The research object investigated in this study is a 300 MW-class tangentially fired coal-fired boiler, with the model number of SG-1025/17.5-M4006. The boiler adopts a subcritical, single intermediate reheat, natural circulation, single-furnace, balanced-draft, and dry bottom ash discharge configuration. A low-NO_X concentric firing system is employed, in which burners are arranged at four corners with tangential firing technology. Tilting burners are additionally equipped to regulate the in-furnace aerodynamic field. The pulverizing system adopts a medium-speed mill positive-pressure direct-firing configuration, equipped with five coal mills, one of which serves as a standby unit. The coal mill model is ZGM95N-I medium-speed mill, coupled with CS2024HP electronic gravimetric belt feeders. The furnace cross-sectional dimensions are 12.800 m × 11.890 m, and the furnace height is 51.422 m. The original boiler combustion system consists of five layers of primary air nozzles, six layers of secondary air nozzles, one layer of compact over-fire air nozzles, and three layers of separated over-fire air nozzles. Based on the characteristics of the primary air-pulverized coal transport system of the investigated boiler, a swirl-straight flow synergistic combustion stabilization system was designed in this study, and the flow characteristics of the louver-type pulverized coal distributor incorporated in the system were investigated. The main design parameters of the boiler are listed in Table 1, and the fuel properties are summarized in Table 2.

Table 1. Main design parameters of the boiler.

Table 2. Coal properties analysis.

To investigate the internal flow characteristics of the louver-type pulverized coal distributor, a computational fluid dynamics (CFD) method was employed to simulate the isothermal airflow behavior under cold-state conditions inside the distributor. Since this study primarily focuses on the velocity distribution, pressure variation, and flow redistribution characteristics of the primary air stream after passing through the louver structure, complex factors associated with pulverized coal particles were simplified. Accordingly, a single-phase incompressible air model was established. During the calculation process, the governing equations were discretized and solved using the finite volume method (FVM). The internal flow field distribution of the distributor was obtained through numerical simulations, and the reliability of the developed model was further validated by comparison with experimental results.

The flow inside the louver-type pulverized coal distributor represents a typical complex turbulent flow within a confined space. After entering the distributor, the primary air flow is redirected under the guidance of the louver blades. Meanwhile, due to the blockage effect of the blades and variations in flow passage cross-sectional area, complex flow phenomena, including velocity gradients, local pressure losses, and secondary flows, are generated inside the distributor. Therefore, an appropriate turbulence model is required to accurately describe the flow characteristics within the distributor.

Turbulence Model:

The fluid motion satisfies the conservation equations of mass, momentum, and energy. The governing equations are expressed as follows [26].

Continuity equation:

where u i represents the mean velocity of the airflow in the i direction (m/s), and u j represents the j direction component of the velocity vector.

Momentum equation:

where x j represents the j direction component of the velocity vector; u j represents the mean velocity of the airflow in the j direction (m/s); p represents the pressure (Pa); μ represents the dynamic viscosity (kg/(m∙s)); ρ s represents the density of the fluid (kg/m³); and F s i represents the external force component (N).

Energy equation:

where C p represents the specific heat capacity of the fluid (J/(kg∙K)); T represents the temperature (K); λ represents the thermal conductivity (W/(m∙K)); ω s represents the component reaction rate; and Q s represents the component heat source (J).

The realizable k-ε model was selected as the turbulence model for the numerical simulations in this study [27]. Compared with the standard k-ε model, the realizable k-ε model exhibits superior prediction capability for complex flows involving swirl, jets, flow separation, and strong velocity gradients. It can more accurately capture the flow separation, secondary flow, and velocity redistribution induced by the louver blades. Therefore, the realizable k-ε model was employed to simulate the internal flow characteristics of the louver-type pulverized coal distributor. The corresponding transport equations are expressed as follows:

where k represents the turbulent kinetic energy (m²/s²); ε represents the turbulent kinetic energy dissipation rate (m²/s³); μ t represents the turbulent viscosity (Pa∙s); σ k and σ ε represent the turbulent Prandtl numbers for the k and ε equations, respectively; G k and G b represent the generation terms of turbulent kinetic energy caused by mean velocity gradients and buoyancy effects, respectively; Y m represents the contribution of fluctuating dilatation in compressible turbulence; S k and S ε represent user-defined source terms; G 1 ε , G 2 ε , G 3 ε , and P r t represent empirical constants; β represents the thermal expansion coefficient; v represents the kinematic viscosity; and g i represents the gravitational acceleration component in the i direction.

The parameters of the actual equipment and the scaled-down equipment are listed in Table 3.

Table 3. Parameters of the actual equipment and the scaled-down equipment.

To evaluate the internal flow characteristics of the actual equipment and the scaled-down equipment, the Reynolds number is adopted as the criterion for flow similarity, and the calculation equation is given as follows [28]:

where ρ represents the air density, kg/m³; U represents the average inlet velocity of the distributor, m/s; D h represents the hydraulic diameter, m; and μ represents the dynamic viscosity of air, Pa∙s.

To ensure that the flow interaction between the swirl side and the lean side in the actual equipment and the scaled-down model remained consistent, the airflow momentum ratio is introduced [28]:

where M s represents the airflow momentum on the swirl side, and M l represents the airflow momentum on the lean side.

The airflow momentum flux was defined as follows:

Scaled-down model momentum ratio [28]:

To verify the flow similarity between the actual equipment and the scaled-down model, further comparisons of the Reynolds number and momentum ratio under the two scales are performed. The Reynolds number is calculated using the inlet hydraulic diameter as the characteristic length. The results show that the Reynolds numbers of the actual equipment and the scaled-down model are 7.99 × 10⁵ and 4 × 10⁵, respectively. Both values are within the high-Reynolds-number flow regime (Re > 10⁵), indicating that the requirements of the second self-modeling region are satisfied. Therefore, the scaled-down model is capable of reproducing the main flow characteristics of the actual equipment. Meanwhile, based on the airflow momentum calculation for the swirl side and lean side, the momentum ratios under the two scales are determined to be 0.185 and 0.185, respectively. The consistency of the momentum ratio demonstrates that the momentum distribution relationship between the two airflow streams in the actual equipment can be maintained in the scaled-down model. Therefore, the scaled-down model is considered suitable for cold-state experiments and numerical simulations (Table 4).

Table 4. Comparison of results between the actual equipment and the scaled-down equipment.

The operating principle of the louver-type pulverized coal distributor is illustrated in Figure 1. After entering the distributor, the primary airflow is redirected by the louver blades and divided between two outlet passages. By adjusting the structural parameters of the distributor, the airflow fractions entering the two passages can be regulated to satisfy the air-distribution requirements of the downstream combustion system. Because only single-phase airflow was investigated in the present study, differences in pulverized coal concentration between the two outlets were not directly measured and therefore cannot be determined from the present results. According to the requirements of practical engineering applications, the outlets of the distributor are connected to the swirl burner and the straight-flow burner, respectively. The outlet passage connected to the swirl burner is defined as the swirl side, while the outlet passage connected to the straight-flow burner is defined as the straight-flow side. The swirl side is mainly used to supply the primary air-pulverized coal flow for the newly developed swirl burner, whereas the straight-flow side provides the primary air-pulverized coal flow for the original straight-flow burner. In the proposed system, the outlet connected to the swirl burner is intended to receive the coal-enriched stream expected from the conventional operating mechanism of a louver-type separator. This arrangement is designed to provide potentially favorable conditions for establishing a locally coal-rich region and improving ignition stability under low-load conditions. However, the pulverized coal concentration distribution and its influence on ignition stability were not measured in the present single-phase study and require verification through gas-solid two-phase and combustion experiments. Meanwhile, the lean pulverized coal flow is transported to the straight-flow burner to maintain the original straight-flow combustion mode. Therefore, in the following sections of this paper, the terms “swirl side” and “straight-flow side” refer to the outlet passages corresponding to different types of burners, whereas “rich side” and “lean side” represent the outlet passages classified according to pulverized coal concentration. These two definitions describe the outlet characteristics of the distributor from different perspectives, namely burner function and pulverized coal concentration distribution.

