SCIENCE CHINA Technological Sciences • RESEARCH PAPER •
September 2011 Vol.54 No.9: 2483–2492 doi: 10.1007/s11431-011-4416-y
Performance of a centrifugal compressor with holed casing treatment in the large flowrate condition XU Wei, WANG Tong* & GU ChuanGang School of Mechanical Engineering, Shanghai Jiaotong University, Shanghai 200240, China Received September 29, 2010; accepted March 15, 2011; published online June 7, 2011
As demonstrated by former work, the holed casing treatment can be used to expand the stall margin of a centrifugal compressor with unshrouded impeller. In addition, the choked margin can also be expanded as experimental results indicated. Moreover, the compressor performance, especially the efficiency, on the whole working range is improved. As shown by experiments, the stall margin and choked margin of the compressor are extended, and the maximum efficiency improvement is 14% at the large flowrate of 1.386. Numerical simulations were carried out to analyze the flow in the impeller and in the holes in the case of large flowrate. The results indicate that in large flowrate conditions, there is a low-pressure region on the throat part of the impeller passage, leading to the bypass flows appearing in the holes, which means the flow area at the inlet of the impeller is increased. The bypass flow can also contribute to the decrease of the Mach number at the throat part near the shroud end-wall which implies that the choked margin is expanded. Besides, as the bypass flow would inhibit the development of the vertexes in the tip clearance and suppress the flow recirculation in the shroud end-wall region, both the pressure ratio and efficiency of the compressor are improved, which agrees well with the experiments. holed casing treatment, centrifugal compressor, numerical simulations, large flowrate condition Citation:
Xu W, Wang T, Gu C G. Performance of a centrifugal compressor with holed casing treatment in the large flowrate condition. Sci China Tech Sci, 2011, 54: 2483−2492, doi: 10.1007/s11431-011-4416-y
Nomenclature b: D: N: β: n: M: T: p: v: Ma: πs:
ηis:
impeller width, mm impeller diameter, mm number of blades blade angle, (°) rotating speed, r min−1 mass flow rate, kg s−1 static temperature static pressure magnitude of absolute velocity absolute Mach number static pressure ratio h is − ht,1 adiabatic efficiency, η is = t,2 ht,2 − ht,1
*Corresponding author (email:
[email protected]) © Science China Press and Springer-Verlag Berlin Heidelberg 2011
hist :
adiabatic total enthalpy
Subscripts 1: inlet position of the impeller 2: outlet position of the impeller t: stagnation point is: adiabatic status design: design condition stall: stall condition choke: choke condition
1 Introduction A compressor has its maximum flow defined as choked flow at a given rotational speed. The pressure ratio of the compressor increases as the mass flow is reduced. Generally, tech.scichina.com
www.springerlink.com
2484
Xu W, et al.
Sci China Tech Sci
the working point at which the pressure ratio reaches a maximum value is defined as the design point. Reduction in mass flow leads to an eventual breakdown of stable flow, which means that the compressor gets into stall or surge. The flow range between choking and stall (surge) is the stable range available for compressor operation, which should be large enough because the compressor or the downstream component may operate at off-design point [1, 2]. Many researchers have been working on obtaining a wider possible operating range for compressor, and most of their studies are focused on expanding the stall margin because of the serious damage of stall. Casing treatment is one of the methods to expand stall margin. It was first put forward by Hartmann et al. [3]. In their experiments, they found that simpler casing treatments, without bleed flow, might provide appreciable improvements in the stall range. Many researchers have been studying the casing treatment on the axial compressor since early 1970s [4–8]. Later on the application to centrifugal compressors was carried out [9,10]. As for turbocharger compressors, a ring-groove structure is used to enhance the stable range [11, 12]. It was first suggested by Fisher [13] and the sketches are shown in Figures 1(a) and 1(b). The holed casing treatment is a kind of casing treatment designed for unshrouded centrifugal compressors. It is shown by experiments that it can expand not only the stall margin but also the choked margin of the compressor obviously. Besides, the compressor pressure rises and efficiencies on the whole working conditions can also be improved [14]. In ref. [11], the stall margin was expanded by 12%, but the impeller efficiency was reduced by 4% at maximum flowrate. The influence of the holed casing treatment on centrifugal compressor in small flowrate conditions has been studied in our previous work [15]. This paper is focused on the
Figure 1
September (2011) Vol.54 No.9
performance of the compressor with the holed casing treatment in the large flowrate conditions. Numerical simulation was adopted to study the flow field in the impellers both with and without the holed casing treatment near the choked point.
