Article pubs.acs.org/IECR
Melt Blowing Dies with Louvers Robert L. Shambaugh,* John D. Krutty, and Shawn M. Singleton School of Chemical, Biological, and Materials Engineering, The University of Oklahoma, 100 East Boyd Street, SEC T301, Norman, Oklahoma 73019, United States ABSTRACT: A slot die is the most common type of spinning device used to make melt blown fibers. A pair of louvers was installed in the air flow field of a melt blowing die. Air velocity measurements were taken in the presence of the louvers and in the absence of the louvers. In some experiments, the louvers were parallel with the airflow. In other experiments, the louvers were angled relative to the airflow at angles up to 6°. If higher air velocities can be achieved with louvers, then the melt blowing process can be improved.
■
and polymer filaments move downward. Horizontal melt blowing is also common. The high-velocity airfield is what drives the melt blowing process. As shown in Tate and Shambaugh,4 there is a rapid decay of the air velocity field below a melt blowing die. The temperature field decays similarly; see Tate and Shambaugh.5 If this decay could be reduced, less air would be needed to fabricate the fibers. Using less air would improve the economics of melt blowing because the cost of compressing, heating, and recycling the air is significant. Shambaugh et al.6,7 used an advanced model for melt blowing to predict the effects of velocity and temperature plateaus in the airfield below a melt blowing die. A plateau was (mathematically defined as) a vertical range over which the air velocity and temperature were constant. For wide plateaus placed near the die face, the effect of the plateau was substantial. Final (product) fiber diameters were reduced (by up to 50%) versus diameters for when there was no plateau. Typical model input conditions involved a 2 cm plateau placed between z = 2 and z = 4 cm (see the coordinate system shown on Figure 1). This paper discusses the use of baffles, or louvers, placed below the melt blowing die; see the lower half of Figure 1. The goal of the louvers was to experimentally achieve the plateaus discussed in the above-described modeling work of Shambaugh et al.7 The louvers were intended to physically prevent spreading of air in the space between the louvers. Without spreading (and ambient air entrainment), the air field velocity would not decay in the controlled space between the louvers. Other investigators have considered ways to modify and improve the Exxon type melt blowing die. These modifications are generally of two types. The first type is a “bolt on” addition to the face of the Exxon type of die; for example, see Chelikani and Sparrow8 and Meyer et al.9 The second type involves modification of the die itself; for example, see Arseneau and Johnston10 and Gerking.11 Modifications of this second type, which usually involve changes in the air channels of the die, are more extensive (and more expensive) than modifications of the
INTRODUCTION Melt blowing is a common process used in industry to produce nonwoven fibers. In melt blowing, heated gas streams impinge upon molten filaments as the filaments emerge from a heated die. The force of the gas upon the filaments causes rapid attenuation of the filaments to fine diameters. The attenuated fibers are cooled and captured upon a mesh screen placed some distance away from the die. The resulting nonwoven mass of fibers can be used as a filter, an absorbent medium, reinforcement, or numerous other uses (see Shambaugh,1 Buntin et al.,2 and Bresee and Ko3). The most common type of melt blowing die is the slot, or “Exxon” die; see Tate and Shambaugh.4 The upper half of Figure 1 is a cross-sectional view of such a die. The center piece is called
Figure 1. Melt blowing die with louvers added.
