AbstractThe primary goal of this research was to determine opti-

mum processing conditions to produce commercially accept-able melt blown (MB) thermoplastic polyurethane (TPU) non-woven fabrics. The 20-inch wide Accurate Products MB pilotline at the Textiles and Nonwovens Development Center(TANDEC), The University of Tennessee, Knoxville, was uti-lized for this study. MB TPU webs having small fiber diame-ters were obtained from film-forming and highly elastic fiber-forming TPUs and the webs were mechanically strong anddurable to abrasion compared to MB PP. The basic MB processwas found to be fundamentally valid for the MB TPU process;however, the MB process was more complicated for TPU thanPP, because web structures and properties of MB TPUs arevery sensitive to MB process conditions. Furthermore, differ-ent TPU polymers responded very differently to MB process-ing and exhibited different web structure and properties,although uniform elastic MB webs were produced with aver-age fiber diameters as small as 5.0μm. Preliminary researchleading to this study was funded the Army Research Office(ARO) and the findings were used by ARO towards the devel-opment of an elastic chemical protective liner with betteroverall comfort attributes and protection against chemicalwarfare agents.

Key WordsMelt blown, melt blowing, thermoplastic polyurethane,

TPU, elastic nonwoven, chemical protective fabric, ultra-finefibers

INTRODUCTIONBackground

Melt blowing is a one-step process and one of the mostpractical processes for producing microfiber nonwovensdirectly from thermoplastic polymers, in which hot/highvelocity air blows the extruded filament from a die tiptowards a moving conveyer belt or a cylinder. Melt blown(MB) nonwovens have an inherent advantage over spunbond(SB) nonwovens and other conventional fabrics made frommelt or solution spun fibers in that MB fabrics typically haveaverage fiber diameters ranging from 2-6μm compared to 12-50μm with SB webs and conventional textiles.

The concept of the MB process was first introduced in 1956through a Naval Research Laboratory project initiated by VanA. Wente [1956]. MB technology was originally developed toproduce filters composed of microfibers to collect radioactiveparticles from the atmosphere during the early years of thecold war. Since then there has been a renaissance of research,development and commercial production for addressing avariety of applications using MB fabrics in nonwoven prod-ucts.

Process Property Studies OfMelt Blown ThermoplasticPolyurethane Polymers ForProtective Apparel

By *Youn Eung Lee, Ph.D. and Larry C. Wadsworth, Ph.D., The University of Tennessee, Departmentof Materials Science and Engineering, Knoxville, TN, USA

ORIGINAL PAPER/PEER-REVIEWED

2 INJ Winter 2005

* Now with Samsung Fine Chemicals Co., Ltd, Daejeon, Korea 305-3

Zhao [2001] reported that only 20 US Patents were grantedfrom 1950s to 1970s related to MB technologies and products,but the number of patents has remarkably increased to 64 and236 during the 1980s and 1990s, respectively. The increasedunderstanding and requirement for high technological end-uses have accelerated MB research and development, result-ing in an increasing number of applications, such as filtration,thermal insulators, battery separators, oil absorbents, hygieneproducts and composites for protective apparel.

The most important advantage of the MB process is thatfundamentally all thermoplastic polymers can be processedby MB technology [Wadsworth, 1991]. Recently, elastomersincluding thermoplastic polyurethane (TPU) have been afocus of MB research because of their unique properties suchas high elasticity in all directions, good shore hardness for agiven modulus, high abrasion/chemical resistance, excellentmechanical/elastic properties, low stress relaxation, and resis-tance to long-term cyclic flex failure. Furthermore, TPUs haveblood/tissue compatibility, structural versatility andhydrophilic compatibility [Hemphill, 2001].

