catalyst,

the total amount of electrical utility was

about 1.22 MW. Since this process

produced several by-products, the addition

of compressor was necessary. If the

compressor had not been utilized for

pressuring the gases, over 80% of total

acetaldehyde production would be lost in

GAS stream.

Regarding the thermal duty, the

partial oxidative dehydrogenation of

ethanol over Ag/HAp catalyst required

tremendous energy compared to another

process, e.g., the partial oxidative

dehydrogenation of ethanol over V-Zr-

La/SBA-15 catalyst. This was because the

low conversion of ethanol that led to the

increased number of distillation columns

for the recycling of raw materials (ethanol

Utility Price Unit

Electricity

Cooling water

Chilled water

Boiler feed water

Low pressure steam

Medium pressure steam

High pressure steam

Natural gas

0.06

0.067

0.185

2.45

12.68

US$/kWh

US$/ton

US$/ton

US$/ton

US$/ton

13.71 US$/ton

16.64 US$/ton

6.0 US$/GJ

IE10


and ethylene glycol). The total amount of

thermal utilities obtained from the

distillation unit was approximately 9.35

MW which accounted for about 61% of the

total thermal requirement.

Table 3. The total amount of electrical and

thermal utilities in each process.

3.4 CO2 emission

Due to the effect of CO2 on global

climate change, the total amounts of CO2

released from the two processes were also

considered. For acetaldehyde production

using Ag/HAp catalyst, CO2 was indirectly

produced from the usage of process utilities

which was estimated to be about 1797

kg/hr.

For acetaldehyde production using V-

Zr-La/SBA-15 catalyst, the total amount of

CO2 produced from the indirect CO2 (utility

usage) was about 942 kg/hr. The chemical

process also produced CO2, releasing into

the atmospheric of about 650 kg/hr. Thus,

the total CO2 emission was 1592 kg/hr.

As mentioned above, the large

amount of CO2 was produced if the large

amount of process utility was used. In the

case of using Ag/HAp catalyst, about 66%

of total CO2 emissions were mainly from

the distillation unit. Since the reaction

provided 18% ethanol conversion, the

distillation columns for recycling ethanol

and ethylene glycol were necessary that

contributed to the use of the large amount

of process utility. For another process (V-

Zr-La/SBA-15), the amount of CO2 emitted

from the distillation column was only 9% of

the total CO2 emissions due to the high

ethanol conversion in the reactor leading to

the small amount of utility usage.

3.5 Economic analysis

The acetaldehyde price range that

could return a profit was estimated to be

about 125-440 $/l for ethanol oxidation

using Ag/HAp catalyst and about 110-285

$/l for ethanol oxidation using V-Zr-

La/SBA-15 catalyst which were indicated

by the existence of Pay-out Period (POP)

and the Profitability Index (PI) of greater

than 1. The selected prices used for

comparative purpose were 110 and 125 $/l,

the minimum selling price of each process.

As seen in Table 4, if the

acetaldehyde price of 110 $/l was

determined in the case of using Ag/HAp

catalyst, the PI value would be lower than 1

which indicated that this process was not

profitable. Until the acetaldehyde price

raised to 125 $/l, the process began to return

a profit in 7.89 years with the IRR of

25.77%. However, both prices provided the

lower of %IRR comparing at the same price

of using V-Zr-La/SBA-15 catalyst. Thus,

the process with V-Zr-La/SBA-15 catalyst

had a potential of returning profit since this

process had a high %IRR within a few years

when compared to another process.

4. Conclusion

Acetaldehyde production via partial

oxidative dehydrogenation of ethanol over

Ag/HAp and V-Zr-La/SBA-15 catalysts

were investigated for comparative

purposes. According to the simulation and

techno-economic results, the amount of

acetaldehyde productions in the case of

using V-Zr-La/SBA-15 catalyst was lower

when compared to another catalyst, namely

Ag/Hap (at fixed acetaldehyde production

at the reactor outlet). However, the total

energy requirement and CO2 emission were

also lower as seen in Table 2 and mentioned

in Section 3.4, respectively. In addition, at

the same price of acetaldehyde, the process

that used V-Zr-La/SBA-15 catalyst could

make more profits indicated by the IRR

value. Therefore, the catalytic reaction

using V-Zr-La/SBA-15 as the catalyst was

Energy

consumption (MW)

Ag/HAp V-Zr-La

/SBA-15

Electrical duty 0.001 1.22

Thermal duty 15.30 6.04

IE11


Table 4. Pay-out Period (POP) and %IRR of each process at different acetaldehyde price

the most appropriate process for the

synthesis of acetaldehyde that used ethanol

as a starting material.

Acknowledgements

The authors would like to

acknowledge the Faculty of Engineering,

Chulalongkorn University for financial

support of this research.

References

1. Renewable Energy: An Overview.

https://www.nrel.gov/docs/fy01osti/27

955.pdf (accessed Jan 23, 2020).

2. Rana, R. H. Bioethanol Oxidation to

Value Added Products. Ph.D. Thesis,

Gujarat Technological University, July

2017.

3. Tunpalboon, N. ETHANOL.

https://www.krungsri.com/bank/getme

dia/aec4cecd-3ab2-4a81-aed9-4b1f2f

90ee32/IO_Ethanol_2018_EN.aspx


4. A Guide to Electric Vehicles;

Sustainable Energy Authority of

Ireland: Dublin.

5. NISHO. Occupational Health

Guidelines for Chemical Hazards;

Centers for Disease Control and

Prevention: Cincinnati, Ohio, 1981; 81-

123.

6. Acetaldehyde.

https://pubchem.ncbi.nlm.nih.gov/com

pound/Acetaldehyde#section=Methods

-of-Manufacturing (accessed Jan 23,

2020).

7. Xu, J.; Xu, X. C.; Han Y. F. Catal.

Today. 2016, 276, 19-27.

8. Autthanit, C.; Praserthdam, P.;

Jongsomjit, B. J. Environ. Chem. Eng.

2018, 6, 6516-6529.

9. Carlson, E.C. Chem. Eng. Prog. 1996,

35-46.

10. Ethyl Alcohol (Industrial Grade) 99%.

https://www.chemipan.com/a/th-th/

244-สิ น ค้ า / 328-เ ค มี ทั่ ว ไ ป / 16552-เ อ ทิ ล -

แอลกอฮอล์-เกรดอุตสาหกรรม-99-ethyl-alcohol-

industrial-grade-99-thai-250ml-m.html


11. Turton, R.; Bailie, R. C.; Whiting, W.

B.; Shaeiwitz, J. A. Analysis, Synthesis,

and Design of Chemical Processes,

fourth ed. Prentice Hall, 1998.

Acetaldehyde

price (US$/l)

Ag/HAp V-Zr-La/SBA-15

PI POP (YEAR) %IRR PI POP (Year) %IRR

110 0.94 - - 1.02 7.59 26.86

125 1.02 7.89 25.77 1.10 3.57 58.13

IE12

https://www.nrel.gov/docs/fy01osti/27955.pdf

https://www.nrel.gov/docs/fy01osti/27955.pdf

https://pubchem.ncbi.nlm.nih.gov/compound/Acetaldehyde#section=Methods-of-Manufacturing




Economic model predictive control with applications to autoclave

polymerization reactors Trittapon Boonchoosri, Rungroj Yadbantung, Pornchai Bumroongsri*

Department of Chemical Engineering, Faculty of Engineering, Mahidol University, Phuttamonthon 73170,

Thailand

*E-mail: [email protected]

Model predictive control (MPC) belongs to a class of an optimal control that is widely

used in industrial processes. It can handle the complicated systems involving with multiple

variables subject to process constraints. The control of the autoclave polymerization reactor is

a major problem in ethylene polymerization process due to complicated behaviors of

polymerization. In this paper, an economic model predictive control (EMPC) algorithm is

developed to control the molecular weight and temperature within the autoclave polymerization

reactors. The economic performance cost is included in order to control the process in an

optimal way. The developed EMPC algorithm is applied to a numerical case study of the low-

density polyethylene (LDPE) autoclave reactor including the comprehensive polymerization

kinetics.

1. Introduction

Low-density polyethylene (LDPE)

is a thermoplastic material that can be

mostly found in daily life. LDPE is widely

used in various industries due to its high

flexibility and relatively low cost. It can be

used in many applications such as

packaging, insulation, coating and film.

LDPE is produced from the polymerization

of ethylene monomer. At high pressure and

temperature, the initiator decomposes and

initiates the polymerization reaction. An

autoclave polymerization reactor is a

continuously mixed vessel with multiple

reaction zones in series.1 The mixing in an

autoclave polymerization reactor is

provided by the impellers. The

development of the advanced process

control algorithm is necessary for the

autoclave polymerization reactors in order

to achieve the highest benefits.

In the industrial processes, the real

time optimization usually provides the

desired targets and the model predictive

control (MPC) is used to regulate the

system to the targets. For the economic

MPC, the economic objective function is

used directly such that the transient profit

can be achieved.2 For this reason, the high

benefit can be obtained with guaranteed

stability.3

In this study, the economic model

predictive control (EMPC) algorithm is

developed to control the molecular weight

and temperature within the LDPE autoclave

reactor. The economic performance cost is

included in order to control the process in

an optimal way.


2.1 Reactor modeling

The ethylene polymerization is

initiated by the peroxide initiator such as

dioctanoyl peroxide.4 The homogenous

single phase with well mixed condition is

considered in this work. Figure 1 shows the

closed-loop control of an LDPE autoclave

reactor. The feed components consist of

ethylene monomer and initiator. The

polymer product is released at the bottom of

the reactor. The feed flow rates of ethylene

monomer and initiator are considered as the

control variables in this system. The

developed EMPC receives the

measurement signals from the autoclave

reactor. Then, the EMPC adjust the inputs

that are the feed flow rates of ethylene

monomer and initiator. IE13


Figure 1. Close-loop control of an LDPE

autoclave reactor.

The material and energy balances

can be written as follows

2fI f

d

q I qIdIk I

dt V

(1)

0

fM f

p

q M qMdMk M

dt V

(2)

fM f ttq M qMdM

dt V

(3)

0( ) f fP P

P

q T qTH k MdT

dt C V

(4)

where I is the initiator concentration, M is

the free monomer concentration, tM is the

total monomer concentration and T is the

reactor temperature. The notation ( )k

are

referred to the kinetic parameters as shown

in Table 1. The symbols fMq and

fIq

denote the feed flow rate of monomer and

initiator, respectively. The symbols fM

and fI denote the monomer concentration

and initiator concentration in the feed,

respectively. The total feed flow rate and

exit flow rate are given by fq and q ,

respectively.

Table 1. Kinetics parameters.a Rate constants Units

14 4

1 2d = 1.83 10 exp( 3.06 10 / ( ) 5.9 / ( ))RT P R Tk 1/s 3 4

1 = 6.04 10 exp( 3.87 10 / ( ))th R Tk L2/mol2 s

5 3

1 2 = 5.12 10 exp( 4.21 10 / ( ) 5.6 / ( ))p R T P R Tk L/mol s

9 3

1 2 = 2.53 10 exp( 3.37 10 / ( ) 9.2 / ( ))tc R T P R Tk L/mol s

4 4

1 2 = 1.20 10 exp( 1.44 10 / ( ) 20 / ( ))fm R T P R Tk L/mol s

5 3

1 = 1.80 10 exp( 9.04 10 / ( ))fp R Tk L/mol s

9 4

1 2 = 1.40 10 exp( 1.93 10 / ( ) 9.9 / ( ))R T P R Tk 1/s

5 3

1 = 3.27 10 exp( 7.47 10 / ( ))b RTk 1/s

11.987

aR cal/mol K, 2

82.06R cm3 atm/mol K.

The heat of polymerization ( PH

)is a function of the reactor temperature (K)

and pressure (bar) given by the following

correlation [4] 84185pH

0.209 273.25 0.105T P . (5)

The density of the reaction mixture is also a function of the reactor

temperature (K) and pressure (bar) that can

be writes as 1

1995.8 261.1logT

1

63.3log log1000

P

T .

(6)

The heat capacity PC of the reaction

mixture is given by 0.518(1 )p MC x

4(1.041 8.3 10 ) MT x (7)

where the monomer conversion Mx can be

computed by the following equation

1 = -M

t

xM

M. (8)

In this study, the method of

moments is used to formulate the dynamic

model. The statistical description is used to

define the weight average molecular

weight. The moment expressions are

obtained by summing the multiplication

between the polymer with length n and its

moment order in as follows IE14


1

i

i n

n

n R

(9)

1

i

i n

n

n P

(10)

where 0,1,2,3i . The notations i and i

refer to the i -order moment of the free

radical nR and the dead polymer nP ,

respectively. Based on the detailed kinetic

scheme, the generalized expressions can be

formulated. The dynamics of the dead

polymer moments i for 0,1,2i can be

written as

2 00

00

1 ( )

2tc fM b

qk k M k

V

d

dt

(11)

0 01

1 1 1 ( )tc fM bk k kd

Mdt

11 1 0 2( )fp

qk

V

(12)

2

0 2 1 22 ( )tc fM

d

tM

dk k

2 1 0( 2 )bk

22 2 0 3( )fp

qk

V

(13)

The quasi-steady-state assumption

on the free radicals nR is made and the

reactor outlet is considered to contain only

the dead polymer nP . The dynamic model

of the free radical moments can be solved

to find the free radical moments as follows

0 I

tc

R

k (14)

2 0

1

0 1

( 2 )

I p fm b fp

tc fm fp b

R k M k M k k

k k k k k

(15)

1 0

2

0 1

(2 )

I p

tc fm fp b

R k M

k k k k k

3 0

0 1

( 2 )fm b fp

tc fm fp b

k M k k

k k k k k

(16)

In the calculation of the high-order

moment 3i , we use the approximation of

Hulburt and Katz [5] which can be

expressed as

223 0 2 1

0 1

2u

. (17)

The weight average molecular

weight wM related to the mean of average

moments can be computed as

2 2 20 0

1 1 1

wM M M

(18)

where 0M is molecular weight of the

single ethylene monomer unit. The reactor

parameters can be shown in Table 2.

Table 2. Reactor parameters. Reactor parameter Value Unit

P 1,700 bar

0M

28.05 g/mol

fI

0.0916 mol/L

fM

20.9 mol/L

fT

313 K

V 1,500 L

2.2 Development of EMPC for LDPE

autoclave polymerization reactors

The differential equations of the

LDPE autoclave reactor is given by

,dx

x t f x t u tdt

(19)

where nx X is the vector of system

states and mu U is the vector of input

variables. They can be written as

0 1 2[ ]T

tx I M M T (20)

[ ]T

fI fMu q q . (21)

The economic objective function to

be minimized is given by

1

1

,N

e e

k

V x k u k

(22)

where e is the economic performance cost.

The monomer conversion Mx is considered

as the economic objective which is relative

to IE15


, (2).e x u wx (23)

where w is the weighting value of

economic function.

In this study, the orthogonal

collocation is used to solve the optimization

problem [6]. The collocation elements can

be illustrated in Figure 2.

Figure 2. Collocation elements.

The orthogonal collocation

technique assumes that the dynamic

equations (19) can be expressed as the

summation of Lagrange polynomials [7].

The Lagrange interpolation polynomials

can be written as

0,

K

kj

k k j j k

t t

t t

(24)

where K is the number of collocation

points, jt and kt are times at j and k

collocation points, respectively. The

collocation equations can be expressed as

,

0

,K

j

i j

j

d tx f x u

dt

(25)

where ,i jx denotes the state at element 𝑖

and collocation point j .

The economic model predictive

control is the MPC method that uses the

economic stage cost in its formulation. The

objective function of EMPC can be written

as

1

0

| , |N

N m

k

V x t k t u t k t

(26)

The optimization problem for

EMPC is subject to the constraints of

orthogonal collocation with initial and

terminal points. The state and input

variables lie within their constraint sets. At

each sampling time t , the input u is

computed by solving the following

optimization problem

,

min Nu x

V

subject to

1| | , |x k t f x k t u k t

|x k t X

|u k t U

| sx k N t x

00|x t x

where the notations ( )( | )t k t represent

the values of the respective variables at time

t k predicted by the information at time t

and N is the prediction horizon. Note that

m is the modified economic objective

function that must satisfy the dissipativity

condition8 defined as the summation

between the economic cost function and the

tracking convex term .m e t (27)

The tracking convex term with respect to

the optimal steady state can be written as

2 21

2s sQt R

x x u u (28)

where Q and R are the weighting

matrices. For the pure tracking MPC

scheme, the objective function can be

written as

1

1

N

T T

k

V x k Qx k u k Ru k

(29)


The performance of the developed

EMPC is compared with the tracking MPC.

The sampling time and the prediction

horizon are chosen to be 10 s and 8N ,

respectively.

