The power required for the process can usually be lowered by reducing velocity variations in the attenuator cross-section. This is difficult to achieve using conventional experimental methods, however, since it is impossible to measure velocity to the level of accuracy that is required. Hills engineers overcame this problem by using computational fluid dynamics (CFD) to determine the effect of design variables such as dimensions, internal air distribution, and compressed air entry angle on the distribution of velocity and other flow variables. Based on the understanding they gained, the engineers developed a large number of new design possibilities using the software, and evaluated the performance of each, finally iterating to a new design that offers dramatic improvements.
Hills, Inc. is a thirty-two year old company devoted to the design, development, and manufacture of machinery and technology for the synthetic fiber industry. The company’s past custom machinery projects include: large scale staple spinning machines; large scale spunbond spinning systems with quench; bicomponent spinning systems to produce virtually any polymer arrangement; spinning machine components including spin packs, polymer filters, meter pump drives, electrical control cabinets, electrical and Dowtherm heated polymer transfer lines; and the modification of existing plants to update machines or convert to different processes.
Spunbond is the most cost effective method of making a fabric from melt-spinnable polymers. In the process, polymer is prepared, melted, extruded, drawn, and quenched into continuous, oriented filaments, which are deposited on a moving screen. The fabric is then bonded in a variety of ways, and formed into products that range from wall coverings to disposable wipes. About 200 production machines are in operation today. John R. Starr, Inc. recently stated that worldwide polypropylene spunbond volume has grown by an average of 12% per year for the last 12 years.
The biggest design challenge
The biggest challenge in attenuator design is that the fiber is accelerated very quickly to a high speed - from zero to 5000 meters per minute or more in a second or less. This means that any measurements taken in the device would alter the flow enough to make the measurements virtually useless. The flow patterns are critical, however, because the compressed air used to accelerate the fabric is the second highest cost input to the process, after materials. Production cost is of enormous importance in this high-volume, low-margin industry, so equipment designers such as Hills are continually trying to improve their product by reducing power consumption, among other things. Without being able to visualize the flow patterns and detect variations that reduce efficiency, they are left with a process that is largely trial and error. Since it is very expensive to build and test a prototype design, the degree to which the design can be optimized is limited.
Hills engineers recognized the potential of CFD to improve the design process. CFD provides fluid velocity, temperature, and other relevant variables throughout the solution domain for problems with complex geometries and boundary conditions. As part of the analysis, a researcher may change the system geometry or the boundary conditions and view the effect on fluid flow patterns, temperatures, or the distributions of other variables. Hills engineers use FLUENT CFD software from Fluent Incorporated, Lebanon, New Hampshire, because its graphical user interface greatly reduces the amount of time and training required to set up the analysis. FLUENT’s parallel processing capabilities take full advantage of workstations with multiple processors to reduce computational time to a degree that is almost linear with the number of processors available.
Validating CFD accuracy
To validate the ability of the software to accurately model conditions within an attenuator, engineers simulated a well-known classical attenuator described in a patent filed in 1972 by Asahi. The CFD model accurately predicted the performance of this well-understood design. For example, the predicted flow patterns in the injection zone, where compressed air enters the fiber channel, suggested the existence of fiber tangling and roping due to eddy backflow. The model also predicted that approximately 25 mm below the injection zone, a smooth velocity profile develops that continues to the exit. These predictions match the performance seen in actual implementations of this design. Hills engineers also simulated more recent slot attenuator designs that improve on the original Asahi patent, and the models predicted smoother entry of the compressed air along the wall, reducing the likelihood of fiber tangles and ropes relative to the early Asahi design. This correlates well with the performance actually seen with these designs.
Confident in the ability of CFD to predict attenuator performance, Hills engineers have begun using the simulation tool to provide insights that allow design refinements that would not be possible otherwise. In one recent design, for example, the scalloped appearance of air velocity contours predicted by CFD revealed a slight periodic nonuniformity of the air velocity due to some upstream features. The effect was so slight -- +/-2% at a constant Z height – that it could not be measured with the best instruments available. Based on this information, Hills engineers optimized the internal airflow distribution passages in order to reduce the velocity variations. After many design iterations performed by altering the CFD model and re-running the analysis, engineers were able to achieve near-perfect uniformity. They also used CFD to evaluate forces due to air pressure acting on the internal surfaces of the aspirator body parts in order to ensure that mechanical deflections are within acceptable levels.
Optimizing the downstream area
Hills engineers also used CFD to optimize the critical area downstream of the attenuator where the air and fiber flow at very high speeds. The objective in this region is to slow the fibers down and lay them on the moving collection screen in a planned manner, to produce the desired fiber orientation and fabric properties. In early design iterations, engineers observed high shear between the jet core and the entrained air, allowing unstable eddies to form and grow. It was apparent that these eddies would cause the fiber to become nonuniform in the flow stream before striking the screen. Being able to visualize the problem made it relatively easy for Hills engineers to change various design parameters to develop a solution.
The appearance of spunbond fabric can also be affected by the distance between the attenuator and the collection screen. As the distance increases, the nonuniformities in the fabric become more pronounced. Very short forming distances can lead to other problems, however. For example, when the high speed jet impacts on the collection screen, it can disturb previously deposited fibers. Hills engineers used CFD to develop and optimize the formation systems in these critical areas.
As a result of being able to visualize fluid flow and pressure throughout the attenuator regions, Hills engineers were able to substantially improve the performance of a recent custom attenuator design. In the past, they would have developed a few prototypes based on previous experience and use the one that worked best. With CFD, they were able to develop dozens of software prototypes, determine the flow velocity and pressure at every point in their flow domain, and iterate to an optimized solution at a much faster pace. The most noticeable improvement in the new design is that it makes better use of the momentum of air to improve energy efficiency and flow uniformity. This lowers the initial cost of the equipment, reduces the cost of energy required to operate it, and also lowers noise pressures, improving working conditions for the operators.
Schematic of a typical spunbond process.
CFD simulation images showing contours of velocity magnitude.