Fibers and Polymers 2014, Vol.15, No.3, 547-552 DOI 10.

1007/s12221-014-0547-7

Polyester Fiber Production Using Virgin and Recycled PET

J. C. Tapia-Picazo*, J. G. Luna-Bárcenas, A. García-Chávez1, R. Gonzalez-Nuñez2,
A. Bonilla-Petriciolet1, and A. Alvarez-Castillo3
Cinvestav-Querétaro, Libramiento Norponiente 2000, Fracc. Real de Juriquilla, Querétaro, Querétaro., C.P. 76060, México
1
Chemical Engineering and Biochemical Department, Technological Institute of Aguascalientes,
Aguascalientes, Aguascalientes, C.P. 20256, México
2
University Center of Exact Sciences and Engineering, Chemical Engineering Department, University of Guadalajara,
Guadalajara, Jalisco, C.P. 44430, México
3
Division of Graduate Studies and Research, Chemical and Biochemical Engineering Department, Electromechanical
Department and Basic Sciences Department, Technological Institute of Zacatepec, Zacatepec,
Morelos, C.P. 62780, México
(Received August 30, 2012; Revised July 9, 2013; Accepted July 30, 2013)

Abstract: In this study, the design and construction of an extrusion equipment with spinning fiber devices has been
developed to produce polyester fiber from virgin and recycled polyethylene terephthalate (PET). Several operating
parameters (i.e., pressure, temperature, feed flow rate, extrusion speed and extruder design) have been analyzed to identify
the best process conditions. In particular, this study has focused on a detailed analysis for the processing of recycled raw
material for polyester textile fiber applications considering the variability of the process and identifying alternatives to
minimize the impact on the quality parameters such as the fiber diameter and mechanical specifications. The experimental
results were compared with the values calculated using a theoretical model, which has been developed for these particular
cases. The mathematical analysis of the mass flow showed a very good agreement with respect to the experimental data,
where there was a percentage difference < 3 %. It was found that the fiber diameter is a function of intrinsic viscosity (VI) or
melt flow index (MFI). Finally, the mechanical properties of the fibers were evaluated and results indicated that the fiber with
higher average molecular weight showed higher tenacity and lower Young’s modulus values.
Keywords: Recycled PET, Extrusion process, Spinning fiber, Textile PET fiber, PET

Introduction functionalized with maleic anhydride [PP-graft-MA]), were

From an economical point of view, the recycling process is
the best way to reduce wastes of poly(ethylene terephthalate)
(PET) [1] and, therefore, many technologies have been
developed for performing this type of processes. The first
effort in the world for recycling PET bottles was in 1977 [2];
but in the following decades, scientific studies were performed
to analyze the properties of recycled PET wastes using
extrusion processes. Several methods have been reported to
obtain recycled recipients and bottles from PET, but in
general PET wastes have been traditionally used for energy
recovery [3-5]. Actually, the recycled PET is used in the
synthesis of special plastic composite materials [6] and to
reinforce concrete [7,8]. For example, Torres et al. [9]
performed a comparative study of the thermal and mechanical
properties of bottles made from waste and virgin PET. These
authors have obtained elongation values of 200 % for virgin
PET and values 10 % smaller for recycled PET. These results
were attributed to the crystallinity differences. Oromiehie
and Mamizadeh [10] used three different methods for recycling
PET bottles, where virgin PET, recycled PET and mixtures
of both PET types, with and without polypropylene
*Corresponding author: tapiajc@hotmail.com
processed. This study reported that the intrinsic viscosity
(IV) and molecular weight (Mw) decreased with the content
of the recycled PET in the mixtures. This behavior was
attributed to thermal effects, as well as, the mechanical
degradation of recycled PET. In addition, the properties of
the functionalized blends were improved due to the chemical
and physicochemical interactions between components in
the blend. On the other hand, Martin and Rojas [11] developed
an extrusion process for the production of recycled PET
filaments using a simple screw. This patent consisted of two
parts. The first one is related to the recycling of PET with a
stirrer and a condenser, while the second one involves the
extrusion process itself. This development was performed to
obtain a constant quality in the extruded product, which
corresponds to the main problem in the traditional PET
extrusion recycling.
It is convenient to remark that several studies have reported
that raw material flakes produced by the size reduction of
PET waste should have certain minimum requirements in
order to achieve a satisfactory PET recycling process [12-
15]. This is because there is a loss of molecular weight in
PET extrusion process due to hydrolytic [16,17] and thermal-
mechanical degradation during melting process [18].
Degradation also causes non uniformity in the flow of fused
material generating negative effects in subsequent processes