Many processing methods such as melt spinning, injectionmolding, coating, film blowing, melt blowing, and heat seal-ing work well with TPU materials. However, the consumptionof TPU elastomers is still much less than conventionalpolyurethanes, which must be solvent-spun to produce fibers,because the melt spinning process of elastic fibers is more dif-ficult than other polymers due to their tendency to snap backduring attenuation of the spin-line [Wadsworth, 2002]. This iseven more challenging with MB since the filaments are atten-uated by aerodynamic forces, and the filaments may be dis-continuous and are not positively held by a take-up spindle ornip while in flight to the collector. But the fact that TPUs stillcan be melt spun into fibers makes them much more versatilefor replacing natural rubber thread, solvent-spunpolyurethane, and other more conventional materials for usein biomedical devices, implants, medical applications, andprotective clothing field [Wadsworth, 2002]. Furthermore, theproducts of MB TPU webs with average fiber diameters in therange of 2-6μm provide high filtration and barrier protection

with the additional advantages such as good mechanicalstrength and stretchability.

This research has concentrated on determining optimumprocessing conditions to produce commercially acceptable MBTPU webs for protective apparel and durable elastic textilesbased on preliminary MB work performed at The Universityof Tennessee in cooperation with the U. S. Army NatickSoldier Center and Noveon Inc. [Wadsworth, 2002]. The MBprocess has many processing parameters such as melt/dietemperature, throughput, die geometry, air flow rate, air tem-perature, die-to-collector distance (DCD) and collector speed.The MB TPU process is even more complex in that high elon-gational viscosity, high shear viscosity and the narrow accept-able processing temperature ranges (to prevent strength lossand possible evolution of toxic gases) requires careful manip-ulation of melt and air temperature and air and polymer flowrates.

EXPERIMENTALThis research was performed to optimize the MB TPU pro-

cessing conditions of the 20-inch MB line through evaluationof the web barrier and strength properties. The 20-inch MBline at TANDEC was used with Estane 58245 (TPU245) andEstane 58280 (TPU280) for continuous and uniform webs hav-ing commercially acceptable mechanical properties. Bothresins are polyether-based TPUs which were obtained fromNoveon Inc. TPU245 was designed to produce a monolithicbreathable film to compete with microporous PTFE film;whereas TPU280 was developed as a stronger fiber forminggrade for producing elastomeric yarns to compete with span-dex (Hemphill, 2001). The 20-inch horizontal spinning MBline, as depicted in Figure 1, provides relatively accurate andprecise control of die/air temperatures and polymer through-put because of a microprocessor controller and a positive dis-placement gear pump. The 20” MB die, which is a single rowof spinneret holes drilled by electrical discharge machining(EDM), was inserted into a nose tip with an angle of 60o angle,a linear hole density of 30 holes/inch and an average nozzlehole diameter of 0.368mm. The MB die was configured with

3 INJ Winter 2005

Figure 1SCHEMATIC DIAGRAM OF THE 20-INCH ACCURATE MB PRODUCTS LINE [WADSWORTH, 2002]

an air gap 0.762mm and setback of 0.762mm, as illustrated inFigure 2.

Experimental processing conditions are given in Table 1.With TPU245, die temperature was varied from 207oC to 210oC,while air temperature in the die manifold was varied from210oC to 221oC. The web collection speed was kept at 3.35m/min. Melt pressures after the metering pump varied from

5,929 to 7,929KPa for TPU245. On the otherhand, die and air temperature of TPU280 wasset higher than TPU245. The die temperaturewas held at 232oC and air temperature in thedie manifold was also maintained at a close tol-erance of 241oC to 242oC. Web collection speedsvaried from 1.52 to 16.15 m/min. Melt pres-sures after the metering pump varied from6,136 to 6,895KPa for TPU280. The MB TPU280webs were not sticky and did not have to bewound with Kraft paper, as did the TPU245.

As shown in Table 1, the polymer throughputwith the TPU245 trials was very low at 0.14g/hole/min in an effort to obtain small fiberdiameters, and then was increased to 0.30g/hole/min. The throughput rate of TPU280was 0.34 g/hole/min in Trials 2.1-2.5. The MBTPU fibers and webs were characterized forweight, thickness, tearing and tensile strength,air permeability and fiber diameter.

Results and DiscussionUniform MB webs were produced from all

TPU trials with TPU245 and TPU280 polymers.As shown in Table 2, TPU245 in Trials #1.1through #1.5 resulted in basis weights of 45 to

88 g/m2 with 5.0 to 13.0μm average fiber diameters, andTPU280 in Trials #2.1 through #2.5 resulted in basis weights of52 to 615 g/m2 with average fiber diameters of 5.3 to 14.5μm.For both TPUs, the thickness increased with basis weight, andwith TPU280 air permeability was decreased with basisweight; whereas, there was no clear trend between basisweight and air permeability of TPU245.

