lable at ScienceDirect

Polymer Degradation and Stability 96 (2011) 363e370

Contents lists avai

Polymer Degradation and Stability

journal homepage: www.elsevier .com/locate /polydegstab

An efficiently halogen-free flame-retardant long-glass-fiber-reinforcedpolypropylene system

Yun Liu, Cheng-Liang Deng, Jing Zhao, Jun-Sheng Wang, Li Chen, Yu-Zhong Wang*

Center for Degradable and Flame-Retardant Polymeric Materials (ERCEPM-MoE), College of Chemistry, State Key Laboratory of Polymer Materials Engineering,Sichuan University, Chengdu 610064, China

a r t i c l e i n f o

Article history:Received 26 January 2010Received in revised form20 February 2010Accepted 26 February 2010Available online 6 March 2010

Keywords:Flame retardanceIntumescenceCandlewick effectLong-glass-fiber-reinforcedPolypropyleneMechanical property

* Corresponding author. Tel./fax: þ86 28 85410259E-mail address: yzwang@scu.edu.cn (Y.-Z. Wang).

0141-3910/$ e see front matter � 2010 Elsevier Ltd.doi:10.1016/j.polymdegradstab.2010.02.033

a b s t r a c t

In order to solve the “candlewick effect” caused by glass fibers, which results in the decrease of flameretardancy of flame-retardant long-glass-fiber-reinforced polypropylene (LGFPP) systems, and thedeterioration of mechanical properties caused by adding an additional amount of flame retardantscompared with flame-retardant non-glass-fiber-reinforced polypropylene systems so as to keep a sameflame retardancy, a novel intumescent flame retardant (IFR) system, which is composed of a charringagent (CA), ammonium polyphosphate (APP) and organically-modified montmorillonite (OMMT), wasused to flame retard LGFPP. The thermal stability, combustion behavior, char formation, flame retardantmechanism and mechanical properties of the IFR-LGFPP samples were investigated by thermogravi-metric analysis (TGA), limiting oxygen index (LOI), UL-94 test, cone calorimeter test, scanning electronicmicroscopy, and mechanical property tests. When the content of IFR is 20 wt%, the LOI value of IFR-LGFPPreaches 31.3, and the vertical burning test reaches UL-94 V-0 rating, solving the “candlewick effect”caused by long glass fiber without additional amount of the IFR. All the relevant cone calorimeterparameters also show that IFR-LGFPP has much better flame-retardant behaviors than LGFPP. Further-more, the mechanical properties of IFR-LGFPP almost remain unchanged in comparison with those ofLGFPP containing no IFR. The flame retardant mechanism was also discussed.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Very recently, there has been a rapid growth in the developmentof long-glass-fiber-reinforced thermoplastic composites and intheir end-use applications, especially in automotive and trans-portation industry [1e3]. Long-glass-fiber-reinforced poly-propylene (LGFPP) has aroused more research activities due to itsadvantages such as its higher strength, stiffness, impact propertiesand recyclability, etc. [1,3], and it has been more and more utilizedas engineering thermoplastics. However, the “candlewick effect”caused by glass fibers generally makes the flame retardation of thethermoplastics become a big challenge [4e8]. Due to the “candle-wick effect”, glass fibers are able to transfer and feed back the fuelfrom the pyrolysis of the polymer matrices to the flame by capillaryaction, speed the heat flowing back to polymers and thus make thepolymers decompose and burn faster. Thus, to achieve UL-94 V-0rating, the glass fiber reinforced thermoplastics need more amountof flame retardants than neat polymers do. However, a largeamount of flame retardants added to polymers will finally affect the

.

All rights reserved.

other properties of polymers, deteriorating largely the mechanicalproperties almost without expectation [9e12].

In this paper, in order to solve the “candlewick effect” caused byglass fibers, which results in the large decrease of flame retardancyof flame-retardant long-glass-fiber-reinforced polypropylene(LGFPP) systems, and the deterioration in mechanical propertiescaused by adding a large amount of flame retardants for enhancingthe flame retardancy of LGFPP, we used a novel intumescent flameretardant (IFR) system, which is composed of a charring agent (CA),ammonium polyphosphate (APP) and organically-modified mont-morillonite (OMMT), and the chemical structure of CA is shown inScheme 1. The flame retardancy, burning behavior, thermalstability, flame retardant mechanism and mechanical properties ofthe LGFPP-IFR systems were investigated systematically.

