Improved Pulsed-Jet Cleaning of Pleated Cone Filter Cartridges Using a Diffusion Nozzle

A pleated filter cartridge with an additional pleated filter cone installed on the cartridge base, i.e., a pleated cone filter cartridge, has been proposed to increase the filtration area and to improve medium regeneration by reverse pulsed-jet cleaning. In this study, the pulsed-jet cleaning performance of a pleated cone filter cartridge in a dust collector was investigated using a diffusion nozzle (as compared to that of a round nozzle). The effects of tank pressure ( TP ), jet distance ( JD ), and inner cone height ( HC ) on the distribution of static pressure acting on the inner surfaces of the cone filter cartridges were evaluated. The filtration pressure drops and emitted dust concentration of cartridges during the dust collector operation were also measured. It was found that, using a diffusion nozzle, the pulsed-jet pressure at the upper portion of the filter cartridges was increased, compared to that of a round nozzle, particularly for cartridges with a low HC and a short JD setting (e.g., a factor of 7.0 was observed when HC = 560 mm and at a JD setting of 150 mm). The pulsed-jet cleaning of a pleated cone filter cartridge using a diffusion nozzle can decrease the residual filtration pressure drop, prolong the cleaning interval, and reduce the average dust concentration emission from cone filter cartridges.


INTRODUCTION
The emission of dust particles from industrial processes (tunneling, mining, cement manufacturing, coal-fired powerplants, etc.) is one of the major sources of air pollution, which poses a serious health threat to the public with extended exposure (Xie et al., 2022;Rice et al., 2019;Sun et al., 2019). For the pollution control purpose, it would be the best to remove dust particles from their emission sources prior to releasing them to the ambient atmosphere. Due to the advantages that dust collectors equipped with filter cartridges have (such as low/moderate flow resistances, high filtration efficiencies, and small footprints), they are widely applied to remove dust particles emitted from industrial processes (Yang et al., 2019). Continuous operation of dust collectors requires periodic cleaning (or regeneration) of the filter cartridges when fully loaded with dust particles, because the pressure drop of filter cartridges increases as the loading of dust particles. For the regeneration of loaded filter cartridges, reverse pulsed-jet cleaning is currently the most popular and efficient method in the industry (Fotovati et al., 2011;Chi et al., 2008;Qian et al., 2014). However, in practice, uniform cleaning of filter cartridges is not an easy task, resulting in patchy cleaning of the cartridges and loss of effective filtration area (Yan et al., 2013;Joubert et al., 2010;Lo et al., 2010). Specifically, patchy cleaning of cartridges is often observed at the upper portion of filter cartridges when they are installed vertically (with the opening facing upwards) in dust collectors (Li et al., 2018;Fotovati et al., 2011). When patchy cleaning occurs and effective filtration area is lost, the filtration pressure increases, resulting in the reduction of filtration efficiency (and making it challenging to meet environmental standards).
Various ideas have been proposed to improve the pulsed-jet cleaning of loaded filter cartridges (e.g., making the peak pressure in the cartridge core more intensive and more uniform during cleaning to minimize the patchy cleaning issue), including the use of different types of cleaning nozzles. Wang et al. (2021) numerically investigated the dynamics of the pulsed-jet flow field in a filter cartridge core by including a diffuser in front of a round nozzle (i.e., diffusion nozzles). They found that the peak pressure in the upper portion of a filter cartridge increased by 53.7%, and in the middle and lower portion of a filter cartridge decreased by 32.5% (compared to that of a round nozzle). The spatial uniformity of peak pressure was also improved. Hu et al. (2019) applied the Laval nozzle to the pulsed-jet cleaning of filter cartridges and showed that the use of a Laval nozzle increased the average peak pressure on the inner surfaces of filter cartridges by 53.2%. Xi et al. (2021) studied the effect of a conical diffuser installed at the exit of a round nozzle on pulsed-jet cleaning and concluded that the diffuser diverges the pulsed-jet airflow, which could improve the cleaning of filter cartridges.
