FORMATION OF SHAPED/MOLDED MELTBLOWING NONWOVEN … · a fabric structure are of prime importance...

JTATMVolume 1, Issue 1, September 2000

Volume 1, Issue 1, September 2000

FORMATION OF SHAPED/MOLDED MELTBLOWING NONWOVEN STRUCTURES

Yogeshwar Velu, Raoul Farer, Tushar Ghosh, Abdelfattah SeyamNorth Carolina State University

ABSTRACT

Three dimensional (3D) fiberweb structures are useful in many applications. The Robotic Fiber Assembly andControl System (RFACS) being developed in this research allows precise control of fiber meltblown fiberdeposition on a 3D mold surface. The effect of various process parameters on a number of polypropylene (PP)web characteristics is reported. Under the experimental range studied, the fiber orientation distribution wassignificantly impacted by the process parameters. The fiber diameter distributions indicate that they are uniqueto a particular process condition. The distributions do not overlap when a parameter is evaluated. In keepingwith the long-term objective of developing chemical/biological barrier fabrics using RFACS technology, thepore distribution of the fiberwebs was characterized. Under the conditions explored, the average pore size ofthe analyzing web has decreased by 60% when the attenuating air pressure was increased from 0.7 bar to 2.8bar. The pore size was decreased by 33% when the take up speed of the web was increased from 20 ft/min to 50ft/min.

1. INTRODUCTIONNonwoven webs can be produced as sheetstructures by using meltblown technology [23].In meltblowing, molten polymer is extrudedthrough a series of orifices in a knife-edge die.The die is jacketed on both sides by highvelocity laminar sheets of air. The polymerstreams from the orifices are elongated by theair-drag to form fibers, which are collected on adrum or other suitable collecting surface. Fiberdiameters can range from 500 microns to assmall as 0.1 microns (µm). The extremeentanglement of fibers, characterizing meltblownfibrous webs, produces coherency and strength.The density of the web is such that it has theproperty to contain and retain particulate matter[12], thus qualifying such structures for filtrationapplications. The entanglement of these longfibers makes it impossible to remove one fiberfrom the web or to trace one fiber frombeginning to end [7]. Meltblown webs arelightweight with a high surface area. Theydisplay a high insulating value and excellentfilter characteristics [3, 17]. Meltblowntechnology can be used to produce efficient filter

materials, filtering particles that are bigger than0.5 µm [24].

The long term objective of this research is todevelop technology to produce shaped protectivegarments, substantially to its final shape andusing minimal seaming or joining. The latterusually constitute the “weak-link” in a protectivesystem. The integration of meltblowingtechnology and robotics can achieve the properformation of molded or shaped seamlessstructure that may be incorporated in a protectiveclothing system. In integrating the twotechnologies the fiber web collector is usually amold object structure. The mold can bemanipulated to rotate continuously so as to formthe molded fabric. The size of the die widthshould be smaller than the size of the mold tohave a good control over the deposition of thefibers on the mold.

The usefulness of molded fabrics so obtaineddepends on the performance characteristics ofthe web structures. The desired performancecharacteristics, globally and locally, can beenumerated as strength, abrasion resistance, tear

2

resistance, burst strength, elastic recovery,air/moisture permeability, moisture/fluidabsorption (rate and capacity), filtrationcharacteristics, etc. Each of these characteristicsis influenced by fiber diameter and itsdistribution, pore size and its distribution, fiberorientation, web consolidation (thickness,bonding), web basis-weight (local),fiber/polymer properties, etc.

The web structure defined by fiber OrientationDistribution Function (ODF) governs theanisotropy of mechanical (strength, tearresistance, bending rigidity, etc.) and physical(wicking/absorption, pore shapes, pore sizedistribution, etc.) properties [9-11, 13-15]. ODFdata for meltblown webs has not been availablein the literature previously, while somepublished ODF results for spunbonded fabricsare generally similar to those found formeltblown webs in this study.

Although meltblowing process produces finerfibers relative to conventional fiber spinningprocesses, the distributions of fiber diameter areusually quite broad. Because lower mean valuesand narrower distributions lead to smaller poresizes (lower mean and narrow distributions) andhigher specific surface, the control of fiberdiameter distribution is considered highlydesirable [18], for a number of importantapplications. The pore size and its distribution ina fabric structure are of prime importance indetermining the transport properties of the fabric.The filtration efficiency, and hence the level ofprotection, is directly related to the pore sizedistribution. Image Analysis techniques havebeen developed to measure the pore size, shape,and orientation of the pores [23, 26]. Most of themethods developed at the Textile ResearchInstitute (TRI) use fluid intrusion or extrusion ona sample to determine the average pore size andits distribution.

2. ROBOTIC FIBER ASSEMBLY ANDCONTROL SYSTEM

To produce the seamless 3D garments a RoboticFiber Assembly and Control System (RFACS)has been set up. The RFACS consists of a modelmeltblown machine and a commercial six-axisrobot that is capable of manipulating themeltblown die in the range of positions andorientations required for use with the complex3D shape molds. Accurate robot positioning isrequired to keep the distance between the die andthe mold at will while maintaining appropriate

orientations with the mold. In addition, toachieve the required level of mold controlneeded for this process, an external seventh axishas been added. The RFACS setup is shown inFigure 1.