2. Experimental

2.1. Materials

The charring agent (CA) was supplied by Weili Flame RetardantChemicals Co. (Chengdu, China). Commercial polypropylene (PP,S1040) was provided by Shanghai Secco Petrochemical Co., Ltd

N N

N

NH2

* NHCH2CH2NHCH2CH2NH *

n

Scheme 1. The chemical structure of CA.

Y. Liu et al. / Polymer Degradation and Stability 96 (2011) 363e370364

(Shanghai, China). Long glass fiber reinforced polypropylene(LGFPP) was supplied by Jiangsu Angete New Materials TechnologyCo., Ltd (Changzhou, China). Ammonium polyphosphate (APP) wassupplied by Changfeng Chemical Co. (Shifang, China). The OMMT,I.44P, was purchased from Nanocor.

Fig. 2. The DTG curves of the IFR-LGFPP samples in air.

2.2. Preparation of flame-retardant PP samples

Flame retardant masterbaches were prepared by mixing 20 wt%of PP with 80 wt% of intumescent flame retardant (IFR) consisted ofAPP, CA and OMMT IFR using a twin-screw extruder (D: 20.5 mm,L/D: 44, model: CTE 20, Kebeilong Keya Nanjing Machinery Co., Ltd,Nanjing, China) at a temperature profile of 170, 190, 195, 200, 195,185 �C. The extruded strands were cut into pellets. Then theintumescent flame-retardant long-glass-fiber-reinforced poly-propylene (IFR-LGFPP) composites were prepared by blending theIFR-PP masterbatches and LGFPP. The renormalized content of IFRwas kept at 25, 22, 20 and 18 wt% in IFR-LGFPP and the content oflong glass fiber was kept as a constant, 20 wt%. The compositeswere injected into standard testing bars for the tests of combusti-bility and mechanical properties.

2.3. Measurements

The LOI values were surveyed on an HC-2C oxygen index meter(Jiangning, China) with sheet dimensions of 130 mm � 6.5 mm �3.2 mm according to ASTM D2863-97.

Vertical burning tests were conducted on a vertical burning testinstrument (CZF-2-type) (Jiangning, China) with sheet dimensionsof 130 mm � 13 mm � 3.2 mm according to ASTM D3801.

Scanning electronic microscopy (SEM) observed on a JEOL JSM-5900LV was used to investigate the surface of char residues of the

Fig. 1. The TG curves of the IFR-LGFPP samples in air.

LGFPP/20% IFR after the cone calorimeter test. SEM graphs of thechar residues were recorded after gold coating surface treatment.

Thermogravimetric analysis (TGA) was performed on a TG 209F1 (NETZSCH, Germany) thermogravimetric analyzer at a heatingrate of 10 �C/min. 5e10mg of the samplewas examined under air ata flowing rate of 60 mL/min at temperatures ranging from 40 to700 �C, respectively.

The cone calorimeter tests were carried out by an FTT conecalorimeter, following the procedures in ISO 5660-1. Square spec-imens (100mm� 100mm� 3mm)were irradiated at a heat flux of50 kW/m2.

Tensile tests were completed in accordance with the proceduresin GB/T 1040-1992 at a crosshead speed of 50 mm/min. Flexuralproperties were carried out in accordance with the procedures inGB/T 9314-2000 at a crosshead speed of 2 mm/min and a spanwidth of 64 mm. The Izod impact properties were tested in accor-dancewith the procedures in GB/T 1843-1996 and the depth of nickis 2 mm.

3. Results and discussion

3.1. Thermal decomposition behaviors

In order to understand the effects of flame retardants on IFR-LGFPP, the thermal degradation behavior and the amount of residuechar obtained from TGA test for the various samples werecompared in air. TG and DTG curves are presented in Figs. 1 and 2,respectively, and some detailed data are given in Table 1. NeatLGFPP starts to decompose at 273 �C in air, and is exhausted atabout 500 �C with only the long glass fiber left. Unlike the neatLGFPP, three IFR-LGFPP samples LGFPP/22% IFR, LGFPP/20% IFR andLGFPP/18% IFR are not decomposed completely, leading to the

Table 1The main data of the samples from TG and DTG in air.