the nosepiece. Two other pieces (the air plates) are located to the left and right of the nosepiece. The spaces between the nosepiece and the air plates are the air slots. Hot air flows through the two slots and impacts polymer filaments as they emerge from the bottom of the nosepiece (the polymer orifices are not shown in Figure 1). The die width (in the direction perpendicular to the plane of Figure 1) can range from centimeters to a meter or more. Figure 1 shows a vertical orientation of the die wherein the air © 2015 American Chemical Society
Received: Revised: Accepted: Published: 12999
September 11, 2015 November 25, 2015 December 3, 2015 December 17, 2015 DOI: 10.1021/acs.iecr.5b03400 Ind. Eng. Chem. Res. 2015, 54, 12999−13004
Article
Industrial & Engineering Chemistry Research
shape has an overall thickness “e” of 1.588 mm (1/16 of an inch) and a chord width “c” of 20 mm. The airfoil is symmetric (the top and bottom are mirror images of each other), has a half-circle leading edge, and has a trailing edge that is a simple wedge. For clarity, Figure 3a has an exaggerated vertical scale. Figure 3b shows the airfoil with the correct e/c ratio (i.e., the figure is true scale). As you can see, the actual airfoils are much thinner than the blunt louvers shown in Figure 1. Of course, much more complicated airfoil designs could have been used (.Abbott and von Doenhoff12), but the shape shown in Figure 3 is simple to fabricate and proved to be effective in melt blowing.
first type. The Results and Discussion section of this paper contains a comparison of our work with past work of the first and second types.
■
EXPERIMENTAL EQUIPMENT AND PROCEDURES A die test stand was used to support the melt blowing die used in the experiments. The setup was similar to that shown in Figure 1 in Tate and Shambaugh.4 The die had a horizontal gap spacing (“b” in Tate and Shambaugh4) equal to 0.75 mm and a slot length (“S ” in Tate and Shambaugh4) of 50 mm. The air velocity was measured with a Pitot tube that had an outer diameter of 0.71 mm, an inner diameter of 0.45 mm, a length of 6.35 cm, and a conical nose with a cone angle of 25°. The pressure in the Pitot tube was measured with a digital pressure gauge. The pressure was referenced to ambient static pressure, and the tube was oriented vertically during the measurements. The Pitot tube was positioned with a Velmex three-dimensional traverse system that permitted x, y, and z motions in 0.01 mm increments. The coordinate system for the experiments is shown in Figure 1. The origin is at the center of the face of the die. The z direction is vertical, and the x direction is parallel to the face of the die. Perpendicular to the plane of Figure 1 is the y direction (not shown). The air flow rate, at standard conditions of 21 °C and 1 atm pressure, was maintained at either 1.25 × 10−3 or 1.67 × 10−3 m3/s (75 or 100 L/min). Two louvers are shown in the bottom half of Figure 1. These louvers were placed in the flow field below the melt blowing die. In Figure 1, the view of the two louvers is “end on”, or crosssectional. The length of the louvers (in the direction perpendicular to the plane of the figure) was 76.2 mm (3 in.), and this length was centered such that the louvers completely covered the length of the die slots (50 mm). For a commercial melt blowing die, where S might be on the order of 1 m, the louver length would be somewhat larger than 1 m. In other words, for the louvers to completely control the flow field over the entire length of a slot die, the louver lengths must always be larger than the slot length. The louvers were fabricated from basswood. As suggested in the work of Shambaugh et al.,6,7 the louvers used in the present study were 2 cm in height and were placed between z = 2 and z = 4 cm. Louvers with simple rectangular cross sections are shown in Figure 1. Because of the turbulence and high drag coefficient of bluff bodies, it is advantageous to replace the rectangular louvers with aerodynamically smooth louvers. Therefore, instead of using blunt louvers of the type shown in Figures 1 and 2, airfoil-shaped louvers were used instead. Specifically, the shape shown in Figure 3 was used. In the parlance of airfoil design (Abbott and von Doenhoff12), this