The economic objective function satisfies

the dissipativity condition with the

weighting matrices diag(1,1,1,1,1,1,1)Q , IE16


diag(1,1)R and 0.1w . For the tracking

MPC, the same weighting matrices

diag(1,1,1,1,1,1,1)Q and diag(1,1)R

are used. The state trajectories of the EMPC

and tracking MPC are shown in Figure 3.

The trajectories of the EMPC and tracking

MPC are shown by red and blue lines,

respectively.

Figure 3. State and output variable profiles:

initiator concentration (a), monomer

concentration (b), total monomer

concentration (c), reactor temperature (d),

0th order (e), 1st order (f), 2nd order

moment of dead polymer (g) and weight

average molecular weight (h).

The monomer concentration which

considered as economic performance is

shown in Figure 4.

Figure 4. Monomer conversion (economic

performance).

The input trajectories are shown in

Figure 5.

Figure 5. Input variable profiles: initiator

feed rate (a) and monomer feed rate (b).

Figure 3 shows the state and output

profiles of TMPC and EMPC. We can see

that TMPC can be driven in the propose of

minimum deviation of the steady-state

values such that the trajectories can be

move to the set points as faster than EMPC.

The state and output trajectories of EMPC

have more oscillation regimes than TMPC.

The monomer concentration as shown in

Figure 3b is obviously seen that EMPC is

lower level than TMPC due to the economic

equation (23) is included in the objective

function. In Figure 4, the monomer

conversion considered as the economic

objective of EMPC is oriented to have

higher value than TMPC. For the input

variables shown in Figure 5, we also can see

the fluctuation of EMPC are corresponding

to state and output variable profiles. The

average values shown in Table 3 can be IE17


confirmed that EMPC gain more the

economic benefit than TMPC.

Table 3. Average economic performance. Average parameter EMPC TMPC

Mx 6.57 5.93

4. Conclusion

The economic MPC is developed in

this work in order to control the autoclave

polymerization reactor with complicated

nonlinear dynamics. The economic

performance cost is included in the

objective function in order to control the

system in an optimal way. The results show

that the developed EMPC gives higher

economic performance as compared with

the conventional tracking MPC.

Acknowledgements

This work is supported by Mahidol

University.

References

1. Havard N.; Osivind M.; Peter S.

Prediction of Molecular Weight J.

Macromol. Sci. 1997, 34 (6), 1017–

1043.

2. Amrit R.; Rawlings J. B.; Biegler J. T.

Comput. Chem. Eng. 2013, 58, 334–

343.

3. Jacob N. C.; Dhib R. J. Process

Control 2011, 21 (9) 1332 –1344.

4. Klwzraei P.; Dbib R. J. Appl. Polym.

Sci. 2008, 109 (6) 3908–3922.

5. Hulburt H. H.; Katz. S. Chem. Eng. Sci.

1964, 555 –574.

6. Biegler L. T.; Nonlinear Programming:

Concepts, Algorithms, and

Applications to Chemical Processes.

2010.

7. Eatonetal J. W.; GNU Octave-a high-

level interactive language for numerical

computations. fifth edition. 2019.

8. Diehl M.; Amrit R.; Rawlings J. B.; A

lyapunov Function for Economic

Optimizing Model Predictive Control

2011. In IEEE Trans. Automatic

Control 2011, 56 (3), 703–707.

IE18


Effect of extrusion variables on sidewall torn edge produced from duplex

extruder Nutchapon Sawatdiwutthipong1, Chaiyan Chaiya2, Pongsak Nimdum3,

Prasert Reubroycharoen1* 1Department of Chemical Technology, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand

2Department of Chemical Engineering, Faculty of Engineering, Rajamangala University of Technology

Thanyaburi, Pathum Thani 12110, Thailand 3Department of Mechanical & Aerospace Engineering, Faculty of Engineering, King Mongkut’s University of

Technology North, Bangkok 10140, Thailand


Work away is a waste and a seriously problem in the tire industry. Based on the data

collections, the 9% of work away was mostly found from sidewall part production. Torn edge

is a kind of the work away from sidewall and derives from the unsteady flow of compounds

through screw of extrusion. In this research, the existence of torn edge from the sidewall was

analyzed by 4M1E analysis tool (man, machine, material, method and environment). It was

found that the screw of extruder may be the key factor to solve the existence of torn edge. The

operating factors as screw clearance, screw speed, and temperature, were studied to clarify this

problem. The results showed that the improper gap of screw clearance was the main factor

affect to existing the torn edge of the sidewall.

1. Introduction The automotive production is one of

the important industry in Thailand.

According to the Ministry of Industry data in

2018, production volume was more than 2.20 million units compared to the similar period

in 2017, with the production volume of 1.98

million units. The production volume was

increasing for 5.59% in between both years.

In 2018, the vehicles were sold

approximately 1.05 million units in Thailand

which is enhancement for 5.00% and 1.15

million units for exportation that is an

increase of 4.55%. These data also affect to

the growth of tire industry and it is one of the

S-curve group. The government has been

promoting and driving tire industry growth to

be a driving force in the economy of the

country.1 In Thailand, tire manufacturers are

separated into 2 groups which are small and

large producer groups. Multinational

companies usually have advanced

technology from abroad and their head

offices are from their own countries. They

have production bases in various countries

due to they need to reduce labor costs, solve

the labor shortage and receive raw materials

from the source.2 Thailand is one of the main

production bases for automotive tires. In

addition, labor cost is low in Thailand. It is

also the source of raw materials for

production of established manufacturing base

in Thailand for long time. Nowadays, there is

economic problems, for example, the

increase of labor cost. Moreover, consumer’s

demand is more diverse and complex. Thus,

the industries have to adapt and reduce

production costs but keep the quality and

standard of products due to the product

quality is important. This will be the key to

develop the tire industry and lead to be

automotive industry leader. There is large

amount of waste which affects internal costs

and product prices. The important things that

customers will consider when they make a

purchase decision are quality and price. In

case that the companies can make high

quality products with low-cost, they will be

able to compete with other companies when

it is high competition nowadays. Therefore,

the important factors that manufacturers need

to pay attention is how to control the products

to gain high standard and product quality,

reduce waste in the production process and IE19


reduce production cost as well. They found

that sidewall tires had the highest loss in the

production process around 80,000 kg

from the production of 843,000 kg,

representing an average percentage generated

work away of about 9%, which exceeded the

standard of production of 6% due to collect

data from October 2018 to March 2019. It

was a high cost of waste due to large quantity

of energy and labor costs. The researchers are

interested to study and solve this problem

which can be seen that whether the company

is able to reduce the amount of waste. The

decrease of waste will be the great impact to

reduce the cost of production per unit, and be

able to compete with other companies.

Figure 1. The amount of work away generated during October 2018 - March 2019.

The goal of this research is to reduce

waste of sidewall tires by using quality tools

to find the cause. Moreover, this research

wants to improve the quality of the

production process from October 2018 –

March 2019. In addition, the inspection

Component of Green-tire4

sheet from production process were

used for measuring the amount of tire waste

and collecting them for classify the problems

by using 4M1E method. Usually, the tire that

is not cured, it is called “green tire”. It

includes several parts as showed in Figure. 2.

Ref: https://www.uniroyal-tyres.com/car/tyre-guide/tyre-knowledge/tyre-components.

Figure 2. The components of Green-tire.

The data from production sheet shows

the highest amount of tire waste. It is from

“sidewall” which is number 6 in figure 2. There is damage form that showed in

sidewall part called “torn edge”. It is the

218,000

181,000

123,000

318,000

238,000

287,000

19,000 26,000 8,000

30,000 25,000 25,000

-

50,000

100,000

150,000

200,000

250,000

300,000

350,000

Oct-18 Nov-18 Dec-18 Jan-19 Feb-19 Mar-19

Wei

gh

t (K

g)

MonthProduct Waste

IE20


tearing at the edges of the extruded

component impact the final steps in

production and detract from the appearance

of the final product. (See Fig.3)

Figure 3. Torn edge of sidewall.


2.1 Materials

The material which was used for this

research are two compounds. They were

extruded by duplex extruder and passed to die

(mold). It was called sidewall and collected

from manufacturing process every week for 7

months.

2.2 Experimental Conditions

This research focuses on the extruder

by two-head single screw extruder, which the

size of each screw head was 6 and 8 inch.

Speed of both screws were fixed with (1) the

screw temperature at 85oC, (2) the

temperature at rubber compound entrance at

40oC, (3) The barrel zone temperature of 6

inch has two zones at 55 and 70oC,

respectively. While (4) the another side has

three zones of 50, 60 and 70oC, respectively.

Figure 4. Duplex extruder temperature

setting position. 2.3 Analysis method

The extrusion process of tire Generally, the green tire production

consists of the three important steps such as

mixing, preparation and building as showed

in figure 5.

Figure 5. Diagram of the production system.

The mixture from mixing department

was transformed by using two types of

extruder. There were Quadplex and duplex

extruders. The two types of extruders injected

the both parts of tire, which were tread and

sidewall. However, the non-conforming

products were generated during these processes. It was called work-away (W/A) or

waste that had to be reused to mixing

department again. The amount of waste was IE21


the important factor for the performance

analysis of the system and its evolution over

the time. This quantity indicator measured

the percentage ratio between the amount of

waste generated in the preparation

department.

Problem identification phrase

Periodically, this work identified the

problems that influence to rubber products.

This involved rubber checks by

brainstorming from people who have

experience in production to gather the causes

that may affect the occurrence of defective

products. In the study of the problem from

Table 1. (Main operations during the duplex-

extrusion process), the production process of

each machine was studied to specify interest

problems.

Table 1. Main operations during the duplex extruder process.

Stage Operation Description

1 Preparation and compound

feeding

The compound was produced in Mixing Department and

transported to the zone that was near the extrusion line and

stored in racks. The process began with the displacement of

the compound-by the operator- to the feed conveyor,

according to production needs.

2 Extrusion At this stage, the compound was heated and homogenized to

gain acquiring plastic state, and transported to the die through

the feed screw. In the extrusion head, the material was forced

through the pre-die and die, which are responsible for

defining the profile of the article to be produced.

3 Cooling and drying The material was cooled through showers or sprayed in

cooling line.

4 Cutting In sidewall extrusion, the cut was made by skiver.

5 Storage Sidewalls were rolled up at wind-up station.

Measure phase

The statistic of waste from sidewall

were collected in figure 6. The data shows the

three highest amount of wastes from sidewall

on January 2019 when it was start-run, end-

run and torn edge, respectively. They all were

produced by duplex-extruder. Thus, this

research focused on this machine to solve this

waste problem. However, it was found that

the waste from start-run and end-run had

already solved by the modification of duplex-

extruder. It seemed that wastes were

decreasing. Thus, the waste from torn edge

became the first priority’s problem that

should be solved suddenly.

Figure 6. Types of waste from sidewall part

on January 2019.

Problem analysis Phase

The 4M1E tool was used to analyze

the cause of main problem for the torn edge

of sidewall. Usually, 4M1E means the

possible factor which affect to the

production, such as man, method, material,

machine and environment. To get the results,

8,254

7,129

6,053

4,492

2,511

1,231 995

-

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

Weig

ht

9K

g)

Types of waste

IE22


we surveyed and collected the data from that

five items, then analyze it.

Figure 7. The 4M1E analysis of Torn edge.

1. Man

Working skills and experience are the

important factor for the operator. We found

that all 4 operators had expertise and

experience in setting different parameters.

This caused defective production due to

machine configuration and employee's

physical condition while working. Working

skills and the operator's experience should

have trained before work This will cause

similarly effective and provide additional

training to enhance the experience and new

knowledge to apply with work. In addition,

testing should be conducted to assess and

motivate them to be active in their work.

2. Method

Defective problems were caused by the

operator without following the work

procedures. There was not set frequency of

inspection and recording of problems and

machine costs, resulting in errors in the

production process and defect of production.

There was a work manual for employees. In

addition, the machine should be inspected to

confirm that the condition was free of any

changes or discrepancies and rules for

employees to be enthusiastic in their work by

keeping the workplace clean.

3. Material

Incubators for storing molds (die) were

unclean. This caused contamination for the

die. The obtained products may have quality

problem which was torn edge.

4. Machine

We found that machinery was the main

problem due to ongoing lack of maintenance,

resulting in unstable machine performance.

The sub-problem was that the gap between

the screw and cylinder exceeded the standard,

resulting in defective product quality. The

machine was checked every 3 months. When

encountering problems should be resolved

immediately.


Improve and control Phase

The results showed that the problems

of torn edge’s sidewall were solved

significantly. The cause of the waste was

analyzed by 4M1E and chosen the ways to

reduce waste in manufacturing process as

follows:

Machine, it was found that the gap between the screw and cylinder usually

exceeded the standard. The method was used

to solve this problem by replacing the new

screw and reducing the gap between screw

and cylinder close to the original. This work

found that the trend of the waste occurrence

was decreased. Regarding to the amount of

total sidewall compound before the

improvement had a percentage of waste equal

to 9%. Then the solution has been proposed,

the waste had decreased to 7% (see Fig.8),

while the amount of torn edge’s sidewall

before the improvement had a percentage of

waste equal to 14 % and then after the

solution had been decreased to 1%. (see

Figure 9).

IE23


Figure 8. The amount of waste by time.

Figure 9. The amount of torn edge’s sidewall.

4. Conclusion

According to study of effect of extrusion

variables on sidewall’s torn edge produced

from duplex extruder to analyze by 4M1E

tools. It was found that gap between screw

and cylinder may be the key factor to solve.

This solution affect to decrease in total waste

from 9 to 7%, while the amount of sidewall’s

torn edge was decreased from 14 to only 1%.

References

1. Pugna, A.; Negrea, R.; Miclea, S.

Procedia-Social Behav. Sci. 2016, 221,

308-316.

2. Loharnchun, P. Tire industry.

http://www.industry.go.th/roiet/index.ph

p/index.php?option=com_k2&view=ite

m&id=10537&rss_id=12618 (accessed

August 29, 2019).

3. Warintornnuwat, S. The viewpoint from

the survey of data in the study of status

and trends of business transfer of

Japanese companies to Thailand.

http://www.tpa.or.th/tpanews/upload/ma

g_content/58/ContentFile1056.pdf

(accessed May 25, 2019).

4. Tyre component. https://www.uniroyal-

tyres.com/car/tyre-guide/tyre-

knowledge/tyre-components (accessed

April 15, 2019).

8.7

9

9.3

2

7.9

5 9.7

5

8.8

7

8.4

2

7.5

5

6.2

9

5.4

7

6.3

4

7.1

5

6.7

5

0.00

2.00

4.00

6.00

8.00

10.00

12.00

Jan-19 Feb-19 Mar-19 Apr-19 May-19 Jun-19 Jul-19 Aug-19 Sep-19 Oct-19 Nov-19 Dec-19

%W

ast

e

14

.29

17

.59

10

.40

2.0

9

0.1

9

1.5

3

0.8

1

5.2

6

0.0

0

0.0

0

2.3

1

2.7

6

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

18.00

20.00

Jan-19 Feb-19 Mar-19 Apr-19 May-19 Jun-19 Jul-19 Aug-19 Sep-19 Oct-19 Nov-19 Dec-19

%(T

orn

ed

ge/W

ast

e t

ota

l)

Before

After

Before

After

IE24


Effect of compound age on work away tire from quadplex extruder

and properties of tire tread Piyatida Chantrangsuwan1, Pongsak Nimdum2, Chaiyan Chaiya3, Napida Hinchiranan1,4*

1Department of Chemical Technology, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand 2Department of Mechanical & Aerospace Engineering, Faculty of Engineering, King Mongkut's University of

Technology North Bangkok, Bangkok 10800, Thailand 3Department of chemical engineering, faculty of engineering, Rajamangala University of technology

Thanyaburi,Pathum Thani 12110, Thailand 4Center of Excellence on Petrochemicals and Materials Technology (PETROMAT), Chulalongkorn University,

Bangkok 10330, Thailand

*E-mail: [email protected] The growth of automotive tire market results in high demand for tire production. After

the end of production, excess compounds used as raw materials were disposed. Long-time

storage of compounds helps to reduce the operating cost. In this study, the effects of compound

aging (0-30 days) and temperature (25oC and 50oC) on properties of compound were

investigated. There were 3 types of compounds used in this research categorized by the type of

reinforcing fillers: carbon black, silica and silica/carbon wt ratio of 1.3. Every 3 days of storage,

the properties of rubber compounds were characterized by using Mooney viscometer and

Moving Die Rheometer (MDR). After vulcanization, the properties of the obtained

vulcanizates in terms of hardness, rebound, tensile properties and crosslink density were

evaluated. The results showed that after storage for 30 days, Mooney viscosity and the min

torque values of the rubber compounds increase, indicating that crosslink occurred in

compounds. For cure characteristics, the optimum cure time of silica and silica/carbon rubber

compounds decreased, while the optimum cure time of the compounds reinforced by carbon

black increased ca. 6.6%.