Samples LGFPP(%)

IFR(%)

Tonset(�C)

Tmax

(�C)The rateof Tmax

(%/min)

The char residues (%)

500 �C 600 �C 700 �C

LGFPP 20 0 273 356 13.22 21.84 21.36 21.28LGFPP/22% IFR 20 22 270 332 4.37 44.29 38.38 34.92LGFPP/20% IFR 20 20 280 331 5.50 39.70 32.92 29.79LGFPP/18% IFR 20 18 276 337 5.46 34.21 27.72 24.96

Table 2The LOI values and UL-94 ratings of samples with different content of IFR underroom temperature.

IFR (%) LOI (%) UL-94

0 18.3 N.R.18 27.1 N.R.20 31.3 V-022 31.7 V-025 35.1 V-0

Fig. 4. The heat release rate (HRR) as a function of time.

Y. Liu et al. / Polymer Degradation and Stability 96 (2011) 363e370 365

formation of carbonaceous residues that are thermally stable athigh temperature zone. The IFR-LGFPP samples have similar TGcurves under the air condition below 350 �C, but the residue charsincrease with the additive amount of IFR above 350 �C, up to 34.92,29.79, and 24.96% of residue chars left at 700 �C, respectively. FromFigs. 1, 2 and Table 1, it can be seen that the addition of IFRsincreases little the onset decomposition temperatures of the IFR-LGFPP samples except the LGFPP/22% IFR sample, and the onsetdecomposition temperature, Tonset, is defined as the temperature atwhich 5% weight loss occurs. However, it can be observed that themaximum decomposition-rate temperatures, Tmaxs, of the IFR-LGFPP samples are about 25 �C lower than that of neat LGFPP, whichcould be primarily attributed to the decomposition of IFRs. As canbe seen from Fig. 2 and Table 1, the addition of IFRs makes themaximum decomposition rate of IFR-LGFPP samples decreaselargely, and the maximum decomposition rates of LGFPP/22% IFR,LGFPP/20% IFR and LGFPP/18% IFR are 4.37, 5.50 and 5.46%/min,respectively.

3.2. Flame retardancy and burning behaviors

3.2.1. LOI and UL-94 testsTo investigate the flame retardancy of IFR-LGFPP samples, the

LOI values and vertical burning ratings (UL-94) of IFR-LGFPPsamples with different contents of IFR were tested under roomtemperature. The LOI values and UL-94 ratings are given in Fig. 3and Table 2. From Fig. 3, it can be seen that the LGFPP is an easilyignited thermoplastic with a low LOI value (18.3) and no rating inUL-94 test, and that its LOI value markedly increases while 18 wt%of IFR is added, which can reach 27.1. However, the UL-94 rating stillremains the same (no rating). While the content of IFR increases to20 wt%, the LOI value and the UL-94 rating of IFR-LGFPP reach 31.3

Fig. 3. The curve of LOI changes with different content of IFR under room temperature.

and V-0, respectively. As can be seen from Fig. 3, the LOI valuesincrease with the increase of IFR content, and the LOI value reaches35.1 when the content of IFR is 25 wt%, showing an obvious positiveeffect on LOI values. A small amount of flame retardant is enough toimprove the flame retardancy of LGFPP after the content of IFRreaches 18 wt%. In the UL-94 test, all flame retardant samples haveno dripping phenomenon, however, the neat LGFPP burns up to theholding clamp from the ignition point until the sample burns outcompletely. The addition of IFR decreases the burning time ofLGFPP largely, and UL-94 rating is improved to V-0 from no ratingwhen the content of IFR reaches 20 wt%, meaning that the burningtime of LGFPP/20% IFR is shorter than 10 s in each test, respectively.

From the above results, it can be concluded that the IFR studiedhas a significant and beneficial flame retardant effect on LGFPP, andit does not need to increase an additional amount of IFR to reach theUL-94 V-0 rating of IFR-LGFPP comparing with the correspondingnon-glass-fiber systems. Thus, the “candlewick effect” caused bythe long glass fiber is solved effectively.

3.2.2. Cone calorimetric analysesThe cone calorimeter has been an efficient tool for the evalua-

tion of flammability, and has been used in quantitative materialcombustion analysis [13e17]. In order to investigate the effects ofadditive amount of IFR on the combustion behaviors and flameretardancy of the IFR-LGFPP, the cone calorimeter has been used inthis study, and a heat flux of 50 kW/m2 was used.