Figure 1. Schematic diagram of the louver-type pulverized coal distributor.

The louver-type pulverized coal distributor consists of a four-stage blade structure. The flow inside the distributor is approximately considered as a two-dimensional flow; therefore, the diffusion of airflow in the z direction is neglected in this study. The main structural parameters of the louver-type pulverized coal distributor include the distributor width D , blade width b , blade inclination angle α , blade spacing c , blade coverage height s , total blade blockage height h , number of blades n , and overall distributor length.

During the design process, the thickness of the distributor in the z direction is set as 0.759 m, and the width is defined as D = 0.696 m. The Reynolds number of the model is calculated to be approximately 1.85 × 10⁵. Therefore, the airflow inside the distributor is located in the second self-similar region, indicating that the gas flow characteristics in the model can effectively represent the actual flow behavior inside the distributor.

The main dimensionless structural parameters of the concentrator are defined as follows [29]:

Blade spacing ratio:

Blade coverage ratio:

Blade blockage ratio:

The performance of the louver-type pulverized coal distributor is mainly evaluated based on the airflow velocity, pulverized coal flow rate, and pressure distribution at the inlet and outlets of the distributor. In this study, seven performance parameters are specifically defined.

Rich-to-lean air ratio:

where Q n represents the volumetric flow rate of the rich-side pulverized coal airflow (m³/s); Q d represents the volumetric flow rate of the lean-side pulverized coal airflow (m³/s). The rich-to-lean air ratio is generally greater than 1 and is expected to approach 1, which ensures a more uniform velocity distribution of the rich and lean primary air streams at the burner outlet. More importantly, it contributes to achieving a higher concentration ratio of pulverized coal.

Rich-side airflow ratio:

where Q represents the volumetric flow rate of the pulverized coal airflow at the concentrator inlet (m³/s).

Lean-side airflow ratio:

Total pressure loss coefficient:

where Δ p j represents the static pressure difference between the inlet and outlet of the distributor (mmH₂O); ρ U i n 2 2 g represents the inlet dynamic pressure of the airflow (mmH₂O). A lower value of ζ is preferred, because an excessively high pressure loss coefficient can adversely affect primary air supply and pulverized coal transportation.

Rich-side pressure loss coefficient:

where Δ p d represents the total pressure difference between the inlet and the rich-side outlet (mmH₂O), Δ P d = Δ P j + ρ 2 g ( U i n 2 − U d 2 ) , and U d represents the airflow velocity at the rich-side outlet (m/s).

Separation efficiency:

where G n represents the pulverized coal mass flow rate on the rich side (kg/s), and G o represents the pulverized coal mass flow rate at the concentrator inlet (kg/s).

In addition, certain relationships exist among the performance indicators of the louver-type pulverized coal distributor. These relationships can be used to facilitate the processing and analysis of experimental data:

To investigate the internal airflow characteristics of the louver-type pulverized coal distributor and the effects of different structural parameters on airflow distribution, a cold-state single-phase experimental method is adopted. The experimental system mainly consists of an air supply system, an airflow measurement device, a louver-type pulverized coal distributor model, a pressure measurement device, and a data acquisition system.

A positive-pressure air supply mode is employed in the experiment. The inlet airflow velocity is controlled by adjusting the valve installed in the air supply pipeline. The total inlet airflow rate, outlet airflow rates on the swirl side and lean side, and system pressure loss are measured. To reduce the influence of random measurement errors, each operating condition is repeatedly tested three times, and the average value of the three measurements is taken as the final experimental result. During the experiment, the inlet airflow rate is maintained at a stable level to minimize the influence of airflow fluctuations on the measurement results. The overall configuration of the experimental setup and the louver-type pulverized coal distributor model are shown in Figure 2.