Pulsed-jet cleaning performance could be also improved by altering the structure of filter cartridges. Li et al. (2015) installed a pleated cone at the base of a pleated filter cartridge and showed that the addition of the pleated cone was beneficial in both increasing the inner surface pressure of filter cartridges and improving the cleaning uniformity. Zhang et al. (2016) numerically investigated the differences in filtration and cleaning performances between cylindrical and cone filter cartridges. This study showed that the built-in filter cone increased the filtration area in a typical filter cartridge, resulting in the reduction of the filtration pressure drop. It also found that the peak pressure along the axis of the filter cartridges was more uniform during pulsed-jet cleaning. Qiu et al. (2021) used numerical modeling to explore the effect of the inner cone height on the pulsed-jet cleaning performance with the use of round nozzles (having straight airflow channels). They showed that a negative pressure zone formed at the upper portion of the cone filter cartridges when the height of the inner cone was between 760 and 860 mm. Their study also found that the low-pressure zone disappeared as the inner cone height further increased, which indicates that increase of the cone height could improve the uniformity of the peak pressure distribution in pleated cone filter cartridges. Chen et al. (2021) numerically explored the feasibility of improving the pulsed-jet cleaning for cone filter cartridges using a diffusion nozzle (via 2-D CFD modeling) and found that the cleaning intensity was increased by a factor of 1.66. However, the dynamics of cleaning cone filter cartridges with a diffusion nozzle have not been fully investigated. Additionally, it is necessary to experimentally study the effects of the cleaning nozzle and inner cone on the cleaning performance in order to optimize the pulsed-jet cleaning of cone filter cartridges with a diffusion nozzle.
In this study, an experimental setup was constructed to investigate the performance of pulsedjet cleaning for pleated cone filter cartridges using a diffusion nozzle. The effects of tank pressure (PT), jet distance (JD) and inner cone height (HC) on the distribution of the static pressure acting on the inner surfaces of the cone filter cartridges were investigated. In addition, the overall performance of a dust collector equipped with pleated cone filter cartridges and a diffusion nozzle for pulsed-jet cleaning (i.e., pressure drop and dust emission concentration) were also measured. The results obtained by this study can provide guidance for future designs and performance optimization of dust collectors equipped with cone filter cartridges. Fig. 1 shows the schematic diagram of the testing rig. The filtration chamber of this dust collector is 1,225 (in width) × 750 (in depth) × 1,550 mm (in height). One pleated cone filter cartridge was vertically installed in the filtration chamber. A typical round nozzle (with a straight airflow channel and having the inner diameter of 12 mm) was installed at a selected distance above the opening of the filter cartridge. The experimental setup includes a dust feeder (LSC-6 type with a range of 0-450 g min -1 , Shanghai Chuanlingjidian Technology Co., Ltd., China), a pressure tank (19.5 L), an electromagnetic pulse valve (DMF-Z-25 type, 1-inch, Shenchi Pneumatic Co., Ltd., China), a pulse controller (LC-PDC-ZC10D type with a pulse duration range of 0.01-0.99 s, Lingchuan Auto Technology Co., Ltd., China), a high-frequency pressure acquisition system (consisting of piezoelectric ceramic high frequency dynamic pressure sensors (MYD-1530A with the size of ϕ7 × 17 mm and sensitivity of 6-13 pC kPa -1 ), charge amplifiers (MCA-02), and the data acquisition cards (MYPCI4526) from Mianyang Minyu Electronics Co., Ltd., China), a thermal air flow meter (CKRSC-D150-F with a range of 0-2000 m 3 h -1 , Shanghai Chi Control Automation Instrument Co., Ltd., China) and a variable frequency fan (TB-150-5 type, Jiangsu All The Wind Environmental Proctection Technology Co., Ltd., China), a pressure drop transducer (DT-8920 with a range of 0-6 kPa, Shenzhen Everbest Machinery Industry Co., Ltd., China), and an online dust detector (ZK-50 with a range of 0-50 mg m -3 , Zhongke Zhengqi (Beijing) Science and Technology Co., Ltd., China). An online dust detector was used to measure the mass concentration of dust particles by detecting the AC signal induced from the collision of dust particles on the surface of sensing probe facing against the airflow direction (because of the tribological effect). The AC signal was then converted to a voltage being recorded by the computer via an A/D card.