3.2.2.1. Heat release rate (HRR). The heat release rate (HRR) isbelieved to have the greatest influence on the fire hazard. Themeasured heat release rate curves for each sample are shown inFig. 4 for comparison. Detailed combustion parameters have beensummarized in Table 3. It can be seen from Fig. 4 that the neat

Table 3Combustion parameters obtained from cone calorimeter at a heat flux of 50 kW/m2.

Sample TTI (s) PHRR(kW/m2)

Average HRRa

(kW/m2)Mass (%) THR

(MJ/m2)

LGFPP 28 605.43 204.36 20.5 130.91LGFPP/22% IFR 14 155.36 121.93 39.0 103.65LGFPP/20% IFR 16 196.71 113.77 36.5 106.46LGFPP/18% IFR 15 183.04 115.13 38.8 101.93

a Average HRR: the average value between time to ignition and end of test,obtained with ISO 5660-1: 2002 standard.

Fig. 5. The mass loss as a function of time.Fig. 6. The total heat release (THR) as a function of time.

Y. Liu et al. / Polymer Degradation and Stability 96 (2011) 363e370366

LGFPP resin burns fast after ignition and two sharp peaks heatrelease rate (PHRR) appear, reaching 404.34 and 605.43 kW/m2,respectively. The lower peak should attribute to the initiativeignition and combustion of the outermost PP matrix from thereinforcing fibers, while the second one is due to the combustionof the whole PP resin. Also, the time to ignition (TTI) of LGFPP is28 s, which is much shorter than that of PP [18]. There are two

Fig. 7. The digital photographs of the residue char after cone calorimeter tes

main reasons for making the LGFPPP more flammable: (1) longglass fibers have a larger heat conduction coefficient than PP does,and they can transmit heat to the LGFPP below the burning zoneeasily; (2) the alignment and contact of the additive long glassfibers can form a continuous mass path and thus speed theflammable mass to the burning area [4]. However, in the case ofIFR-LGFPP, the values of HRR caused by the outer PP around the

t. a: neat LGFPP; b: LGFPP/22% IFR; c: LGFPP/20% IFR; d: LGFPP/18% IFR.

Y. Liu et al. / Polymer Degradation and Stability 96 (2011) 363e370 367

long glass fibers decrease at least half, but a much longer burningtime appears and the second peak of HRR almost vanishes, whichbecomes a little lower than the first peak value and muchdifferent from the combustion behavior of the original LGFPP.Also, the TTIs of the IFR-LGFPP samples with different IFR contentsare much shorter than that of LGFPP, which is mainly caused bythe initial combustion of the intumescent flame retardants beforethese could play their role in the composites [19]. It can beobserved from Fig. 4 that the HRR curves of LGFPP/22% IFR, LGFPP/20% IFR and LGFPP/18% IFR are similar with each other, especiallythe curves of LGFPP/20% IFR and LGFPP/18% IFR, and the PHRRvalue of LGFPP/22% IFR is a little lower than those of the others,about 155.36 kW/m2, however, the value is much lower than thatof LGFPP, 605.43 kW/m2.

The heat radiation on the surface results in degradation of theLGFPP resin, leading to its ignition. Further penetration of the heatbelow the long glass fiber causes degradation of the underlyingresin. The degradation products transmit to the burning zonesthrough the long glass fibers, and a few chars retain in the LGFPP.This process goes on until all resin is burnt. However, if the charformation is enhanced and can act as a thermal barrier, it can slowdown this migration, resulting in slowing down or even stoppingburning [20]. The addition of IFR increases the amount of residuechar, which can slow down the heat migration to the underlyingmatrix and protect the PP matrix; meanwhile, the residue charformed slows down or even stops the burning of PP, making the

Fig. 8. Scanning electron micrographs of the out

decrease in HRR and the improvement in flame retardancy of IFR-LGFPP samples.

3.2.2.2. Mass loss (ML). Fig. 5 shows mass loss as a function ofcombustion time for the samples. At the end of burning, there are20.5, 39.0, 36.5 and 38.8% of residue chars left corresponding to theneat LGFPP, LGFPP/22% IFR, LGFPP/20% IFR and LGFPP/18% IFR,respectively. From Fig. 5, it can be seen that the addition of IFRenhances the residue char of the IFR-LGFPP significantly, and themass curves of the IFR-LGFPP samples are also similar with eachother, which is in accordance with the HRR curves. It is observedthat the residue char amount of LGFPP/22% IFR is a little higher thanthose of LGFPP/20% IFR and LGFPP/18% IFR at the same burningtime, and the residue char amount of LGFPP/20% IFR and LGFPP/18%IFR is similar.

