Professional Documents
Culture Documents
www.elsevier.com/locate/compscitech
Chemical Engineering Department and Stadler Minerva Center for Mesoscopic Macromolecular Engineering,
Ben Gurion University, Beer Sheva 84105, Israel
b
Dead-Sea Bromine Group Ltd, POB 180 Beer Sheva, Israel
Received 25 April 2002; received in revised form 3 March 2003; accepted 11 March 2003
Abstract
Composite materials demand constant improvements in mechanical and ame retardant (FR) properties. The goal of this project
is to study the eect of additives on these properties in polypropylene composites containing glass bers, pentabromobenzyl acrylate (PBBMA) as a primary FR and magnesium hydroxide as a secondary FR. Optimal composition is reached by means of statistical design of experiments (DOE) rather than by trial and error approach. The DOE approach allows minimization of the
number of experiments, investigation of the inuence of each additive and the mutual interactions between additives. It also allows
here, prediction of optimal sample properties better than 8% from experimental values. The optimal composition exhibits improved
mechanical and FR properties. Both FRs reduce the impact strength while enhancing ame retardancy. Glass bers increase the
modulus, but have only a moderate eect on the impact strength due to poor adhesion with PP. The interpretation of the eect of
glass bers on the ammability is inconclusive.
# 2003 Elsevier Ltd. All rights reserved.
Keywords: A. Glass bers; A. Polymer-matrix composites; B. Mechanical properties; C. Statistics; Flame retardancy
1. Introduction
Polypropylene (PP) is a polymer of great importance
in the industrial sector due to its low density, high water
and chemical resistance, ease of processability and being
one of the most cost-eective polymers available today.
It is employed in numerous applications from nonwoven bers to the automotive industry. Its relatively
low price is the driving force to improve its performance, to broaden its versatility and to enable it to
compete with other polymers used nowadays for more
demanding applications. PP as the polymer host
requires the incorporation of additives and reinforcement in order to enhance its mechanical properties. One
of the most commonly used reinforcement agents are
chopped strand silane surface treated glass bers. The
glass bers are surface modied since PP has a nonpolar nature, which hinders its interaction and adhesion
to polar llers such as glass bers. The surface treatment improves the compatibility of the glass with the PP
* Corresponding author. Tel.: +972-8-646-1486; fax: +972-86472916.
E-mail address: moshe@inca.bgu.ac.il (M. Gottlieb).
0266-3538/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0266-3538(03)00170-2
1866
2. Experimental
2.1. Materials
The materials are all technical grade, commercially
available and were used as received. A typical formulation
was prepared by compounding the following materials:
Polymer matrix:
Polypropylene (3120MN1, Appryl, France) d=0.9 g/
cm3, fusion temp=163 C.
Additives:
1. A. PBBMA (FR1025M, DSBG, Israel) primary
FR;
B. Antimony trioxide (ATO). A masterbatch of
80% Sb2O3 and 20% Low Density Poly(ethylene), characteristic diameter 1.5 mm
(L0112 Kafrit, Israel) is added as FR synergist. For optimal FR activity, it is added in
1:3 ratio of ATO:PBBMA.
1867
1868
Table 2
Initial formulations
Component
Formulation No.
Units
Upper limit
[+] (% w/w)
Lower limit
[] (% w/w)
Glass bers
PBBMA+ATO (3:1)
MAH
Irganox
Mg(OH)2
25
20
2
0.3
25
15
13.3
1
0
14.15
PP
Glass bers
PBBMA
ATO
MAH
PE Wax
Irganox
Mg(OH)2
%
%
%
%
%
%
%
%
A1
38.5
20.0
14.0
4.6
1.0
0.2
0.3
21.4
A2
A3
A4
40.0
20.0
14.0
4.6
1.0
0.2
0.3
19.9
42.0
20.0
14.0
4.6
1.0
0.2
0.3
17.9
44.0
20.0
14.0
4.6
1.0
0.2
0.3
15.9
V0
NRa
10
19
49
86
No
No
25.025.5 25.025.5
NRa
55
178
No
25.025.5
NRa
50
242
No
24.024.5
q
q
X
X
bi xi bij xi xj
i1
i<j
1869
1870
Table 3
Experimental matrix for the mixture design
Component
Formulation No.