Experimental Setup
The operation of this dust collector has two phases, i.e., filtration and cleaning: in the filtration phase, the dust-laden airflow entered the filtration chamber from its inlet duct under the fan suction. The fan speed was controlled by a frequency converter to maintain a constant airflow rate in the dust collection chamber (via the feedback from a flowrate meter). Dusty airflow was forced through a pleated cone filter cartridge from the outside of the cartridge to the inside. Dust particles were first collected inside the filtration media and then on the upstream surface of media. After passing through the filter cartridge, the cleaned airflow exited the dust collection chamber through the fan. The operation of the dust collector moved into the cleaning (or regeneration) phase when either the measured cartridge pressure drop or operation time exceeded an operator-defined value, which triggers the pulsed-jet cleaning. Once triggered, the electromagnetic pulse valve was opened. Clean air, originally stored in a high-pressure tank, was released, forced through the transport tube and exited from a cleaning nozzle installed right above the cartridge opening. A jet flow was launched upon exiting the cleaning nozzle. The clean jet flow together with the airflow entrained from the ambient entered the core of a filter cartridge. The buildup of high pressure in the cartridge core forced the cleaning air through the cartridge filtration media, and dislodged the built-up dust cake on the filtration surface of cartridge media.
Cone filter cartridge under the study were constructed by adding pleated filter cones to the bases of cylinder filter cartridges. The height, outer and inner diameters of the cylinder filter cartridges are 660, 320, and 240 mm, respectively. The filtration media is made of non-woven polyester and its thickness is 0.6 mm. In this investigation, HC, was varied. i.e., 560, 660, and 760 mm. The schematic diagram and key dimensions of cone filter cartridges can be found in Fig. 2(a) and Table 1, respectively. The photo of a pleated cone filter cartridge is also shown in Fig. 2 The diffusion nozzle was constructed by adding a conical diffuser at the exit of a typical round nozzle. The angle of the conical diffuser is 71° and the apex of the cone is 40 mm (D) apart from the round nozzle exit. The inner diameter of the diffuser exit is 90 mm. The schematic diagram and photo of the diffusion nozzle are shown in Figs. 2(a), 2(c), respectively.
Talc particles were used in this study. The selection of Talc particles is because of the experimental safety as it is not flammable or explosive. The D10, D50, and D90 measured by a laser particle size analyzer (Malvern 3000+Hydor EV) were 0.42, 0.83, and 2.06 µm, respectively (Shown in Fig. 3). Note that talc dust is inhalable. To prevent the personal exposure from the dust hazard, the outlet of the dust collector was placed outside the laboratory building. A separate suction hood with its outlet placed outside of the laboratory building was installed over the dust feeder. All the operators wore masks during the experimental runs.
Three high-frequency pressure sensors were installed on the inner surface of the pleated filter cylinder of a cone filter cartridge, at the P1. P2 and P3 located at the distances of 100, 330, and 560 mm from the cartridge opening ( Fig. 1). These sensors were used to measure the static pressure change during the pulsed-jet cleaning. The recording frequency of the high-frequency pressure sensors was set at 500 Hz.

For evaluating the evolution of the pulsed-jet pressure:
In this part of the study, experiments were carried out on clean filter cartridges at the PT of 0.3, 0.45, and 0.6 MPa, and JD of 150 mm. The evolution of pulsed-jet pressures at the selected observation points were measured in the cases with round and diffusion nozzles with HC = 560 mm and 760 mm.