Tbmratrscvmpomcaao

3

Tmwdapmau

1

2 3

4

Figure 1. Robotic Fiber Assembly andControl System (RFACS)1. Melt-Blowing Die Housed in a Cage.2. Polymer and Air Flexible Supply Hoses.3. Melt-Blowing Extruder Unit. 4. Collector.


he shaped nonwoven structures are producedy depositing meltblown fibers on a collapsibleold that is placed on the seventh axis of the

obot. To develop the contour-followinglgorithms for mold shapes point coordinates onhe mold shape were determined at regularotational increments. To find these points in 3D-pace for the 2D model, a pointer has beenonstructed such that the pointer tip assumes theirtual position of the right most orifice in theeltblowing die body. For the 3D model, a

ointer mirroring the position of the polymerrifices has been constructed. The pointer uses toark points on a mold relative to the world

oordinate system. This procedure has beendopted as the means of developing the positionnd speed of the meltblowing die to the rotationf the mold body.

. EFFECT OF PROCESS PARAMETERSON WEB CHARACTERISTICS

he research documented in this paper deals withold structures being coated with a meltblowneb, hence, experiments were conducted toevelop appropriate control algorithms. Theselgorithms would control the RFACS in a way torecisely control fiber dispensation on to theold surface. The objective of the control

lgorithms is not necessarily to controlniformity of web characteristics, but to control


3

these locally. The web characteristics of interestin the context of the present research are (1)Basis weight, (2) Fiber orientation distribution,(3) Fiber diameter distribution, and (4) Pore sizedistribution. These characteristics determine thelocal physical, mechanical and fluid flowbehavior of the web. The relevant processparameters are:

• Polymer throughput• Die / Melt / Attenuating air temperature• Attenuating air pressure,• Fiber stream approach angle,• Take-up speed, and• Relative orientation and movement of

die

In the following, the parametric studies arereported and categorized according to theirinfluence on a specific web characteristics.

Basis Weight

The basis weight (g/m2) of the PP fabric wasevaluated by measuring the weight of knownarea of fabric samples at regular intervals, alongthe surface of the mold. The variation inmeasured basis weight is expressed as percentcoefficient of variation (%CV). Initialexperiments were conducted such that the diewas moving up and down the mold, while themold was rotating on the 7th axis at a uniformspeed (2D no correction Model). The 2D modelwas modified to correct for variations in the dieto collector distance (DCD correction) and thelinear speeds of the mold on the 7th axis (linearrotational speed correction model).

Figures 2 and 3 show the basis weightdistributions using the patterns of motionvariation for the 2D model. The differences in

basis weight distributions are related to theinteraction between the geometrical features ofthe mold and the characteristics of the fiberflaring profile and the die orientation. CV valuesfor samples formed using the no correctionmodel were 8% higher than CV values forsamples obtained using the linear correctionmodel. The difficulty in achieving more uniformbasis weight distribution is inherently related to

0

20

40

60

80

100

120

140

1 3 5 7 9 11 13 15 17 19 21 2315 degree interval

Bas

is W

eigh

t (g/

sq. m

)

no correction - CV: 23%DCD correction - CV: 26%linear rot. Spd. Correction - CV: 15%DCD and linear rot. Spd. Correction - CV: 16%

Figure 2. Two-Layer Basis-WeightDistribution Using the 2D Model

0

20

40

60

80

100

120

140

1 3 5 7 9 11 13 15 17 19 21 2315 degree interval

Basi

s W

eigh

t (g/

sq. m

)

DCD and nonlinear rot. Spd. Correction (single up) - CV: 20%DCD and nonlinear rot. Spd. Correction (single down) - CV: 24%DCD and nonlinear rot. Spd. Correction (two-layer) - CV: 21%

Figure 3. Single-Layer/Two-layer Basis-Weight Distribution using the 2D Model andNonlinear Rotation Speeds

Figure 4. 3D Positioning Sequence


4

the die orientation during the production. Whenthe die is not reoriented in relation to the surfaceof the mold during fiber application, overshot offibers occurs. This causes loss of control for thedeposition of fibers on the mold.

The 3D model was developed by theimplementation of 3D-triangulation positioningand feedback control of the mold positioningdata using the LVDT. This included the controlof the movement of the die and the rotationalspeed of the mold mounted on the 7th axis(Figure 4). The position of the die is controlledsuch that the fiber streams are normal to thesurface of the mold at the point of deposition ofthe fiber. A LVDT feed back control was usedto control the speed of the 7th axis, such that thelinear speed of the mold surface was constant.The overall control performance improved withthe use of the 3D model [7]. Using the rule-basedcontrol resulted in the lowest overall CV values(11%) for basis-weight uniformity as it can beseen from the results of Figure 5. Since previous

experimentation had shown basis weight effectsto be symmetric around 180 degrees, only oneside of the mold was evaluated.