1. Introduction

Nowadays, cars are widely used as a

vehicle for transportation. the sales of both

commercial and passenger vehicles increased

from several factors such as changing

lifestyles, mounting income levels and rising

population. The growth in automobiles

consumption also leads to an increase in

rubber tire consumption and production. For

over the competition, tire industry produced

many types of tires and developed their

products such as ecological tires, flat run tires

and nitrogen-based tires that are

environment-friendly.

Tire manufacturers used several types

of raw material to cover all requirements

from customers. All components of tire e.g.

tread, sidewall, inner liner, bead, ply, belt

used compound as raw material. For one tire,

several types of compound depend on tire

structure and usability were used, the

quantity of the compound was estimated by

using monthly order as references. However,

compounds were created more than

estimates. There are some lost in the

production line because the component

which has defection was disposed before

sending to the building machine to prevent

tire malfunctions. Each compound was

produced many lots at the same time to

reduce time and cost then placed in storage

area at room temperature.

Compound which consist of rubber,

sulfur, accelerator, activator and additives

were produced by Bunbury mixer. Raw

materials were mixed in a mixing chamber

and sheeted by two roll mills1. For tire tread,

there were 3 types of compound categorized IE25


by using reinforcing fillers: carbon black,

silica/carbon wt. ratio of 1.3, and silica. Other

compounding ingredients included N-

Cyclohexyl-2-benzothiazole sulfenamide

(CBS) and Diphenyl Guanidine (DPG) were

used as accelerators in this research.

Compound had time limit within 14 days

from industrial policy. After long time

storage, excess compounds from production

were disposed because properties of

compound were out from tolerances. Cost

from produced and disposed excess

compound are the problem of manufacturing.

In this research, sample of

compounds were storage 0-30 days at 25oC

and 0-6 days at 50oC. Effect of time and

temperature were characterized by using

Mooney viscometer, Moving Die Rheometer

(MDR). After vulcanization, the properties of

the obtained vulcanizates in terms of

hardness, rebound, tensile properties and

crosslink density were investigated.


2.1 Materials

Compounds were received from tire

company. Butadiene rubber (BR) and

Styrene-Butadiene Rubber (SBR) were used

in this research.

Other additives were all industrial

grade products. Compound formula is shown

in Table 1. the samples of compounds were

kept at 25oC and 50oC for 30 and 6 days

respectively.

Table 1. The formulations of compounds

(phr). Raw

materials Sample1 Sample2 Sample3

BR

SBR

15

85

-

100

15

85

Carbon Silica

Silane

Processing aid

Plasticizer

Antioxidant Stearic acid

ZnO

79 -

-

1

19

2 2

2

33 40

4

2

14

2 2

2

12 69

7

2

19

3 2

2

Accelerator 2 3 3 Sulfur 2 1 1

2.2 Characterization and testing

Every 3 days of storage, the

properties of rubber compounds were

characterized by Mooney viscometer (MV-2000) with ML (1 + 4) at 100oC.

Rheometric characteristics were

carried out using an Alpha technologist

Rheometer (MDR 2000) according at 150oC.

For vulcanization, compounds were

cured by compression molding for 1 hr. at

150oC. Tensile properties were investigated

by Instron universal testing machine follow

ASTM D412 at 25°C, using tension load at 2

kN and crosshead speed at 500 mm/min. Durometer hardness tester (shore A) and

Zwick were used for hardness and rebound.

For crosslink density V, sample were

swollen in toluene solvent for 3 days at room

temperature. crosslink density was calculate by

the Flory-Rehner Equation.

𝑉 =− ln(1 − 𝑉2𝑚) + 𝑉2𝑚 + 𝜒𝑉2

2𝑚

𝑉1(𝑉132𝑚 −

𝑉2𝑚

2)

with the volume fraction of rubber in the

swollen network (V2m), the solvent molar

volume (V1) and Flory-Huggins interaction

parameter for toluene and rubber (χ).2


3.1 Mooney viscosity and Rheometric

characteristics

The effect of time on Mooney

viscosity in was show in Fig.1. The results

showed that increasing of Mooney viscosity

sorted in descending order as follows:

Sample 3 > Sample 2 > Sample 1. The rise of

Mooney viscosity caused by filler-filler

interaction. If filler-filler interactions were

increase, shearing force also increased to

break filler networks. Sample 3 with contains

silica as the main filler had the highest

Mooney viscosity because silica had amount

of hydroxyl groups on the surface, which

result in strong filler–filler interactions. IE26


However, sample 1 and 2 with contained

carbon black slightly change in Mooney

viscosity because carbon black had also some

polar functional groups on the surface, but

the quantity is very small. Thus, the filler–

filler interaction from silica are higher than

carbon black3.

Figure 1. Effect of storage time on Mooney

viscosity (0-30 days).

The results from MDR (scorch time,

cure time, delta torque) were shown in Table

2. Scorch time from compound were

decreased, slightly dropped in sample 1.

Compounds contains sulfonamide groups as

accelerator, lead to fast-delayed in

vulcanization. After storage time, compound

contact with moisture in atmosphere.

Humidity and moisture effect to cure

behavior by hydrolysis of the accelerator led

to inhibit delayed action4.

For cure time, sample 1 was increased

while sample 2 & 3 were decrease. Increasing

in cure time was the effect of accelerator in

sulfonamide groups. Hydrolysis reaction

inhibited scorch-delay action and produced

2-mercaptobenzothiazole (MBT), which

would increase the cure rate4. In silica

content, silanol groups on surface of silica

were acidic that absorbed vulcanization

accelerators, thus retarding the cure time5. In

case of delta torque, the results shown that

delta torque decreased with in long time

storage. Delta torque presented the quantity

of molecule that could occur crosslinked

reaction. Thus, increase of crosslinked

reaction was reduced delta torque.

3.2 Tensile testing

Tensile strength and elongation at

break were shown in Fig. 2 and 3

respectively. Tensile strength of sample 1 and

3 were decreased, while sample 2 was

increased. Sample 2 use SBR rubber which

has higher Payne effect than BR. Payne effect

was occurred in filler-filler interaction.

Sample 1 and 3 tended to reduce with

increasing BR content6. This formulation led

to the increasing of tensile strength. Some

part of decreasing in tensile strength and

elongation at break caused by over cured

compound. Compound with over curing

occurred some small grains in dumbbell

samples. When samples were tested by

Instron machine, small defect obstructed

stretching in sample.

3.3 Rebound and hardness

In Fig. 4 and 5, data showed the

results of rebound and hardness. All of

sample were insignificantly changed. It

indicated that long time storage of compound

was not affect in vulcanized rubber.

Figure 2. Effect of storage time on Tensile

strength (0-30 days).

35

37

39

41

43

45

47

49

51

53

55

0 6 12 18 24 30

Mo

on

ey v

isco

sity

(M

U)

Storage time (days)

Sample 1 Sample 2 Sample 3

10

12

14

16

18

20

22

0 6 12 18 24 30

Ten

sile

str

eagth

(M

Pa)

Storage time (days)Sample 1 Sample 2 Sample 3

IE27


Table 2. The results from Moving die rheometer (MDR). Storage time (days)

3 12 21 30

Scorch time

Sample 1 5.96 6.18 5.87 5.87

Sample 2 3.76 3.66 3.57 3.48

Sample 3 6.67 5.67 5.38 4.98

Cure time

Sample 1 14.47 14.68 14.58 15.18

Sample 2 8.76 8.76 8.37 8.18

Sample 3 13.68 11.98 11.38 10.88

Delta Torque

Sample 1 12.83 12.76 12.67 12.47

Sample 2 12.12 11.88 11.86 11.59

Sample 3 15.53 14.92 14.75 13.79

Figure 3. Effect of storage time on

Elongation at break (0-30 days).

Figure 4. Effect of storage time on hardness

(0-30 days).

Figure 5. Effect of storage time on hardness

(0-30 days).

3.4 Heating compound

The results were shown in Table 3.

All compounds obtained the higher of

Mooney viscosity at 25oC and 50oC. At the

same storage time, the higher temperature

had the higher of Mooney viscosity because

moisture at the surface of compound was

removed, led the compounds stiffer than

storage at 25oC. Cure time was not changed

significantly with temperature and time,

while scorch times after storage in high

temperature were decreased. Although

compounds did not contact the moisture, but

some accelerators were destroyed by heat. At

50oC, tensile properties were decreased from

over cured compound. Rebound properties

slightly reduced by time, not effected by

temperature. The crosslink densities were

slightly changed by time and temperature,

indicating that some parts of vulcanized

rubber were degraded while cure in

compression mold.

400

450

500

550

600

650

700

0 6 12 18 24 30

Elo

ngat

ion

at

bre

ak (

%)

Storage time (days)

Sample 1 Sample 2 Sample 3

15

20

25

30

35

0 6 12 18 24 30

Reb

ou

nd

Storage time (days)Sample 1 Sample 2 Sample 3

40

45

50

55

60

65

70

75

0 6 12 18 24 30

Har

dn

ess

Storage time (days)Sample 1 Sample 2

IE28


Table 3. The results of characterization.

4. Conclusion The effect of storage time and

temperature on properties of compound were

investigated. After storage time of pre-

vulcanized rubber, compound with silica

content had filler-filler interactions stronger

than carbon black. Accelerator with

sulfonamide groups reacted with moisture

content and produced MBT to increase the

cure rate. The surface of silica had silanol

groups that absorbed vulcanization

accelerators lead to retarding the cure time.

For vulcanized rubber, compounds were over

curing, occurred post curing effect and

created small grains samples led to reducing

in tensile strength and elongation at break.

Acknowledgements

This research was supported by

Department of Chemical Technology,

Faculty of Science, Chulalongkorn

University, Bangkok.

References

1. Yang, H.H.; Manas-Zloczower, Mixing

(Second Edition). 2009, 267, 269-297.

2. Leandro, H. E.; Angel, J. M. Radiat.

Phys. Chem. 2020, 108651

3. Sung, S.C.; Changwoon, N.; Seung, G.L.;

Chang W.J. Polym. Int. 2003, 52, 23-28

4. Butler J. and Freakley P. K.. Rubber

Chem. Technol. 1992, 65, No. 2, 374-384

5. Sattayanurak, S.; Jacques, N.; Sahakaro,

K.; Wisut, K.; Dierkes, W. K.; Blume, A.

Adv. Mater. Sci. Eng.. 2019, 1-8.

6. Ramesan, M. T. J. Polym. Res. 2005, 11,

333-340.

Figure 6. Small grains in dumbbell samples.

Characterization

Temperature

Storage

time

(days)

Mooney

viscosity

Scorch

time

Cured

time

Tensile

strength Elongation Rebound

Crosslink

density

Sample

1

25oC

0 43.5 6.0 13.3 17.5 574 24.4 2.0E-04

3 43.6 6.0 13.6 16.5 570 24.4 2.1E-04

6 44.0 6.2 14.0 15.8 561 24.2 2.1E-04

50oC 3 45.0 5.7 13.9 17.3 567 24.4 1.8E-04

6 45.4 5.6 13.8 16.9 553 24.4 2.2E-04

Sample

2

25oC

0 43.8 4.8 9.1 16.6 524 27.0 1.4E-04

3 43.2 4.6 9.2 16.9 532 26.4 1.4E-04

6 43.6 4.6 9.5 17.2 559 26.0 1.5E-04

50oC 3 46.2 4.1 9.2 18.5 625 26.2 1.4E-04

6 47.2 4.0 9.3 16.9 573 25.8 1.4E-04

Sample

3

25oC

0 42.6 7.5 15.5 16.7 531 29.6 1.6E-04

3 44.2 7.2 14.9 16.2 510 29.2 1.7E-04

6 46.3 6.8 13.7 15.7 509 29.0 1.7E-04

50oC 3 47.6 6.9 14.7 16.1 523 29.2 1.7E-04

6 50.0 6.1 13.3 14.2 472 29.0 1.7E-04

IE29


Effect of Silica Filler on Sidewall Properties Thosapon Khumngern1, Pongsak Nimdum2, Napida Hinchiranan1, 3*, Chaiyan Chaiya4

1Department of Chemical Technology, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand 2Department of Mechanical & Aerospace Engineering, Faculty of Engineering, King Mongkut's University of

Technology North Bangkok, Bangkok 10800, Thailand 3Center of Excellence on Petrochemicals and Materials Technology (PETROMAT), Chulalongkorn University,

Bangkok 10330, Thailand 4Department of chemical engineering, faculty of engineering, Rajamangala University of technology

Thanyaburi, Pathum Thani 12110, Thailand

*E-mail: [email protected], [email protected]

In the tire industry, there are competitions in both of price and tire quality. Most of the

produced tire components are sidewall. To reduce the production cost, the aim of this research

is the use of silica (Si) to replace the carbon black (CB) partially and produce the sidewall

compound since Si was cheaper than CB. At the constant filler loading of 51 parts per hundred

of rubber (phr), the sidewall compound mixed with Si- filler at 4.90%, 9.80%, and 14.71%

without the assistance of any coupling agents. The properties of the sidewall in terms of

Mooney viscosity, cure characteristics, tensile properties, crosslink density, thermal aging

properties, rebound, and hardness of the vulcanizates were investigated. The results showed

that Mooney viscosity of the sidewall compound mixed with small amount of Si was not

significantly changed. However, the increased amount of Si-filler to 14.71% increased the

scorch time, cure time and elongation at break of the vulcanizates. Moreover, the addition of

Si decreased the hardness, rebound and tensile strength of the vulcanizates compared to the one

which contained only CB. This might be attributed to crosslink density that declined by the

interaction between silanol groups on the Si-filler surface with accelerators.

1. Introduction

For quality and price competition of

tire manufacturing industry, the sidewall is

an important component of automobile tires. The previous research has been studied to

improve the properties of compounded

rubber in many ways.1, 2 One of them is the

use of silica (Si) filler with hydroxyl group

on their surface to improve reinforcement

and mechanical properties. However, carbon

black (CB) filler which is able to improve

tensile strength, tear strength and abrasion

resistance is commonly used in the tire

industry. The combination of carbon black

and silica filler with appropriate ratio can

improve the properties and reduce the

production cost of the sidewall compound.

In this research, Si is a cheaper filler.

It was blended with CB in three different

formulations. The effect of Si on the

properties of sidewall compound was

studied in terms of Mooney viscosity, cure

characteristics, tensile properties, crosslink

density, thermal aging properties, rebound,

and hardness.


2.1 Materials

Rubber and all mixing ingredients

were obtained from tire industry. Mooney

viscosity of natural rubber (ML1+2.5 at 100 ◦C)

was 80.2. Carbon black from tire industry

(N330, iodine absorption = 43 mg/g and pour

density = 360 g/dm3) was used. Silica

(Newsil®125 GR) with BET surface area of

125 m2/g was used in this research. 2.2 Mixing and processing

The formulation of sidewall

compound was given in Table 1. The filler

loading kept constant at 51 phr and the

silica filler was varied at 4.90%, 9.80%, and

14.71%. All of sidewall compounds were

blended by using a two-roll mill at 70 °C.

NR and BR were masticated for 3 min.

Then, CB, Si, processing aid, plasticizer,

antioxidant, stearic acid, zinc oxide, IE30


accelerators and sulfur were added. The

total mixing time was 41 min. After mixing,

the sidewall compounds were kept in

storage room at 25 ◦C overnight before

testing.

Table 1. Sidewall compound formulation

(phr).

Materials Content (phr)

SW0.0S SW2.5S SW5.0S SW7.5S

NR 35 35 35 35

BR 65 65 65 65

Carbon

black N330

51 48.5 46 43.5

Silica 0 2.5 5 7.5

Processing

aid

18.42 18.42 18.42 18.42

Plasticizer 1 1 1 1

Antioxidant 5.25 5.25 5.25 5.25

Stearic acid 1 1 1 1

ZnO 2 2 2 2

Accelerator 0.6 0.6 0.6 0.6

Sulfur 1.92 1.92 1.92 1.92

2.3 Characterizations sidewall

compound

Mooney viscosity (ML1+2.5 at 100 ◦C) was investigated by using Mooney

viscometer, MV-2000. Cure characteristics

was evaluated by using moving die

rheomerter (MDR), MDR-2000 at 150 °C.

