Abstract
In order to decrease the turbulence influence and the backflow during weft insertion in an air-jet loom, a new main nozzle composed of two nozzle needles connected in series is designed. The internal flow fields of the commonly used nozzle and the new nozzle are numerically studied by means of a three-dimensional model implemented by the computational fluid dynamics technique. It is observed that the turbulence strength is effectively controlled with different air supply pressures and the backflow phenomenon at the upstream passage generated by high supply pressures can be eliminated when the supply pressure is reasonably distributed into two inlets of the new nozzle. The drag force exerted on the weft yarn is calculated when the velocity distribution of the internal airflow is obtained, and results show that the drag force can be improved. Also, the numerical simulation can give some guidelines for future prototyping and experiments.
The main nozzle is a key component in the weft insertion system of an air-jet loom; it works as an ejector, sucking the weft yarn and driving it through the warp shed. It is the first step for the weft yarn to be sucked into the main nozzle. Therefore, the performance of the internal airflow, especially the centerline airflow, will directly influence the efficiency of an air-jet loom and the quality of the fabric. The airflow varies under different air supply pressures, which may cause a stronger turbulence area and generate backflow phenomenon.
Studies on the research and optimization of the main nozzle have been intermittently dealt with over the past few decades. Xue et al. 1 and Liang et al. 2 studied the internal flow field of the main nozzle by use of computational fluid dynamics (CFD) softwares; they obtained the velocity distribution of the airflow along the axial and radial direction, and observed the turbulence and backflow phenomena. Belforte et al. 3 also carried out a numerical simulation to determine which physical model could predict a better behavior of the main nozzle, and they also discussed the influence on the drag force arising from the acceleration tube length, the shape, and the size. Zhu and Zhou 4 analyzed the structure, the velocity, and the static pressure characteristics of the airflow inside the main nozzle. Mohamed and Salam 5 and Salama et al. 6 researched nozzles with various geometries and the corresponding weft insertion behaviors through the tubes both in experiment and theory. They also dispose the influence of air supply pressure and the presence of a yarn tube on the performance of the main nozzle. Recently, more attention has been paid to the flow field analysis and structural optimization of the weft insertion system by our research group at Soochow University. Concretely, Yuan 7 built up a two-dimensional flow field model, and simulated the flow field of weft insertion by the change of parameters, such as the diameters, throat sectional areas, acceleration tube lengths of the main nozzle, and air supply pressures. Guo 8 modeled a three-dimensional flow field, and investigated the main nozzles with different numbers of slots, diameters of the holes for weft insertion, distances between the outlet of the needle and the slots and the air supply pressures. The advantage of different parameter combinations for the structure of the main nozzle and the air supply pressures was also analyzed. Lu 9 provided a method to control the velocity and pressure of the outlet of the main nozzle by changing the throat area of the main nozzle. Xu et al. 10 combined the numerical simulation of the airflow velocity and the density with the experimental test of the drag force to obtain the friction coefficients of different weft yarns. Dong et al. 11 arrived at the conclusion that the numerical simulation results for the flow field obtained by Fluent software are in good agreement with those obtained experimentally.
In this paper, a new main nozzle composed of two nozzle needles connected in series is designed to address the need to control the turbulence strength and to eliminate the backflow phenomenon. The three-dimensional models of the internal flow fields of the commonly used nozzle and the new nozzle are built up by use of the three-dimensional software named Unigraphics NX 8.5. Then the models are meshed by the professional meshing software named ICEM CFD. Finally, the numerical simulation is carried out using Fluent. In order to evaluate the internal airflow performance, the drag force exerted on the weft is calculated. It is observed that the performance of the internal airflow of the new nozzle is optimized compared with that of the commonly used nozzle. Naturally, these numerical simulation results can give some guidelines for future prototyping and experiments.
Structures and flow field models of the main nozzles
The commonly used main nozzle
Structure of the commonly used main nozzle
Figure 1 shows the structure of a commonly used nozzle. The nozzle is composed of a body (1), a needle (2), a cylinder (3) and an acceleration tube (4). Figure 2 shows the internal flow passage of the commonly used nozzle. The compressed air flows from air supply inlet (A) into the chamber (B) of the body. The air is rectified and accelerated when it flows through the polar array of slots (C) on the nozzle needle and the cone-shaped interior passage (D) of the cylinder; the velocity reaches its maximum at the needle tip (E), which generates a negative pressure area to cause the yarn to be sucked into the interior flow passage (F) of the needle. When the air passes away from the needle tip, a strong flow expansion occurs, and the air flow is accelerated in the interior passage (G) of the acceleration tube.