Fiber Orientation Distribution

The orientation distribution function (ODF) ofthe fibers was measured on samples that wereprepared using the 90o fiber approach angle(Figure 6) relative to the collector drum. Animage analysis package developed at theNonwovens Cooperative Research Center wasused to make ODF measurements on thesamples. Influence of various processparameters on fiber orientation distribution isdiscussed below.

Take-Up Speed

Figure 7 gives the ODF for fabrics formed attake-up speeds of 18, and 24 m/min, while otherprocess parameters are held constant. Note thatwith extruder output kept constant the basisweight decreases with increasing take-up speed.The results show an increase in orientation withan increase in take-up speeds.

Maximum observed fiber fraction data along themachine direction (MD) in Figure 8, depictsincreases in orientation along MD with increasesin take-up speeds for all fabric types evaluated.Webs do not initially exhibit significant changesin its ODF at take-up speeds below 18.3 m/min.This may be due to aerodynamic interaction with

Figure 2. Fiber-Stream Approach-Angle

Orientation Angle - Degrees (mid pt. of the bin

7.5 22.5

37.5

42.5

67.5

82.5

97.5

112.5

127.5

142.5

157.5

172.5

Rel

ativ

e Fr

eque

ncy

(%)

2468

101214161820

B3B4

Figure 7. ODF for MB PP fabrics at differenttake-up speeds; 7 x 10-2 g/min/hole polymerthroughput rate; 1.4 bar attenuating airpressure; 14 cm DCD and 282 ºC air temp. ;B3: 18 m/min and B4: 24 m/min.

01020304050607080

12 13 14 15 16 17 18 19 20 21 22 23

15 degree interval

LVDT - Rule-Based Control System - CV: 11%

Figure 5. Rule-Based Control Results UsingLVDT Feedback and 3D Triangularization

5

fibers. For take-up speeds above 18.3 m/minfabrics show higher orientation due to higherdrum surface velocity induced fiber alignmentalong the machine direction.

Die-to-Collector-Distance (DCD)

Figure 9 shows the ODF for fabrics formed atDCD settings ranging from 14 to 31 cm. Fiberorientation is shown to continuously increasefrom DCD settings 31 to 14 cm. All webs

evaluated are shown to exhibit less MDorientation when formed at larger DCD settings.With longer distances the turbulent fiber flow,undulating, or flapping motion of the fibers isknown to increase [6, 16 ]. As the air velocityreduces at longer distances from the die, andloses some of its initial planar characteristics, theair volume disturbed and the dimensions of theair stream increase [18, 19]. With higherfreedom of movement, and lower fiber velocitythe fibers appear to take on a more randomorientation in the air stream. This effectsubsequently translates to reduced MDorientation in webs formed, upon collection atlarger distances (higher DCD settings) from thedie body.

At a DCD setting of 7 cm, webs exhibitedreduced “MD” orientation relative to thoseformed at 14 cm. This effect is attributed to somefibers being unable to successfully develop aforward flow pattern below DCD settings of 14cm, and rapid successive, more random,accumulation of short segments of fiber on top ofeach other. Some fibers are suspected to beessentially blown into themselves, therebycausing less overall orientation in the structuresformed. It is also stipulated that air velocities arestill much higher at small DCD values, thusbouncing back from the forming drum andthereby disturbing the MD orientation of fibers.

Fiber-Stream Approach-Angle

Orientation Angle - Degrees (mid pt. of the bin)

7.5 22.5

37.5

42.5

67.5

82.5

97.5

112.5

127.5

142.5

157.5

172.5

Rel

ativ

e Fr

eque

ncy

(%)

4

8

12

16

20B10B9B8B7

Figure 9. ODF for MB PP fabrics at differentDCD; 7 x 10-2 g/min/hole polymerthroughput rate, 1.4 bar attenuating airpressure, 21 m/min take-up speed, and 282 ºCair temp.; B7: 14 cm; B8: 21 cm; B9: 26 cm;and B10: 31 cm.

Take-Up Speed (m/min)0 5 10 15 20 25 30 35

Rel

ativ

e Fr

eque

ncy

- 75-

105o

(%)

30

32

34

36

38

40

B1 - B5A1 & A2A11 & A13

Figure 8. Maximum observed fiber fractionalong “MD” for MB PP fabrics at different take-up speeds; 282 ºC air temp.; B1-B5: 7 x 10-2

g/min/hole polymer throughput rate, 1.4 barattenuating air pressure, and 14 cm DCD; A1 &A2: 5.4 x 10-2 g/min/hole polymer throughputrate, 1.4 bar attenuating air pressure, and 14 cmDCD; Ap11 & Ap13: 5.4 x 10-2 g/min/holepolymer throughput rate, 0.7 bar attenuating air-pressure, and 18 cm DCD.


7.5 22.5

37.5

42.5

67.5

82.5

97.5

112.5

127.5

142.5

157.5

172.5

Rel

ativ

e Fr

eque

ncy

(%)

0

5

10

15

20

25

30Aa2fAa2eAa2dAa2cAa2bAa2a

Figure 10. ODF for MB PP fabrics at differentfiber-stream approach-angles; 5.4 x 10-2

g/min/hole polymer throughput rate, 1.4 barattenuating air pressure, 21 m/min take-upspeed, 14 cm DCD, and air temp. 305 ºC;Aa2a: 90º; Aa2b: 79º; Aa2c: 61º; Aa2d: 46º;Aa2e: 36º; and Aa2e: 26�.