(a) (b)

Figure 2. Single-phase cold-state experimental test rig. (a) Photograph of the experimental test rig, (b) Photograph of the distributor model.

The airflow velocity is measured using an IFA300 constant-temperature hot-wire/hot-film anemometer manufactured by TSI, USA. The instrument provides a high temporal response and measurement accuracy, and therefore satisfies the requirements for measuring the outlet velocity distribution of the louver-type pulverized coal distributor. The measurement cross-sections at the swirl side and lean side outlets are arranged in the stable flow region downstream of the outlet. A multi-point sampling method is adopted for velocity measurement. During the measurement process, the probe is moved point by point along the outlet cross-section at equal intervals, ensuring that the measurement points cover the main flow region of the outlet. The local velocity at each measurement point is obtained, and the average velocity over the outlet cross-section is calculated based on the statistical analysis of multiple measurement results, thereby reducing the influence of local flow non-uniformity on the measurement accuracy. The hot-wire anemometry system and data acquisition device are shown in Figure 3.

To ensure that the flow conditions within the experimental model adequately represent the operating characteristics of the actual louver-type pulverized coal distributor, the cold-state modeling experiments are conducted according to the principles of geometric similarity, Reynolds-number similarity, and momentum-ratio similarity. The experimental model is scaled from the actual distributor at a geometric scale ratio of 1:2, while geometric similarity is maintained for all major structural parameters. According to similarity theory, when the Reynolds number Re > 10⁵, the flow enters the second self-similar regime. Under this condition, the flow resistance and velocity distribution are governed primarily by the geometric

Figure 3. Hot-wire anemometry and data acquisition system.

configuration and are relatively insensitive to variations in the Reynolds number. In the present experiments, the Reynolds numbers of the airflow in the model ducts are all greater than 10⁵, thereby satisfying the requirement for Reynolds-number similarity. In addition, according to the principle of momentum-ratio similarity, the inlet airflow rate of the model is adjusted so that the momentum ratios among the internal flow passages are consistent with those of the actual device. Consequently, satisfactory dynamic similarity is achieved between the experimental model and the actual louver-type pulverized coal distributor.

During the experiments, the total inlet airflow rate is measured using an averaging Pitot tube (Annubar-type flow probe) installed in the fully developed section of the air supply duct. The inlet airflow velocity is determined based on the cross-sectional average dynamic pressure obtained from the measurement section, and the total inlet airflow rate is calculated according to the cross-sectional area of the duct. Since the averaging Pitot tube is a non-standard flow measurement device, calibration is performed using a standard Pitot tube before the experiments. The measured cross-sectional flow rate obtained by the standard Pitot tube is used as the reference value to determine the calibration coefficient of the averaging Pitot tube, thereby reducing the systematic error in airflow measurement. The installation arrangement of the averaging Pitot tube is shown in Figure 4. The pressure loss across the louver-type pulverized coal distributor is measured using a static-pressure U-tube manometer. Because the outlets of the cold-state experimental system are exposed to ambient pressure, the liquid-column height difference measured by the U-tube manometer is used to determine the static pressure difference between the distributor inlet and outlets, thereby obtaining the system pressure loss. The outlet airflow rates on the swirl and lean sides are measured using Pitot tubes. The measurement sections are arranged in the stable flow regions downstream of the distributor outlets. The average dynamic pressure over each cross-section is obtained through multipoint measurements, and the mean airflow velocities and volumetric flow rates at the swirl-side and lean-side outlets are further calculated.

Figure 4. Schematic diagram of the arrangement of the Pitot tube in the air supply duct.

For the i-th measurement point obtained using the Pitot tube, the local velocity can be expressed as follows [30]:

where U i represents the airflow velocity at the i-th measurement point, m/s; and Δ p d , i represents the dynamic pressure at the i-th measurement point, Pa.