Units
PP
Fibers
PBBMA
ATO
MAH
PE Wax
Irganox
Mg(OH)2
%
%
%
%
%
%
%
%
Units
PP
Fibers
PBBMA
ATO
MAH
PE Wax
Irganox
Mg(OH)2
%
%
%
%
%
%
%
%
10
38.50
25.00
10.00
3.30
1.50
0.20
0
21.50
38.50
21.68
12.50
4.15
1.00
0.20
0.30
21.67
38.50
25.00
12.50
4.15
1.00
0.20
0.15
18.50
38.50
17.78
13.90
4.64
1.77
0.20
0.22
22.98
38.50
15.00
14.63
4.87
2.00
0.20
0.30
24.50
38.50
25.00
15.00
5.00
1.50
0.20
0
14.80
38.50
15.00
14.74
4.91
2.00
0.20
0
24.65
38.50
23.35
10.00
3.30
1.00
0.20
0.30
23.35
38.50
15.00
14.74
4.91
2.00
0.20
0
24.65
38.50
22.85
10.00
3.30
2.00
0.20
0.30
22.85
11
12
13
14
15
16
17
18
19
20
38.50
20.00
15.00
5.00
1.00
0.20
0.30
20.00
38.50
20.85
10.00
3.30
2.00
0.20
0.15
25.00
38.50
25.00
12.50
4.15
2.00
0.20
0
17.65
38.50
18.65
12.50
4.15
1.00
0.20
0
25.00
38.50
25.00
15.00
5.00
1.50
0.20
0.30
14.50
38.50
18.10
12.30
4.10
1.50
0.20
0.30
25.00
38.50
25.00
15.00
5.00
1.50
0.20
0.30
14.50
38.50
15.15
15.00
5.00
1.00
0.20
0.15
25.00
38.50
20.00
15.00
5.00
2.00
0.20
0.15
19.15
38.50
22.85
10.00
3.30
2.00
0.20
0.30
22.85
Table 4
Flame retardancy results of the mixture design matrix
Formulation No.
Units
UL-94 Rating (3.2 mm)
Max. aming time
Total aming time
Cotton ignition
UL-94 Rating (1.6 mm)
Max. aming time
Total aming time
Cotton ignition
LOI
V1
s 26
s 143
No
NRa
s 32
s 227
Yes
% 24.024.5
Units
10
V0
9
27
No
V1
19
71
No
25.526.0
V1
19
83
No
NRa
40
142
Yes
25.025.5
V0
6
9
No
V0
5
25
No
26.026.5
V1
23
64
No
V2
26
85
Yes
26.527.0
V0
2
4
No
V0
4
12
No
25.025.5
V0
0
0
No
V0
5
13
No
26.026.5
NRa
48
182
Yes
NRa
55
266
Yes
24.525.0
V0
0
0
No
V0
5
15
No
25.025.5
NRa
32
120
No
NRa
39
161
Yes
24.525.0
12
13
14
15
16
17
18
19
20
NRa
40
134
No
NRa
53
241
Yes
24.024.5
V1
13
41
No
V2
20
100
Yes
24.024.5
V0
1
1
No
V0
2
5
No
25.526.0
V1
13
44
No
V1
19
102
No
25.025.5
V1
12
42
No
NRa
47
128
No
25.025.5
V1
24
65
No
NRa
31
86
No
25.526.0
V0
4
6
No
V0
3
8
No
26.026.5
V0
10
45
No
V1
10
54
No
26.026.5
NRa
60
257
Yes
NRa
45
222
Yes
24.024.5
11
V0
7
30
No
V0
s
9
s 48
No
% 25.526.0
s
s
1871
Modulus
Max. strength
Impact
Modulus
Max. Strength
Impact
Units
10
MPa
MPa
J/m
J/m
6297
39.2
78.7
2.4
6288
40.9
87.8
5.4
6488
41.7
101.5
2.5
4777
37.6
100.0
8.9
4874
37.9
103.4
2.8
6521
42.3
98.7
4.2
4508
36.4
90.6
9.1
6396
41.5
97.5
8.4
4245
36.2
89.4
2.8
5690
39.6
90.3
6.8
Units
11
12
13
14
15
16
17
18
19
20
MPa
MPa
J/m
J/m
5683
40.4
95.0
7.2
5373
37.6
92.5
10.2
5803
39.7
82.2
5.9
5082
36.6
83.7
5.6
6032
40.4
101.2
10.0
4890
36.7
93.1
4.6
6021
40.3
93.1
4.6
4354
36.7
95.6
6.5
5047
38.2
85.6
5.2
5444
39.1
96.9
4.4
1872
1873
1874
PP
Fibers
PBBMA
ATO
MAH
PE Wax
Irganox
Mg(OH)2
UL-94 Rating (3.2mm)
Max. aming time
Total aming time
Cotton ignition
UL-94 Rating (1.6mm)
Max. aming time
Total aming time
Cotton ignition
LOI
Modulus
Max. strength
Impact
Units
B1