4 INJ Winter 2005

Figure 2THE SCHEMATIC DIAGRAM OF AIR GAP AND SETBACK

IN MB DIE [ZHAO, 2001]

Table 1MELT BLOWN PROCESSING CONDITIONS OF TPUS ON THE 20-INCH LINE

Figure 3 shows the relationship betw een MB processingtime and pressure after the pump for TPU245 and TPU280.TPU245 initially resulted in lower pressure after the gearpump as would be expected with a film-forming grade TPUas compared to the fiber-forming grade TPU280, even though

the latter was processed at notably higherdie and air temperatures. However, after 90minutes, the pressure after the gear pump ofthe TPU245 began to increase whereas thepressure with TPU280 leveled off after havingdecreased from start-up to that time. Whenthe die nosepiece was removed after meltblowing TPU245 for 150 min., there was evi-dence of thermal degradation in the melt.This indicated that some degradation ofTPU245 was occurring with time; whereas,no apparent degradation of TPU280 wasobserved. Nevertheless, the MB webs pro-duced with both TPUs were very uniformand showed no sign of degradation.Furthermore, TPU245 MB webs had thesmoothest texture and softest hand. TPUshave very high viscosity compared to PP,and are more subject to thermal degradationwhen the processing temperatures are raisedto reduce the melt viscosity for extrusion offilaments. Thus more optimization studieswill be required to determine if degradationof TPU245 can be minimized.

On the other hand, Figure 4 shows the rela-tionship between the volumetric air flow rate and fiber diam-eter of TPU245 with the same DCD of 64 cm and collectorspeed of 3.35 m/min (#1.3, #1.4 and #1.5) and TPU280 withthe same DCD of 58 cm and collector speed of 7.01 m/min(#2.2, #2.4 and #2.5). Thus, fiber diameters of MB TPUs

5 INJ Winter 2005

0 20 40 60 80 100 120 140 1605000

5500

6000

6500

7000

7500

8000

8500

9000

TPU245

TPU280

Time (min)

Pre

ss a

fter

pum

p o

f TP

U245

(K

Pa)

5000

5500

6000

6500

7000

7500

8000

8500

9000

2.52.42.3

2.2

2.1

1.51.4

1.31.21.1

Pre

ss afte

r pum

p o

f TP

U280 (K

Pa)

Figure 3RELATIONSHIP BETWEEN TIME AND PRESSURE AFTER

PUMP FOR TPU245 AND TPU280

Table 2MELT BLOWN WEB PROPERTIES TPUS PRODUCED ON THE 20-INCH LINE

Note: Standard Deviations are in parentheses. Values in table are the average of five specimens, except for fiber diameters in which 30measurements were made from the two edges and middle of the sample.e.

decreased with an increase of air flow ratewith the same processing conditions for eachpolymer for the same DCD, collector speed,metering pump speed, extrusion rate andnearly the same die and air temperatures.The difference in the fiber diameters betweentwo TPU polymers is primarily considered toresult from different polymer processingtemperatures. The fiber diameters of TPU280are smaller than film grade TPU245 below 140scfm air flow rate (4.00 m3/min), althoughTPU280 has fiber grade rheological proper-ties, but the fiber diameters of TPU245decreased more rapidly than TPU280 with anincrease of air flow rate.

Figure 5 shows longitudinal photomicro-graphs obtained from optical microscopy(YS1-T Nikon) and scanning electronmicroscopy (SEM, Hitachi S-3500) of theTPU245 and TPU280, respectively. Some indi-vidual MB TPU fibers from Trial #1.5(TPU245) and Trial #2.5 (TPU280) were assmall as ~2μm with volumetric air flow rateof 144 scfm (4.11 m3/min) and 150 scfm (4.29

m3/min), respectively.The web strength values of

TPU245, under the same through-put, DCD, and collector speed, aregiven in Figure 6(a). Tear and ten-sile strength increased first, andthen decreased with basis weight.Figure 6(b) shows the web strengthof TPU280; tear strength shows thesame tendency as TPU245, but ten-sile strength increased withincreasing basis weight. The peakelongation of TPU280 web wasalmost three times higher thanelongation of TPU245 web and thepeak tensile force of MB TPU280web almost twice of TPU245 web.