■
RESULTS AND DISCUSSION For the base case when no louvers are present, Figure 4 shows the air velocity distribution below the melt blowing die. Velocity profiles are given for z positions ranging from 5 to 30 mm below the die. Near the die, the profiles are quite sharp, but the profiles get flatter and more spread out as the distance from the die increases. For positions that are away from the die, significant amounts of ambient air are entrained into the air jet. The profiles shown in Figure 4 are typical for melt blowing dies (see Tate and Shambaugh4). Figure 5 shows air velocities for when louvers are present. Two louvers were placed below the melt blowing die in the manner shown in Figure 2. Each of these louvers had the airfoil shape that is shown in Figure 3. All velocity profiles were taken with f1 = 12 mm. Profiles are shown for f 2 = 12, 10, and 8 mm. For comparison, a “no-louver” profile is also shown. The airflow rate was 100 L/min. The airfoil louvers had their leading edges at z = 20 mm (z1 = 20 mm), and the velocity measurements were taken at the trailing edge of the louvers (at z = 40 mm). Taking the velocity measurements at this trailing position shows definitively whether the presence of the louvers has prevented decay in the air velocity within (between) the louvers. From continuity (and for when f1 = f 2), the average velocity between the trailing edges is equal to the average velocity in the zone between the louvers (from z = 20 mm to z = 40 mm). Thus, the louvers cause a plateau in the air velocity. This type of plateau was hypothesized as a boundary condition in the theoretical work of Shambaugh et al.7 For the positioning of the airfoils (Figure 3), values of f1 and f 2 were based on measurements for a rectangular approximation of the airfoil shape. Specifically, the f1 and f 2 values were set based on a rectangular louver with a thickness of 1.588 mm and a chord width of 20 mm (i.e., the overall dimensions of the airfoil shown in Figure 3). Because of the rounding of the leading edges of the airfoils, the actual distance between the tops of the airfoil pair is slightly greater than f1. Likewise, because of the tapering of the trailing edges of the airfoils, the actual spacing between the trailing edges is slightly greater than f 2. For the various velocity profiles on Figure 5, the values of f1 and f 2 are indicated by nomenclature of the form “f1−f 2“. Thus, “12−10” is shorthand for f1 = 12 mm and f 2 = 10 mm. Figure 5 shows that, for 12−12 (when the louvers are parallel and vertical with f1 = f 2), the centerline (x = 0) air velocity is 9.53 m/s. This is about 10% higher than the 8.67 m/s velocity for when no louvers are present. Therefore, the presence of louvers is significant. For 12−10 and 12−8 (when f 2 = 10 mm and f 2 = 8 mm, respectively), the centerline velocities are essentially the same as for 12−12. Therefore, when f1 = 12 mm, angling the louvers has little effect on the centerline velocity (for the range of f 2 values considered).
Figure 2. Melt blowing die with angled louvers. 13000
DOI: 10.1021/acs.iecr.5b03400 Ind. Eng. Chem. Res. 2015, 54, 12999−13004
Article
Industrial & Engineering Chemistry Research
Figure 3. Airfoil louver. (a) In this cross-sectional view, the vertical scale is exaggerated for clarity (the vertical scale is twice as large as the horizontal scale). For this airfoil, e = 1.588 mm and c = 20.0 mm (see Figures 1 and 2). (b) True scale version of the louver shown in panel a.
Figure 6. Air velocity below louvers with f1 = 10 mm. The airfoil louvers (see Figure 3) had their leading edges at z = 20 mm (z1 = 20 mm), and the velocity measurements were taken at the trailing edge of the louvers (at z = 40 mm). Velocity profiles are shown for f 2 = 10, 8, and 6 mm. For comparison, a “no-louver” profile is also shown. The airflow rate was 100 L/min.
Figure 4. Air velocity profile below the melt blowing die when no louvers are present. The airflow rate was 75 L/min.