B2
B3
%
%
%
%
%
%
%
%
38.50
22.31
15.00
5.00
1.00
0.20
0
17.99
38.50
24.10
15.00
5.00
1.30
0.20
0
15.90
38.50
24.77
15.00
5.00
1.41
0.20
0
15.12
s
s
%
V0
0
0
No
V0
1
1
No
V0
1
2
No
V0
1
6
No
25.526.0
V0
2
4
No
25.526.0
V0
2
3
No
25.025.5
MPa 6953
MPa
41.8
J/m
90.6
J/m
3.8
Table 7
Comparison between model prediction and experimental results of
formulation B1
Response
Modulus
Predicted results
Experimental results
Mean
A 95%
condence
interval
Experimental
value
A 95%
condence
interval
5908
55786237
6953 (av.)
64707436
Impact
90.0
82.797.2
90.6 (av.)
8794.2
Max. strength
42.3
40.344.4
41.8 (av.)
40.942.8
(48) 55
(68)a25
24.925.9
25.5 - 26
Total aming
time (1.6 mm)
Formulation No.
s
s
6834
42.1
91.9
4.7
7203
41.6
89.4
5.7
Total aming
time (3.2 mm)
7
(22)a
LOI
a
25.4
Table 8
Properties of the optimal formulations
Initial
formulation
A1
B1
1033
90 4
99 4
1009
Modulus
6333
6953
6521
4777
UL94
(1.6 mm)
T=109 V1
T=6 V0
T=12 V0
T=25 V0
UL94
(3.2 mm)
T=49 V0
T=0 V0
T=4 V0
T=9 V0
LOI (%)
25.025.5
25.526.0
25.025.5
26.026.5
Impact
high impact and V-0 rating (i.e. low total aming time).
The comparison of the latter formulation with the initial
formulation demonstrates the compromise character of
the process optimization: by accepting a slight decrease
in impact value a dramatic improvement in ame retardancy is obtained. It is instructive to examine in detail the
two additional compositions selected as superior formulations. Both of them are found to be inferior to (B1)
the best predicted formulation. Total aming time (1.6
mm) in these formulations is 12 and 25 s, respectively, as
opposed to 6 s for the optimal formulation. However, if
we take into account the high value of the standard error
of the predicted mean for this response (22 s), we are
obliged to reject these formulations, since the total aming
time in UL94V=(V-0) is not allowed to exceed 50 s.
Indeed, these two formulations fail the UL94V test
regarding V-0 at the =0.05 condence level, if we take
the total aming time as a sum of the predicted mean (12
or 25 s) and twice the standard error (222=44 s).
5. Conclusions
The technique of statistical experimental design has
been employed to obtain an optimal formulation of
high loading polypropylene composite with suitable
mechanical and ame retardancy properties. The DOE
technique helps optimize a set of ame retardant additives, which guarantee acceptable ame retardant properties with minimal negative eect on mechanical
properties. Moreover, it allows the study of the inuence of individual additives and the mutual interactions
in the composite within the chosen concentration
working range. DOE provides a route to obtain the
optimal formulations, and to get the abovementioned
information with the minimum number of experiments.