547
548 Fibers and Polymers 2014, Vol.15, No.3
and the properties of final product are also affected [4].
Gurudatt et al. [19] obtained chips from PET waste for the
filament extrusion process. Virgin PET and waste PET from
bottles were used in mixtures of different compositions.
These authors analyzed the extrusion and stretching stages
of the spinning process and found that the variations of
molecular orientation is very important for determining the
properties, efficiency and production of final fiber. Abbasi et
al. [20] showed that the crystallinity of recycled materials
was higher than those obtained for virgin materials.
Consequently, the tenacity of samples from used material
was higher and the elongation was smaller. Herein, it is
convenient to highlight that there are other scientific and
industrial reports about the production of fibers from
recycled PET [21-24]; unfortunately, the technical details
are not described because of the industrial secrecy.
In this study, an analysis of the critical variables that
impact on the flow uniformity of the melt polymer and the
degradation of recycled PET using an extruder was performed.
This study includes the following aspects: a) the design of
the PET extrusion and spinning process for textile applications,
b) the mathematical analysis of the operation curves in the
extrusion process, c) experimental results and the material
characterization in the extrusion and spinning stages, and
finally, d) the validation of the operation curves using the
experimental data.
Methodology
Process Design
The process design was based on the process specifications
and final product (i.e., textile fiber) and they include: the
operation temperature: 270-320 oC, the drying temperature:
70-140 oC, the maximum PET humidity percentage: 0.02 %,
the fiber tenacity: 2 gf/denier, the elongation: 20 %, the
residual retraction: < 5 %, and the product denier: 1.5-3. For
the screw design, the properties of the raw material and the
design parameters (e.g., compression ratio, residence time,
angles of the helix, the relationship length/diameter of the
extruder) were considered and they values were established
according to the literature [11,25].
Mathematical Analysis
Mathematical equations for the process design analysis
were adapted to the extruder and raw materials used in the
present communication. These equations were based on the
PET degradation during reprocessing and the results reported in
[25]. Specifically, the total flow equation (Q) inside the
extruder was obtained by solving the moment, heat and mass
balances. This problem was considered as a fluid movement
between two surfaces, in which one of them is movable
(screw) and using a rectangular coordinate in z direction
(channel of the screw). Then, the total flow is given by
J. C. Tapia-Picazo et al.
2 2 3 2
π D NHsenθcosθ πDH sen θ ∆P
Q = ------------------------------------------- − ---------------------------- ------- (1)
2 12η L
where N is the screw rotation speed, ∆P is the pressure drop,
η is the melt viscosity of the material; while D, H, L and θ
are the diameter, depth, length and angle of the screw,
respectively. The melt viscosity η is considered a function of
the inherent viscosity (IV) and absolute temperature (T) and
it is given by
( 6800/T ) 5.1
η = 0.098e ( VI ) (2)
During processing, the degradation of the PET material
occurs due to chain scission, causing a decrement of the
average molecular weight, which is measured by the inherent
viscosity at any residence time in the melt (IVt). This
function is given by the following equation
17080 1/1.47 −1
26.9 ---------------
1 ⎞ 1.47
⎛ ------- T
IVt = + e t (3)
⎝ IV0⎠
where T is the absolute temperature, IV0 is the initial
inherent viscosity, and t is the residence time in the melt,
respectively.
On the other hand, if the PET polymer contains some
water, the hydrolytic degradation may also occur and can be
calculated by
IV0
IVH = --------------------------------------------------------
1.47 0.68
(4)
[ 1 + 21.9( x ) ( IV0 ) ]
where IVH is the inherent viscosity after reaction with water,
IV0 is the initial inherent viscosity of supplied PET polymer
and x is the % weight of water in supplied PET polymer,
respectively.
The analysis was performed by taking into account the
degradation and the material uniformity in the process. For
this purpose, we have used the strategy and method reported
by Franceschini and Macchietto [14]. Therefore, a complete
factorial experimental design and mathematical calculations
of the operation curves were performed. The analysis was
performed considering the variation levels for recycled PET
in the following critical ranges [9,11,13,25-27]: a) mass flow
variation: ±10 %, b) pressure drop: ±10 %, c) intrinsic
viscosity: 0.8-0.5, and d) raw material humidity: <0.02 %.
Experimental
The extruder was maintained at constant temperature for
10 minutes before the initial feeding. The raw material was
fed using one batch charge of 100 g and the operation
conditions (i.e., temperature, pressure and mass flow) were
constant and continuously monitored during an operating
time corresponding to three times of the residence time
Polyester Fiber Production Using PET