Based on the measurement results at each point, the cross-sectional average velocity U ¯ is obtained, and the outlet volumetric flow rate is calculated as follows:

where A represents the effective flow area of the measurement cross-section, m²; and U ¯ represents the cross-sectional average velocity, m/s.

To evaluate the uncertainty propagation of the experimental measurement error to the airflow rate and pressure loss coefficient, an uncertainty analysis is performed. Under the condition that the input quantities are mutually independent, the combined relative standard uncertainty of the air volumetric flow rate can be expressed as follows [31]:

where u ( A ) represents the standard uncertainty of the flow area; and u ( U ¯ ) represents the standard uncertainty of the cross-sectional average velocity.

For the velocity measured by the Pitot tube, its relative standard uncertainty can be further expressed as follows:

where u ( Δ p d ) represents the standard uncertainty of dynamic pressure measurement; and u ( ρ ) represents the standard uncertainty of air density.

According to the first-order uncertainty propagation relationship and the definition equation of the pressure loss coefficient, the relative standard uncertainty is obtained as follows:

where u ( Δ p ) , u ( ρ ) , and u ( U 0 ) represent the standard uncertainties of pressure loss, air density, and inlet average velocity, respectively.

To investigate the influence of the swirl-side inlet geometry on the airflow distribution performance of the louver-type pulverized coal distributor, the main distributor structure is kept unchanged while different test conditions are established by varying the swirl-side inlet height. Six inlet heights are considered: 146 mm, 137 mm, 132 mm, 127 mm, 120 mm, and 116 mm. The corresponding structural configurations are shown in Figure 5.

Figure 5. Configurations of the louver-type pulverized coal distributor with different swirl-side inlet heights (from left to right and from top to bottom: Cases 1 - 6).

Based on the cold-state experimental results, the operating condition with a swirl-side inlet height of 137 mm is selected for the numerical simulation. The computational model is established according to the dimensions of the experimental model, and the test section of the distributor is 2.672 m in length. To minimize the influence of the inlet and outlet boundary conditions on the internal flow field, a 1 m inlet development section and a 0.9 m outlet development section are added to the computational domain. The computational domain is discretized using a three-dimensional multi-block structured mesh composed of hexahedral cells. To ensure sufficient numerical accuracy, local mesh refinement is applied in the louver-blade region, the swirl-side inlet region, and regions where the flow-passage cross-section changes. Grid-independence analysis is performed using three meshes with coarse, medium, and fine resolutions, containing approximately 1.50 million, 2.55 million, and 4.00 million cells, respectively. The swirl-side outlet velocity and the overall pressure loss coefficient are selected as the evaluation parameters. The results show that increasing the mesh size from approximately 2.55 million to 4.00 million cells results in only minor variations in the key evaluation parameters, with the maximum deviation remaining below 2%. Therefore, the mesh containing approximately 2.55 million hexahedral cells is adopted for the subsequent calculations, providing an appropriate balance between computational accuracy and efficiency. The computational model and mesh configuration are shown in Figure 6.

Figure 6. Computational model and mesh configuration of the louver-type pulverized coal distributor.

The inlet boundary conditions for the numerical simulations are specified according to the cold-state experimental conditions. The inlet adopts a velocity-inlet boundary condition, with the average inlet velocity set to 22 m/s. The inlet turbulence parameters are defined using the turbulence intensity and hydraulic diameter, with the turbulence intensity set to 10%. The hydraulic diameter of the actual equipment is 0.546 m, while the hydraulic diameter of the cold-state experimental model and numerical simulation model is 0.273 m, which is established according to the geometric scale ratio of 1:2 between the model and the actual equipment. A pressure-outlet boundary condition is imposed at the outlet to represent fully developed pipe-flow conditions. All solid walls are treated as no-slip boundaries to account for the restriction imposed by the distributor walls on the airflow. During the numerical calculations, the governing equations are spatially discretized using the third-order MUSCL scheme to reduce numerical diffusion and improve spatial accuracy. Pressure-velocity coupling is achieved using the SIMPLE algorithm, while the PRESTO scheme is employed for pressure interpolation. Convergence is considered to be achieved when the residuals of all governing equations decrease below 10⁻⁴.

