02-antec-2005-removing the mystery from rotomoulding

8/12/2019 02-Antec-2005-Removing the Mystery From Rotomoulding

1/5

REMOVING THE MYSTERY FROM ROTOMOULDING:

New insights into the physiochemical processes involved leading to improved quality control

NG Henwood, MA Roberts, A Quaratino, S Collins (Matrix Polymers Limited, Northampton, UK),

P Sharifi, C Liauw, and GC Lees (Dalton Research Institute, Manchester Metropolitan University, Manchester, UK)

Abstract

This paper presents a first report from a long-term collaborative programme between MatrixPolymers Limited and the ManchesterMetropolitan University. The purpose of the

programme is to examine physiochemicalmechanisms of the rotational moulding processusing a variety of analytical techniques.

The effect on the performance of polyethylene(PE) caused by variation of the rotomouldingcooking cycle is investigated using a combinationof infrared spectroscopy and melt rheology.Analytical results are correlated with large scale

performance characteristics, measured byestablished industrial assessment techniques suchas low temperature impact strength, brittleness,

part density development and yellowness index.

Introduction

Rotational moulding, or rotomoulding, is awell-established process for the manufacture oflarge, hollow items. It differs from other types of

plastic processing in a number of important ways:

Rotomoulding is characterised by an absenceof shear in the melt phase of the process. Thismeans that many potential rotomouldingmaterials cannot be used because theirviscosity characteristics do not lendthemselves to what is essentially a sintering

process that takes place under atmospheric pressure.

The cooking cycle, during which the materialenters a melt phase, tends to be of relativelylong duration - typically >20 min in anindustrial rotomoulding situation. This createsa relatively harsh environment and the

possibility of severe degradation of the polymer.

The apparent simplicity of rotomoulding beliesa process of significant complexity and many

process variables (eg the effect of ambientconditions) can be difficult to control.

Over the past two decades, significant progress has been made in characterising the two process stages of therotomoulding process, cooking and cooling. Much of the

previous work has concentrated on investigating the process at a macroscopic level.

In this paper we seek to extend our existingknowledge base, by examining the process and its effectson the polymer being moulded at a more microscopiclevel. The results from industrial material

characterisation techniques have been compared to thoseobtained from laboratory based polymer analysismethods.

Materials and Experimental Methods

Material TestedThe material tested was a widely available European

roto grade of polyethylene (PE) from Matrix Polymers,Revolve N-307. This is a linear medium density

polyethylene (LMDPE), typically used for tanks and largemouldings, with the following basic specification:

Melt Flow Rate (MFR): 3.5 g/10 minDensity: 0.939 g/cm 3

Catalyst Type: Ziegler-NattaReactor Type: Unipol Gas PhaseComonomer Type: Hex-1-enePowder Specification: Standard 500 mStabilisation: Full heat and UV

Material Characterisation TechniquesMaterial characterisation techniques are used to show,

on a macroscopic scale, the effects of process variables onfinal product performance. Current techniques are

particularly useful as a method of comparing the performance of different grades of LMDPE.

Of particular reference to the current work, the effectsof cooking time on the impact strength of the final

product can be tracked, along with the failure mode of thesample, whether brittle or ductile [1].

Typically, a series of mouldings are made at astandard oven temperature and with varying cook times.Sample plaques cut from the mouldings (approx. 20) are

ANTEC 2005 / 1003


2/5

conditioned at 40C and then tested using a dropdart impact test established by the Association ofRotational Molders International [2].

The mode of failure of samples is examinedand failed samples are classified as either brittle orductile. In practice, the failure type is easy todetermine visually (see Figure 1).

A brittleness failure criterion is defined asfollows:

The %BF provides a useful empiricaldefinition of the processing window for a particularrotomoulding product, which can be defined as therange of cook times where the %BF is less than

20%.Samples from characterisation tests can also be

evaluated for part density and yellowness. These parameters, tracked against cook time, can givevaluable additional insights into the state of cure ofthe materials being rotomoulded.

Part density was measured using a balance setup to weigh a small sample both in air andsubmerged in water.

Yellowness was evaluated using a visible lightspectrophotometer and expressed numerically as anindex, YIE .

Rotomoulded samples were made on a FerryRotospeed RS2-190 industrial sizedrotomoulding machine, using two identicalfabricated steel hexagon test moulds. Shot weightwas 1.25 kg of powder for each mould, which

produced mouldings of wall thickness 3 mm.