The crosslink density was measured by

immersing the sample (10 mm × 10 mm × 2

mm) in toluene for 72 hr. Then, the crosslink

density was calculated following Flory–

Rehner equation1. The hardness of the

vulcanizates was analyzed by using

durometer hardness tester, Shore A

according to ASTM D2240. Rebound was

investigated by using Zwick according to

Din5312. The tensile properties of sidewall

were obtained from Universal Testing

Machine following ASTM D412 with a

crosshead speed of 500 mm/min a tension

loading of 2 kN. The carbon dispersion was

also evaluated by using DisperGraderTM

reflect light microscope following ASTM

D7723. The thermal aging was tested by

heating the samples in the form of dumbbell

shape at 100 °C for 2 days. After taking the

samples out of the oven and leaving it

overnight, the tensile properties of the

specimens were tested using Universal

Testing Machine. The retention of the tensile

properties was calculated from Eq. 1.2

Retention (%) = Value after aging

Value before aging× 100 (1)

3. Results & Discussions

3.1 Mooney viscosity and cure

characteristics

Table 2 shows Mooney viscosity

and cure characteristics of sidewall

compounds. The Mooney viscosity of the

sidewall compound did not significantly

change when the amount of Si increased

due to the small amount of Si which

replaced CB. Although Si had hydroxyl

groups on the surface and it was expected

that it could form strong filler-filler

interactions and aggregate tightly, the small

amount of Si filler could not provide the

interaction filler and rubber3.

Table 2 shows the rheological

behavior. Scorch time and cure time rose up

with the increase in the amount of Si filler. This might be attributed to the effect of

silanol group on the Si surface, which could

adsorb the accelerators via the formation of

strong hydrogen bonds with amine group of

accelerator resulting in the reduction of

vulcanization rate.4, 5

Table 2. Mooney viscosity and cure

characteristic of sidewall compounds Properties Sample code

SW0.0

S

SW2.5

S

SW5.0

S

SW7.5

S

Mooney

viscosity

26.9 25.5 26.2 26.3

Scorch

time (min)

5.07 6.14 6.57 7.24

Cure time

(min)

10.9 13.0 14.0 16.1

Delta

torque

(dNm)

8.05 7.94 7.15 6.21

For the torque difference (Delta

torque), it is related to the crosslink density IE31


of the sidewall compound. It was observed

that the value of torque difference

decreased with increasing the amount of Si.

This indicated that the use of high amount

of Si filler without the addition of coupling

agent resulted poor contribution to the

formation of the filler-rubber bonds and

provided the low crosslink density.4, 5

3.2 Crosslink density

Figure 1 shows the crosslink density

of the sidewall compound. The crosslink

density of the sidewall compound

decreased when the amount of Si filler

increased because the silanol group on the

silica filler surface adsorbed some

accelerator which caused a low crosslink

density.4, 5, 8

Figure 1. Crosslink density of sidewall

compound.

3.3 Tensile Test The tensile properties of the

sidewall are showed in Figure 2. It was

observed that the maximum tensile strength

(14.51 MPa) and 300% modulus (4.81

MPa) of the sidewall compound was

obtained for the samples containing only

CB. However, these properties slightly

decreased with increasing the Si content in

the sidewall compound due to the lower

rubber-filler interaction results.4-6

Figure 3 shows the effect of Si

content on the elongation at break of the

sidewall compounds. It exhibited that the

higher of Si filler content increased the

value of elongation at break. This behavior

also confirmed that the reduction of the

crosslink density of the sidewall compound

was similar as tensile strength and 300%

modulus.3-6

Figure 2. Tensile strength and 300 %

modulus of sidewall.

Figure 3. Elongation at break of sidewall.

3.4 Hardness and rebound

The hardness of the sidewall

compounds is showed in Figure 4. This

value decreased when the amount of Si

filler increased resulting that the decrease in

the crosslink density of the sidewall

compound, which was caused by the

interaction between the silanol group of Si

filler and the accelerators.6-8

The rebound properties of the

sidewall was also showed in Figure 4. It can

be seen that the rebound value decreased

with increasing the amount of Si filler. The

lower rebound value indicated that the

higher amount of energy transformed into

heat. Thus, it was indicated that the more

amount of Si filler, the more energy

1.75

1.611.54

1.35

1.20

1.30

1.40

1.50

1.60

1.70

1.80

1.90


Cro

sslin

k d

ensi

ty (

×10

-4m

ol/

cm3)

14.51 14.4613.87

12.66

4.834.47

4.30

3.443.00

3.50

4.00

4.50

5.00

5.50

6.00

6.00

8.00

10.00

12.00

14.00

16.00


30

0%

mo

du

lus

(MP

a)

Ten

sile

str

engt

h (

MP

a)

Tensile strength 300% modulus

664673

693

717

640

660

680

700

720

740


Elo

nga

tio

n a

t b

reak

(%)

IE32


transformed into heat, which was not

appropriate for the sidewall compound.4

Figure 4. Hardness and rebound of

sidewall.

3.5 Filler dispersion

The dispersion of filler could be

examined by the size of filler aggregates in

the compounds. Figure 5 and 6 indicate that

the increase of Si content did not

significantly affect to the degree of filler

dispersion. The added filler into the

sidewall compounds showed high

dispersion degree with higher than 95%

which was considered the desired

distribution as shown in Figure 5, 6.7

Figure 5. Filler dispersion image of the

sidewall (1 div = 100 µm): a.) SW0.0S, b.)

SW2.5S, c.) SW5.0S, and d.) SW7.5S.

Figure 6. Filler dispersion properties of

sidewall.

3.6 Tensile properties and thermal aging

properties

From figure 7, 8, 9 show the value

before and after thermal aging of tensile

strength, 300% modulus, and elongation at

break of the sidewall. After two days of

thermal aging, tensile strength and

elongation at break of the sidewall were

decreasing because of the degradation of

vulcanizates by the oxidation that

encourages the polymer chain scission.

Another reason was due to the stiffening of

vulcanizates resulting in stress could not

transfer from rubber network to filler and

led to the decrease of tensile stress and

elongation at break.9, 10 Figure 8 shows

300% modulus tends to increase after

thermal aging for 2 Days. However, this

behavior can be seen in a similar result of

previous research which is explained by the

stiffening that results in a higher 300%

modulus after thermal aging.11, 12

Figure 7, 8, 9 and Table 3, the

retention properties of tensile strength, and

300% modulus increased when the amount

of Si increased. The reason would be the

post-curing effect which related to the

proportion of Si loads.12 Another reason

that could be explained was the dilution

effect13 which improved the tensile

properties. However, the retention

properties of elongation at break also

increased because of thermo-oxidative

degradation and post vulcanization.14

54.653.9

53.252.3

46.8

44.343.7 42.1

40.0

42.0

44.0

46.0

48.0

50.0

46.0

48.0

50.0

52.0

54.0

56.0

SW0.0S SW2.5S SW5.0S SW7.5SH

ard

nes

s (S

ho

re A

)

Reb

ou

nd

Rebound Hardness (Shore A)

98.58 98.47 98.44 98.56

90.00

92.00

94.00

96.00

98.00

100.00


Dis

per

sio

n (

%)

b.)

c.) d.)

a.) b.)

c.) d.)

IE33


Figure 7. Effect of Si on tensile strength of

the sidewall after thermal aging for 2 days.

Figure 8. Effect of Si on 300% modulus of

the sidewall after thermal aging for 2 days.

Figure 9. Effect of Si on elongation at

break of the sidewall after thermal aging for

2 days.

Table 3. The retention properties of

sidewall. Retention

(%)

Sample code


Tensile

strength

(MPa)

86.94 87.67 91.57 99.04

300%

modulus

(MPa)

155.7 156.3 159.8 186.6

Elongation

at break

(%)

74.56 76.53 79.21 80.39

4. Conclusion

The effect of silica filler on sidewall

properties by using silica as a replacement

for carbon black had been investigated in

this research. The higher amount of silica

did not change Mooney viscosity of the

sidewall compound. On another hand, it

resulted in increased scorch time, and cure

time due to the silanol group on the silica

filler surface adsorbed some accelerator in

the sidewall compound. The decrease of

crosslink density with increasing of the

amount of silica filler was also caused by

the adsorption of the vulcanization

accelerator. This behavior of silica was not

only caused the decreasing of crosslink

density but also caused the decreasing of

tensile strength, 300% modulus, hardness,

and rebound. The elongation at break

increased as increased amount of Si due to

low crosslink density. However,

replacement of carbon black with silica

could also increase the retention after

thermal aging on the sidewall. Therefore,

when silica is used to replace carbon black

for the tire industry in order to improve the

sidewall quality, the appropriate amount of

silica should be used.

Acknowledgements

This research was supported by

Department of Chemical Technology,

Faculty of Science, Chulalongkorn

University, Bangkok.

0

20

40

60

80

100

120

0.0

5.0

10.0

15.0

20.0


Ret

enti

on

(%

)

Ten

sile

Str

engt

h (

MP

a)

Before Age After Age Retension (%)

140

150

160

170

180

190

0.0

2.0

4.0

6.0

8.0


Ret

enti

on

(%

)

30

0%

mo

du

lus

(MP

a)


0

20

40

60

80

100

0

200

400

600

800


Ret

enti

on

(%

)

Elo

nga

tio

n a

t b

reak

(%

)


IE34


References 1. Xin, X.; Jingyi, W.; Hongbing, J.;

Lifeng, D.; Xiu, D.; Xian, F. Polym.

Compos. 2014, 35, 1466-1472.

2. Norjulia, A. M.; Hanafi, I.; Nadras, O.

J. Polym. Mater. 2016, 33, 233-243.

3. Ika, M. U.; Riastuti, F.; Sri, R.; Diah, A.

F.; Dita, A. S.; Dody, A. W.; Lies, A.

W. Procedia Chem.. 2015, 16, 258-264.

4. Viviane, D. V.; Taline, M. R.; Janaina,

S. C.; Larissa, N. C. J. Appl. Polym. Sci.

2017, 134(39), 45334.

5. Sanjay, B.; Vivenkanda, L.; Saikat, D.;

Rabindra, M.; Abhilash, G.; Preetom,

S.; Tuhin, S.; Anil, K. B. J. Appl.

Polym. Sci.. 2019, 136(18), 47560.

6. Ramin, Z.; Saeed, T. G.; Mir, H. R. G.;

Mehran, D. E-J. Chem. 2012, 9(3),

1102-1112.

7. Teku, Z. Z.; Ahmad, K. C. A.; Mazlina,

M. K. Malaysian J. Anal. Sci. 2014, 18,

604-611.

8. Fei, Z.; Weina, B.; Shugao, Z. J.

Macromol. Sci. B: Phys. 2011, 50,

1460-1469.

9. Khalil, A.; Shaikh, S. N.; Nudrat, Z. R.;

Khalid, M. Int. J. Indust. Chem. 2012,

3, 21.

10. Zhong, X. O.; Hanafi, I.; Azhar, A. B.

J. Appl. Polym. Sci. 2013, 130(6),

39649.

11. Arayapranee, W.; Rempel, G. L. Int. J.

Mater. Struct. Reliability. 2007, 5, 1-12.

12. Pongdhorn, S.; Chakrit, S.; Thanandon,

W.; Kannika, H. Journal of Apply

Polymer Science. 2007, 104, 3478-

3483.

13. Sarawut, P.; Nattaya, R. Polym. Test.

2011, 30, 515-526.

14. Khalil, A.; Shaikh, S. N.; Nudrat, Z. R.;

Farzana, H. J. King Saud University-

Science. 2013, 25, 331-339.

IE35


Effects of Lampang pottery stone, lampang clay, feldspar and quartz on

the physical and mechanical properties of artificial porcelain tableware Soravich Mulinta1, 2*, Sakdiphon Thiansem1, Worapong Thiemsorn1, Apinon Nantiya1

1Department of Industrial Chemistry, Faculty of Science, Chiang Mai University, Chiang Mai, 50200, Thailand 2Department of Ceramics Technology, Faculty of Industrial Technology, Lampang Rajabhat University,

Lampang 52100 Thailand


The aim of this research was to study and characterize the properties of raw materials

for artificial porcelain in Lampang province, Thailand. The raw materials used in the study

were from local sources comprising Lampang pottery stone, Lampang clay, potassium feldspar

and quartz. The characterization of the raw materials was analyzed by a particle analyzer, X-

ray fluorescence (XRF) and X-ray diffraction (XRD). The mixing ratio of the artificial

porcelain was an addition to the Lampang pottery stone of 50 – 100%, Lampang clay of 5 –

30%, potassium feldspar of 5 – 30%, and quartz of 10 – 50% on the physical-mechanical

properties of the artificial porcelain. The properties of the artificial porcelain body after firing

at a temperature of 1,200°C in an oxidation atmosphere were studied, and the shrinkage, water

absorption, bulk density, appearance porosity, crazing and bending strength of the artificial

porcelain body were also tested. The results showed that the Formula 7 components of 70% of

Lampang pottery stone, 10% of Lampang clay, 10% of potassium feldspar and 10% of quartz

had optimum properties. Formula 7 had a shrinkage of 12.3%, water absorption of 0.26%, bulk

density of 1.42 g/cm3, appearance porosity of 6.0 %, non–crazing and resistance to the bending

strength at 680 kg/cm2. The artificial porcelain tableware fulfilled the requirements of the Thai

industrial standard (TIS 564-2546).

1. Introduction

Lampang province, Thailand is a noted

ceramic source of the country, which has

large, medium and small-sized entrepreneurs,

including small and medium-sized

enterprises (SMEs). However, most

entrepreneurs lack basic knowledge and the

techniques for applying innovation.

Lampang’s ceramic products are mainly

made of earthenware and stoneware porcelain

clay that have particular characteristics. They

are in general very siliceous, aluminous and

contain significant proportions of potassium,

which acts as flux. These clays have the

property of gradually vitrifying without

becoming deformed with a rise in temperature,

and they yield an opaque material, often brown

in color, though this varies according to the

impurities contained in the initial mixture.1

From the collection of the data from the Ban

Nam Cho Factory, there was a production

process of semi-porcelain, mainly chicken

bowls, that was pottery painted by hand with a

chickens, botan flowers and banana trees

pattern. Their colors were white with

transparent or turbid glaze, or creamy with

transparent glaze with an estimated production

of 100,000 units/month. Stoneware products

were normally fired at a temperature of around

1,200°C in a shuttle kiln using LPG as the fuel

and were fired for eight hours. The quality of

the semi-porcelain production is shown in

Table 1.

Table 1. The properties of the semi-porcelain

products in the Ban Nam Cho Factory.

Properties Products

Water absorption (%) 6.2

Bending strength (kg/cm2) 420

The mechanical strength of the semi-

porcelain product was resistant from bending at IE36


420 kg/cm2. Thus, the mechanical property of

the porcelainware was not exactly within the

requirements of the Thai community’s product

standard (Table 2).

Table 2. The properties of porcelainware as per

the requirements of the Thai industrial

standard.2

Properties (TIS 564-2546). Quality

Water absorption (%) 0.2

Bending strength (kg/cm2) 650

The objective of this research was to

study and characterize the properties of the

Lampang pottery stone, Lampang clay,

feldspar and quartz for artificial porcelain

tableware in Lampang province, Thailand.

Artificial porcelain is a type of ceramic material

in pottery usually accepted as a type of

porcelain, It is weaker than true porcelain and

does not require high firing temperatures or

pottery stone, Lampang clay and mineral

ingredients.


2.1 Raw materials for standard clay

The raw materials used in the study

came from local sources in Lampang

province, Thailand. They were Lampang

clay, pottery stone from Patra Ratana Clays

and Minerals (1992) Company Limited, and

potassium feldspar and quartz from Sibelco

Minerals (Thailand) Co., Ltd. The raw

materials were controlled by grinding and

passed through a 325-mesh sieve (particles

smaller than 0.42 m).

2.2 Characterization analysis of the raw

materials.

The particle size of the raw materials was

analyzed by laser diffraction (Mastersizer S,

Malvern, USA), and the chemical composition

of the raw materials was analyzed by X-ray

fluorescence (XRF, Megix Pro MUA/ USEP

T84005, Philips, USA). The mineralogical

composition of the raw materials was identified

by X-ray diffraction (XRD, D500, Siemens,

Germany).

2.3 Preparation of the specimens.

The component ratios of the porcelain

clay used in this study were Lampang pottery

stone of 60 – 90%, Lampang clay of 5 – 20%,

potassium feldspar of 5 – 25% and quartz of

10–40% (Table 3). All compositions were

wet mixed and milled by a high-speed mill

for 16 hours and passed through a 325-mesh

sieve. The specimens were formed by a

casting method and dried at 110°C then fired

at a temperature of 1,200°C in a shutter kiln

with LPG fuel.

2.4 Properties measurement of the

porcelainware

The linear firing shrinkage was

assessed as the difference between the diameter

of the dried and fired specimens divided by the

diameter of the dried specimen. The wet–to–

fired shrinkage value was quoted on the basis

of the wet size; wet–to–fired shrinkage is given

by:2

Shrinkage = wet length – fired length

Wet length × 100

The water absorption was calculated

by premeasuring the weight of the test

specimens after placing them in boiling water

at 110°C for five hours. The samples were

cooled in desiccators and weighed to an

accuracy of 0.01g to give the dried weight

(D). Then, the moisture was removed from

the surface of the samples with a moist cloth,

and the samples’ weight was measured to

give the soaked weight (S). The value of (D)

and (S) are used in the formula2:

Water absorption = S - D

D × 100

IE37


Table 3. Composition of the clay mixture in this study.