before the test evaluation. The pressure was fixed for every
experimental test using different screw speeds. The mass
flow was determined at different pressure conditions (from
2.68 to 4.14 kg/cm2) and temperatures (i.e., 260, 280, 290
and 300 oC) for the two different raw materials: recycled
PET (VI=0.65) and pharma grade PET (VI=0.71). The results
obtained from the mathematical modeling were compared
with those obtained from the experimental tests.
Finally, the fibers obtained using virgin and recycled PET
were mechanically characterized. Specifically, the tensile
properties were determined using an Instron machine (model
4011) in accordance with ASTM D 638-10. Also, the average
molecular weight was evaluated using the viscosimetric
approach and the micrographs were obtained using a
scanning electron microscope HITACHI TM-1000.
Results and Discussion
Process Design
The extruder is based on a normal design but it has an
adjustable support for the motor that allows the screw to be
removed in just one minute for analyzing the changes of
material inside the barrel at different process conditions [28].
To remove the screw, the gear box has two dissembling
gears than permit to separate them with a single bolt. Latter,
the motor is moved on the adjustable support for taking out
the screw in a direction opposite to the material flow. In the
spinning stage, first a slow stretching and a second stretching
are applied to the polyester fiber using three groups of five
rollers. This operation improves the molecular orientation
and crystallinity in order to increase the mechanical properties
of the polyester [19,25]. Two drying systems were used to
improve the crystalinity, dimensional uniformity and to
reduce fiber porosity [25]. One drying system used heat rolls
and the other included three heat plates operating at
temperatures between 10 and 100 oC over the glass transition
temperature of PET. The same operating conditions were
used in extrusion and spinning stages and the virgin and
recycled PET were processed at the values reported in Table 1.
Tensile Mechanical Properties
The tensile mechanical properties of the fibers are
Table 1. Operation conditions for processing virgin and recycled
PET
Condition Value
Residence time 10 minutes
Extrusion temperature 280 oC
First stretching 1.5 times
Second stretching 1.5 times
Drying temperature 120 oC
Spinning speed 50 m/minute
Fibers and Polymers 2014, Vol.15, No.3 549
reported in Table 2. The break stress of the virgin PET fiber
was higher than the value obtained for the recycled PET
fiber; while its Young’s modulus was lower (almost 50 %).
The break elongation values found in this study are very
near to the values reported in other studies [29,30]; while the
breaking stress and Young’s modulus are lower than the
values reported by Wróbel and Bagsik [30] and Frounchi et
al. [31] but in the same order of magnitude. To explain these
differences, the average molecular weights of virgin and
recycled fiber were calculated using the intrinsic viscosity
approach. The average molecular weights of virgin and
recycled fiber were 19342 g/mol and 15813 g/mol, respectively.
According to literature, the presence of smaller polymer
chains, due to the polymer degradation, may be are not
accommodated than the larger chains. This can cause a
decrement of the crystallinity, which has a significant influence
on the mechanical properties. Therefore, the fibers with
lower average molecular weight may show lower break
elongation, higher Young’s modulus [29,30] and a lower
breaking stress [30]. Flow index of raw materials used in the
extrusion process was obtained, see Table 3. The difference
in flow index is mainly caused by the hydrolytic degradation
that the extrusion process produced to the recycled PET;
therefore, an increase in the melt flow is an indicator of the
extent of the thermo-mechanical degradation [32].
Figure 1 shows the micrographs of virgin and recycled
PET fibers obtained at different magnifications. The surface
morphology of the virgin fibers consists of smooth cylindrical
fibers. It appears that this smooth surface is caused by the
partial solidification that occurs as soon as the fiber leaves
the extruder [33]. Experimental studies have reported that
the viscosity of the PET polymer has a significant impact on
the fiber diameter than any other process variable [34].
Therefore, different average diameters should be obtained
for virgin and recycled PET (see Figure 1). In fact, virgin
fibers showed a lower diameter than that obtained for
recycled fibers. This is because the lower flow mass of the
virgin PET at the process temperature, caused principally by
the lower flow index of the raw material in comparison to
Table 2. Mechanical properties of virgin and recycled PET fibers
Property Virgin PET fiber Recycled PET fiber
Tensile strength (kg/cm2) 140.5 220.0
2
Breaking stress (kg/cm ) 82.2 42.20
Break elongation (%) 6.96 5.00
Young’s modulus (kg/cm2) 5690 10500
Table 3. Melt Flow index of virgin and recycled raw materials
Raw material Melt flow index (g/10 minutes)
Virgin PET 13.21
Recycled PET 21.34
550 Fibers and Polymers 2014, Vol.15, No.3 J. C. Tapia-Picazo et al.