Configuration of the structure of the commonly used main nozzle. Internal flow passage of the commonly used main nozzle.

Flow field model of the commonly used main nozzle
Figure 3 shows the three-dimensional model and the boundary condition of the flow field of the nozzle. The external flow fields were added to the inlet and outlet of the nozzle. The ejection flow of the outlet is cone-shaped, in order to mesh the model more easily and decrease the number of the elements; the model of the ejection flow field of the outlet is cylinder-shaped in steps, including the actual ejection flow. Figure 4 shows the finite element model of the flow field of the nozzle. There are about 1,134,149 elements and 972,828 nodes. Figure 5, obtained by the use of Fluent, shows the velocity distribution of the centerline of internal airflow under different air supply pressures of the commonly used main nozzle. Figure 6, obtained by the use of Fluent, shows the velocity contour under 0.2 MPa air supply pressure of the symmetric surface of the internal flow field of the commonly used nozzle.
Three-dimensional model and boundary condition of the flow field of the commonly used main nozzle. Finite element model of the flow field of the commonly used main nozzle. Velocity distribution of the centerline of internal airflow under different air supply pressures of the commonly used main nozzle. Velocity contour under 0.2 MPa air supply pressure of the symmetric surface of the internal flow field of the commonly used nozzle.



A new main nozzle
Structure of the new main nozzle
In Figure 5, one can find that there is a strong turbulence area at point “S2”, and when the air supply pressure is 0.3 MPa or more, backflow exists in the internal upstream passage of the commonly used nozzle. Obviously, such phenomena are harmful for the weft yarn insertion. In order to overcome such shortcomings, a new main nozzle is designed and is shown in Figure 7. This nozzle consists of a body1 (1), a needle1 (2), two cylinders (3), a body2 (4), a needle2 (5) and an acceleration tube (6). The length of the two needles and the acceleration tube is the same as that of the commonly used nozzle. Figure 8 shows the internal flow passage of the new nozzle. The compressed air flows from inlet2 (G) into the chamber2 (H). The air gets rectified and accelerated, and then rises to the maximum velocity when the air flows through slots (I), the cone-shaped interior passage (J) and gets to the needle tip2 (K) in the new nozzle. When the air passes away from the needle tip2, there is a negative pressure area and a strong turbulence area; thus, a strong flow expansion occurs. The compressed air flows from inlet1 (B) into the chamber1 (C). Just like the airflow described above, the air gets rectified and accelerated, and then arrives at the maximum velocity when the air flows through slots (D), the cone-shaped interior passage (E) and gets to the needle tip1 (F) in the new nozzle; when the air passes away from the needle tip1, there is also a negative pressure area and a strong turbulence area, and a flow expansion also occurs. After the air flows into the nozzle needle2, the interior flow passage (L) works as an acceleration area. After the air from inlet1 and inlet2 meets, the mixed air will be accelerated in the interior passage (M) of the acceleration tube, and the weft yarn is sucked into interior flow passage (A) of the needle1. Since the air supply pressures in inlet1 and inlet2 are adjustable, different airflow can be obtained. Thus, different pressure distribution can be chosen to obtain a suitable airflow to meet the different needs of the manufacturing process.
Configuration of the structure of the new main nozzle. Internal flow passage of the new main nozzle.

Flow field model of the new main nozzle
Figure 9 shows the three-dimensional model and the boundary condition of the flow field of the new nozzle. The flow field of the new nozzle is disposed like that of the commonly used nozzle. Figure 10 shows the finite element model of the flow field of the new nozzle. There are about 754,895 elements and 546,606 nodes.
Three-dimensional model and boundary condition of the flow field of the new main nozzle. Finite element model of the flow field of the new main nozzle.

Numerical simulation of two main nozzles
Simulation setting
To implement the simulation, some assumptions are made as given below:
the air is assumed to comply with ideal gas law; the air viscosity is calculated as a function of absolute temperature T, that is, the outlet static pressure is specified as atmosphere pressure.
The k-ε model is a semi-empirical model developed for high-Reynolds-number flow; there are three variants (standard, RNG and Realizable). The RNG k-ε model can deal with the flow with high strain rate and higher curvature degree.
Due to the variation of the airflow at the wall, turbulent stress almost has no effect on the airflow, especially at the viscous sublayer. Fluent provides three different wall treatments: standard wall function; non-equilibrium wall function; and enhanced wall treatment. Standard wall function is employed to deal with the airflow at the wall.