Raw Material Formula

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Pottery stone 90 90 80 80 80 70 70 70 70 60 60 60 60 60

Lampang clay 5 0 10 5 0 15 10 5 0 20 15 10 5 0

Potassium feldspar 5 0 10 5 0 15 10 5 0 20 15 10 5 0

Quartz 0 10 0 10 20 0 10 10 30 0 10 20 30 40

The bending strength for the clay was

determined by a universal testing machine

(Jinan testing, UTM Instrument model;

WDW-20) represented by the modulus of the

rupture equation:

MOR = 3LD

2bd2

where L is the breaking load.

D is the distance between the supports.

b is the breadth of the rectangular rod.

and d is the depth of the rod 3.

A three-point loading system was

employed, and the supper surface of the

specimen was in compression while the

lower surface was in tension (Figure 1).

Figure 1. Three-point loading for the MOR

test.


3.1 Characterization of the raw material The particle size of the raw materials

was analyzed by laser diffraction of X-ray

fluorescence (Table 4). It was found that

Lampang clay had an average particle size of

10.59 m whereas pottery stone had an

average particle size of 22.41 m. Potassium

feldspar had an average particle size of 9.21

m, and quartz had an average particle size of

12.24 m.

Table 4. Particle size of the raw materials. Raw Material Particle Size (m)

Pottery stone 22.41

Lampang clay 10.59

Potassium feldspar 9.21

Quartz 12.24

The chemical composition of the raw

materials was analyzed by X-ray

fluorescence (Table 5). The major

components of Lampang clay included

67.47% of silicon oxide (SiO2), 20.092% of

aluminum oxide (Al2O3), and up to 1.81wt%

of ferric oxide (Fe2O3). The amount of SiO2

was much lower than Al2O3 and Fe2O3,

respectively. CaO, K2O, BaO MnO, ZrO2,

TiO2, P2O5, MgO, Y2O3 and Rb2O were also

found in small quantities. The major

components of the pottery stone included

66.47wt% of silicon oxide, 11.52% of

aluminum oxide, and 13.26% of ferric oxide,

which would mostly affect the color of the

fired product (Fe2O3, MnO, and TiO2). Other

elements (MgO, K2O, and Na2O) would act

as fluxes and may have a strong effect during

sintering.4 Another important component of

the studied pottery stone was the total

amounts of the alkali and alkaline earth

oxides (Na2O, K2O, MgO, and CaO) that

acted as flux materials and were slightly high

in the pottery stone5. The major components

of the Lampang clay included 66.9% of

silicon oxide (SiO2), 18.14% of aluminum

oxide (Al2O3), and up to 11.17wt% of

potassium oxide (K2O). The major

components of quartz included 97.2wt%.

specimen

IE38


Table 5. Chemical composition of the raw materials.

Raw Material Composition (wt%)

SiO2 Al2O3 Fe2O3 K2O MgO MnO Na2O P2O5 TiO2 CaO LOI

Pottery stone 60.30 19.52 1.26 3.64 1.14 <0.01 0.20 0.18 0.04 0.05 5.96

Lampang clay 67.47 20.99 1.81 1.72 0.29 <0.01 0.12 0.04 1.23 0.27 6.77

Potassium feldspar 66.9 18.14 N/D 11.17 N/D N/D 2.96 N/D 0.2 N/D 4.17

Quartz 99.16 0.45 N/D 0.03 0.004 N/D 0.06 0.01 N/D 0.01 0.276

N/D: non detector

0 10 20 30 40 50 60 700

100

200

300

400

500

600

700

Figure 2. X-ray diffraction patterns of the raw materials: Lampang clay (A) and pottery stone (B).

The mineralogical composition of the

standard clay was analyzed by X-ray

diffraction (XRD). The results of the

mineralogical analysis of the raw materials in

Figure 2(A), show the XRD pattern of the

Lampang clay. The predominant peaks are

associated with quartz (Q) while the

secondary peaks identify the crystalline

mineral phase of kaolinite (K)

(Al2Si2O5(OH)4) and muscovite (M) (K2Al4

(Si6 Al2) O20 (OH,F)4). Figure 2(B) shows the

XRD pattern of the pottery stone. The

predominant peaks are associated with quartz

(Q) while the secondary peaks identify the

crystalline mineral phase of muscovite (M)

(K2Al4 (Si6 Al2) O20 (OH,F)4.

3.2 Properties of the porcelain body.

From the linear shrinkage of the fired

porcelain body, the shrinkage of the body at

100-180°C was due to hygroscopic water that

was associated with weight loss. When the

feldspar content was increased and over-fired

at temperatures of 900-1,100°C, this resulted in

the formation of mullite (3Al2O32SiO2) and

another phase.6 The primary mullite forms

originated from the decomposition of the clay

content whereas the secondary mullite forms

resulted from the reaction of the alkaline (CaO,

K2O or Na2O) and pottery stone or Lampang

clay7. The results showed that after firing, the

clay body had low shrinkage of 8.9% (Formula

1 in Figure 3).

Figure 3. Fired shrinkage of porcelain body.

For the water absorption of the

porcelain body, this depended on whether the

pottery stone had a potassium oxide (K2O)

1 2 3 4 5 6 7 8 9 10 11 12 13 140

2

4

6

8

10

12

14

16

18

20

Fir

ed

Sh

rin

ka

ge (

%)

Formula

Inte

nsi

t

y

2-theta (degree)

A

Q

K

M Q Q K

K+M K Q

2-theta (degree) 10 20 30 40 50

Q Q

Q

Q M M

200

400

B

IE39


content that had an effect on the higher

density and glassy phase than the Lampang

clay content. The appropriate glassy phase

and pore change size results of the feldspar

content could be obtained under firing

temperatures of 1,200˚C. The results showed

that after firing, the body had low water

absorption of 0.26% (Formula 7 in Figure 4).

The porcelain body was fired at 1,200°C

to obtain the appearance porosity. The results

showed that after firing, the body had a low

appearance porosity of 0.6% (Formula 7 in

Figure 5. The bulk density was optimum with

the addition of 70% of pottery stone. Formula

7 had a bulk density of 1.42 g/cm3; the result

is shown in Figure 6.

The bending strength of the fired

porcelain body showed that before firing, the

firing strength was associated with the

development of strong crystals; such as,

secondary mullite (3Al2O3.2SiO2) and glassy

matrix. Heating pottery stone and Lampang

clay to a high temperature in a porcelain ware

product of mullite contributed markedly to

the strength of the fired body system. For the

mechanical properties of firing in the

porcelain body, the porcelain composition

was prepared by mixing Lampang clay,

alkaline earth and silica. The analysis

revealed that the representative composition

of the stoneware was produced by the means

of the firing process on the mechanical

properties with several physical features and

phase compositions including secondary

mullite needles8. The porcelain body had a

high bending strength of 680 kg/cm2

(Formula 7 in Figure 7).

Figure 4. Water absorption of the porcelain body.

Figure 5. Appearance porosity of the

porcelain body.

Figure 6. Bulk density of the porcelain body.

1 2 3 4 5 6 7 8 9 10 11 12 13 140.0

0.5

1.0

1.5

2.0

Wate

r a

bso

rp

tion

(%

)

Formula

1 2 3 4 5 6 7 8 9 10 11 12 13 140.0

2.5

5.0

7.5

10.0

Ap

pea

ren

ce p

oro

sity

(%

)

Formula

1 2 3 4 5 6 7 8 9 10 11 12 13 140.0

0.5

1.0

1.5

2.0

2.5

3.0

Bu

lk d

en

sity

(g/c

m3)

FormulaIE40


Figure 7. Bending strength of the porcelain

body.

Figure 8. Microstructure of the porcelain

body of Formula 7.

The crystallographic analysis of the

phases present in the porcelain body in Formula

7, as shown in Figure 8, indicated the presence

of mullite and quartz. Such analysis confirmed

the complete reaction of pottery stone, feldspar

and silica in the microstructure of the porcelain

body fired at 1,200°C (10% HF acid etching

within 15 seconds). The phases formed in the

clay body were much different than those

formed in artificial porcelain, and this

explained the difference of the physical and

mechanical properties of the porcelain body.

4. Conclusion The results concluded that Formula 7 (70%

of Lampang pottery stone, 10% of Lampang

clay, 10% of potassium feldspar and 10% of

quartz) had optimum properties. The properties

of the porcelain body after firing at 1,200°C

were studied. Shrinkage of 12.3%, water

absorption of 0.26%, bulk density of 1.42

g/cm3, appearance porosity of 0.6%, non–

crazing and resistance to the bending strength

at 680 kg/cm2 were found in the specimens.

This porcelain body achieved the requirements

of the Thai industrial standard (TIS 564-2546).

Acknowledgements

The authors would like to thank Chiang

Mai University and Lampang Rajabhat

University for providing the experiment and

support for the research.

References 1. Philippe B. and N., 2007., Ceramic

Materials: Processes, Properties and

Application., ISTE Ltd, UK.

2. “Standard for Porcelain Tableware.” Thai

Industrial standard. (TIS-564-2546)

3. Ray W. and Radford C. Pergamon press,

London, 1987.

4. Jone. J.T. and Berard. M.F., Ceramics

Industrial Processing and Testing, second

ed., Ames; Iowa; 1993.

5. Thiansem, S., W, S., K, K., T, P. and P, S.,

ScienceAsia 2002, 28, 45-152.

6. Chitwaree, S., T, J., T, N. and P, L.,

Thermal Engineering, 11 (2018) 81–88.

7. Lee, W.E., I., Y. J. Eur. Ceram. Soc. 2001,

21(14), 2583-2586.

8. Martín-Márquez, J., R.M., J. and R., M. J.

Eur. Ceram. Soc. 2010, 30(15), 3063–

3069.

1 2 3 4 5 6 7 8 9 10 11 12 13 140

100

200

300

400

500

600

700

Ben

din

g S

tren

gth

(k

g/c

m2)

Formula

IE41


Comparisons of the effect of the firing temperature of red clay and sponge

waste on the physical-mechanical properties of planting materials Sukanya Pukpobsuk*, Soravich Mulinta, Thitima Khunyotying, Winai Tasang,

Kanokkanya Ruammaitree, Tamonwat Hirunchatanan Department of Production Innovation and Ceramics Design, Faculty of Industrial Technology, Lampang

Rajabhat University, Lampang 52100, Thailand.


The aim of this research is to study the characteristics and physical properties of the

raw materials used for planting materials. The experiment started with the preparation of local

red clay and sponge waste that were obtained from local sources in Lampang province,

Thailand. The characteristics of the local red clay were analyzed by particle analysis, X-ray

fluorescence (XRF) and X-ray diffraction (XRD) techniques. The physical properties of the

planting materials after being fired at 900 °C, 1000 °C, 1100 °C, and 1200 °C, respectively,

were studied and compared. The water absorption, linear shrinkage, bulk density, apparent

porosity, microstructure and bending strength of the planting materials were tested. The major

components of the chemical composition of the local red clay were found to be 67.05 wt%

silicon oxide, 12.01 wt% ferric oxide and 11.3 wt% aluminum oxide. The red clay and sponge

waste slip mixtures were prepared with three specific gravities of 1.3, 1.5 and 1.8. The results

showed that the optimum physical and mechanical properties of the planting materials were

those that, after being fired at 900 °C, consisted of the slip with specific gravity of 1.5. This

material had water absorption of 72.3%, linear shrinkage of 2.9%, bulk density of 0.85 g/cm3,

apparent porosity of 47.3%, and resistance to bending strength at 21.5 kg/cm2.

1. Introduction

A planting medium is a substance

placed into pots for the purpose of providing

material for the roots to hold and to keep the

trunk of the tree upright, rather than leaning

or falling. It is also necessary for covering the

roots and important for the growth of the root

system. The roots are responsible for

transporting water and minerals to the trunk

to allow the plant to grow and develop, bloom

and produce fruit. Planting materials made

from local clay are thus essential for the

growth of trees. The material should be

durable and not a source of accumulation of

fungi and plant diseases affecting trees, from

which the red clay can be used as the main

raw material for an earthenware body.

The term ‘earthenware body’ usually

refers to a porous clay body maturing

between 900 - 1,180 °C (1,873 °F ‐ 2,152 °F).

Water absorption generally varies between 5

- 20%, so earthenware clay is usually not

fired to vitrification. This means that pieces

with crazed glaze may seep liquids.

Terracotta applied to the base also helps

decrease absorption and reduces delayed

crazing. Moreover, low-fired fluxes melting

over a shorter range than high-fired materials

and firing an earthenware body to near

vitrification usually result in a dense, brittle

body with poor thermal shock resistance and

increased potential for warping and dunting.

Although it is possible to fire terracotta in a

gas kiln in an oxidized atmosphere, this is

often difficult to control. 1,2

In this study, the quantitative

collection of the sponge wastes was obtained

from a car seat repair shop located in Mueang

district, Lampang province, Thailand. The

sponge wastes were processed by cutting

them into the cubic size of 111 inch (refer

to Table 1).

IE42


Table 1. Amount of waste from car seat

production Materials for car seats Waste (%)

Foam 60

Leather cushion 25

Leather upholstery 22

Velvet cushion 3

Sixty percent of the wastes from the

community were collected each day in the form

of sponge wastes that can be used to enhance

the properties of planting materials, and the

planting materials containing a mixture of red

clay and sponge waste can provide the physical

and mechanical properties that achieve

compliance with the Thai Community Product

Standard (TCPS 46-2556).3

The purpose of the present study was

to characterize the physical properties of raw

materials to be used as planting materials.

The aim of this work was to study and

compare the effect of the firing temperature

of red clay and sponge waste on the bulk

density, water absorption and apparent

porosity of the potential planting materials.


2.1 Preparation of the red clay.

The raw material used in the study,

which was Sri Khum red clay, came from

local sources in Lampang province, Thailand.

The raw material was processed by grinding

and being passed through a 40-mesh sieve

(particles smaller than 0.42 mm).

2.2 Characterization analysis of the red clay.

The particle size of the raw material

was analyzed by laser diffraction

(Mastersizer S, Malvern, USA), the chemical

composition of the raw material was

analyzed by X-ray fluorescence (XRF,

Megix Pro MUA/USEP T84005, Philips,

USA), and the mineralogical composition

was identified by X-ray diffraction (XRD,

D500, Siemens, Germany).

2.3 Preparation of the specimens.

The red clay was crushed and dried at a

temperature of 110 °C for five hours in order to

remove any moisture. The feasibility of using

red clay in brick making was investigated. The

preparation of specimens by mixing the red

clay with water to have a range of specific

gravities of 1.3, 1.5 and 1.8, respectively, was

conducted. Following this, the pieces of sponge

with a size of 1 cubic inch were dipped in the

red clay slip for 24 hours. Finally, the

specimens were fired at temperatures of 900,

1000, 1100 and 1200 °C in a shutter kiln with

LPG fuel.

2.4 Properties measurement of the

planting materials.

Bulk density is the porous solid

volume and includes the volume of the solid

components, open pores and sealed pores.

The density of clay is defined as the

relationship between its mass (weight) and its

volume as applied in Formula 1:

Density = Mass

Volume (1)

The water absorption was calculated

by measuring the weight gain of the test

specimens after placing them in boiling water

at 110 °C for five hours. The samples were

cooled in desiccators and weighed to an

accuracy of 0.01 g in order to obtain the dry

weight. Then, the moisture on the samples’

surface was wiped off with a moist cloth, and

they were weighed to obtain the soaked

weight. The value of the dry weight and

soaked weight is shown in Formula 2:

soaked weight - dried weight

dried weight × 100 (2)

Apparent porosity in the test specimens for

the determination of the water absorption of the

dry weight (D) and the soaked weight (S) was

assessed by following exactly the same

procedure. However, the immersed weight of

the soaked test piece (I) also needed to be

determined. The obtained weight (I) was the

immersed weight, which was used in Formula 3: IE43


soak weight - dry weight

dry weight- immersed weight × 100 (3)

The fired bending strength of the clay

samples was determined by a universal

testing machine with a maximum capacity of

1000 kN. The bending load was applied onto

the face of the samples having the dimensions

of 110 × 110 mm2. The bending strength was

determined by dividing the maximum load by

the applied load area of the clay samples. The

fired bending strength of the clay samples

was determined by the three-point bending

test with a supporting span of 300 mm, a

height of 75 mm and a width of 105 mm. The

bending strength for the clay was determined

by the universal testing machine as represented

by the modulus of the rupture equation.4

MOR = 3LD

2bd2 (4)

where L is the breaking load,

D is the distance between the supports,

b is the breadth of the rectangular rod,

and d is the depth of the rod.