For the study of jet distance
The experiments were carried out on a clean filter at PT = 0.45 MPa and the JD = 150-650 mm (at the interval of 50 mm). For each combinational setting of PT and JD, both round and diffusion nozzles were tested. Note that the setting of JD = 150 mm was the minimal JD, which could be set in the cases testing filter cartridges with HC = 760 mm. It is because, under the above JD setting, the apex of the inner cone reached the exit of the diffusion nozzle. The JD = 150 mm was thus selected as the lower limit of the JD studied in our experiments. Due to highly turbulent jet airflow, the average of 3-6 runs for each combinational setting is reported.  In this part of the study, experiments were carried out to evaluate the overall performance of a dust collector equipped with the pulsed-jet cleaning by the round/diffusion nozzle. The testing was only performed once in each case. The dust concentration at the dust collector inlet was set at 10 g m -3 , which corresponds to an inlet dust mass flowrate of 78.9 g min -1 (under the speed of the dust feeder of 950 rad min -1 ), a dust collector airflow volume of 7.89 m 3 min -1 (i.e., the fan frequency was 43 Hz), and an average filtration velocity of 1.67 cm s -1 in the duct collector. The PT was set at 0.45 MPa and the pulse duration was 0.15 s. The clean-on-demand cleaning mode was adopted in our study with the allowable pressure limit set at 500 Pa. The cleaning using either a round or diffusion nozzle was individually tested. Both the filtration pressure drop and emitted dust concentration of cone filter cartridges in the filtration phase of the duct collector were monitored and recorded over time with a measuring cycle of 1 second.

Temporal and Spatial Distribution of Pulsed-Jet Pressure
The time-dependent static pressure at the P1, P2 and P3 observation points in the cases using round and diffusion nozzles for cone filter cartridges with HC = 560 and 760 mm are shown in Fig. 4 under the settings of JD = 150 mm, and PT = 0.3, 0.45, and 0.6 MPa.
Upon the initiation of cleaning flow, the static pressure at all the observation points first experienced a brief pressure fluctuation except for the P1 point when the cartridges with HC = 560 mm were cleaned by the diffusion nozzle, and when the cartridges with HC = 760 were cleaned by the round nozzle. The above pressure fluctuation might be due to the sampling of the signals from pressure sensors by the data acquisition, particularly when the sensors experience an extremely fast change of airflow pressure and vibration of the cartridge. After the first fluctuation, the static pressure at all observation points rose rapidly to a positive peak, which was considered as the primary pressure peak in the pulsed-jet cleaning. Since the pulse duration was kept constant, the same duration of the primary pressure peak at each observation point was observed. In general, the pressure peak at the lower portion of a cone filter cartridge (i.e., P3) was the highest, and the peak value at P1 was the lowest. The variation of the pressure peak along the axis of the filter cartridge was similar to that reported in the literature (Qian et al., 2014). It is because that the cleaning jet flow impinged on the base of the filter cartridge and was forced to reverse its flow direction, resulting in the significant conversion from the flow momentum to the static pressure.
After the positive peaks, the pressures peaked negatively (possibly due to the rebound action of cleaning airflow (Li et al., 2015)). The compressed air flow in the cartridge core would rebound and rush out the cartridge through the opening after the valve was closed. As a result, a negative pressure zone was created.
The peak pulsed-jet pressures at observation points of P1-P3 under the variation of PT were shown in Fig. 5. The peak pressures at P1, P2 and P3 increased with PT and were significantly higher for the diffusion nozzle than the round nozzle.
Compared to that in the cases using the round nozzle, the peak pressure at the P1 improved the most when using the diffusion nozzle. The highest improvement was by a factor of 7.0, from 519 to 3,632 Pa, when testing cartridges with HC = 560 mm at the setting of PT = 0.45 MPa. The lowest improvement was only by a factor of 1.3, i.e., from 1,618 to 2,014 Pa when cleaning cartridges with HC = 560 mm at the setting of PT = 0.3 MPa. The improvement by using the diffusion nozzle for the pulsed-jet cleaning is attributed to its divergence (or de-centering) of the pulsed-jet airflow , which reduced the forward momentum of jet airflow, facilitating the conversion from the dynamic pressure to the static pressure. The conversion was further enhanced by the flow collision and diversion due to the presence of inner filter cones in the cartridge core.
It was also found that the difference in the pulsed-jet peak pressure for cleaning both cartridges with HC = 560 and 760 mm using the round nozzle was less than that for cleaning cartridges using the diffusion nozzle. The pulsed-jet intensity for cartridges with HC = 560 mm was found to be slightly higher than that for cartridges with HC = 760 mm. It is because that, for cartridges with HC = 760 mm, the presence of high inner cone diverted a portion of cleaning airflow outside cone filter cartridges.