Air velocity at the inlet under different air supply pressures
Simulation profile
Air supply pressures for the commonly used main nozzle
Air supply pressures for the new main nozzle
Note: Free indicates that the inlet is exposed to the atmosphere, and Closed indicates that the inlet is treated as wall.
Simulation results
Simulation classification
Figure 11 shows the comparison among different conditions under 0.1 MPa total supply pressure. From the figure, one can find that type “I-1” possesses a better performance, the turbulence strength in the downstream area at point “S2” in all types of the new main nozzle is controlled better than that of the commonly used main nozzle, and the maximum backflow velocity caused by the strong turbulence area in the downstream area at point “S2” of the commonly used main nozzle and the new main nozzle is about 108 and 46 m/s, respectively, which implies that the new nozzle is suitable for the weft insertion. The velocity increases at the internal flow passage before point “S1” because of the cone-shaped internal passage. There is not a stronger turbulence area in the types “B-1” and “C-1”, in which the fact of velocity drop is due to the increase of the cross-section area of the airflow passage. However, there is also a velocity drop area in the downstream area at point “S3” in the new main nozzle; for types “B-1” and “C-1”, it is due to the strong turbulence area, whereas for other types it is due to the increase of the cross-section area of the airflow passage. In addition, the fact that inlet2 is free to the atmosphere in type “H-1” also contributes to the velocity drop in such a degree. Generally speaking, the airflow in the types where an inlet is closed plays better than that of the types where an inlet is free, because when an inlet is free, the internal airflow is obviously influenced by the atmosphere through the inlet.
Comparison among different conditions under 0.1 MPa total supply pressure.
Figure 12 shows the comparison among different types under 0.2 MPa total supply pressure. As with the types under 0.1 MPa total supply pressure, the velocity increases in the upstream area at point “S1”, and also drops down in the downstream areas at points “S2” and “S3”. The airflow in types “C-2” and “I-2” plays better than that of other types. The maximum backflow velocity caused by the strong turbulence area in the downstream area at point “S2” of the commonly used main nozzle and the new main nozzle is about 173 and 97 m/s, respectively. However, because the air is distributed into two inlets in type “D-1”, the influence of the strong turbulence area is lower than other ones; the backflow velocity in the downstream area at point “S2” is only about 52 m/s, and drops in the downstream areas at points “S2” and “S3” due to the combination result of the increase of cross-section area and strong turbulence area. The difference between type “D-1” and other types in the new main nozzle shows that the air distribution may improve the internal air performance at the strong turbulence area.
Comparison among different conditions under 0.2 MPa total supply pressure.
Figure 13 shows the comparison among different conditions under 0.3 MPa total supply pressure. Just as the types with 0.1 and 0.2 MPa total supply pressures, the air velocity increases in the upstream area at point “S1”. However, there still exists a backflow phenomenon in the whole upstream area at point “S2” in the commonly used main nozzle. However, there is not a similar phenomenon in types “E-1”, “H-3”, and “I-3”. The maximum backflow velocity caused by the strong turbulence area in the downstream area at point “S2” of the commonly used main nozzle and the new main nozzle is about 282 and 133 m/s, respectively. There are two air distribution methods (Types “E-1” and “D-2”), and the airflow in type “E-1” plays better than that in type “D-2” because the internal flow passage in needle2 works as an acceleration area, and the accelerated airflow can decrease the turbulence strength in the downstream area at point “S3”. Thus, this shows that the preferred airflow may be obtained only in a proper air distribution method.
Comparison among different conditions under 0.3 MPa total supply pressure.
Figure 14 shows the comparison among different conditions under 0.4 MPa total supply pressure. As the air supply pressure increases, the backflow phenomenon is more serious, which does not happen in types “C-4”, “H-4”, and “I-4”. There are three air distribution methods (Types “D-3”, “E-2”, and “F-1”): the airflow in type “F-1” plays better than that in type “E-2” and the airflow in type “E-2” plays better than that in type “D-3”. This is similar to the types under 0.3 MPa total supply pressure. As the air supply pressure of inlet1 is much higher than that of inlet2, the airflow may play much better, because the compressed air with 0.4 MPa supply pressure distributed to inlet1 does not generate a stronger turbulence area to obviously influence the airflow in the whole upstream area at point “S2”, and to seriously influence the airflow in the downstream area at point “S2”. The maximum backflow velocity caused by the strong turbulence area in the downstream area at point “S2” of the commonly used nozzle and the new nozzle are about 409 and 245 m/s, respectively. However, the maximum backflow velocity in some types of the new nozzle is about 151 m/s, except type “D-3”.