6

Figure 10, shows the fiber ODF for fabricsformed at approach-angles ranging from 90o to26o. Fiber fraction along MD is shown toincrease by 60 % when the fiber-streamapproach-angle changes from normal to 36 o . A0 o approach-angle would identify a fiber-streampath parallel to the tangent of the fiber forming-drum (essentially blowing past it). By reducingthe fiber-stream approach-angle, some fibers willtravel a longer path before being collected on thedrum’s surface (see Figure 4). Those fibers,which travel a longer path, are also traveling at apath more parallel to the tangent of the forming-drum, and exhibit this orientation uponcollection. This effect contributes to an increasein orientation in the machine direction. Thispattern continues up to a point where some fibersare actually blowing past the drum, but arepulled back into the fabric because their fiberends are anchored inside the fiber stream. Thiseffect was visually observed when collectingsamples at approach-angles of 26 o. Fibers wereobserved blowing past the drum’s surface, butdue to their continuous nature were still trappedinside of the fiber stream. This either causedthem to stick out of the formed fabric, or to bepulled back into it. In either case the structure ofthe formed fabric was disrupted, which can beobserved in the data shown by the respectivedecrease in fiber orientation at approach-anglesof 26 o.

PolymerThroughput Rate

Fiber fraction along MD decreases withincreasing throughput rate, as is shown in Figure11. In meltblowing, analogous to what occurs inmelt-spinning, higher polymer throughput rate isequated with larger average fiber diameter,keeping all other conditions constant [2]. Whenthis is the case, larger size fibers present in thefiber stream will resist aligning themselvessubstantially in the air flow direction and assumea more random orientation. These fibers undulatemore slowly (at lower frequencies) [16], as itbecomes more difficult for a thicker and heavierstructure to change direction as quickly as asmaller diameter structure. Lower undulatingfrequencies are shown to directly result information of more isotropic structures.

Attenuating Air Pressure


7.5 22.5

37.5

42.5

67.5

82.5

97.5

112.5

127.5

142.5

157.5

172.5

Rel

ativ

e Fr

eque

ncy

(%)

4

8

12

16

20At1At2At3

Figure 11. ODF for MB PP fabrics at differentpolymer throughput rate; 2.1 bar attenuating airpressure, 15 m/min take-up speed, 18 cm DCD,and 282 ºC air temp.; At1: 5.4 x 10-2 g/min/hole,At2: 8.1 x 10-2g/min/hole, and At3: 9.6 x 10-2

g/min/hole.


7.5 22.5

37.5

42.5

67.5

82.5

97.5

112.5

127.5

142.5

157.5

172.5

Rel

ativ

e Fr

eque

ncy

(%)

2468

101214161820

Dp1Dp2Dp3

Figure 12. ODF for MB PP fabrics at differentattenuating air pressures; 9.6 x 10-2 g/min/holepolymer throughput rate, 15 m/min take-upspeed, 18 cm DCD, and 282 ºC air temp.; Dp1:0.7 bar; Dp2: 1.4 bar, and Dp3: 2.1 bar.


Figure 12 shows ODF results from fabricsformed by varying attenuating air pressure.Fabrics are shown to be less oriented along“MD” when higher attenuating air pressures areused. If one considers that higher take-up speedshave been shown to cause more orientation, thesame argument can be made here. At lowerattenuating air pressures the fiber speed reducesas compared to fiber formed at higher attenuatingair pressures [25]. The ratio of take-up speed tofiber speed will increase with lower attenuatingair-pressures. Subsequently an argument can bemade that a decrease in fiber speed, i.e. lower


7

attenuating air-pressure, has the same effect asan increase in take-up speed, thereby causinghigher orientation to form in the fabric structure.

Fiber Diameter Distribution

SEM images at 600X were used to make theFDD measurements. Each image contained atleast 20 fiber images. A well developed protocolon the image analysis software was used foradjusting the SEM image [8]. The resultingblack and white fiber image was evaluated forfiber diameter. Influence of various processparameters on fiber diameter distribution isdiscussed below.

Polymer Throughput Rate and AttenuatingAir Pressure

Figure 13 shows cumulative frequencies of fiberdiameter distributions of PP meltblown fabricsformed at different levels of polymer throughputrates and at a selected level of attenuating airpressure. Figure 14 shows cumulativefrequencies of fiber diameters in fabricsproduced at varying attenuating air pressure anda selected polymer throughput rate.