When using the diffusion nozzle, the increase of the pulsed-jet pressure for cleaning the cartridges with HC = 560 mm was significantly better than that for cleaning the cartridges with HC = 760 mm, particularly at the P1 point. It is because, for cartridges with HC = 560 mm, the cleaning airflow impinged at the top cap of the inner filter cone and diverted locally. The above resulted in the  formation of a relatively high-pressure zone in the upper core of cone filter cartridges. For cartridges with HC = 760 mm, the cap of the inner filter cone was very close to the diffusion nozzle exit. A portion of cleaning airflow was diverted to the outside of filter cartridges. Note that, the static pressure in Fig. 4 was found to fluctuate more strongly in the cases using the diffusion nozzle (when compared to that in the cases using the round nozzle). It could be due to the fast conversion from the dynamic to static pressures. Fig. 6 is the comparison of the peak pressures at P1, P2, and P3 points when cartridges with HC = 560, 660, and 760 mm were cleaned by either the round or diffusion nozzles under the JD setting of 150-650 mm.

Shown in
In the cases cleaned by the round nozzle, the measured static pressure peaks at all observation points increased with the increase of jet distance, JD. As shown in Fig. 6, the peak pressure at the lower portion of the cartridge (i.e., P3) was always the highest. It is because that, when the jet distance increased, more surrounding airflow was entrained into the cartridge core by the jet flow issued from the round nozzle ( Fig. 7(a)). Note that we did not perform experiments at JD above 650 mm due to limitation on the ceiling space.
In the cases cleaned by the diffusion nozzle, the peak static pressure in general decreased with the increase of JD. Note that the opposite trend on the JD effect was observed in the cases with the round nozzle. In other words, the use of the diffusion nozzle can save the vertical head space for dust collectors. It is because the cleaning airflow exiting the diffusion nozzle was divergent. The cross-section of jet airflow was increased when the JD was increased. Under the fixed cartridge opening, the percentage of cleaning airflow entering the cartridge core was reduced. The above was further illustrated in Figs. 7(b), 7(c). In the case testing cartridges with HC = 560 mm, the peak pressure at P1 was the highest, and that at P2 was the lowest. It is because of the impingement of the cleaning jet airflow on the cap of the inner filter cone. In the cases testing cartridges with HC = 760 mm, the peak pressure at the P3 was the highest and that at P1 was the lowest. It is due to the collision of cleaning airflow at the cartridge base, and the build-up of static pressure from the cartridge base. In the cases testing cartridges with HC = 660 mm, the highest peak pressure was at P3 for the JD setting of 150-450 mm, and at P1 for the JD setting of 450-650 mm. It was attributed to the lessening of the jet airflow momentum after exiting from the diffusion nozzle. The expansion of cleaning airflow happened in the upper and middle portions of cone filter cartridges when the jet distance was set far away from the cartridge opening, resulting in the weakening of the flow impingement to the cartridge base.   Fig. 8 is the enhancement of static pressure peaks by using the diffusion nozzle compared to the corresponding cases using the round nozzle (when cleaning cartridges with HC = 560, 660, and 760 mm at the JD setting of 150-450 mm). It was observed that, when the JD increased or HC reduced, the enhancement (defined as the ratio of peak pressure in the diffusion nozzle case to that in the corresponding round nozzle case) at P1 obviously decreased.