Figure 1. Micrographs of PET fibers obtained with, (a),(c) virgin PET and (b),(d) recycled PET at the operating conditions given in Table 1.

the recycled material. Additionally, the surface of virgin Table 4. Geometry and extruder operating conditions
fibers has lower defects and higher shining than those Extruder zone Feeding Compression Dosing
obtained for the recycled fibers.
Diameter (cm) 3.33 3.33 3.33
Mathematical Analysis of the Process Length (cm) 33 33 34
Equation (1) was used for the calculation of the flow rate Depth (cm) 0.6 0.38 (average) 0.16
(Q) using the geometry and extruder operating conditions Angle (º) 17.2 17.7 15
given in Table 4. We have considered a standard PET Operating temperature (ºC) 280 285 290
extrusion process with an initial intrinsic viscosity of 0.6, a
mass flow in the feed of 99 g/minute, L/D extruder ratio of
30 and an extruder diameter of 0.033 m. Figure 2 shows the
behavior of the flow along the extruder length. The changes
of density or flow rate were produced by the temperature
and the geometry of the extruder. Note that these changes
were more significant in the first two sections at the same
speed of change even though these sections showed differences
in the geometry and temperature. On the other hand, the
pressure along the screw was affected mainly by its
geometry in the last extrusion zone. For the case of the
extrusion process of recycled PET (Figure 3), this increasing
pressure was mainly caused by the total fusion of the
material and the geometry in the last 0.34 m of the extruder.
If an increment of 10 % of mass flow variation in the feed is Figure 2. Flow behavior inside the extruder obtained with
considered, the pressure does not present a considerable equation (1).
Polyester Fiber Production Using PET Fibers and Polymers 2014, Vol.15, No.3 551

Figure 3. Pressure inside the extruder of a standard PET extrusion

process with a flow rate variation of +10 %.

Figure 4. Mass flow as a function of pressure and temperature for

change inside the extruder for the first two sections, but for recycled pharma-grade PET.
the last zone (compression) the pressure may present a
significant variation, see Figure 3. This is an important issue
to be considered in the control of the system because this
level of mass variation is a normal value in the recycling
process and it will affect the final diameter of the filaments
produced using this type of extruders. The normal variations
for the filament diameter are ±10 % and this is the reason for
minimizing the mass variation using the pressure control of
the screw.
Based on the results of the operation curves, we can define
that the variability of VI0 in the range of 0.5-0.8 must be
reduced using the appropriate raw materials to produce a
quality product. On the other hand, it is necessary to handle
the following operating variables: a) screw speed: 30-
45 rpm, b) temperature: 270-300 oC and c) humidity: 0.01-
0.02 %.