All web samples in Figure 13 show fiberdiameters to decrease and their distribution tonarrow with a decrease in throughput rate. Figure14 shows cumulative frequencies for fiberdiameter distributions in fabrics formed atvarying attenuating air pressures, but a constantpolymer throughput rate of 9.6 x 10-2 g/min/hole,and the same temperature settings as in Figure

13. Fiber diameters decrease with increase inattenuating air pressures in all cases; broaderdistributions of fiber diameters are observed withdecreases in attenuating air pressures. Allsamples depict the same trend of increasingfraction of small diameter fibers with decreasingthroughput rates; the increase in attenuating air

pressure also increases the fraction of fine (lessthan 10 µm) fibers. An increase in attenuatingair pressure results in higher velocity of formingair, exerting higher drag forces on the polymermass as it is being pushed out of the die orifices,as well as resulting in higher fiber velocities [2,20]. Higher drag apparently attenuates thepolymer mass to finer diameters, analogous tohigher take-up roller speeds resulting in finerdiameter filaments in conventional spinning.

Attenuating Air Temperature

Figure 15 shows cumulative frequencies for fiberdiameter distributions obtained using varyingattenuating air temperature. Fiber diameter washardly affected by the attenuating air temperaturein the range of (282 ºC – 327 ºC) studied. Similartrends were observed in a study by Rao andShambaugh, where a 1000 C increase in airtemperature did not show much effect on fiberdiameter [16]. In both Rao and Shambaugh andour case discussed, the polymer temperature (dietemperature) settings were lower than theattenuating air temperature. The air temperaturemeasured at the die orifice does not changesignificantly even when the employed die

Fiber Diameter (µµµµm)0 5 10 15 20 25 30 35

Cum

ulat

ive

Freq

uenc

y (%

)

0

20

40

60

80

100

Fp1Ep1Ap13Cp1Dp1

Figure 13. Cumulative frequency distribution offiber diameters at different polymer throughputrates; 0.7 bar air pressure, 282 ºC air temp., and260 ºC die temp.; Fp1: 1.7 x 10-2 g/min hole;Ep1: 3.7 x 10-2 g/min/hole; Ap13: 5.4 x 10-2

g/min/hole; Cp1: 8.1 x 10-2 g/min/hole; and Dp1:9.6 x 10-2 g/min/hole.


Cum

ulat

ive

Freq

uenc

y (%

)

0

20

40

60

80

100

Dp1Dp2Dp3Dp4Dp5

Figure 14. . Cumulative frequency distribution offiber diameters at different attenuating airpressure; 9.6 x 10-2 g/min/hole throughput rate,282 ºC air temp., and 260 ºC die temp.; Dp1: 0.7bar; Dp2: 1.4 bar; Dp3: 2.1 bar; Dp4: 2.8 bar;and Dp5: 3.5 bar.

temperature is lower than the air temperature. Itis observed that the attenuating air temperaturesstudied in this research were not high enough tocause significant changes in diameters of thefibers.

Figure 16 shows cumulative frequencies for fiberdiameter distributions of PP meltblown fabricsformed due to varying die temperatures. As isapparent, fiber diameter decreased with increasesin die temperature. All samples show increasesin fine fiber content with increasing dietemperature.

The dominant resultant effect of an increase indie temperature is lower polymer viscosity in thedie body. A lower viscosity liquid will show lessphysical resistance to high velocity attenuatingair, and allow finer fiber diameters to form. Thereduction in polymer viscosity appears to bemore significant over the temperature rangeevaluated, as similar changes in air temperatureshave in this study not been able to show a lastingaffect. Reduction in polymer viscosity wouldfurther show a larger effect at higher polymerthroughput rates, where finer diameter fibers areformed with more difficulty.

Pore Size Distribution

The pore size distribution was measured using anautomated perm porometer designed and sold bythe Porous Materials Inc. This equipment workson the principle of capillary flow. Detailsregarding the theory and the working principle ofthe instrument are found else where [8, 22].Influence of various process parameters on poresize distribution is discussed below.

Fiber Diameter (µµµµm)0 5 10 15

Cum

ulat

ive

Freq

uenc

y (%

)

0

20

40

60

80

100

Ad1Aa2a Aa3a

0 5 10 15

E2E3E

Figure 15. Cumulative frequency distribution offiber diameters at different attenuating airtemperatures; 5.4 x 10-2 g/min/hole throughputrate, 1.4 bar attenuating air pressure, and 260 ºCdie temp.: Ad1: 282 ºC, Aa2a: 305 ºC, and Aa3a:327 ºC; 3.7 x 10-2 g/min/hole throughput rate,1.4 bar attenuating air pressure, and 260 ºC dietemp.: E: 282 ºC, E2: 304 ºC, and E3: 327 ºC.

Die Temperature


Cum

ulat

ive

Freq

uenc

y (%

)

0

20

40

60

80

100

E3E4E5

Figure 16. Cumulative frequency distribution offiber diameters at different die temperature; 9.6 x10-2 g/min/hole polymer throughput rate; 1.4 barattenuating air pressure, and 327 ºC attenuatingair temp.; E3: 260 ºC; E4: 293 ºC; and E5: 327ºC die temperature.