9.12, 9.69, and 10.0 m/s, respectively. These compare with a nolouvers velocity of 8.67 m/s. Unlike the situation in Figure 5 (for f1 = 12 mm), angling the louvers (decreasing f 2) does increase the velocity. Figure 7 shows results for louvers that are placed even closer together (with f1 = 8 mm). For 8−8, 8−6, and 8−4, the centerline velocities are 9.81, 10.8, and 11.0 m/s, respectively. The nolouvers velocity is 8.67 m/s. As with the situation for f1 = 10 mm (as shown in Figure 6), angling the louvers causes further increases in velocity. Also, the centerline velocities (for all f 2 values) are higher for f1 = 8 mm than for f1 = 10 or f1 = 12 mm. Figure 8 shows the effects of a further decrease of f1 to 6 mm. For 6−6, 6−4, and 6−2, the centerline velocities are 10.0, 11.3, and 12.3 m/s, respectively. These compare with a no-louvers velocity of 8.67 m/s. Again, overall velocities are higher than for conditions where f1 is larger (see Figures 5−7). Also, angling the louvers causes further increases in velocity. Figure 9 shows velocity profiles with the louver separation further decreased to f1 = 4 mm. For 4−4 and 4−2 the centerline velocities are 10.6 and 11.7 m/s, respectively. The no-louvers velocity is 8.67 m/s. As observed previously, angling the louvers increases the velocity.
Figure 5. Air velocity below louvers with f1 = 12 mm. The airfoil louvers (see Figure 3) had their leading edges at z = 20 mm (z1 = 20 mm), and the velocity measurements were taken at the trailing edge of the louvers (at z = 40 mm). Velocity profiles are shown for f 2 = 12, 10, and 8 mm. For comparison, a “no-louver” profile is also shown. The airflow rate was 100 L/min.
Figure 6 shows a family of profiles for when the louvers are closer together. For these profiles, f1 = 10 mm, while f 2 is 10, 8, or 6 mm. For 10−10, 10−8, and 10−6, the centerline velocities are 13001
DOI: 10.1021/acs.iecr.5b03400 Ind. Eng. Chem. Res. 2015, 54, 12999−13004
Article
Industrial & Engineering Chemistry Research
Figure 7. Air velocity below louvers with f1 = 8 mm. The airfoil louvers (see Figure 3) had their leading edges at z = 20 mm (z1 = 20 mm), and the velocity measurements were taken at the trailing edge of the louvers (at z = 40 mm). Velocity profiles are shown for f 2 = 8, 6, and 4 mm. For comparison, a “no-louver” profile is also shown. The airflow rate was 100 L/min.
Figure 9. Air velocity below louvers with f1 = 4 mm. The airfoil louvers (see Figure 3) had their leading edges at z = 20 mm (z1 = 20 mm), and the velocity measurements were taken at the trailing edge of the louvers (at z = 40 mm). Velocity profiles are shown for f 2 = 4 and 2 mm. For comparison, a “no-louver” profile is also shown. The airflow rate was 100 L/min.
Figure 8. Air velocity below louvers with f1 = 6 mm. The airfoil louvers (see Figure 3) had their leading edges at z = 20 mm (z1 = 20 mm), and the velocity measurements were taken at the trailing edge of the louvers (at z = 40 mm). Velocity profiles are shown for f 2 = 6, 4, and 2 mm. For comparison, a “no-louver” profile is also shown. The airflow rate was 100 L/min.
Figure 10. Air velocity below louvers with f1 = 3 mm. The airfoil louvers (see Figure 3) had their leading edges at z = 20 mm (z1 = 20 mm), and the velocity measurements were taken at the trailing edge of the louvers (at z = 40 mm). The velocity profile is shown for f 2 = 0 mm. For comparison, a “no-louver” profile is also shown. The airflow rate was 100 L/min.
Figure 10 shows a velocity profile for when the louvers are quite close together with f1 = 3 mm and f 2 = 0 mm. (As explained above in the discussion of Figure 5, f 2 = 0 mm does not mean that there is zero distance between the trailing edges of the louvers.) For 3−0, the velocity is 13.9 m/s. This velocity is 60% higher than the no-louvers velocity of 8.67 m/s. Tate and Shambaugh4 defined discharge velocity (face velocity) as the air flow rate divided by the area of the discharge slots in a melt blowing die. For an airflow of 100 L/min, our die had a discharge velocity of vjo = 22.3 m/s. Figure 11 shows the effect of louver angle β on v/vjo, the dimensionless centerline velocity. As shown in Figure 2, β is the complement of the angle α (i.e., β = 90 − α). Vertical (parallel) louvers correspond to β = 0. Figure 11 shows that, as the louvers are moved closer together (when f1 is decreased), the effect of increasing β becomes more pronounced. Because of the use of dimensionless velocity on the ordinate, Figure 11 can be used to estimate the effect of louvers for a very wide range of face velocities. The louvers act as baffles (barriers) to prevent ambient air entrainment along the length of the louvers. From continuity, air
Figure 11. The dimensionless air velocity as a function of f1 and Beta (β). When β = 0, f 2 = f1. When β > 0, f 2 < f1. The airflow rate was 100 L/ min.