Shown in
The cleaning performance can be gauged by the pulsed-jet intensity and uniformity, which are defined as the average and the variation coefficient (C.V.) of the positive pressure peak values at all the observation points, respectively Wu et al., 2020). Shown in Fig. 9 is the pulsed-jet intensity and uniformity as the function of jet distance, JD. Note that each data shown in Fig. 9 was obtained by averaging the peak pressure intensities measured 3-6 times under the same setting. In the cases using the round nozzle, the pulsed-jet intensity increased with the increase of jet distance, and reached the maximal values of 2,798, 1,343, and 1,407 Pa at the JD setting of 650 mm for cleaning cone filter cartridges with HC = 560, 660, and 760 mm, respectively. When using the diffusion nozzle, the pulsed-jet intensity decreased with the increase of jet distance, and reached the maximal values of 3,103, 2,370, and 2,396 Pa at the JD setting of 150 mm for cartridges with HC = 560, 660, and 760 mm, respectively. A factor of 2.72, 2.81, and 2.68 increase for the three cartridges with HC = 560, 660 and 760 mm was found when compared to that in the corresponding round nozzle cases. This was due to the fact that, at the setting of low HCs, the pulsed-jet airflow collided with the cartridge cone and the collision benefited to form a high static pressure (also illustrated in Figs. 7(b) and 7(d)).  In addition, the pulsed-jet uniformity given by the diffusion nozzle was better than that given by the round nozzle at the JD setting of 150-450 mm for cartridges with HC = 560 mm. The pulsedjet uniformity given by the diffusion nozzle was also better than that offered by the round nozzle for cartridges with HC = 660 and 760 mm. The above observations were because the use of diffusion nozzle primarily increased the pressure at the upper portion of cone filter cartridges, making it closer to the pressures peaked at the middle and lower portions of the cartridge core.

On the Filtration Pressure Drop and Dust Emission Concentration
To illustrate the overall performance of the dust collector equipped with the round/diffusion nozzle for pulsed-jet cleaning, the filtration pressure drop and dust emission concentration during the collector operation were monitored at the JD setting of 150 mm for cone filter cartridges with HC = 560 and 760 mm. Note that, in this part of the study, we only operated the dust collector until the obvious difference on the filtration pressure drop and emitted dust concentrations among different cases was observed for the illustration purpose. For the reference, the static pressures at the observation points during the cleaning in the above cases are given in Fig. 10. Shown in Fig. 11 is the filtration pressure drop of the above filter cartridges as the function of dust collector operation time. During the filtration stage, the filtration pressure drop of cone filter cartridges gradually increased as particles were continuously collected, and reached the operatordefined pressure value. The increasing rates of filtration pressure drops, and filtration stage periods were almost the same for the three cases. It indicates the good stability of the generated dust concentration and a good repeatability of our experimental runs.
Using the round nozzle, the residual pressure drop of cartridges with HC = 760 mm (i.e., the pressure measured right after the pulsed-jet cleaning) were 347, 385, 392, 408 and 421 Pa (with an average of 391 Pa) for the first five cycles (Fig. 11(a)). The average residual pressure drop of the above cartridges for the first five cycles was 265 Pa when using the diffusion nozzle (Fig. 11(c)),  which was 70% of the value obtained when using the round nozzle. For cartridges with HC = 560 mm, the average residual pressure drop was 240 Pa when using the diffusion nozzle, which was lower than that for cartridges with HC = 760 mm. The first five filtration cycles were 162, 103, 73, 64, and 48s, with an average of 90 s each cycle for filter cartridges with HC = 760 mm when cleaned by the round nozzle. When the diffusion nozzle was used, the average first five filtration cycles was 421s, which was 4.7 times longer than that by using the round nozzle. The average filtration time of the first five cycles for cartridges with HC = 560 mm cleaned by the diffusion nozzle was 490s, which was longer than that for filter cartridges with HC = 760 mm. It is evidenced that the pulsed-jet cleaning using the diffusion nozzle is beneficial to reduce the residual filtration pressure drop and to prolong the filtration time of each operational cycle of cone filter cartridges. Furthermore, it is found that better pulsed-jet cleaning performance could be achieved in the cases of filter cartridges with short inner cones.