It should be noted that the ten-sile strength of the MB webs islargely decided by tensile strengthof the fiber and is also affected byfiber entanglements and bondingin the MB webs; however, the tearstrength of a MB web is mainlydecided by fiber entanglementsand the ability of the fibers tomove and accumulate in the tear-ing area. The large decrease of tearstrength for Trial #2.4 compared to#2.2, in spite of the increase ofbasis weight, could possibly beexplained by fiber orientation andentanglements since air flow rate

6 INJ Winter 2005

120 125 130 135 140 145 1500

2

4

6

8

10

12

14

1.5

1.4

1.3

TPU245

TPU280

Air flow rate (scfm)

Fib

er

dia

mete

r (u

m)

of T

PU

245

0

2

4

6

8

10

12

14

2.5

2.4

2.2

Fib

er d

iam

ete

r (um

) of T

PU280

Figure 4RELATIONSHIP BETWEEN THE VOLUMETRIC AIR FLOW

RATE AND FIBER DIAMETER OF TPU245 AND TPU280

Optical microscopy (x500)

Optical microscopy (x500)

SEM (x350)

SEM (x350)

(a) TPU245 (Trial #1.5)

(b) TPU280 (Trial #2.5)

Figure 5OPTICAL AND SEM PHOTOMICROGRAPHS OF MB TPUS

of #2.4 was increased from 118 scfm (3.37 m3/min) to 138scfm (3.94 m3/min).

The abrasion resistance of MB TPU fabrics was measuredby a rotary platform table abrasion tester, which had twoheads. 50, 100 and 200 revolutions of the rotating head wereapplied with 250g per wheel of load for each specimen. Then,those specimens were compared to non-cycled samples and toa 40 g/m2 MB PP web.

Figure 7 shows that the MB TPU280 Trial #2.3 (52 g/m2) and40 g/m2 of MB PP webs after applied revolutions of wheel,respectively. The 40 g/m2 MB PP was a non-electrostaticallycharged air filtration-grade web with a median fiber diameterof 2-3 μm, which was obtained from Clean & Science Co., Ltd.,Seoul. Korea. The MB sample had a smooth, uniform surface

texture before abrasion testing. Apparently, theincrease of revolutions caused a clear green circlemark and fragments to MB TPU280 webs, but therewas no development of TPU web breakage. Overall,MB TPU280 webs exhibited substantially more abra-sion resistance than similar PP MB webs.

The light weight webs of TPU245 such as Trial #1.1and #1.2 showed some fiber breakages at a highernumber of revolutions around 200. However, the MBPP webs totally lost mechanical strength after 50 rev-olutions. Table 4-3 shows the tensile properties of theMB TPU web after abrasion. Peak force and peakelongation decreased with increase of revolution ofthe wheels for all TPUs, but the rate of loss in break-ing load was different depending on the type ofTPUs.

Figure 8 represents the rate of tensile strength lossof MB TPUs webs after abrasion. The tensile strengthof MB TPU245 webs decreased with increasing revo-lutions. MB TPU245 webs lost notable tensile strengthfrom 50 to 100 revolutions, and tensile strength losswas more gradual to the test cycle of 200 revolutions.

The tensile strength of fiber-grade MB TPU280 websshowed lower strength loss even after the high revo-lutions. Strength loss was higher for the light weightweb and lower for heavy weight web. The loss ofpeak tensile force for the Trial #2.1 was less than 4%even after 200 revolutions. However, the other trialsof MB TPU280 webs showed relatively higher tensilestrength loss than Trial #2.1, but those webs still con-tained more than 60% of tensile strength compared tothe peak force before abrasion.