13002
DOI: 10.1021/acs.iecr.5b03400 Ind. Eng. Chem. Res. 2015, 54, 12999−13004
Article
Industrial & Engineering Chemistry Research
(4) Large amounts of air are used in melt blowing. Because of the high costs associated with compressing, heating, and recycling this air, the economic value of louvers could be significant.
which enters the top of the louvers equals the air which leaves the bottom of the louver pair. Hence, the velocity between the louvers is maintained at a higher level than when there are no louvers present. Also, the air temperature is maintained within the louvers. In the Introduction, other types of modifications to melt blowing dies were mentioned. Type 1 modifications involve relatively simple, bolt-on additions to dies.Type 2 modifications are more complex. Chelikani and Sparrow’s8 type 1 modification involves controlling edge effects during fiber laydown by using flat or convex plates. Through the use of the Coanda effect, plates placed at the ends of the die cause the endmost fibers to be drawn outward such that there is a more uniform laydown pattern at the edge of the collected sheet. Unlike our work, which has the plates parallel to the major axis of the die nosepiece, their plates are at the end of the die and perpendicular to the axis of the nosepiece. Their plates are quite far apart. For example, their plates would be about 1 m apart for a 1 m die, while our louver plates are separated by a centimeter or less. Furthermore, their work does not address improving melt blown fiber formation by reducing air requirements. Meyer et al.9 suggest a type 1 modification in which a boxlike chamber is mounted below a melt blowing die. Typical chamber dimensions are as follows: an x-direction (see Figure 1) of 1.3 cm, a z-direction length of 46 cm, and a width that is greater than the die width (e.g., greater than 1 m for a 1 m die). This thickness of 1.3 cm is larger than our maximum thickness (louver spacing) of f1 = 0.8 cm. Also, their chamber length of 46 cm is much larger than our 2 cm length. Their modification requires secondary air to draw the filaments through the chamber. The purpose of their box is to produce fibers of greatly improved strength. In contrast with our louvers, their equipment is not directed toward reducing air consumption during melt blowing. Arseneau and Johnston10 patented a type 2 modification wherein the melt blowing die has S-shaped air inlet channels with Coanda bends. This configuration allows uniform, parallel (shear) flow of air against filaments. The goal is a more uniform nonwoven product. In another embodiment of their invention, a second pair of air inlets is provided near the die tip (nosepiece). In comparison with our work, their goal is not to reduce air consumption. Gerking11 patented a type 2 modification wherein the die is fabricated as a Laval nozzle. Claims include the ability to produce fine, continuous threads without excessive temperatures (which cause polymer degradation) and with less expenditure of energy. Supersonic flow can occur in the Laval nozzles.11,13 Our louvers are thin, or, more specifically, they have a small e/c value (see Figure 3b), and they have a tight (small) radius on their leading edges. Thus, for when the louvers are oriented vertically (f1 = f 2), the louvers are much like sharp-edged flat plates (that separate the flow). When f1 > f 2, the louver pair can act as a nozzle.