8


9

Die to Collector Distance

Figure 17 shows normalized pore sizedistribution and cumulative filter flow of PPmeltblown fibers formed due to varying die tocollector distances. The pore size initiallydecreases with increase in the distance. Thediameter of the pore at 15 cm is 18 microns. Atthis distance the web has a filtration efficiencygreater than 95%, and hence particles larger than18 microns can be filtered. Further increase inthe distance of the collector, increases thediameter of the pore. An increase in die tocollector distance decreases the air velocity,alters the planar nature of the air, and thetemperature of the air profile. It leads to theformation of shots and hence an increase in thepore size of the web.

Number of Layers

The Pore size distribution and the cumulativefilter flow for the influence of the number oflayers are given in Figure 18. These graphsindicate that with increase in the number oflayers the pore size decreases and the filtrationefficiency increases. The technique adoptedmeasures the smallest diameter of a given poreand thereby if a layer of fibers with relativelysmall pore size distribution is incorporated in theweb, the overall pore size becomes smaller. Withthe decrease in the average pore size, thefiltration efficiency increases. The data furtherindicates that increase in the number of layersfrom three to four does not have a significanteffect on the pore size distribution and thefiltration efficiency.

5 cm

25 cm

15 cm10 cm

20 cm

Average Diameter (µµµµm)0 20 40 60 80

Pore

Siz

e D

istr

ibut

ion

0

20

40

60

80

100

120

(a)

5 cm

25 cm

15 cm10 cm

20 cm

Average Diameter (µµµµm)0 20 40 60 80 100

Cum

ulat

ive

Filte

r Flo

w (%

)

0

20

40

60

80

100

120

(b)

Figures 17 a & b. Pore Size Distribution andCumulative Filter Flow (%) for MB PP Fabrics atdifferent Die to Collector Distances. (Throughput :3.7 x 10-2 g/min/hole, Air Pressure: 1.4 bar,Number of Layers: 2, Take Up Speed: 20 ’/min)

3 Layers

2 Layers

4 Layers

Average Diameter (µµµµm)0 10 20 30 40 50 60

Pore

Siz

e D

istri

butio

n

0

20

40

60

80

100

120

1 Layer

(a)

2 Layers


Cum

ulat

ive

Filte

r Flo

w (%

)

0

20

40

60

80

100

120

1 Layer

3 Layers4 Layers

(b)

Figures 18 a & b. Pore Size Distribution andCumulative Filter Flow (%) for MB PP Fabricsfor variation in the number of layers. (Throughput: 3.7 x 10-2 g/min/hole, Air Pressure: 1.4 bar,DCD 15 cm, Take Up Speed: 20 ’/min)


10

Attenuating Air Pressure

As was mentioned in our earlier discussion, withincreasing attenuating air pressure the fraction ofsmaller fibers increases. An increase in theattenuating air pressure results in an increase inthe velocity of the forming air and hence, anincrease in the drag force exerted on the fiber.This leads to the formation of fine fibers. Finerfibers produced arrange themselves closertogether and hence finer pores are formed in theweb. Figure 19 shows the pore size distributionand the filtration efficiency of the webs producedat different attenuating air pressures.Complementing the pore size distribution, thefiltration efficiency increases with increasingattenuating air pressure. A higher percentage of

finer pores increase the filtration efficiency ofthe web [1, 4, 5].

Take-Up Speed and Through Put

As has been mentioned in the earlier discussions,increase in the take up speed of the webdecreases the basis weight of the web. Theincrease in take up speed also increases the fiberorientation in the machine direction. Theincrease in the orientation in the machinedirection and the decrease in basis weight due toincrease in the take up speed decreases theaverage pore size and forms a narrowerdistribution, as also reported by Bhatia [4]. As isshown in Figure 20, the increase in take up speeddecreases the pore size and increases thefiltration efficiency of the web. But, at thecurrent throughput rate, basis weight and

Average Diameter (µµµµm)

0 20 40 60 80 100

Pore

Siz

e D

istr

ibut

ion

0

20

40

60

80

100

120

0.7 bar1.4 bar2.1 bar2.8 bar

(a)

0 20 40 60 80 100

Cum

ulat

ive

Filte

r Flo

w (%

)

0

20

40

60

80

100

120

2.8 bar

0.7 bar1.4 bar2.1 bar

(b)

Figures 19 a & b. Pore Size Distribution andCumulative Filter Flow(%) for MB PP Fabrics forvariation in the attenuating air pressure. (Throughput: 3.7 x 10-2 g/min/hole, Number of Layers: 2, DCD15 cm, Take Up Speed: 20 ’/min)

0 10 20 30 40 50 60

Pore

Siz

e D

istr

ibut

ion

0

20

40

60

80

100

120

20 '/min30 '/min40 '/min50 '/min


(a)


0 10 20 30 40 50 60

Cum

ulat

ive

Filte

r Flo

w (%

)

0

20

40

60

80

100

120

20 '/min

40 '/min

30 '/min

50 '/min

(b)

Figures 20 a & b. Pore Size Distribution andCumulative Filter Flow(%) for MB PP Fabrics forvariation in the web Take up Speed (Throughput :3.7 x 10-2 g/min/hole, Number of Layers: 2,DCDC: 15 cm, Air Pressure: 1.4 bar)

attenuating air pressure there is no change in thepore size and filtration efficiency when the takeup .