Analytical TechniquesAn infrared microscope (Nicolet Continuum

with silicon ATR objective) was used to detect the presence of non-volatile carbonyl degradationspecies in sample plaques cut from therotomoulded samples. All spectra were ATRcorrected to compensate for the effect ofwavelength on penetration depth.

The viscosity of polyethylene samples takenfrom rotomoulded plaques was investigated using aPhysica MCR300 oscillatory rheometer.

Results and Discussion

Material CharacterisationFigure 2 is a plot of low temperature impact strength

vs. cook time. It can be seen that, as cook time isincreased, there is a progressive development of impact

strength.

The Peak Inner Air Temperature (PIAT) inside themould is superimposed on the impact strength plot toserve as a cross-reference to the degree of cure. Theconcept of measuring PIAT has become well establishedin rotomoulding [3]; it has been demonstrated that, for PEroto grades, a PIAT of between 190C and 200C isindicative of optimum processing conditions. It will beseen that, on this basis, optimum cure has occurred atapprox. 14 min in this case.

In a seriously under-cured state (in this case, below 9min cook time) very little impact strength can bedeveloped because a full sinter has not been achieved.Above 9 min the impact strength rises and at optimumconditions an impact strength of approx. 100J has beenachieved. As the processing conditions go into over-cook(above 15 min, above 200C PIAT) the impact strengthcontinues to rise until, at 17 min cook, the impact strengthdrops rapidly. This behaviour is typical of materials ofthis type; there is a definite hump after optimum cure

before catastrophic loss of properties occurs.

The behaviour described above is furtherdemonstrated in a plot of %BF vs. cook time. This isillustrated in Figure 3. It can be seen that samples fail in awholly brittle fashion in both under- and over-curedstates, in this case below 12 min and above17 min. In

both cases, the change of failure mode is easy to see and,for this reason, %BF is a superior indicator of the

processing window for the material being tested.

Long experience of this type of testing shows that it isan excellent way of demonstrating how easy a particularPE roto grade will be to mould consistently in anindustrial environment. Grades with wide processingwindows will find favour in a practical moulding situation

because there will be good retention of properties over arange of conditions.

Figure 4 illustrates the variation of part density withcook time. Below 14 min the density of the part isreduced by the presence within the wall structure of gas

bubbles. The bubbles are remnants of incompletesintering. In the initial stages of cooking, bubbles of airare trapped between the interstices of the powder particlesas they lay down and melt on to the mould surface. Givensufficient time at an elevated temperature, the gaseswithin these bubbles will dissolve into the polymer melt.

Brittle Failure Criterion (%BF)=100 x number of brittle failures (1)

total number of failures

1004 / ANTEC 2005


3/5

Figure 4 also shows how yellowness index,(YIE), varies with cook time. Once optimum cureis achieved, the YIE starts to rise. This increase inYIE is mainly due to transformation of the primaryantioxidant into quinone methides [4] that havelight absorption characteristics that give a yellowcolour.

Infrared Spectroscopy StudiesInfrared spectroscopy was used to further study

the processes of degradation indicated by theincrease of YIE as cook time increases.

A range of chemical species is formed duringthe thermal oxidative degradation of PE. These aremostly alcohols and ketones, together with smallamounts of carboxylic acids and aldehydes; thecarbonyl species absorb from 1700 1750 cm -1 [5,6].

In order to evaluate the presence ofdegradation species, plaques cut from the walls ofmouldings prepared at different cook times wereexamined by IR spectroscopy. Both the side of the

plaque in contact with the mould surface (themould side) and the opposite side in contact withair (the airside) were examined.

A typical IR-spectrum for PE is given inFigure 8 and shows characteristic C-H stretchingand deformation absorptions at 2919 cm -1, 2851cm -1and 1463cm -1. At 20 minutes cook time theairside of the rotomoulded plaques yieldedsubstantial carbonyl species, whereas reduced cooktimes afforded much reduced carbonyl growth.Equivalent data for the mould side showedinsignificant carbonyl development, even at longcook times.

The relative amounts of non-volatile carbonylspecies were quantified by measuring the carbonylindex (CI) [6] using the C-H stretching absorptionsas an internal standard. Peak area rather thanabsorbance was used in order to accommodate therange of carbonyl species formed.

Figure 5 shows a plot of CI vs. cook time, for both the airside and the mould side of the sample plaques, superimposed on the earlier plot relatingto YIE. This clearly illustrates that the observedinitial rise in YIE (approx. 14 17 min cook time)is not due to the presence of carbonyl species. Thisindicates that PE oxidation only commences whenstabiliser breakdown has reached a critical level.The plot also illustrates that degradation species arefar more prevalent on the airside of overcookedmouldings than on the mould side. This confirms

the important role of oxygen in the degradation processfor PE.