■
AUTHOR INFORMATION
Corresponding Author
*Tel.: (405) 325-6070. Fax: (405) 325-5813. E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
NOMENCLATURE c = louver chord width as defined in Figures 1 and 2, mm e = louver thickness as defined in Figures 1 and 2, mm f = louver spacing for parallel louvers (see Figure 1), mm f1 = louver spacing at top of louvers (see Figure 2), mm f 2 = louver spacing at bottom of louvers (see Figure 2), mm v = velocity, m/s vjo = discharge velocity, m/s x = Cartesian coordinate defined in Figures 1 and 2, mm y = Cartesian coordinate direction perpendicular to the planes of Figures 1 and 2, mm z = distance below the die as defined in Figures 1 and 2, mm z1 = distance from the die face to the leading edge of the louvers, mm z2 = distance from the die face to the trailing edge of the louvers, mm
Greek Symbols
■
α = angle as defined in Figure 2, degrees β = 90 − α, degrees
REFERENCES
(1) Shambaugh, R. L. A Macroscopic View of the Melt Blowing Process for Producing Microfibers. Ind. Eng. Chem. Res. 1988, 27, 2363. (2) Buntin, R. R.; Keller, J. P.; Harding, J. W. Nonwoven Mats by Melt Blowing. U.S. Patent 3,849,241, Nov 19, 1974. (3) Bresee, R. R.; Ko, W. C. Fiber Formation During Melt Blowing. Int. Nonwovens J. 2003, 12, 21. (4) Tate, B. D.; Shambaugh, R. L. Modified Dual Rectangular Jets for Fiber Production. Ind. Eng. Chem. Res. 1998, 37 (9), 3772−3779. (5) Tate, B. D.; Shambaugh, R. L. Temperature Fields below Melt Blowing Dies of Various Geometries. Ind. Eng. Chem. Res. 2004, 43 (17), 5405−5410. (6) Shambaugh, B. R.; Papavassiliou, D. V.; Shambaugh, R. L. NextGeneration Modeling of Melt Blowing. Ind. Eng. Chem. Res. 2011, 50 (21), 12233−12245. (7) Shambaugh, B. R.; Papavassiliou, D. V.; Shambaugh, R. L. Modifying Air Fields to Improve Melt Blowing. Ind. Eng. Chem. Res. 2012, 51, 3472−3483. (8) Chelikani, S.; Sparrow, E. M. Numerical Simulations of Plane-Wall Coanda Effects for Control of Fiber Trajectories in the Melt-Blown Process. Ind. Eng. Chem. Res. 2013, 52, 11639−11645. (9) Meyer, D. E.; Krueger, D. L.; Bodaghi, H. Oriented Melt-Blown Fibers, Processes for Making Such Fibers, and Webs Made from Such Fibers. U.S. Patent 4,988,560, Jan 29, 1991. (10) Arseneau, W.; Johnston, G. W. Method of Melt Blowing Polymer Filaments through Alternating Slots. U.S. Patent 6,562,282, May 13, 2003. (11) Gerking, L. Method and device for the production of an essentially continuous fine fiber. U.S. Patent 6,800,226, October 5, 2004. (12) Abbott, I. H.; von Doenhoff, A. E. Theory of Wing Sections; Dover Publications: New York, 1959.
■
CONCLUSIONS (1) The air flow field below a melt blowing die can be modified by placing louvers in the airfield. (2) These louvers can maintain the centerline air velocity at higher values than for melt blowing without louvers. For the configurations tested in the experiments, the centerline air velocity at the trailing edge of the louvers was as much as 60% higher than for when no louvers were present. (3) Higher centerline velocities mean that less air is needed to melt blow a given amount of nonwoven fiber. 13003
DOI: 10.1021/acs.iecr.5b03400 Ind. Eng. Chem. Res. 2015, 54, 12999−13004
Article
Industrial & Engineering Chemistry Research (13) Tan, D. H.; Herman, P. K.; Janakiraman, A.; Bates, F. S.; Kumar, S.; Macosko, C. W. Influence of Laval Nozzles on the Air Flow Field in Melt Blowing Apparatus. Chem. Eng. Sci. 2012, 80, 342−348.
13004
DOI: 10.1021/acs.iecr.5b03400 Ind. Eng. Chem. Res. 2015, 54, 12999−13004