Figure 1. Three-point loading for MOR test.


3.1 Characterization of the raw material.

The particle size of the standard clay was

analyzed by laser diffraction, which indicated that

the Sri Khum red clay has an average particle size

of 24.69 m. The chemical composition of Sri

Khum red clay was analyzed by XRF as shown

in Table 2. The major components included

67.05% of silicon oxide (SiO2), 11.32% of

aluminum oxide (Al2O3) and up to 12.01 wt% of

ferric oxide (Fe2O3). The amount of SiO2 was

found to be significantly higher than that of Al2O3

and Fe2O3, respectively. CaO, K2O, MnO, TiO2,

P2O5, and MgO acted as fluxes; thus, there was a

strong effect during sintering.4 The mineralogical

composition of the Sri Khum red clay was

analyzed by XRD. The results of the

mineralogical analysis showed the XRD pattern

of the Sri Khum red clay (refer to Figure 2). The

main peaks were associated with quartz (Q).

Figure 2. X-ray diffraction of the Sri Khum

red clay.

Table 2. Chemical composition of the raw

materials.

Composition Raw material

Sri Khum red clay

SiO2 67.05

Al2O3 11.32

Fe2O3 12.01

K2O 3.640

MgO 1.144

MnO <0.01

Na2O 0.202

P2O5 0.177

TiO2 0.036

CaO 0.051

LOI+SO3 0.962

3.2 Properties of planting materials.

The shrinkage of the planting materials

fired at 900 - 1200 °C was studied. The results

showed that the test specimens underwent

escalated shrinkage when the sintering

temperature was increased. The sample was

fired at 350 °C, and carbon combustion

occurred, causing closed and open pores in the

test specimens.5 The shrinkage was optimized

by the addition of 20% of water content, and

the specific gravity 1.5 composite fired at 900

C showed a linear shrinkage of 2.9% (refer to

Figure 3).

IE44


Figure 3. Linear shrinkage of planting materials.

Figure 4. Bulk density of planting materials.

The results following the planting

materials being fired at 900 - 1200 C showed

that the bulk density of the planting material

of the specimens at temperatures 900 - 1000

C indicated that when the amount of the

specific gravity was increased, the bulk

density decreased. However, at the firing

temperature of 1000 C, when the firing

temperature was increased, the test

specimens had an decreased bulk density.

This occurred because the red clay melted

due to the presence of iron oxides of up to

12.01%. The bulk density of the planting

materials was improved by the specific

gravity 1.5 composite being fired at 900 C

with a bulk density of 0.85 g/cm3 (refer to

Figure 4).

The water absorption rates of the

planting materials fired at 900 - 1200 °C are

shown in Fig 5. The results indicate that the

specimens had lower water absorption when

the specific gravity and firing temperature were

increased. The firing temperature affects the

densification of the specimens, resulting in

reduced water absorption. The firing of test

pieces at 900 - 1000 C still resulted in open

pores and connected pores. If fired at 1100 -

1200 C, the water absorption was reduced.

The water absorption was optimum with the

specific gravity 1.5 composite having water

absorption of 72.32% (refer to Figure 5).

Figure 5. Water absorption of the planting

materials.



the apparent porosity of the specimens. The

test specimens passed the firing process when

the firing temperature increased, resulting in

the decreased apparent porosity of the

specimens. In addition, the melting point of

the red clay was also affected by firing

temperature of the planting materials. The

apparent porosity of the planting materials

was improved by the specific gravity 1.5

composite being fired at 900 C with an

apparent porosity of 47.3% (refer to Figure 6).

0

2

4

6

8

10

1 .3 1 .5 1 .8

Fir

ed S

hri

nk

ag

e (%

)

Specific gravity

900 1000 1100 1200

0

0.5

1

1.5

2

2.5

3

1 .3 1 .5 1 .8

Bu

lk d

ensi

ty (

g/c

m3)

Specific gravity

900 1000 1100 1200

0

20

40

60

80

100

1 .3 1 .5 1 .8

Wa

ter

ab

sorp

tio

n (

%)

Specific gravity

900 1000 1100 1200

IE45


Figure 6. Appearance porosity of planting

materials.



that there was a relationship between the

strength of the planting material and the

amount of the apparent porosity. When the

porosity was high, the bending strength of the

test specimen decreased. This was because

the porous material obstructed the particles of

the red clay and prevented the creation of a

void in the structure of the planting

materials.7 Figure 7 shows that the

mechanical properties were improved by the

specific gravity 1.5 composite being fired at

900 C, resulting in the resistance to the

bending at 21.5 kg/cm2.

Figure 7. Bending strength of planting

materials.

The micrograph of the fracture

surface of the specimens displays a rough

fractured surface with loose particles and

interconnected pores8. The sponge waste

appears deep black in color after being fired

at 700 - 800 ºC. According to the data of the

optical microscopic analyses, the particles of

the sponge waste are composed of gray and

brown color porous grains of irregular shape

following the firing at 900 - 1200 ºC, as

shown in Figure 8.

Figure 8. Microstructure of planting materials.

0

10

20

30

40

50

60

70

80

1 .3 1 .5 1 .8

Ap

pea

ren

ce p

oro

sity

(%

)

Specific gravity

900 1000 1100 1200

0

10

20

30

40

50

1 .3 1 .5 1 .8

Ben

din

g s

tren

gth

(k

g /

cm

2)

Specific gravity

900 1000 1100 1200

IE46


4. Conclusion

From the results, it was concluded that

the material that comprised red clay and

sponge waste had the optimum properties in

the specimen of specific gravity 1.5 that

was fired at 900 °C. The properties of the

red clay and sponge waste after firing at 900

°C were studied, which showed that the

specimens had a water absorption of 72.3%,

linear shrinkage of 2.9%, bulk density of

0.85 g/cm3, apparent porosity of 47.3%, and

resistance to the bending strength at 21.5

kg/cm2. Thus, the planting materials

produced by the community fulfilled the

requirements of the Thai Community

Product Standard (TCPS 46-2556).

Acknowledgements

The author would like to thank

Lampang Rajabhat University for its

support of this research.

References 1. A. Dinsdale. Pottery Science. Ellis

Horwood Limited, USA, 1986.

2. W. Ray and C. Radford, Whitewares:

Production, Testing and Quality

Control, Pergamon Press, London,

1987.

3. Thai Community Product Standard

(TCPS 46-2556).

4. J. T. Jones and M. F. Berard, Ceramics

Industrial Processing and Testing,

second ed., Ames, Iowa; 1993.

5. Mst. Shanjida Sultana, A. N. A., M. N.Z.

and Md. A. R., Pradip K. B., P. K.N. J

Asian Ceramic Societies 3 (2015), pp.

22 – 26.

6. Izwan Johari., R. P.J. and Zainal A.A.,

B.H. A.B., S.S., International Conf. on

Env. Sci and Eng. Singapore, 2011, pp.

141 – 174.

7. Vieira C.M.F., De S. E., S. N. M.,

Cerâmica, vol. 50, pp. 254 – 260, 2004.

8. Vieira C.M.F., M. S.N., Cons. Buil.

Mat. 2007, 21(8), 1754 – 1759.

IE47


The study of citrogypsum phase transformation process

via recycling saline water Navin Boonsuwan1, Thanakit Sirimahasal1, Siriporn Pranee2, Samitthichai Seeyangnok1*

1Department of Industrial Chemistry, Faculty of Applied Science, King Mongkut’s University of Technology

North Bangkok, Bangkok 10800, Thailand 2Department of Veterinary Technology, Faculty of Veterinary Technology, Kasetsart University, Bangkok

10900, Thailand


Nowadays, citric acid industry became an important role in food and beverage, Citric

acid is used for food preservation. Unfortunately, citric acid production caused environmental

issue such as waste water, industrial scale/dross including to gypsum waste, also known as

citrogypsym. The main component of citrogypsum is calcium sulfate dihydrate (DH) with low

physical properties and non-valuable material. Therefore, this research aims to study the phase

transformation process of the citrogypsum via hydrothermal process at 95°C for 5 hrs under

atmospheric pressure to form alpha calcium sulfate hemihydrate (α-HH) by using recycling

calcium chloride solution. The final products were separated, solid product was characterized

by FT-IR spectroscopy, Thermo gravimetric analysis (TGA), Differential scanning calorimetry

(DSC) and Scanning Electron Microscopy (SEM) and a final recycling solution was titrated and

pH measured. The results shown -OH Bending at wave numbers at 1600 cm-1 in FT-IR spectra

indicate that α-HH contain crystal water with 0.5 molecules. In addition, the TGA thermogram

of α-HH have a crystal water containing 4.5 wt% and the DSC thermograms of α-HH shows

endothermic peak at 151.2°C and exothermic peak at 183.14°C which corresponding to the

loss of 0.5 water molecules in CaSO4. Moreover, the morphology of α-HH are hexagonal

structure which confirmed by SEM images.

1. Introduction Citric acid is important in many

industries because it is a product with high

solubility and a flavoring agent that is widely

used in the pharmaceutical and food

industries.1 Today, citric acid is mainly

produced from cassava root by fermentation

with Aspergillus niger. After that, neutralize

and precipitated with calcium oxide (CaO) to

calcium citrate (Ca3(C6H8O7)2) then sulfuric

acid (H2SO4) was added to purify the citric

acid. From the citric acid producing process,

it will create waste or byproduct called

citrogypsum (CG) shown in Figure 1 that up

to 70 tons/years.2, 3 To managing these

wasted will initially be eliminated by sea

reclamation or land reclamation. Which may

cause environmental problems in many ways.4

For the above reasons, this research

aims to study the phase change of Calcium

sulfate dihydrate (CaSO4·2H2O) to α-Calcium

sulfate hemihydrate (CaSO4·0.5H2O) through

heat treatment (Hydrothermal process) as

equation 1.5 To reduce amount of solvents

waste from the CG production process, the

recycling calcium chloride solution (CaCl2)

was employed in the hydrothermal process.6

In general, gypsum is divided into 2

phases; alpha- (α-HH) and beta- (β-HH) that

have the same main components. But differs

in the structure, arrangement, and amount of

water in the crystalline. This characteristic

effect the properties of β-HH lower in

strength and brittle when compared to the

white-colored α-HH. It acceptable in the

industry that α-HH with high strength and

less water absorption will be higher cost than

β-HH.7 Therefore, this research focuses on

the recovery of citrogypsum and the recycling

of CaCl2 solution.. IE48


CaSO4·2H2O (s) Ksp DH > Ksp α-HH Ca2+

(aq) + SO42+

(aq) + 2H2O (l) (1)

95°C / under atmospheric pressure

α-CaSO4·0.5H2O (s) + 1.5H2O (l)

CaCl2 (aq) Ca2+ (aq) + 2Cl- (aq) (2)

Figure 1. The formation process of citrogypsum.

To reduce the amount of solvents in

the system and wastewater from the

hydrothermal process. With increasing the

value of the citrogypsum as well as reduce

environmental problems and can also replace

gypsum from nature as well.


2.1 Materials

Citrogypsum from the citric acid

production process was supported by the

Thai Citric Acid, Co., Ltd., Thailand, washed

with distilled water to wash out the impurities

prior used. Then bake at 120°C for 2 hrs to

reduce humidity. CaCl2 solution is solvent-

gypsum citrus. Due to the low solubility in the

citrogypsum (Ksp(CaSO4) = 2.49·10-5).8

2.2 Methods

The study of phase transformation

process of CG to α-HH in recycling CaCl2

solution at 95°C under atmospheric pressure

for 5 hrs as follows. The CG (50g) was added

into 4M CaCl2 aqueous solution in 1000 ml

Batch reactor with 600 rpm stirring rate at

95°C for 5 hrs. After that, the mixture was

filtered to separate a solid (α-HH) from liquid

(CaCl2 solution) by reduced pressure. The

filtrate was kept in a 500 ml volumetric flask

and the white solid was washed with hot

water 3 times to remove inorganic impurities

(Ca2+ and Na+) then rinse with methanol and

dry it at 60°C for 2 hrs. The α-HH was

obtained. The functional groups of products

were determined by Fourier transform

infrared spectroscopy (FT-IR, Perkin-Elmer,

Spectrum 2000). The phase morphology of

samples was analyzed by Differential scanning

calorimetry (DSC, TA Instruments Inc., DSC-

Q2000) at the scanned 10°C / min under N2

atmosphere. Thermal analysis was performed

on a simultaneous thermogravimetry (TGA,

METTLER Toledo, TGA/ DSC 1 Star System)

was used to investigate crystal water. The

morphology characteristic of samples was

observed by Scanning Electron Microscopy

(SEM, Jeol, JSM-6610LV and Oxford,

XmaxN50). The concentrations of calcium

and impurities of the reaction media

to analyze for the remaining concentration

by an Automatic Titrator (SI Analytics,

TitroLine®7000) and then the acid-base is

measured with a pH meter (Sartorius, pH

Basic Meter). The tested liquid samples

were kept in (25 ± 2) °C. Each of the test

was conducted for 3 times and the results

were shown as ‘‘mean value ± Sx/root(N)”.


3.1 Characterization of CG and α-HH

The Physical characteristics and

surface morphology of CG (shown in Figures

2A and 1B) and α-HH (shown in Figures 2C

and 2D) are characterized by the SEM

technique respectively. The CG is apparently

gray and black powder caused by the addition

of carbon black in production line of citric

acid (shown in Figure 2A) and uneven plate-IE49


liked morphology (shown in Figure 2B),

resulting in low strength property. α-HH has

(A) (B)

(C) (D)

Figure 2. (A) Physical characteristics of CG, (B) SEM image of CG, (C) physical

characteristics of α-HH (D) SEM image of α-HH.

a fine white powder with commercial

acceptable (shown in Figure 2C) and the

morphology is hexagonal (shown in Figure

2D), resulting in high strength property.9

3.2 The dehydration of DH to α-HH by

CaCl2 solution recycle.

FT-IR spectra of DH and α-HH are

shown in Figure 3, the absorption band at

3250-3800 cm-1 and 1600 and 1700 cm-1

corresponding to -OH stretching and -OH

Bending respectively, indicating that there

are 0.5 and 2 water molecule contain in

Figure 3. FT-IR spectra comparing CG and

α-HH of Control, 1st, 2nd and 3rd using

recycled calcium chloride solution.

calcium sulfate crystal structure. The

absorption band 1280 cm-1 is belonged to

S=O stretching of -SO42- group in calcium

sulfate (CaSO4) composition. After DH has

been treated with the hydrothermal process,

found that the absorption band at 1700 cm-1 is

disappeared from FT-IR spectra of the

product from Control process and recycling

saline solution process at 1st, 2nd and 3rd

cycles indicating that the water content in

CaSO4 crystal is releasing to the solution and

α-HH is formed.10

The analysis of the water content in

crystals structure of DH and α-HH are

analyzed by TGA as shown in Figure 4. It can

be seen that the water content in DH crystals

begin to dehydrate rapidly in the beginning of

hydrothermal process. Since the crystals of

DH is plate-liked morphology, the amount of

water in the crystal is easily dehydrated and

then the dehydration of crystal begins slowly

because the crystals begin to rearrange into IE50


a hexagonal shape, resulting in the water

content in the crystal stable after passing 1.30

hrs until the end of the hydrothermal process

DH completely changed to α-HH. The water

content in the crystals decreased from about

19.43 to about 4.5 wt%, slightly lower than

Figure 4. Influence of salt solution recycling

on the dehydration rate of citrogypsum in

recycling CaCl2 solution.

the water content in crystal theoretically DH

(20.92 wt%) and α-HH (6.21 wt%)

respectively. The sample was dried before

analyze.11

Thermal analysis of DH and α-HH are

characterized by DSC technique as shown in

Figure 5. From the DSC thermograms of DH

showing endothermic peaks at 151.2°C and

183.14°C corresponding to the amount of

water loss in DH crystals 1.5 and 0.5

respectively, which consists of 2 water

molecules. After DH has threat by

hydrothermal process, as can be seen from

the DSC thermograms of the product of

control process and the recycling saline

solution process at 1st, 2nd and 3rd cycle are

shown an exothermic peak of around 190 °C,

indicating that DH is completely changed to

α-HH.12

The investigation of the CaCl2

solution concentration effect on 3 times

recycling CaCl2 solution as a CG solvent

in the phase transformation process of DH

to α-HH. The Fig 6A and 6B shown that

the concentration of calcium ion (Ca2+) is

decreased from 4 M to 2.6 M and go to

steady state about 1 h after reaction start (

Fig. 6A), confirm that the concentration of

Ca2+ is reduced by the releasing of water

molecules in the CaSO4 crystals to the

solution. The final concentration of Ca2+ of

Figure 5. DSC Thermogram comparing CG

and α-HH of Control, 1st, 2nd and 3rd using

recycled calcium chloride solution.

product from Control process and

recycling process at 1st, 2nd and 3rd cycle

(Fig 6B) are reduced from 4 M to 2.1 M.