Figure 6 further illustrates the effect by comparingimpact strength and CI (airside) vs. cook time. Thecoincidence between loss of properties and increase in CI

is clearly demonstrated.

Rheological EvaluationA typical rheometer frequency sweep is a plot of

complex viscosity ( *) against angular frequency ( ).Small samples of material for rheological evaluation wereobtained by shaving the mould side and airside ofmoulded plaques.

Figure 7 shows frequency sweeps for mould side andairside at 13 min and 18 min cook times.

The sweeps for mould side samples at both cooktimes and airside at the lower cook time virtuallycoincide. They exhibit visco-elastic behaviour typical ofa semi-crystalline material such as PE. The sweep forairside at higher cook time exhibits an entirely differenttype of behaviour; there is no viscous component. Thistype of curve is typical of a crosslinked material.

Conclusions

This investigation has successfully unified standardindustrial material characterisation techniques withlaboratory evaluation methods.

Undercooked mouldings (PIAT < 190C) exhibitreduced density due to the presence of entrapped air bubbles, have high brittleness and underdeveloped impactstrength.

Optimally cooked mouldings develop a satisfactorylevel of impact strength and fail in ductile mode. The fulldevelopment of their part density is another sign ofadequate cure.

The onset of over-cure can be identified in a numberof ways. Sample plaques will continue to fail in a ductilemode, provided the material is well stabilised.

The first sign of over-cure is an increase in YIE in thesample plaques. Further overcure results in PE oxidationas manifested by an increase in CI on the airside of the

plaque. The CI in material on the mould side of sample plaques shows a relatively modest rise, even in conditionsof severe overcook.

In the later stages of overcook, rheological dataindicates that the material on the airside becomescrosslinked

ANTEC 2005 / 1005


4/5

At the onset of overcook, the crosslinking ofthe material on the airside of the mouldingincreases the impact strength up to a maximum.However, with increasing overcook, chemicalchanges in the PE structure reduce its ability toabsorb kinetic energy, causing a catastrophic lossof impact strength.

Further investigations examining stabiliserconsumption and changes in crystal structure of thePE are being currently being carried out and will be

published soon.

Acknowledgements

The work described in this paper is part of along-term research project being carried out jointly

by Manchester Metropolitan University(Manchester, UK) and Matrix Polymers Limited(Northampton, UK). The authors would like tothank Revolve Group Limited for their

permission to publish this paper.

Trademark of Revolve Group Limited Trademark of Ferry Industries, Inc.

References

1. Henwood, N.G., Proc., ARM 21 st Fall Meeting, 1996.

2. Association of Rotational MoldersInternational, Low Temperature Impact Test

Method.3. Crawford, RJ et al, Proc., BPF RotomouldersConference, 1991.

4. PP Klemchuk, PL Horng, Polym. Degrad.Stability , 1991, 34 , 333.

5. F Gugamus, Polym. Degrad. Stability , 2002,77 , 147.F Gugamus, Polym. Degrad. Stability ,2002, 77 , 147.

6. N.Grassie G. Scott, Polymer Degradation and stabilisation , Cambridge, 1985.

Key Words

Rotomoulding, polyethylene, degradation,rheometry

Figure 1

0

20

40

60

80

100

120

140

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Cook Time (min)

0

50

100

150

200

250

300

Impact

PIAT

Figure 2

1006 / ANTEC 2005


5/5

%BF

0

20

40

60

80

100

120

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Cook time (min)

Figure 3

0.900

0.910

0.920

0.930

0.940

0.950

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

cook time (min)

0

5

10

15

20

25

30

Y I E

YIE

Density

Figure 4

0

5

10

15

20

25

30

8 9 1011 12 13 14 15 1617 18 1920 21 22

cook time

0

0.05

0.1

0.15

0.2

0.250.3

0.35

C I

YIE

CI-AS

CI-MS

Figure 5

0

20

40

60

80

100

120

140

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22Cook time (min)

0

0.1

0.2

0.3

0.4

C a r

b o n y

l I n d e x

ImpactCI-MS

Figure 6

102

103

104

105

Pas

| *|

0.1 1 10 1001/sAngular Frequency

Figure 7

3000 2000 1000

Wave Number (cm -1)

A b s o r b a n c e

Figure 8

ANTEC 2005 / 1007

02-antec-2005-removing the mystery from rotomoulding

Documents