The experimental results obtained for different swirl-side inlet configurations are presented in Table 5. The swirl-side inlet height is a key structural parameter governing the redistribution of the primary air-pulverized coal flow. Variations in this parameter not only alter the effective flow area at the swirl-side inlet, but also affect the local pressure distribution in the louver region and the extent of airflow deflection. Consequently, the swirl-side inlet height directly determines the proportion of airflow entering the swirl-side passage and the matching of jet momentum between the two outlets. As shown in Table 5, variations in the swirl-side inlet height lead to pronounced changes in the swirl-side outlet velocity, airflow distribution ratio, and system pressure loss. These results indicate that the inlet geometry exerts a significant influence on the internal flow organization of the distributor.

Table 5. Experimental results of the louver-type pulverized coal distributor.

To satisfy the primary air flow organization requirements of the swirl-straight flow synergistic combustion stabilization system, the airflow distribution ratio on the swirl side, outlet velocity matching, and system pressure loss are selected as the main evaluation indicators for optimizing the inlet height of the swirl-side passage. According to the system design parameters, the designed airflow ratio on the swirl side accounts for 11.74% of the total airflow. The engineering design specified a nominal swirl-side airflow fraction of 11.74%, but no strict acceptance ranges are prescribed for the airflow fraction, outlet velocity ratio, or pressure loss coefficient. The nominal swirl-side airflow fraction of 11.74% is used as an engineering reference rather than as a strict set point, because no predetermined acceptance ranges are available for the airflow fraction, outlet velocity ratio, or pressure-loss coefficient. The outlet velocity ratio is treated as the primary comparative criterion, followed by the overall pressure-loss coefficient and the swirl-side airflow-supply capability. The 137 and 127 mm configurations produced velocity ratios of 1.03 and 0.97, respectively, and are therefore equally close to unity. However, the 137 mm configuration exhibited a lower overall pressure-loss coefficient than the 127 mm configuration (0.853 versus 1.078). Although the 132 mm configuration produced a slightly lower pressure-loss coefficient of 0.817, its outlet velocity ratio of 0.93 deviated further from unity. Therefore, among the six investigated configurations, the 137 mm configuration is selected as the preferred engineering compromise. This selection is comparative and should not be interpreted as a universal optimum. The selection of the preferred configuration represents a trade-off among these indicators rather than compliance with predetermined acceptance limits.

When the swirl-side inlet height decreases from 146 mm to 137 mm, the swirl-side outlet velocity increases from 14.16 m/s to 15.56 m/s, corresponding to an increase of approximately 9.9%. Meanwhile, the swirl-side airflow rate increases from 646.03 m³/h to 709.46 m³/h, and its proportion of the total airflow rises from 13.63% to 15.17%. This behavior is mainly attributed to the reduction in the effective inlet area on the swirl side. As the inlet height is moderately reduced, part of the airflow that originally travels along the main flow direction is more strongly constrained by the inlet boundary and is therefore more readily deflected laterally under the guidance of the louver blades. As a result, the transverse momentum directed toward the swirl-side passage is enhanced. In addition, a moderate reduction in inlet height strengthens the pressure gradient near the swirl-side inlet, further promoting lateral flow deflection and improving the air extraction capability of the swirl-side passage. Thus, as the inlet height decreases from 146 mm to 137 mm, the airflow redistribution capability of the distributor is progressively enhanced.