As this study requires 500 ml of water in

the system and the recycling CaCl2 solution

to be 4M CaCl2. It can be concluded that

the process of changing phase from DH to

α-HH completely by recycling CaCl2 solution

and also the concentration of CaCl2 solution is

decreased.13

IE51


Figure 6. Variations of concentration of Ca2+

ion of CaCl2 solution recycle.

The acidity of CaCl2 solution from

the recycling saline solution process is shown

in Figure 7, the acidity is reduced because the

dehydration of water content in the crystals

of DH to form α-HH.14

Figure 7. The pH value variations of the CaCl2

solution recycle with different recycling times.

4. Conclusion The phase transformation of the CG

to α-HH in CaCl2 solution recycle are

investigated via the hydrothermal process at

temperature 95°C for 5 hrs. The α-HH

hexagonal shape product are obtained and the

recycling CaCl2 solution can be used up to

4 cycles leading to reduction of solvents and

waste water in the phase transformation

process of CG to α-HH.

Acknowledgements The authors would like acknowledge

Capacity Building Program for New

Researcher 2020 from National Research

Council of Thailand (NRCT) and Department

of Industrial Chemistry, Faculty of Applied

Science, Graduate College King Mongkut’s

University of Technology North Bangkok.

References 1. Tongwen, X. U.; Weihua, Y., Chem. Eng.

Process. Process Intensification 2002, 41 (6),

519-524.

2. Yigitoglu, M., J. Islamic Acad. Sci. 1992,

5 (2), 100-106.

3. Sirimahasal, T.; Kalhong, Y.;

Simasatitkul, L.; Pranee, S.; Seeyangnok, S.,

Key Eng. Mater. 2019, 803, 351-355.

4. Pirdashti, M.; Omidi, M.; Pirdashti, H.;

Hassim, M. H., An AHP-Delphi multi-

criteria decision making model with

application to environmental decision-making.

2011.

5. Pavlov, K. A.; Dmitrevskii, B. A., Russian

J. Appl. Chem. 2003, 76 (9), 1408-1413.

6. Wang, P.; Lee, E. J.; Park, C. S.; Yoon,

B. H.; Shin, D. S.; Kim, H. E.; Koh, Y. H.;

Park, S. H., J. Am. Ceram. Soc. 2008, 91 (6),

2039-2042.

7. Miao, M.; Feng, X.; Wang, G.; Cao, S.;

Shi, W.; Shi, L., Particuology 2015, 19, 53-

59.

8. Barba, D.; Brandani, V.; di Giacomo, G.,

Chem. Eng. J. 1982, 24 (2), 191-200.

9. Kong, B.; Guan, B.; Yates, M. Z.; Wu, Z.,

Langmuir 2012, 28 (40), 14137-42.

10. Felten, C. J., Food materials science:

Effects of polyphenols on sucrose

crystallization and characterization and

creation of alternative salts of thiamine.

2019.

11. Guan, B.; Yang, L.; Wu, Z.; Shen, Z.;

Ma, X.; Ye, Q., Fuel 2009, 88 (7), 1286-

1293.

12. Guan, B.; Kong, B.; Fu, H.; Yu, J.;

Jiang, G.; Yang, L., Fuel 2012, 98, 48-54.

13. Brunner, J. K., Process for recovery of

carboxylic acids from the waste salt solutions

of cyclohexanone manufacture. Google

Patents: 1976.

14. Lu, W.; Ma, B.; Su, Y.; He, X.; Jin, Z.;

Qi, H., Constr. Build. Mater. 2019, 214, 399-

412.

IE52


Synthesis of anionic surfactant from oleic acid Laongdaw Vitee, Supatta Midpanon, Theerachart Leepasert, Potjanart Suwanruji*

Department of Chemistry, Faculty of Science, Kasetsart University, Bangkok 10900, Thailand


The anionic surfactant was prepared from oleic acid by oxidation and sulfosuccination

reaction. The first step was the oxidation reaction using 1:1.5 molar ratio of oleic acid to

potassium permanganate in a basic solution, pH = 13. Potassium permanganate (1% w/v) was

added to oleic acid at 10°C and the oxidation reaction was maintained at 15°C for 1 h, yielding

65.22% of 9,10-dihydroxystearic acid. The sulfosuccination reaction between 9,10-

dihydroxystearic acid and maleic anhydride was performed with the molar ratio of 1:1.5 at 95-

100°C for 10 h followed by sodium hydroxide and sodium bisulfite in isopropanol at 70°C for

2 h gave about 77% yield of the anionic surfactant. The product was confirmed by HR-MS,

FT-IR and NMR spectroscopy. The surface tension of the anionic surfactant at CMC (0.30

mM) investigated by Wilhelmy Plate method was found to be 38.31 mN/m.

1. Introduction

Surfactants are so-called surface

active agents because they can reduce the

interfacial tension of two-phase systems.

Surfactants are amphiphilic molecules

containing a hydrophilic head and a

hydrophobic tail. The hydrophobic tail is

typically a hydrocarbon, fluorocarbon or

siloxane. Surfactants are classified based on

the nature of the hydrophilic head into 4 main

groups: anionic, cationic, non-ionic and

amphoteric surfactants. In general,

surfactants adsorb at the interface and reduce

the interfacial interaction (Fanton et al.,

1997).

One of many roles of surfactants is a

detergent. By facing the hydrophilic head to

water and the hydrophobic tail to oil or dirt,

surfactant can remove oil or dirt from the

fabric surface. Another role of surfactants is

an emulsifier such as oil in water. The

hydrophobic part is adsorbed by the oil

droplet and the hydrophilic head is oriented

outward, giving a hydrophilic sheath around

the droplet. Surfactants can also act as a

dispersant. Dispersants ease the

incorporation of the solid particulates into its

medium such as pigments in water (Zhang

et al., 2014).

Most surfactants are synthesized from

petroleum which is a non-renewable

resource. Many research groups have been

interested in using renewable sources to

produce surfactants that are biodegradable

(Plat and Linhardt, 2001). Common

agricultural products that are used as raw

materials in the synthesis of surfactants are

sugar, starch, rice straw, corn cob and

vegetable oil. Palm oil is the most widely

used vegetable oil in the world. The major

fatty acids of palm oil are palmitic acid

(C16:0) and oleic acid (C18:1) (Foley et al.,

2012). Fatty acids could be used as a

hydrophobic part of surfactant. For example,

sodium n-octyl sulfosuccinate diester as an

anionic surfactant was prepared by

sulfonation and esterification reaction

(Jun wei et al., 2016) and sulfosuccinate

monoester was synthesized from oleyl

alcohol and maleic anhydride by sulfonation

of sodium hydrogen sulfite and sodium

hydroxide solution (Patil et al., 2013).

The main focus of this study is to

synthesize anionic surfactants using oleic

acid from palm oil as a reactant. The

synthesis was carried out using oxidation and

sulfosuccination reactions. The chemical

IE53



Figure 1. Synthesis of the anionic surfactant (III).

15°C, 1 h

95-100°C, 10 h

70°C, 2 h

structure of the anionic surfactant was

characterized using Fourier transform

infrared spectroscopy (FT-IR), nuclear

magnetic resonance (NMR) and mass

spectrometry (MS). Surfactant properties

including surface tension and γcmc were also

investigated.


2.1 Materials

Maleic anhydride was obtained from

Sigma-Aldrich (France). Oleic acid and

sodium sulfite were supplied by Merck

(Germany). Potassium permanganate was

from Elago enterprise (Australia) and

sodium hydrogen sulfite was from Himedia

(India). All the chemicals were of analytical

grade and used without further purification.

Deionized water and distilled water were

used for preparing solutions

2.2 Synthesis of 9,10-dihydroxystearic

acid.

Figure 1 shows the preparation of (II).

A reaction mixture of (I) 2.00 g (7 mmol) and

NaOH 2.00 g (50 mmol) in water 40 mL was

stirred at room temperature for 30 min, pH =

13. Water 300 mL and ice water 100 g were

added to the mixture. Then the mixture was

cooled down to 10°C, 1% w/v KMnO4

solution 188 mL (11 mmol) was added

immediately. After 5 min, Na2SO3 7.00 g was

added to reduce the excess KMnO4 from the

oxidation reaction and continued stirring for

1 h. Concentrated HCl 20 mL was added to

reduce the pH from 14 to 3. The precipitate

was collected by vacuum filtration. The

product was washed with boiled hexane 10

mL and digested by hexane 20 mL to remove

the remained (I). The purified product (III)

was checked by TLC having 2:1 ethyl

acetate/hexane as eluent (Wahyuningsih et

al., 2017).

IE54


2.3 Synthesis of anionic surfactant

(III) was synthesized according to

Figure 1. Firstly, (II) 5.00 g (0.0158 mol) was

melted in a 100 mL three neck reactor vessel

at 95-100°C with stirring. Maleic anhydride

2.34 g (0.0237 mol) was added and the

reaction was maintained at 90 °C for 10 h.

After that, 50% aqueous solution of NaOH

1.27 mL (0.0158 mol) and isopropanol 7 mL

were added to the mixture. After 15 min, 30%

aqueous solution of NaHSO3 5.5 mL (0.0158

mol) was added to the reaction. The reaction

was completed within 2 h. The product (III)

was dissolved in chloroform and filtered by

vacuum filtration to removed the remained

maleic anhydride. After chloroform was

removed, (III) was checked by TLC having

3:1 ethyl acetate/methanol as eluent (Patil et

al., 2013; Kumar et al., 2016; Xu et al., 2016).

2.4 Characterization of (II) and (III)

The chemical structure of (II) was

confirmed by 1H NMR and 13C NMR using

DMSO-d6 as solvent. The functional groups

of (II) and (III) were analyzed by FT-IR. The

sample in KBr pellet was scanned in the

wavenumber range between 4500-500 cm-1.

HR-MS was used to determine molecular

masses of (II) and (III) using APCI mode.

2.5 Surface tension and critical micelle

concentration (CMC) measurements

Surface tension of the synthesized (III) was

measured with a Wilhelmy tensiometer (DY-

300) at 25 °C. (III) was dissolved in

deionized water with 8 concentrations: 0.005,

0.010, 0.050, 0.10, 0.20, 0.30, 0.60 and 0.90

mM. The CMC and γcmc values were

analyzed from the plots of the surface tension

versus the logarithm of the concentration

(Reddy et al., 2015).


Oxidation at a double bond of (I) to give (II)

was successfully performed with KMnO4 in a

basic solution at 10°C. The yield was

65.22%.

Figure 2. FT-IR spectra of I (a), II (b) and III

(c).

From Figure 2b, the formation of the

hydroxyl group was verified by the absence

of the absorption band at 3006 cm-1 from

C-H stretching of alkene and the presence of

two absorption bands at 3323 and 3282 cm-1

from O-H stretching of alcohol hydroxyl

group and hydroxyl carboxylic acid,

respectively. Five absorption bands of (I)

were observed at 2917 and 2950 cm-1 from

C-H symmetric and asymmetric stretching of

aliphatic hydrocarbon, at 1714 cm-1 from

C=O stretching of carboxylic acid, and at

1463 together with 1411 cm-1 from C-H

bending of aliphatic CH2 and CH3,

respectively.

(a)

(b)

(c)

3006

(C=C)

2900

(C-H) 2950

(C-H)

1714

(C=O)

3323

(O-H)

3282

(O-H)

2950 (C-H)

2917 (C-H)

1750

(C=O)

2917 (C-H)

2950 (C-H) 1714

(C=O)

1350

(S=O)

1170

(S=O)

1043

(S-O)

1463

(C-H)

1411

(C-H)

IE55


Figure 3. MS spectrums of II (a) and III (b).

Figure 4. 1H NMR (a) and 13C NMR (b) spectrums of II.

(a)

(b)

(a)

(b)

[M+H] +

317

[M+Na] +

607

IE56


The MS spectra in Figure 3a shows

[M+H+] peak of (II) [C18H36O4] at m/z = 317

g/mol. From Figure 4a, 1H NMR of (II) in

DMSO-d6 shows the signals at 0.86 ppm (t,

3H, CH3), 1.1-1.6 ppm (CH2), 2.18 (CH2-

COOH), 3.34 (CH-OH), 4.18 ppm (OH of

alcohol) and 11.95 ppm (OH of acid). 13C

NMR of (II) in DMSO-d6 (Figure 4b) reveals

the signals at 13.96 ppm (CH3), 22-29.5 ppm

(CH2), 33.31 ppm (CH2-COOH), 73.75 and

73.76 ppm (CH-OH at C9 and C10).

(III) was synthesized by esterification

of (II) with maleic anhydride, followed by

sulfonation over the double bond of maleic

ester. The yellow wax-liked product was

obtained about 77% yield. The presence of

(III) was confirmed by FT-IR and HR-MS.

The FT-IR spectrum (Figure 2c) presents the

C=O stretching band of ester sulfosuccinate

at 1714 cm-1 and two S=O stretching bands

of the sulfonate group at 1350 and 1173

cm-1. The MS spectra (Figure 3b) shows the

[M+Na] peak of anionic surfactant

[C22H36O10SNa4] at m/z = 607 g/mol.

Figure 5 exhibits the surface tension of (III)

at various concentrations. The surface

tension continuously decreased as the

concentrations of (III) increased until it

reached the critical micelle concentration

(CMC). The CMC of this surfactant was

determined by the intersection between the

descending slope and the constant horizonal

line in the surface tension plots against

concentration, showing the value of 0.30 mM

and the surface tension of 38.31 mNm-1

(Fekarcha et al., 2012). Comparing with

sodium lauryl sulfate (SLS), a commercial

anionic surfactant, the CMC of SLS is at 0.6

mM and γcmc is 33.7 mN/m (Esteves et al.,

2016). (III) can form micelles at a lower

concentration and can reduce surface tension

better than SLS.

Figure 5. Surface tension of III aqueous solutions at different concentrations.

38.31

IE57


4. Conclusion

(III) was successfully synthesized

from (I) through oxidation and

sulfosuccination reactions, having

sulfonate groups as a hydrophilic part. The

yield of yellow waxy product was around

77%. The chemical structure was

confirmed by HR-MS and FT-IR. (III) can

reduce surface tension as indicated by CMC

value and γcmc at 0.30 mM and 38.31 mN/m,

respectively.

Acknowledgements

Department of Chemistry, Faculty

of Science Kasetsart University Bangkhen

Campus and National Research Council of

Thailand, for the financial support of this

research.

References 1. Esteves, R., Dikici, B., Lehman, M.,

Mazumder, Q. & Onukwuba, N.

Undergraduate Research Journal.

2016, 1(4).

2. Fanton, E., Fayet, C. & Gelas, J.

Carbohydr. Res. 1997, 298(1), 85-92.

3. Fekarcha, L. & Tazerouti, AJ. Surfact.

Deterg. 2012, 15, 419–4.

4. Foley, P., Beach, E. S. & Zimmerman,

J. B. Chem. Soc. Rev. 2012, 41(4),

1499-1518.

5. Junwei, D., Bin, S. & Jun, X J. Surfact.

Deterg. 2011, 14, 43–49.

6. Kumar, P. P., Ramesh, P. & Kanjilal, S.

J. Surfact. Deterg. 2016, 19, 447–454.

7. Patil, H. V., Kuulkarni, R. D. & Mishra,

S. Inter. J. Chem. Chem. Eng. 2013, 3,

69-74.

8. Plat, T. & Linhardt, R. J. J. Surfact.

Deterg. 2001, 4, 415-421.

9. Reddy, T., V. K., Rani, G. S., Prasad, R.

& Devi, B. P. RSC Adv. 2015, 5(51),

40997-41005.

10. Wahyuningsih, T. D. & Kurniawan, Y.

S. AIP Conference Proceedings. 2017.

11. Xu, J., Cao, F., Li, T., Zhang, S., Gao,

C. & Wu,Y.J. J. Surfact. Deterg. 2016,

19, 373-379.

12. Zhang, G., Zhu, S., Zhang, W., Wang,

Y. & Campus, L. MEIC-14 Conference

Proceedings, 2014.