3.2.2.3. Total heat release (THR). The samples studied showconsiderable differences in the total heat release (THR) curves aspresented in Fig. 6. The reduction of HRR resulted in a corre-sponding reduction in THR. At the end of burning, neat LGFPP hasreleased a total heat of 130.91 MJ/m2, while LGFPP/22% IFR, LGFPP/20% IFR and LGFPP/18% IFR have released 103.65, 106.46 and101.93 MJ/m2, respectively. As can be seen from Fig. 6, THR value ofLGFPP/22% IFR is a little lower than those of LGFPP/20% IFR andLGFPP/18% IFR at the same burning time, which is in accordancewith the HRR curves andmass loss curves. However, the decrease in

er surface of LGFPP/20% IFR after cone test.

Fig. 9. Scanning electron micrographs of the inner surface of LGFPP/20% IFR after cone test.

Y. Liu et al. / Polymer Degradation and Stability 96 (2011) 363e370368

HRR and THR of IFR-LGFPP is much lower than those of IFR-PP withor without the presence of OMMT [18], which is mainly caused bythe long glass fibers filled. The long glass fibers filled destroy thecontinuity of the char, and it can be seen obviously from Fig. 7. Fig. 7is the digital photographs of the residue chars after cone

Fig. 10. The experimental and theoretical TG curves of APP/CA in air.

calorimeter test. It can be observed that the LGFPP is burnt outexpect the long glass fibers left, on a contrary, there are residuechars left in the IFR-LGFPP samples; however, the long glass fibersfilled destroy the intumescence of the intumescent flame retar-dants and the continuity of the residue char, which decrease the

Fig. 11. The experimental and theoretical TG curves of APP/CA/OMMT in air.

Fig. 12. The experimental and theoretical TG curves of IFR-PP masterbatch in air.

Y. Liu et al. / Polymer Degradation and Stability 96 (2011) 363e370 369

flame retardancy of IFR in IFR-LGFPP samples. The detailedmorphologies of the residue chars will be discussed thereinafter viathe results of scanning electron microscope.

3.2.3. The morphology of residue chars after the cone testScanning electron micrographs of the outer surface and inner

surface of LGFPP/20% IFR after cone test are presented in Figs. 8 and9, respectively. It can be seen obviously from Figs. 8 and 9 that theaddition of long glass fibers destroys the continuity of the intu-mescent residue chars, and the residue chars formed are inserted tothe surface of the long glass fibers in the inner surface of LGFPP/20%

Fig. 13. The mechanical properties of the neat LGFPP and IFR-LGFPP samples, Note: a, n

IFR. As can be seen in Figs. 8 and 9, the amount of long glass fibers inouter surface of the composite is much more than that in innersurface, indicating that the long glass fibers transfer to the surfaceof the IFR-LGFPP composites, which makes the combustion worsebecause of their larger heat conduction coefficient than PP. It is alsoobserved that the final length of long glass fibers after injectionmolding is above 1 mm, which decides the mechanical propertiesof the IFR-LGFPP composites [21].

3.3. The flame retardant mechanism of IFR

Figs. 10e12 show the experimental and theoretical TGA curvesof APP/CA, APP/CA/OMMT and IFR-PP masterbatch, respectively.The theoretical data are a linear combination of the TGA curves ofthe individual components of the mixture, thus the theoretical dataare representative of a non-interacting phenomenon.

The thermal decomposition behavior is different from theexpectation based on the decomposition behavior of the APP/CAcomposite, comparing the experimental with theoretical thermo-grams. As can be seen from Fig. 10, the amount of residue chars inAPP/CA composites generated at temperatures higher than about370 �C is much higher than that of their independent thermaldecomposition, which is the important evidence that APP and CAinteract with each other in the high temperature zone. Further-more, the addition of OMMT in APP/CA/OMMT composites makesthe difference in the experimental and theoretical curves moreobvious (see Fig. 11), indicating that the additive OMMT couldcatalyze the interaction between APP and CA andmake the amountof residue chars left more in the high temperature zone. Theamount of residue chars at high temperatures is related to the flameretardancy of the intumescent flame retardant, so the addition ofOMMT increases the flame retardancy of APP/CA composites.

eat LGFPP; b, LGFPP/25% IFR; c, LGFPP/22% IFR; d, LGFPP/20% IFR; e, LGFPP/18% IFR.