However, further reduction in the swirl-side inlet height does not lead to a continued increase in the extracted airflow. When the inlet height decreases from 137 mm to 116 mm, the swirl-side outlet velocity decreases from 15.56 m/s to 12.60 m/s, while the swirl-side fraction of the total airflow decreases from 15.17% to 12.10%. This decline is mainly associated with the intensified local contraction effect caused by an excessively small inlet. As the airflow passes through the narrowed region, larger velocity gradients and greater local pressure losses are generated, thereby reducing the effective flow rate through the swirl-side passage. Moreover, with a further decrease in inlet height, the restriction imposed by the flow passage becomes stronger and the local high-velocity region expands. A greater proportion of the available pressure difference is consequently dissipated in overcoming local flow resistance rather than being converted into an increase in swirl-side airflow.

Therefore, a transition exists from enhanced airflow deflection to increased flow blockage as the swirl-side inlet height is reduced. An excessively small inlet also limits the effective area available for airflow entering the swirl-side passage, thereby weakening the redistribution effect induced by the louver structure. Accordingly, the relationship between the swirl-side inlet height and the air extraction capability is not simply monotonic; instead, an optimum matching range exists.

Therefore, a transition exists from enhanced airflow deflection to increased flow blockage as the swirl-side inlet height is reduced. An excessively small inlet also limits the effective area available for airflow entering the swirl-side passage, thereby weakening the redistribution effect induced by the louver structure. Accordingly, the relationship between the swirl-side inlet height and the air extraction capability is not simply monotonic; instead, an optimum matching range exists.

Further analysis of the velocity matching between the swirl-side and lean-side outlets indicates that, when the swirl-side inlet height is 137 mm, the outlet velocity on the swirl side is 15.56 m/s, whereas that on the lean side is 15.08 m/s. The resulting velocity ratio reaches 1.03, which is close to the ideal condition. A high degree of velocity matching ensures a coordinated momentum distribution of the inlet airflow to the subsequent swirl burner and straight-flow burner, thereby avoiding deviations in the in-furnace aerodynamic field caused by excessive differences in jet velocity between the two sides. For tangentially fired boilers, the jet momentum distribution of the burners directly affects the stability of the rotating flow field in the furnace. If the outlet velocities on the swirl side and straight-flow side differ excessively, local jet penetration capability may be altered, and the in-furnace aerodynamic field may be shifted accordingly. Therefore, while the enhancement of pulverized coal concentration is required, a reasonable velocity matching relationship between the two outlet streams should also be ensured.

By contrast, when the inlet height is reduced to 116 mm, although the overall pressure loss coefficient of the system decreases to 0.602, the velocity ratio between the swirl side and the lean side is only 0.79. This indicates that the outlet velocity on the swirl side is substantially lower than that on the straight-flow side, resulting in unfavorable outlet velocity matching between the two passages. Therefore, when evaluating the performance of the louver-type pulverized coal distributor, pressure loss alone cannot be regarded as the sole criterion. The swirl-side airflow supply and outlet velocity ratio should also be considered. The airflow redistribution capability and outlet velocity matching should also be comprehensively considered. Among the six configurations, the 137 mm case provides the closest outlet velocity matching, with a swirl-side/straight-flow-side velocity ratio of 1.03. Its swirl-side airflow fraction of 15.17% exceeds the design value of 11.74%, while its overall pressure loss coefficient of 0.853 is not the lowest among the investigated cases. Therefore, the 137 mm configuration is selected primarily because it provides the best outlet velocity matching while maintaining sufficient swirl-side airflow supply and a moderate pressure loss. This selection represents a multi-criteria engineering trade-off rather than the optimization of any single indicator. Overall, the 137 mm configuration provides the most favorable compromise among outlet velocity matching, swirl-side airflow supply, and system pressure loss.

To further elucidate the influence of the louver structure on the flow behavior of the primary air-pulverized coal stream, numerical simulations are performed to analyze the internal velocity field of the distributor under the condition with a swirl-side inlet height of 137 mm. Figure 7 presents the velocity distribution on the central section of the louver-type pulverized coal distributor. The velocity distributions at the swirl-side and lean-side inlet sections are also shown in Figure 7.