ConclusionsMB TPU webs having fiber diameters as low as 5ìm

were obtained from TPU245 (film-grade) andTPU280 (fiber-grade) resins. The webs were mechan-ically strong and durable to abrasion compared to MBPP, and TPU245 MB webs had a smoother texture andsofter hand than webs of TPU280. However, TPU280webs generally had twice the peak load and elonga-tion as did TPU245. Overall, the basic MB processwas fundamentally valid for the TPUs; however, the

MB TPU process was more complicated than MB PP, becauseweb structures and properties of MB TPUs are very sensitiveto process conditions, especially for the die/air temperatureand DCD. Furthermore, the different TPU grades respondedquite differently to MB processing and exhibited differentweb structure and properties.

Both TPU resins resulted in MB webs with high elasticity,high peak tensile strength, and very high peak elongation.Fiber diameter decreased with air flow rate. For both TPUsthe thickness increased with basis weight, and with TPU280air permeability was decreased with basis weight; whereas,there was no clear trend between basis weight and air perme-ability of TPU245.Tensile strength of TPU280 web increasedwith basis weight, but tear strength increased with basis

7 INJ Winter 2005

74 76 78 80 82 84 86 88 90

60

90

120

150

1.4

1.3

1.5

Tear strength Tensile strength

Basis weight (g/m2)

Tea

r st

reng

th (

gf)

0.0

0.3

0.6

0.9

1.2

1.5

Peak tensile force (kgf)

Figure 6RELATIONSHIP BETWEEN BASIS WEIGHT, TEAR

STRENGTH AND PEAK TENSILE FORCE

60 80 100 120 140 160 180

50

100

150

200

250

300

350

400

450

Tear strength Tensile strength

Basis weight (g/m2)

Tea

r st

reng

th (

gf)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

2.42.2

2.5

Peak tensile force (kgf)

(a) TPU245

(b) TPU280

weight from 80 to 140g/m2, and then decreased.

AcknowledgementsThe authors would like to acknowledge Drs. Heidi

Schreduder-Gibson and Phil Gibson, U.S. Soldier SystemsCenter for technical assistance and the melt blown techniciansat TANDEC for their help with the trials. The USDAAgricultural Experiment Station Multi-state Project S-1002and the University of Tennessee Center of Excellence forMaterials Processing are gratefully acknowledged for fundingthis project. Noveon, Inc. is also appreciated for providing theTPU polymers, as well as technical assistance.

References1. Wente, V. (1956). Superfine thermoplastic fibers. Industrial

and Engineering Chemistry, 48(8), 1342-1346.2. Zhao, R. (2001). An investigation of bicomponent

polypropylene/poly(ethylene terephthalate) melt blownmicrofiber nonwovens (Doctoral dissertation, The Universityof Tennessee at Knoxville).

3. Wadsworth, L., Lee, Y., Bresee, R., Gibson, S., & Gibson, P.(2002). Melt blown thermoplastic polyurethane for elastic mil-itary protective chemical liners. International NonwovensTechnical Conferences, Atlanta, USA.

4. Private communication with Susan Hemphill, Noveon,2001.

5. Wadsworth, L., & Malkan, S. (1991). A review on meltblowing technology. International Nonwovens Bulletin (INB)Nonwovens 2 46-52.

6. Wadsworth, L., & Malkan, S. (1991). A review on meltblowing technology. INB Nonwovens 3 22-28. — INJ

8 INJ Winter 2005

Figure 7MB TPU280 AND MB PP SPECIMENS AFTER

ABRASION

0 revolution

0 revolution

50 revolutions

100 revolutions

200 revolutions

(a) TPU280 (52 g/m2) (b) PP (40 g/m2)

50 revolutions

25 revolutions

9 INJ Winter 2005

NA*: Specimen was not available due to failure of abrasion resistance.

Table 3ABRASION RESISTANCE OF MB TPU WEBS

0 50 100 150 200 250

10

20

30

40

50

Loss

in p

eak

tensi

le forc

e (

%)

Revolutions of abrasion

1.1 1.2 1.3 1.4 1.5

0 50 100 150 200 250

10

20

30

40

50

Loss

in p

eak

tens

ile fo

rce

(%)

Revolutions of abrasion

2.1 2.2 2.3 2.4 2.5

(a) TPU245 (b) TPU280

Figure 8LOSS IN PEAK TENSILE FORCE OF MB TPU WEBS AFTER ABRASION