An optimization of dierent additive loads was performed and the nal polypropylene composite exhibited
improved mechanical and ame retardant properties. The
DOE scheme allows a prediction of most sample properties within 8% from the actual experimental values. In a
typical PP composite under investigation, we observed
that maleic anhydride lowers the modulus and impact
strength. This was attributed to the failure of maleic
anhydride to enhance adhesion at the interface between
PP and the dierent llers and its eect was reduced to a
plastizing eect in the bulk. As expected, PBBMA and
Mg(OH)2 decrease the composite impact strength and
enhance ame retardant properties. Glass bers increase
the modulus as a reinforcing agent, but have only a
moderate eect on impact strength within the investigated range, probably due to poor adhesion with PP.
Contradictory results have been obtained for the eect of
glass bers on ame retardant properties. Dierent
explanations for these results have been suggested, but
no denitive conclusion is possible at this point.
1875
Acknowledgements
The Ariel program for French-Israeli cooperation has
nancially supported this work. MG acknowledges the
nancial support of the German-Israel Foundation and
the Israel Basic Science Foundation. We are grateful to
the sta of the Plastics Application Laboratory at
DSBG for their help in carrying out the mechanical and
ammability tests.
References
[1] Berger ES, Petty EH. Organofunctional silanes. In: Katz HS,
Milewski JV, editors. Handbook of llers for plastics. New York:
Van Nostrand Reinhold Company; 1987.
[2] Muskatel M, Utevski L, Shenker M, Daren S, Peled M, Charit Y.
Flame retardant polypropylene bers with good dyeability. J App
Polym Sci 1997;64(3):6016.
[3] Dvir H, Gottlieb M, Daren S. Thermal polymerization of a
brominated ame retardant in a glass ber reinforced polypropylenequantitative analysis. J App Polym Sci 2003;88(6):1506
15.
[4] Dvir H, Goldraich M, Gottlieb M, Daren S, Lopez Cuesta J.
Distribution of a brominated acrylate ame retardant in polypropylene. Polymer Degradation and Stability 2001;74:46574.
[5] Dvir H, Goldraich M, Gottlieb M, Daren S. Imaging the distribution of a brominated acrylate FR in polypropylene. In:
Conference ProceedingsConference on Polymer Modication,
Degradation and Stabilization, Palermo, Italy, September 2000.
Available: http//modest.unipa.it/conferences/2000/html/symp2/
O_2_Tu_1230.pdf.
[6] Montezin F, Lopez-Cuesta JM, Gaudon P, Crespy A. Flame
retardant and mechanical properties of a copolymer PP/PE containing brominated compounds/antimony trioxide blends and
magnesium hydroxide or talc. In: Conference Proceeding
Eurollers 97, Manchester, UK, 1997, p. 399402.
[7] Touval I. antimony oxide. In: Katz HS, Milewski JV, editors.
Handbook of llers for plastics. New York: Van Nostrand Reinhold Company; 1987.
[8] Cornell JA. experiments with mixtures. New York: John Wiley &
Sons; 1990.
[9] Bowles ML, Montgomery DC. How to formulate the ultimate
margarita: a tutorial on experiments with mixtures. Quality
Engineering 19971998; 10(2): 239253.
[10] Constable RC, Humenix JA, Adur AM. Glass ber reinforced
polypropylene- performance enhancement with acid modied polypropylene. In: Conference ProceedingANTEC 89, 13891393.
[11] Karian HG, editor. Handbook of polypropylene and polypropylene composites; 1999.
[12] Mareri P, Bastide S, Binda N, Crespy A. Mechanical behaviour
of polypropylene composites containing ne mineral ller: eect
of ller surface treatment. Composites Science and Technology
1998;58(5):74752.
[13] Karger-Kocsis J, Harmia T, Czigany T. Comparison of the fracture and failure behavior of polypropylene composites reinforced
by long glass bers and by glass mats. Composites Science and
Technology 1995;54(3):28798.
[14] Troitzsch J. International plastics ammability handbook. New
York: Hanser Publishers; 1990.
[15] Hornsby PR, Mthupha A. Analysis of re retardancy in magnesium
hydroxide lled polypropylene composites. Plastics and Rubber
and composites processing and Applications 1996;25(7):34755.