Experimental Test Figure 5. Comparison of the experimental and calculated values

Figure 4 shows the relationship of the flow with respect to
the pressure applied to the extruder for different temperatures
of the extrusion process using pharma grade recycled PET.
At 290 oC, the mass flow variation is not so sensitive to
pressure changes in the extruder. For the case of bottle grade
recycled PET, the mass flow showed a minimum variation at
280 oC regarding the analyzed operating pressures. Finally,
the experimental data were used for validating the mathe-
matical model and Figure 5 shows the experimental mass
flow of the pharma grade recycled PET and the respective
calculated mass flows. The predicted results are very close
to experimental data with a maximum error of 3 %. This
result indicated that the model used to describe the performance
of experimental set up is adequate.
of mass flow.
Conclusion
In this study, we have reported a screw design that
includes an appropriate handling of flows and pressure drop
inside the extruder and adequate design parameters for
recycling PET. This design includes a spinning system,
which is based on current technologies for the development
of special fibers. Results of the mathematical analysis were
used to identify the quantitative impact of the characteristics
of the raw material, pressure, temperature, feeding flow,
extrusion speed and design of the extruder on the performance
552 Fibers and Polymers 2014, Vol.15, No.3
of the extrusion process of recycled PET. The mathematical
prediction of the extruded mass flow is in good agreement
with the experimental data with a maximum error of 3 %.
Finally, the recommended operating conditions for textile
applications and the principal characteristics of raw recycled
material are: VI0: 0.5-0.8, screw speed: 30-45 rpm, temperature:
270-300 oC and humidity: 0.01-0.02 %. These operating
conditions are important to control the variation levels than
the recycled PET normally shows in extrusion process.
References
1. B. Kuczenski and R. Geyer, Resour. Conserv. Recycl., 54,
1161 (2010).
2. F. La Mantia and M. Vinci, Polym. Degrad. Stab., 45, 121
(1994).
3. F. Daw-Ming, S. K. Huang, and L. Jiunn-Yih, J. Appl.
Polym. Sci., 61, 261 (1996).
4. I. A. Abu-Isa, C. B. Jaynes, and J. F. O'Gara, J. Appl.
Polym. Sci., 59, 1957 (1996).
5. B. Saha and A. K. Ghoshal, Chem. Eng. J., 111, 39 (2005).
6. Y. Zou, N. Reddy, and Y. Yang, Compos. Part B-Eng., 42,
763 (2011).
7. F. Fraternali, V. Ciancia, R. Chechile, G. Rizzano, L. Feo,
and L. Incarnato, Compos. Struct., in press, Corrected
proof (2011).
8. D. Foti, Constr. Build. Mater., 25, 1906 (2011).
9. N. Torres, J. J. Robin, and B. Boutevin, Eur. Polym. J., 36,
2075 (2000).
10. A. Oromiehie and A. Mamizadeh, Polym. Int., 53, 728
(2004).
11. J. Martín and M. Rojas, Patent and Trademark Spanish
office (OEPM), 2226512 (2005).
12. K. S. Seo and J. D. Cloyd, J. Appl. Polym. Sci., 42, 845
(1991).
13. A. Cata, G. Bandur, I. Balcu, D. Buzatu, C. Tanasie, and D.
Rosu, Chem. Bull. “POLITEHNICA” Univ. (Timişoara),
52, 143 (2007).
J. C. Tapia-Picazo et al.
14. G. Franceschini and S. Macchietto, Ind. Eng. Chem. Res.,
46, 220 (2007).
15. S. D. Mancini, J. A. S. Schwartzman, A. R. Nogueira, D.
A. Kagohara, and M. Zanin, J. Clean Prod., 18, 92 (2010).
16. E. Pirzadeh, A. Zadhoush, and M. Haghighat, J. Appl.
Polym. Sci., 106, 1544 (2007).
17. H. Saeid, S. Taheri, S. Zadhoush, and A. Mehrabani-
Zeinabad, J. Appl. Polym. Sci., 103, 2304 (2007).
18. F. Samperi, C. Puglisi, R. Alicata, and G. Montaudo,
Polym. Degrad. Stab., 83, 3 (2004).
19. K. Gurudatt, P. de A. K. Rakshit, and M. K. Bardhan, J.
Appl. Polym. Sci., 90, 3536 (2003).
20. M. Abbasi, M. R. M. Mojtahedi, and A. Khosroshahi, J.
Appl. Polym. Sci., 103, 3972 (2007).
21. D. M. Juriga and R. E. Allen, U. S. Patent and Trademark
Office Pre-grant Publication, US20080292831 (2008).
22. J. Brownstein and K. Brownstein, Patent Cooperation
Treaty Application, US20110030557A1 (2011).
23. M. W. Coates, M. H. Kierzkowski, P. J. Gibbons, and A. J.
Eiden, Patent Cooperation Treaty Application, WO10063079
(2010).
24. M. C. Thiry, AATCC Rev., 20 (2009).
25. R. G. Dale, Int. Nonwovens J., 9, 15 (2000).
26. L. Incarnato, P. Scarfato, L. Di Maio, and D. Acierno,
Polymer, 41, 6825 (2000).
27. F. Awaja and D. Pavel, Eur. Polym. J., 41, 1453 (2005).
28. A. García and J. C. Tapia, Master Thesis, Technological
Institute of Aguascalientes, Aguascalientes, México, 2008.
29. S. D. Mancini and M. Zanin, Mater. Res., 2, 33 (1999).
30. G. Wróbel and R. Bagsik, J. Achiev. Mater. Manuf. Eng.,
43, 178 (2010).
31. M. Frounchi, M. Mehrabzadeh, and R. Ghiaee, Iran.
Polym. J., 6, 269 (1997).
32. M. A. Silva-Spinacé and M. A. De Paoli, J. Appl. Polym.
Sci., 80, 20 (2000).
33. J. Lyons, C. Li, and F. Ko, Polymer, 45, 7597 (2004).
34. H. Rajabinejad, R. Khajavi, A. Rashidi, N. Mansouri, and
M. E. Yazdanshenas, Int. J. Envir. Res., 3, 663 (2009).