The variation in dust emission concentration at the dust collector outlet was also measured (shown in Fig. 12). In the filtration stage, the dust emission concentration was kept below 1.0 mg m -3 , which was relatively steady over the filtration time in each operational cycle. In the case of using the round nozzle for cleaning filter cartridges with HC = 760 mm, the dust emission concentrations in the first five cycles were peaked at 2.62, 2.10, 2.11, 1.85, and 1.63 mg m -3 , respectively (with an average of 2.06 mg m -3 ). When using the diffusion nozzle to clean filter cartridges with HC = 560 and 760 mm, the average peak dust emission concentrations for the first five cycles were 5.33 and 3.92 mg m -3 , respectively. The peak dust emissions measured in the cases with the diffusion nozzle was higher than that in the corresponding cases with the round nozzle. It is because the high pulsed-jet intensity generated by using the diffusion nozzle stripped off more dust cakes from the filtration surface of filter cartridges during the puled-jet cleaning, resulting in the increase of the upstream dust concentration after each cleaning. In additional, it is speculated that the filtration media might temporarily became more opened due to the high pulsed-jet intensity during the cleaning (at least temporarily if not permanently). However, the average concentration of the dust emission in the first 1,200 seconds were 0.623 and 0.434, 0.434 mg m -3 in the above-discussed cases, respectively (Fig. 12). It concludes that the total dust emission from cone filter cartridges when cleaned by the diffusion nozzle was lower than that when cleaned by the round nozzle. The cleaning frequency was also decreased when using the diffusion nozzle. From the viewpoint of total dust emission, the cleaning by the diffusion nozzle is favored for cone filter cartridges. The images of the filtration surface of cone filter cartridges with HC = 760 mm at the end of the experimental runs when cleaned by either the round or diffusion nozzle are shown in Fig. 13. It is evident that, compared to the cases using the round nozzle, less residual dust remained on the cartridge filtration surface in the cases using the diffusion nozzle (especially at the upper portion of cartridges). Moreover, the residual dust was more uniformly distributed along the cartridge length in the cases with the diffusion nozzle. These above observations evidence that the use of the diffusion nozzle enhanced the pulsed-jet intensity and uniformity for cleaning cone filter cartridges.

CONCLUSIONS
In this study, the pulsed-jet cleaning of pleated cone filter cartridges using a diffusion nozzle (as compared to that using a round nozzle) were investigated. The temporal and spatial distribution of pulsed-jet pressure on the inner surfaces of cone filter cartridges was measured by varying the air tank pressure, PT, jet distance, JD, and pleated filter cone height, HC. In addition, the filtration pressure drops and emitted dust concentration of the cone filter cartridges while the dust collector was in operation were also measured. Based on our investigation, the following can be concluded: (1) The pulsed-jet pressure improved with the use of a diffusion nozzle: Compared to that of using a round nozzle, the forward momentum of the pulsed-jet airflow was diverted (or de-centered) using a diffusion nozzle. Due to the presence of the pleated inner cone, the flow is diverted when the cleaning flow enters the cartridge core and the kinetic energy of the pulsed-jet flow is effectively transformed to static pressure, which is beneficial for the pulsed-jet cleaning. The pulsed-jet pressure increased the most at the upper portions of the cone filter cartridges, increasing from 519 Pa to 3,632 Pa, with an enhancement of 7.0 times when using filter cartridges with HC = 560 mm and the JD setting of 150 mm. (2) The diffusion nozzle's improved performance over a round nozzle is best observed at a low JD setting: With higher values of the JD (within the studied JD range of 150-650 mm), it was observed that a portion of the cleaning flow might run off from the cartridge opening. Thus, the performance of pulsed-jet cleaning using a diffusion nozzle was better observed in filter cartridges tested at a low JD setting. At a JD setting between 150-400 mm, the pulsed-jet pressures were improved, with the highest improvement being observed in cases corresponding to a JD setting of 150 mm. The pulsed-jet uniformity from a diffusion nozzle was also shown to be improved compared to that of a round nozzle. (3) The filtration performance of the pleated cone filter cartridges during dust collection using a diffusion nozzle was better than that of a round nozzle: As a result of the improved pulsed-jet pressure by using a diffusion nozzle, the residual pressure drop of the cleaned cone filter cartridges was reduced (to 0.68 times on average for the first five operation cycles when testing filter cartridges with HC = 760 mm). The residual dust cake on the filtration surfaces of filter cartridges was also observed to be smaller, and the filtration interval in each cycle was extended (to 4.7 times the interval when using the round nozzle). The average dust emission concentration was also decreased (to 0.69 times for the first 1,200 seconds). Moreover, it was found that the cleaning performance was better observed for filter cartridges with low inner cones.