Ffiththfipinsydind

4

Aha

Fiber Assembly & Control System (RFACS) canbe implemented for production of seamless 3Dmeltblown nonwoven fabric structures. The set-up developed here allows for arbitrarypositioning of the melt-blowing die in an initialreference frame to an arbitrary collecting surfacewithout difficulty and accuracy of 1/10 of amillimeter.

Experiments have demonstrated that rotationalmotion of a mold needs to be controlledaccording to its shape (curvature, etc.), whenuniform basis-weight distribution is required.Various motion correction models (i.e. linear andnonlinear) may be successfully employed tosome extent. Relative movement and orientationof die and collecting surface must be consideredcarefully to achieve desirable results. For thecase of fiber deposition on a mold, it is desirableto have the center of the tool carrying die alignednormal to the surface of the mold. Such anorientation allows for more uniform fiberapplication to the mold surface and results in a4%-10% improvement in basis-weightuniformity. Implementation of a rule-basecontrol algorithm that compensates for the moldposition and speed as well as tool position wassuccessfully done. The approach improves thebasis-weight uniformity and significantlysimplified the robot programming operation.

Fiber ODF is shown to be a useful parameter fordetailed analysis of changes taking place inmeltblown nonwoven structures. Changesinduced in structures by variation of processparameters were appropriately related torespective changes observed in the fabric’s ODF.

Changes in fabric take-up speed, die-to-

speed is changed from 40 ft/min to 50 ft/min

Average Diameter (µµµµm)0 5 10 15 20 25 30 35 40 45 50 55 60

Pore

Siz

e D

istr

ibut

ion

0

20

40

60

80

100

120

9.6 x 10-2 g/min/hole


7.0 x 10-2

g/min/hole



1.7 x 10-2

g/min/hole

(a)


Cum

ulat

ive

Filte

r Flo

w (%

)

0

20

40

60

80

100

120







(b)

Figures 21 a & b. Pore Size Distribution andCumulative Filter Flow (%) for MB PPFabrics for variation in the PolymerThroughput (Take Up Speed: 20’/min.,Number of Layers: 2, DCD: 15 cm, AirPressure: 1.4 bar)


11

igure 21 shows the pore size distribution andltration efficiency of the webs for differentroughput levels. Similar to the influence ofrough put on the diameter distribution of thebers in the web [8], the average diameter of theore size of the web decreases with the decrease the through put through the meltblowingstem. The diameter distribution range

ecreases. The filtration efficiency of the webscreases with the decrease in the average

iameter of the pore size.

. CONCLUSIONS

novel 3D fiberweb manufacturing technologys been successfully developed. This Robotic

collector-distance, fiber-stream approach-angle,polymer throughput rates, and attenuating air-pressures were shown to significantly affect fiberODF in webs. Generally all fiber ODFs exhibiteda bell-shaped pattern, with the highest frequencyof orientation in the vicinity of MD. Changes infiber-stream approach-angles were demonstratedto affect respective changes in fiber orientationdistribution to the highest extent. For a change infiber-stream approach-angle from normal (90°)to 36°, a 60 % increase in fiber alignment alongthe machine direction was shown.

Fiber diameter distributions were shown tocorrelate well to processing conditions employedin meltblowing of polypropylene resin. Fiberdiameters were demonstrated to reduce withreductions in throughput rate, and increases in


12

attenuating air pressures and die temperatures.Air temperatures in the range studied, below 210ºC at the die-orifice exit, were shown to notaffect fiber diameter distributions. Goodagreement for all results was also found toaverage fiber diameter data observed inpublished literature.

The pore size distribution and the filtrationefficiency of the melt blown polypropylene webswere characterized. The effect of the processvariables on the average pore size distributionwas investigated. The changes in the averagepore diameter were related to the diameterchanges that take place in the fibers formed andtheir orientation distribution. The change in die-to-collector-distance, number of layers,attenuating air pressure, polymer throughputrates, and web take up speeds were shown tosignificantly affect the average pore size and thefiltration efficiency of the webs formed. In thecase of die to collector distance, the least poresize and the best filtration efficiency was formedat a distance of 15 cm. as the temperature of theattenuating air is dropped to around temperaturebeyond the 15 cm. This does not allow furtherchanges in the internal structure of the fiber andalso effects the consolidation of the fiber web onthe collecting surface. The increase inattenuating air pressure from 0.7 bar to 2.8 barreduces the average predominant pore size by60%, while when the take up speed is increasedfrom 20’/min to 50’/min, the average pore sizereduces by 33%.

ACKNOWLEDGEMENTSThis research was supported in part by an ARO-MURI grant from the Army Research Office,and in part by support from the NonwovensCooperative Research Center, North CarolinaState University. We gratefully acknowledgetheir generous support of this project.

REFERENCES

1. Allen,T., “Particle Size Measurement”, 4th

Ed., Chapman and Hall, New York, 1990.

2. Bansal, Vishal, R. L. Shambaugh, On-lineDetermination of Diameter and Temperatureduring Melt Blowing of Polypropylene, Ind.Eng. Chem. Res. 37, 1799-1806 (1998).