IE58


CFD simulation of flow behavior and heat transfer in high density

circulating fluidized bed riser Waritnan Wanchan1, Parinya Khongprom1,2*, Sunun Limtrakul3

1Department of Industrial Chemistry, Faculty of Applied Science, King Mongkut’s University of Technology

North Bangkok, Bangkok, 10800, Thailand 2Integrated Nanoscience Research Center, Science and Technology Research Institute, King Mongkut’s

University of Technology North Bangkok, Bangkok 10800, Thailand 3Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok, 10900, Thailand


A high density circulating fluidized bed riser reactor was widely used to operate the

fast-chemical reaction of gas-solid flow system. The excellent wall-to-bed heat transfer is

required for the effective operation. Thus, this research aims to study the hydrodynamics and

wall-to-bed heat transfer in high density fluidized bed riser using CFD simulation. The flow

behavior and heat transfer characteristic were predicted using a two-fluid model with the

kinetic theory of granular flow (KTGF). A fluidizing gas was an air at ambient condition with

superficial gas velocity of 3-5 m/s. Fluid catalytic cracking (FCC) particles were fed to the

reactor with solid circulation rate in the range of 200-300 kg/m2s. The solid circulation rate

slightly effects on gas and particle axial velocities, but the solid volume is remarkably increased

with the solid circulation rate leading to the higher wall-to-bed heat transfer coefficient. The

superficial gas velocity strongly influences on the gas and particle axial velocities, and solid

volume fraction. Thus, the wall-to-bed heat transfer coefficient decreases with increasing

superficial gas velocity.

1. Introduction

Circulating fluidized bed (CFB) riser

is a good choice in gas-solid operation such

as fluid catalytic cracking, combustion,

drying, and coal gasification.1 This process

involved heat transfer due to the heat of

reaction occurring in the system.2 Thus, not

only the hydrodynamic but also the heat

transfer significantly influences on the

performance of the reactive fluidized bed

reactor. So, an understanding of the

hydrodynamics and heat transfer is essential

for proper design and effective operation.

The operations of CFB reactor are

mainly classified by many researchers as two

regimes, which are a low density regime and

high density regime. Zhu and Bi3 defined the

operating condition for the low density

regime at solid circulation rate less than 100

kg/m2s, superficial gas velocity between 2-8

m/s, and solid volume fraction less than 0.1

in the developed region. This condition is

mainly applied for gas-solid process such as

coal combustion.1,4 The axial flow structure

exhibited the uniform pressure gradient. S-

shape profile of solid fraction was observed.

A dilute core with dense annulus was found

in radial flow structure. This flow pattern is

called core-annulus flow structure. The back-

mixing of gas and solid was observed.4 For

the high density regime, the process was

operated at high solid flux which used for

gas-catalytic process such as fluid catalytic

cracking.1,3 The solid circulation rate more

than 200 kg/m2s and solid volume fraction

more than 0.1 in the entire riser were

operated, as proposed by Zhu and Bi.3 S-

shape profile of solid fraction was not

observed.1,5 Less gas back-mixing was

observed in high density regime.5 In addition,

the higher solid circulation rate tends to

reduce solid-back mixing and increased solid

particle throughput in the bed.4

The hydrodynamics and heat transfer

in CFB reactors were investigated by several

research groups. Numerous investigators IE59


revealed that the performance of heat transfer

in the bed was controlled by hydrodynamic

behavior.7 Ma and Zhu2 studied the heat

transfer in downer and riser fluidized bed

reactors. They found that heat transfer

behavior of both reactors related with the

flow characteristic. In the riser, heat transfer

coefficient increases with the increasing solid

circulation rate and the decreasing of

superficial gas velocity. Basu et al.8 studied

the heat transfer in a pressurized circulating

fluidized bed. The effects of pressure,

suspension density, particle size, and

superficial gas velocity on the heat transfer

coefficient were investigated. The heat

transfer coefficient increased with the system

pressure and suspension density. The

decreasing of particle size and superficial gas

velocity tends to increase the heat transfer

coefficient. However, a lack of

hydrodynamics and heat transfer

investigation under high density riser was

carried out, which is favored for industrial

application in the literature.

Computational fluid dynamics (CFD)

has become a useful tool for predicting the

complex hydrodynamics and heat transfer

behavior in multiphase flow system owing to

its advantages over experimental

measurement such as a short calculation time,

ability to simulate of the complex reactor

geometry with severe conditions, and

providing the detail hydrodynamic behaviors.

A two-fluid model (TMF) has been

developed to predict flow behavior of gas

solid flow system. The kinetic theory of

granular flow (KTGF) model is used to

compute the fluid properties of solid phase,

which are related with the granular

temperature.9 CFD method based on two-

fluid model gave accurate prediction, which

shows good agreement of quantitative and

qualitative validations with experimental

results.10 Hence, the purpose of this work was

to applied CFD for investigating the

hydrodynamics and heat transfer

characteristic in high density fluidized bed

riser.

2. Simulation setup

2.1 Reactor geometry

CFB reactor based on Koksal’s

experiment11 was used in this study, as

illustrated in Figure 1a. The riser reactor with

a height of 7.6 m and an inner diameter of

0.23 m, was modeled as shown in Figure 1b.

Heating surface was located at 2.06 m high

above the inlet.

2.2 Mathematical model

A two–fluid model coupling with

KTGF was adopted to simulate the

hydrodynamic and heat transfer. Both gas

and solid phases were considered as

interpenetrating continuum phases. The

turbulence flow in the system was predicted

by using the standard k– turbulence model.

The governing equations was tabulated in

Table 1. The constitutive equations are

summarized by Khongprom et al.12

Figure 1. Schematic of CFB unit (a) and

Riser 2D simulation.

IE60


Table 1. Governing equations.

Table 2. Gas and solid properties and model

parameters. Properties Value

Gas

kg (W/m-K)

Cpg (J/kg-K)

Particle

Type of particle

ρs (kg/m3)

ds (m)

kg (W/m-K)

Cpg (J/kg-K)

Parameters

Restitution coefficient between

particle (es)

Restitution coefficient between

particle and wall (ew)

Specularity coefficient (φ)

Granular temperature (Θ, m2/s2)

0.025

1006.43

FCC catalyst

1600

09

0.25

800

0.9

0.9

0.0003

0.0001

2.3 Numerical simulation

A commercial mesh-generator,

Workbench 15.0 was used to create the

geometry and computational domain. The

commercial software Ansys-FLUENT 15.0.

was used to solve the governing and the

constitutive equations, which were based on

the finite volume method. The SIMPLE

pressure-velocity coupling method was

carried out. A discretization of the

convection term was based on a second order

upwind scheme. The simulation was

performed with a time step size of 0.001s.

2.4 Boundary condition

The stagnant air at ambient condition

was set as an initial condition. At the inlet,

both gas and solid particle at ambient

temperature were fed to the system. The

velocity inlet boundary condition was

applied, which the velocities, volume

fraction, and temperature were specified. The

pressure outlet boundary condition was set

for the outlet. The wall and heater surface

were considered as no-slip and partial-slip

conditions, respectively. The heater surface

was set to a constant heat flux. The wall was

considered as a perfect insulated. The gas and

solid properties and the model parameters are

tabulated in Table 2.

2.5 Heat transfer coefficient

The heat transfer coefficient is

defined as the ratio of the heat flux to the

different temperature between the wall and

the bed. This coefficient can be calculated

from eq (10). 13

h =q

Tw−Tb (10)

where q stands for the heat flux (W/m2), and

Tw and Tb are wall and bed temperatures,

respectively.


3.1 Hydrodynamics behavior

The radial distribution of gas and

solid velocities under different operating

conditions is presented in Figure 2.

Comparatively, the gas and solid velocities

1. Continuity equation

Gas phase ∂

∂t(gρg) + ∇ ∙ (gρgv⃗ g) = 0 (1)

Solid phase ∂

∂t(sρs) + ∇ ∙ (sρsv⃗ s) = 0 (2)

2. Momentum conservation equation

Gas phase

∂

∂t(gρgv⃗ g) + ∇ ∙ (gρgv⃗ g

2) =

−g ∙ ∇p + ∇ ∙ τg̿ + gρgg⃗ + Kgs(v⃗ g − v⃗ s)

(3)

Solid phase

∂

∂t(sρsv⃗ s) + ∇ ∙ (sρsv⃗ s

2) = −s ∙ ∇p −

∇ ∙ ps + ∇ ∙ τs̿ + sρsg⃗ + Kgs(v⃗ s − v⃗ g)

(4)

4. Granular temperature conservation

3

2[∂

∂t(sρsΘs) + ∇ ∙ (sρsv⃗ sΘs)] =

(−PsI̿ + τ̿s): ∇v⃗ s + ∇ ∙ (kΘs∇Θs) − γΘs

(5)

5. Energy equation

Gas phase

∂

∂t(gρgHg) + ∇ ∙ (gρgu⃗ gHg) =

∇.gkg∇Tg − hgs(Tg − Ts)

(6)

Solid phase

∂

∂t(sρsHs) + ∇ ∙ (sρsu⃗ sHs) =

∇.sks∇Ts − hgs(Ts − Tg)

(7)

7. k- turbulence model

∂

∂t(jρjkj) + ∇ ∙ (jρjkjUj) =

∇ ∙ (εjμt

σk∇kj) + (jGk − jρjεj)

(8)

∂

∂t(jρjj) + ∇ ∙ (gρgεjUj) =

∇ ∙ (εj

μt

σk∇εj) +

εj

k(C1εjGk − C2jρjεj)

(9)

IE61


are slightly different owing to the different

gravitational force acting on gas and solid

phases. At a low superficial gas velocity

(Ug=3 m/s), both gas and solid velocities

show an almost uniform distribution in radial

direction. The velocity slightly decreases

toward the wall. The uniform distribution

indicated the fully developed flow. At a high

superficial gas velocity (Ug=5 m/s), both gas

and solid velocities exhibited less uniform

distribution, which velocity significantly

decreased from the center toward the wall.

Under this operating condition, the

investigated axial position was in the

acceleration zone. The acceleration length

increases with the increasing of superficial

gas velocity.14 Obviously, the gas and solid

velocities remarkably increased with the

superficial gas velocity. However, the solid

circulation rate exhibited less influence on

gas and solid velocities. Such trend was

observed by Hung et al. 15 However, Qi et al.

16 and Huang et al. 17 found that the solid

circulation rate significantly impacted on

velocity in the bed because a wider range of

the solid circulation rate was investigated in

their work.

The lateral distribution of solid

volume fraction was shown in Figure 3. A

non-uniform flow structure was observed.

The large particle cluster was formed near the

wall. The high gas velocity in the center

region leads to low solid fraction. This flow

pattern was defined as core-annulus flow

structure, which is a key characteristic of

such reactor type that can be observed by

experimental and simulation studies.9,18 In

addition, both parameters significantly

influenced on solid fraction. The solid

volume fraction increases with the increasing

of solid circulation rate and the decreasing of

superficial gas velocity. This trend was found

in several researches.2,18,19

Figure 2. The radial distribution of gas and solid velocity.

Figure 3. The radial distribution of solid volume fraction.

IE62


3.2 Heat transfer

In this work, the gas phase temperature

was selected to represent the wall and bed

temperatures for calculating heat transfer

coefficient owing to both gas and solid

temperatures are insignificantly different as

shown in Fig. 4. It is indicated that high heat

transfer between phases due to the efficient

gas-solid mixing. Fig. 5 shows the radial

distribution of the temperature near the wall

under various operating conditions. The

increasing of solid circulation rate causes the

reducing of the temperature near the wall.

Hence, the wall-to-bed heat transfer coefficient

increases with Gs as shown in Fig. 6. Although,

the solid circulation rate is insignificantly

influenced on gas velocity (see Figure 2).

Indicating that the gas convection does not

depend on the solid circulation rate. But the

solid fraction near the wall significantly

increases with solid circulation rate (see Figure

3) resulting in high contact efficiency between

solid particles and the heating surface. Thus,

the particle convection increases with the solid

circulation rate leading to high the wall-to-bed

heat transfer. Figure 5 also shows the effect of

Ug on the temperature profile. The increasing

of the temperature near the wall with an

increasing of the Ug was obtained resulting in

the decreasing of the wall-to-bed heat transfer

coefficient as shown in Fig. 6. Although, the

gas velocity increases with the superficial gas

velocity, which enhances the gas convection,

but reducing the particle convection because of

Figure 4. Gas and solid temperatures near the

wall.

low solid volume fraction in the bed (see Fig.

3). Thus, the wall-to-bed heat transfer

coefficient decreases with the increasing of

superficial gas velocity. It is confirmed that the

particle convection is the main mechanism

affecting on the overall heat transfer in gas-

solid flow system. This observation was also

observed by several investigators.3,7,8

4. Conclusion

In the present work, the hydrodynamic

and heat transfer in a high-density fluidized

bed riser reactor were studied. A two-fluid

model with the kinetic theory of granular flow

was used to predict hydrodynamic and heat

transfer. The solid circulation rate slightly

impacts on gas and particle velocities, but

significantly increased solid

fraction resulting in

Figure 5. The distribution of gas temperature near the wall.

IE63


Figure 6. Effects of Ug and Gs and on the

wall-to-bed heat transfer coefficient. high wall-to-bed heat transfer coefficient.

The increasing of superficial gas velocity

tends to increase gas and solid velocities

but decreases solid volume fraction in the

bed. However, the wall-to-bed heat transfer

coefficient decreases with the increasing of

superficial gas velocity indicating that the

particle convection is the main mechanism

of heat transfer.

Nomenclature

Cp Specific heat capacity, J kg-1 K-1

C1,C2 turbulence constant

ds particle diameter, m

es restitution coefficient between

particle

ew restitution coefficient between

particle and wall

g⃗ gravitational acceleration, m s-2

Gs solid circulation rate, kg m-2s-1

Gk production of turbulent kinetic

energy,k m-1 s-3

h heat transfer coefficient, W m-

2K-1

hgs interphase heat transfer

coefficient

H enthalpy, J kg-1

k thermal conductivity, W m-1K-1

kj turbulent kinetic energy,J kg-1

t Time, S

T Temperature, K

Ug superficial gas velocity, m s-1

v Velocity ms-1

ρ density, kg m-3

μ viscosity, kg m-1s-1

I ̿ unit tensor

𝜏̿ stress tensor, Pa

Subscript

b bed

g gas phase

i species i

j phase j

s solid phase

w wall

Acknowledgements This research was supported by

Department of Industrial Chemistry,

Faculty of Applied Science, King

Mongkut’s University of Technology North

Bangkok.

References 1. Berruti, F.; Chaouki, J.; Godfroy, L.;

Pugsley, T.S.; Patience, G.S.; Can. J.

Chem. Eng. 1995, 73, 579-602.

2. Ma. Y.L.; Zhu, J.X. Chem. Eng.

Technol. 2001, 24, 85-90.

3. Zhu, J.X.; Bi, H.T. Can. J. Chem. Eng.

1995, 73, 644-649.

4. Marcus, R.R.; Leung, L.S.; Klinzing,

G.E.; Rizk, F. Chapter 4-6 in Pneumatic

Conveying of Solid, Chapman and Hall,

New York, 1990; pp 83-258.

5. Liu, J.; Ph.D. Thesis, The University of

British Columbia, 1995.

6. Qi, M.; Zhu, J.; Barghi, S. Chem. Eng.

Sci. 2012, 84, 437-448.

7. Glicksman, L.R.; in Grace, J.R.;

Avidan, A.A.; Knowlton, T.m.; (Eds).

Circulating fluidized beds. Blackie,

London, U.K. 1997, 261-311.

8. Basu, P.; Cheng, L.; Cen, K. Int. J. Heat

Mass Transfer. 1996, 39, 2711-2722.

9. Upadhyay, M.; Park, J.H. Powder

Technol. 2015, 272, 260-268.

10. Almuttahar, A.; Taghipour, F. Powder

Technol. 2008, 185, 11-23.

11. Koksal, M.; Golriz, M.R.;

Hamdullahpur, F. Int. J. Energy Res.

2005, 29, 923-935.

IE64


12. Khongprom, P.; Aimdilokwong, A.;

Limtrakul, S.; Vatanatham, T.;

Ramachandan, P.A. Chem. Eng. Sci.

2012, 72, 8-19.

13. Glicksman, L.R., 1997. (Eds).

Circulating fluidized beds. Blackie,

London, U.K.1997, 261-311.

14. Sabbaghan, H.; Gharebagh, R.S.;

Mostoufi, N. Powder. Tehcnol. 2004,

142, 129-135.

15. Huang, W.; Yan, A.; Zhu, J. Chem.

Eng. Technol. 2007, 30, 460-466.

16. Qi, M.; Zhu, J.; Barghi, S.; Chem. Eng.

Sci. 2012, 84, 437-448.

17. Wang, C.; Zhu, J.; Barghi, S.; PhD.

Thesis, The University of Western

Ontario London, Ontario, Canada,

2013.

18. Mastellone, M. L.; Arena, U.; Chem.

Eng. Sci. 1999, 54, 5383-5391.

19. Parssinen, J.H.; Zhu, J.X.; AIChE J.

2001, 47, 2197-2205.

IE65

Date post:	27-Dec-2021
Category:	Documents
Upload:	others
View:	29 times
Download:	1 times

Process analysis of acetaldehyde production from ethanol ...

Documents