Y. Liu et al. / Polymer Degradation and Stability 96 (2011) 363e370370

It can be observed from Fig. 12 that the experimental and theo-retical TG curves of IFR-PP masterbatch in air are different from eachother. The experimental amount of residue chars of IFR-PP master-batch is lower than the theoretical of IFR-PP masterbatch in thetemperature zone from 165 �C to 304 �C, and the reason for thisphenomenon is that APP and CA interact with each other in theprocess of extrusion, releasing NH3 and H2O, which results in themass loss and can also be proved by the smell of NH3 recognizedduring the extrusion. From 304 �C to 412 �C, the experimentalamount of residue chars of IFR-PP masterbatch is higher than thetheoretical of IFR-PPmasterbatch, which is attributed to the compactchar formed from the interaction between APP and CA. However, theexperimental decomposing rate of residue chars of IFR-PP is fasterthan the theoretical of IFR-PP masterbatch at the temperaturesranging from 412 �C to 520 �C, which results in the similar amount ofresidue chars for the experimental and the theoretical. In the highertemperature zone above 520 �C, the experimental amount of residuechars of IFR-PP masterbatch is much higher than the theoretical,suggesting that the components of IFR-PP masterbatch interact witheach other, making the more residue chars left in the high temper-ature zone.

It can be seen from Figs. 10e12 that the OMMT added couldcatalyze the interaction of APP and CA, make the residue char stableand leave more at high temperatures. The stable residue chars canprotect the polymer matrix, prevent the mass/heat transfer effec-tively, and improve the flame retardancy of APP/CA composites.

3.4. Mechanical properties

The mechanical properties of the neat LGFPP and IFR-LGFPPsamples are presented in Fig. 13, including tensile properties,flexural properties and notched Izod impact strength. It can be seenfrom Fig. 13 that the addition of IFRs decreases slightly the tensilestrength of the IFR-LGFPP composites, and increases the elongationat break of the composites. With the increase of the IFR content inIFR-LGFPP composites, the tensile strength of IFR-LGFPP compos-ites gradually increases, and the flexural strength and notched Izodimpact strength of IFR-LGFPP composites decrease. The tensilestrength, elongation at break, flexural strength and notched Izodimpact strength of LGFPP/20% IFR composite are 76.28 MPa, 8.73%,99.92 MPa and 14.11 kJ/m2, respectively, and the mechanicalproperties decrease by 6.97% for tensile strength, 3.00% for flexuralstrength and 12.09% for Notched Izod impact strength comparedwith the neat LGFPP. From the above analysis, it is concluded thatthe additive amount of 20 wt% is suitable according to the flameretardancy and mechanical properties of the composite comparedwith the neat LGFPP.

4. Conclusion

The problems of “candle effect” and deterioration in mechanicalproperties for long-glass-fiber-reinforced polypropylene systems(LGFPP) are effectively solved by the used IFR, which is composed ofAPP, CA and OMMT. When the content of IFR increases to 20 wt%,without an additionally increased amount of the IFR comparedwith PP systems containing no glass fibers, the LOI value of the IFR-LGFPP reaches 31.3, and the UL-94 rating is also enhanced to V-0;the results of the cone calorimeter indicate that the addition of IFRsignificantly decreases the HRR and THR of the IFR-LGFPP. Theflame retardant mechanism suggests that APP and CA can interact

at high temperatures, and OMMT can catalyze this interaction,making more residue chars leave. Furthermore, the mechanicalproperties of IFR-LGFPP almost keep the same as the system withno flame retardants, only a little decreasing. All the results indicatethat the IFR improves largely the flame retardancy of IFR-LGFPPsamples.

Acknowledgements

The authors wish to acknowledge the financial support of theKey Project of the National Science Foundation of China(50933005).

References

[1] Kumar KS, Ghosh AK, Bhatnagar N. Mechanical properties of injection moldedlong fiber polypropylene composites, part 1: tensile and flexural properties.Polym Compos 2007;28:259e66.