3. Bergmann, L., "Microfiber AndElectrostatically Charged Nonwovens inFiltration", Nonwovens Industry 22, (2), 28-34 (1991).

4. Bhatia, S. K., and Smith, J.L., “GeotextileCharacterisation and Pore Size Distribution:Part I. A Review of ManufacturingProcesses”, Geosynthetics International, 3(1), 85-105, 1996.

5. Bhatia, S. K., Smith, J. L., and Christopher,B. R., “Geotextile Characterisation and PoreSize Distribution: Part III. Comparison ofMethods and Application to Design”,Geosynthetics International, 3 (3), 301 –328, 1996.

6. Chhabra, Rajeev, S, R. L. Shambaugh,Experimental Measurements of FiberThreadline Vibrations in the Melt-BlowingProcess, Ind. Eng. Chem. Res. 35, 45366-4374 (1996).

7. Eian, G., "Durable Melt-Blown FibrousSheet Material, U.S. patent 4,681,801, 1987.

8. Farer, Raoul., Formation of 3D MeltblownStructures Using Robotic Control of FiberDeposition, Doctoral Thesis, NCSU,Raleigh, NC, 1999.

9. Hearle, J. W. S., and Sultan, M. A. I., AStudy of Needled Fabrics, Part IV, J. TextileInst. 59, 161 (1968).

10. Hearle, J.W. S., and Stevenson, P.J.,Nonwoven Fabric Studies, Part III: TheAnisotropy of Nonwoven Fabrics, TextileRes. J. 34, 3 (1964).

11. Huang, X. C., and Breese, R. R.,Characterizing Nonwoven Webs UsingImage AnalysisTechniques, Part I: PoreAnalysis in Thin Webs, INDA J. NonwovensRes. 5 (2), 14 (1983).

12. Insley, T., "Sorbent Sheet Material",European patent application 0 156 649 A2,1985.

13. Matsuoka, T., Takabatake, J., Inoue, Y., andTakahashi, H., Prediction of FiberOrientation in Injection Molded Parts ofShort-Fiber Reinforced Thermoplastics,Polym. Eng. Sci. 30, 16, 957 (1990).

14. Petterson, D. R., Doctoral thesis, M.I.T.,Cambridge, MA, 1958.

15. Pourdeyhimi, B., Ramanathan, R.,Measuring Fiber Orientation in Nonwovens,Part I; Simulation, Textile Res. J. 66 (11),713-722 (1996).


13

16. Rao, Rajiv. S, R. L. Shambaugh, Vibrationand Stability in the Melt Blowing Process,Ind. Eng.Chem. Res. 32, 3100-3111 (1993).

17. Schwarz, E. C. A., “Apparatus and Processfor Melt-Blowing a Fiber formingThermoplastic and Product ProducedThereby”, U.S. Patent 4,380570, 1983.

18. Tate, Brian D., R. L. Shambaugh, ModifiedDual Rectangular Jets for Fiber Production,Ind. Eng. Chem. Res. 37, 3772-3779 (1998).

19. Uyttendale, Marc A. J., R. L. Shambaugh,The Flow Field of Annular Jets at ModerateReynolds Numbers, Ind. Eng. Chem. Res.28, 1735-1740 (1989).

20. Uyttendale, Marc A. J., R. L. Shambaugh,Melt Blowing: General EquationDevelopment and ExperimentalVerification, AIChE Journal. 36, 175-186(199

21. Veerabadran, K. Studies on Uniformity ofFiberweb Basis Weight and Structure,Doctoral Thesis, NCSU, Raleigh, NC, 1994.

22. Velu, Yogeshwar., Ghosh, Tushar., andSeyam., Abdelfattah., “Formation of Melt-Blowing Shaped/Molded Fabric Structures,Part II: Influence of Process Parameters onPore Size Distribution” , Submitted forpublication to Textile Research Journal.

23. Wadsworth, L., and Fagan, J., “Melt blownprocessing and Characterisation ofFluoropolymer Resins”, INDA J. NonwovensRes, 4 (4), 27 – 31.

24. Wadsworth, L. C., “Novel Melt BlownResearch Findings”, Nonwovens Industry,17, (11), 44-51, 1986.

25. Wisneski, T. J., and Morman, M. T.,"Polyolefin Containing ExtrudableCompositions and Methods for theirFormation into Elastomeric ProductsIncluding Microfibers", U.S. patent4,663,220, 1987.

26. Xu, B., Measurement of Pore Characteristicsin Nonwoven Fabrics Using Image Analysis,Clothing and Textile Research Journal., 14,81 – 88, 1996.

AUTHORS’ INFORMATION

Yogeshwar Velu, Raoul Farer, Tushar Ghosh,Abdelfattah SeyamCollege of TextilesNorth Carolina State UniversityRaleigh, NC 27695, USATelephone: (919) 515-6568FAX: (919) 515-3733

FORMATION OF SHAPED/MOLDED MELTBLOWING NONWOVEN … · a fabric structure are of prime importance...

Documents

Transcript of FORMATION OF SHAPED/MOLDED MELTBLOWING NONWOVEN … · a fabric structure are of prime importance...