Comparison among different conditions under 0.4 MPa total supply pressure.
Figure 15 shows the comparison among different conditions under 0.5 MPa total supply pressure. The maximum backflow velocity caused by the strong turbulence area in the downstream area at point “S2” of the commonly used nozzle and the new nozzle are about 400 and 370 m/s, respectively, whereas the maximum backflow velocity in some types of the new nozzle is about 185 m/s, except types “B-5”, “D-4”, and “E-3”. Furthermore, there is no backflow phenomenon in the upstream area at point “S1” in types “H-5” and “I-5”. The airflow in type “F-2” plays better than that in type “E-3”, and the airflow in type “E-3” plays better than that in type “D-4”, which is similar to that in the types with 0.4 MPa air supply pressure. However, the airflow in type “G-1” is not better than that in type “F-2”, because the air supply pressure distributed to inlet1 has already generated a stronger turbulence area, which influences the airflow in the downstream area at point “S2”.
Comparison among different conditions under 0.5 MPa total supply pressure.
Drag force evaluation
In order to evaluate the internal airflow performance, the drag force exerted on the weft is calculated. The drag force Fdl exerted on one element of the weft yarn is defined as
Physical parameters on a series of points, such as ρ and v, can be obtained from numerical simulation. Figure 16 shows a model of an element connected by two points.
A model of an element connected by two points.
π is a constant, and d is a constant for a kind of weft in theory. Although C
f
varies with the relative velocity between the air velocity and the weft yarn velocity, it only has a small influence on the final drag force according to Guo and Chen.
12
Therefore, C
f
is treated as a constant. In order to simplify the evaluation, it is reasonable to treat C
f
, π and d as 1, and the drag force exerted on an element is defined as
Physical parameters of the element, such as ρ and v, are disposed as
In order to compare the air performance of the new nozzle with that of the commonly used nozzle fully, two kinds of flying status of weft yarn are taken into consideration:
u = 0, which means one side of the weft yarn is fixed; and u = 80 m/s, which means the weft yarn maintains normal flying status.
Drag force of the new nozzle and commonly used nozzle under different total air supply pressures
In order to describe the advantages of the new nozzle, some better and special situations are chosen to compare with those of the commonly used nozzle in Table 5. Considering Equation (1), it is observed that the air density ρ and the air velocity v are the key factors increasing the drag force, whereas the air velocity v plays a more important role. As a consequence, when the air supply pressure is less than 0.3 MPa, the air velocity grows faster and so does the drag force; however, when the air supply pressure is more than 0.3 MPa, the air velocity varies towards a gentle gradient, and the air density grows faster, which results in the drag force increasing more gently. When the drag forces in different types under different air supply pressures are compared with each other, it is observed that a stronger turbulence area, which causes generation of the backflow phenomenon in the upstream area at point “S2”, is generated under high air supply pressure; it obviously influences the increase of the total drag force. It also indicates that when the air is distributed reasonably into two inlets of the new nozzle, a better airflow performance can be obtained; different performance can be obtained due to different air distribution.
Conclusions
In this paper, a new main nozzle composed of two needles connected in series is designed, and the internal airflow field is simulated. Compared with that of the commonly used nozzle, one can conclude the following.
When the air supply pressure is more than 0.3 MPa, there exists a backflow phenomenon in the upstream area at point “S1”, both in the commonly used main nozzle and the new main nozzle. However, when the air is distributed reasonably, the turbulence strength can be effectively controlled under different air supply pressures, and the backflow phenomenon generated by high air supply pressure can be eliminated. Thus, when the air is distributed reasonably, a preferred airflow performance can be obtained. Although the air supply pressure increases, the backflow in the upstream area at point “S2” can influence the increase of the total drag force. In the new main nozzle, the internal flow passage of the needle2 works as an acceleration area, and it plays an important role in the internal airflow performance. Thus, the higher the air supply pressure of inlet1 compared with that of inlet2 is, the better the internal airflow plays when the air supply pressure of inlet1 is less than 0.4 MPa. However, the strong turbulence area generated by high air supply pressure may also influence the airflow performance in the downstream area at point “S2”. Although there exists the backflow phenomenon in the upstream area at point “S2”, a better airflow performance may also be obtained when the backflow is too weak to influence the airflow performance.
Footnotes
Funding
This work was supported by the Priority Academic Program Development of Jiangsu Higher Education Institutions.