[2] Thattaiparthasarathy KB, Pillay S, Ning H, Vaidya UK. Process simulation,design and manufacturing of a long fiber thermoplastic composite for masstransit application. Composites Part A 2008;39:1512e21.

[3] Bartus SD, Vaidya UK. Performance of long fiber reinforced thermoplasticssubjected to transverse intermediate velocity blunt object impact. ComposStruct 2005;67:263e77.

[4] Zhao CS, Huang FL, Xiong WC, Wang YZ. A novel halogen-free flame retardantfor glass-fiber-reinforced poly(ethylene terephthalate). Polym Degrad Stab2008;93:1188e93.

[5] Chen YH, Wang Q. Preparation, properties and characterizations of halogen-free nitrogenephosphorous flame-retarded glass fiber reinforced polyamide 6composite. Polym Degrad Stab 2006;91:2003e13.

[6] Wang ZY, Feng ZQ, Liu Y, Wang Q. Flame retarding glass fibers reinforcedpolyamide 6 by melamine polyphosphate/polyurethane-encapsulated solidacid. J Appl Polym Sci 2007;105:3317e22.

[7] Liu Y, Wang Q. Melamine cyanurate-microencapsulated red phosphorus flameretardant unreinforced and glass fiber reinforced polyamide 66. Polym DegradStab 2006;91:3103e9.

[8] Huang XH, Li B, Shi BL, Li LP. Investigation on interfacial interaction of flameretarded and glass fiber reinforced PA66 composites by IGC/DSC/SEM. Polym2008;49:1049e55.

[9] Li B, Xu MJ. Effect of a novel charringefoaming agent on flame retardancy andthermal degradation of intumescent flame retardant polypropylene. PolymDegrad Stab 2006;91:1380e6.

[10] Gao F, Tong LF, Fang ZP. Effect of a novel phosphorousenitrogen containingintumescent flame retardant on the fire retardancy and the thermal behaviourof poly(butylene terephthalate). Polym Degrad Stab 2006;91:1295e9.

[11] Xie F, Wang YZ, Yang B, Liu Y. A novel intumescent flame-retardant poly-ethylene system. Macromol Mater Eng 2006;291:247e53.

[12] Hu XP, Li WY, Wang YZ. Synthesis and characterization of a novel nitrogen-containing flame retardant. J Appl Polym Sci 2004;94:1556e61.

[13] Wang XY, Li Y, Liao WW, Gu J, Li D. A new intumescent flame-retardant:preparation, surface modification, and its application in polypropylene. PolymAdv Technol 2008;19:1055e61.

[14] Morgan AB, Bundy M. Cone calorimeter analysis of UL-94 V-rated plastics. FireMater 2007;31:257e83.

[15] Siat C, Bras ML, Bourbigot S. Combustion behaviour of ethylene vinyl acetatecopolymer-based intumescent formulations using oxygen consumption calo-rimetry. Fire Mater 1998;22:119e28.

[16] Schartel B, Hull TR. Development of fire-retarded materials e interpretation ofcone calorimeter data. Fire Mater 2007;31:327e54.

[17] Wang DY, Liu Y, Wang YZ, Artiles CP, Hull TR, Price D. Fire retardancy ofa reactively extruded intumescent flame retardant polyethylene systemenhanced by metal chelates. Polym Degrad Stab 2007;92:1592e8.

[18] Liu Y, Wang JS, Deng CL, Wang DY, Song YP, Wang YZ. The synergistic flame-retardant effect of O-MMT on the intumescent flame-retardant PP/CA/APPsystems. Polym Adv Technol 20, in press, doi:10.1002/pat.1502.

[19] Gallina G, Bravin E, Badalucco C, Audisio G, Armanini M, Chirico AD, et al.Application of cone calorimeter for the assessment of class of flame retardantsfor polypropylene. Fire Mater 1998;22:15e8.

[20] Kandola BK, Horrocks AR, Myler P, Blair D. Mechanical performance of heat/fire damaged novel flame retardant glass-reinforced epoxy composites.Composites Part A 2003;34:863e73.

[21] Kumar KS, Bhatnagar N, Ghosh AK. Mechanical properties of injection moldedlong fiber polypropylene composites, part 2: impact and fracture toughness.Polym Compos 2008;29:525e33.