Figure 7. Velocity distribution on the central widthwise section of the louver-type pulverized coal distributor.

As shown in Figure 7 and Figure 8, the primary airflow undergoes pronounced deflection after entering the distributor under the guiding effect of the louver blades, resulting in the formation of two major flow regions. The airflow adjacent to the blades is redirected by the blade blockage and gradually enters the swirl-side passage, whereas the remaining airflow continues approximately along its original direction and ultimately enters the lean-side passage. Near the leading edges of the blades, the abrupt change in the flow-passage cross-section accelerates the airflow and produces local high-velocity regions. Meanwhile, flow separation occurs on the leeward sides of the blades, giving rise to low-velocity and locally negative-pressure regions. This non-uniform velocity distribution constitutes an important mechanism by which the louver structure induces airflow deflection and redistribution.

Figure 8. Velocity distributions at the swirl-side and lean-side inlet sections.

As the airflow proceeds toward the outlets, pronounced velocity gradients develop in the swirl-side passage because of the curved flow path and variations in the passage cross-section. Owing to the centrifugal effect in the bend, the airflow velocity on the outer side of the curved passage is higher than that on the inner side. After passing through the downstream development section, however, the velocity distribution gradually becomes more uniform. This indicates that the outlet development section plays an important role in improving airflow uniformity.

From the perspective of the overall flow process, the louver blades do not simply divide the airflow mechanically. Instead, the redistribution of the primary air flow is achieved by changing the local pressure distribution and flow direction. Therefore, the proportion of primary air flow entering the swirl side and lean side can be regulated through reasonable design of the inlet structural parameters, providing suitable airflow matching conditions for the subsequent swirl-straight flow synergistic combustion system. It should be noted that the single-phase air model and cold-state experimental method adopted in this study are mainly used to verify the influence of the louver structure on airflow distribution characteristics. The redistribution of pulverized coal concentration and its effects on the combustion process require further investigation through gas-solid two-phase experiments and combustion tests.

To validate the reliability of the numerical model in predicting the flow characteristics of the distributor, the simulation results are compared with the cold-state experimental data, as presented in Table 6.

Table 6. Comparison between experimental and numerical results for the louver-type pulverized coal distributor.

As shown in Table 6, the numerical simulation results are generally consistent with the experimental measurements in terms of the overall trends. To quantitatively evaluate the prediction accuracy of the numerical model, the relative errors of each evaluation parameter are further calculated. The relative error is defined as the absolute difference between the simulated and experimental values divided by the experimental value. The simulated and experimental swirl-side outlet velocities are 14.98 m/s and 15.56 m/s, respectively, with a relative error of 3.73%. The simulated and experimental lean-side outlet velocities are 14.01 m/s and 15.08 m/s, respectively, with a relative error of 7.10%. The simulated velocity ratio between the two outlets is 1.07, compared with the experimental value of 1.03, resulting in a relative error of 3.88%.

For airflow distribution, the simulated swirl-side airflow ratio is 15.28% of the total airflow, while the experimental value is 15.17%. The absolute deviation between the two values is only 0.11 percentage points, corresponding to a relative error of 0.73%. In terms of pressure loss characteristics, the simulated pressure loss coefficient of the swirl-side circular duct is 0.32, compared with the experimental value of 0.30, resulting in a relative error of 6.67%. The simulated overall pressure loss coefficient is 0.87, while the experimental value is 0.853, with a relative error of 1.99%. The differences between the numerical and experimental results can mainly be attributed to several factors. The numerical model neglects the two-phase coupling between pulverized coal particles and the airflow; small deviations in local flow-passage dimensions may also arise from manufacturing tolerances in the experimental model. In addition, measurement uncertainties are inevitably introduced during hot-wire velocity and pressure measurements. Therefore, the deviations between the numerical and experimental results are considered to be within a reasonable range.