2 Study object and numerical simulation method 2.1
Numerical model
A centrifugal compressor with unshrouded impeller is studied in this paper and its design parameters are shown in Table 1. The flow field in this compressor with the holed casing treatment is simulated and compared with the untreated original casing compressor. The sketch of the holed casing treatment for the centrifugal compressor is shown in Figure 2. It consists of a set of bleeding recirculation passage including bleeding recirculation holes and a channel chamber. The bleeding recirculation holes are located in the stationary shroud circumferentially with endpoints called bleed ports as shown in Figure 2. The other endpoint of the bleeding recirculation hole is connected to the channel chamber, which has an outlet called reinjecting port. In small flowrate conditions, the gas from all holes would gather, mix and become uniform in the channel chamber and then reenter the upper region of the impeller passage inlet, mix with the main flow in the impeller passage and improve the flow in the compressor. The configuration of the compressor with the holed casing treatment is shown in Figure 3. The additional parts on the impellers shown in Figure 3 are the recirculation passages including the bleeding recirculation holes and the channel chamber. In this study, the diameter of the holes is 3 mm, the radial position of the holes is
Sketches of casing treatment in references. (a) Sketch in ref. [13]; (b) sketch in ref. [12].
Xu W, et al.
Table 1
Sci China Tech Sci
Design parameters of the compressor Item b1/D2 b2/D2 N
β1 β2 ndesign Mdesign
September (2011) Vol.54 No.9
2485
wall is adiabatic and non-slip. Value 0.166 0.06 16 30 90 22790 0.64
2.2
Performance
The performance curves of static pressure ratio and adiabatic efficiency versus mass flow rate at the rotating speed of ndesign are shown in Figure 4. It can be seen that the working range of the compressor is expanded effectively by the holed casing treatment both in experiments and numerical simulations. Stall margin is defined as SM = (Mdesign−Mstall)/Mdesign.
(1)
Stall margin increment is then defined as SMI = Ms/Mdesign,
(2)
Ms=|Mstall, holed casing treatment−Mstall, original casing|.
(3)
Similar to the above definition, choked margin and choked margin increment are defined as
Figure 2
Sketch of the holed casing treatment.
Figure 3 Compressor with the holed casing treatment.
Dhole/D2=0.686 and the number of the holes is 64. The commercial software package NUMECA/FINE Turbo v. 8.4-1 was used to simulate the flow fields in a single impeller passage with and without the holed casing treatment. The H-type and C-type structured grids were applied for the main flow passage and butterfly type grids for the cross section of the holes in the holed casing treatment compressor as shown in Figure 3. The generated grid qualities are as follows: the minimum orthogonality is 16°, the maximum aspect ratio is 298 and the maximum expansion ratio is 7. The numbers of grid nodes are 478289 and 323936 respectively for compressor with and without the holed casing treatment. The 3-D Reynolds averaged Navier-Stokes equations are integrated in time by a fully implicit formulation of the second-order scheme for the real gas in conjunction with the Spalart-Allmaras turbulence model. The inlet boundary conditions include total temperature and total pressure. Variable outlet static pressures are given for simulations of different flow conditions. The solid
CM = (Mchoke−Mdesign)/Mdesign,
(4)
CMI = Mc/Mdesign,
(5)
Mc=|Mchoke, holed casing treatment−Mchoke, original casing|.
(6)
It can be easily obtained from Figure 4 that the value of SMI is 24.04% and 17.90%, and CMI is 4.84% and 4.31% for experimental result and numerical simulation respectively. Furthermore, it is found that the compressor performance is improved by the holed casing treatment on the whole working conditions. The maximum improvement of ηis is 14% for experimental result at the large flowrate of 1.386 and 23% for numerical result at the large flowrate of 1.289. The static pressure ratio of the compressor on condition that M/Mdesign>1.0 is also improved a little by the holed casing treatment. The SMI obtained from numerical simulation is smaller than the experimental value. This is probably because that the step length of the outlet static pressure in experiment is not as small as that in numerical simulation. The quantity of experimental result is also a little different from the numerical value. The pressure ratio obtained from experiment result is larger while the efficiency is lower. This is mainly due to that there is a diffuser at the downstream of the impeller in the experiment, but not in the simulation. The numerical results agree well with the experimental ones. The numerical results can be used to analyze the impeller performance with the holed casing treatment. The performance curves of static pressure ratio and adiabatic efficiency versus mass flow rate at different rotating speed obtained from numerical simulations are shown in Figure 5. It is indicated that the holed casing treatment is effective at different rotating speed. The effectiveness of the holed casing treatment is more significant at larger rotating speed because of the larger flowrate in the holes.
2486
Xu W, et al.
Sci China Tech Sci
September (2011) Vol.54 No.9
Figure 4 Performance curves of experiments and numerical simulations at ndesign. (a) Experimental static pressure ratio; (b) numerical static pressure ratio; (c) experimental adiabatic efficiency; (d) numerical adiabatic efficiency.
Figure 5
Performance curves of numerical simulations at different rotating speed. (a) Static pressure ratio; (b) adiabatic efficiency.
3 Simulation results and analysis Flow fields in both compressors with and without the holed casing treatment in the large flowrate condition at ndesign were obtained from numerical simulations. The large flowrate condition, which is M/Mdesign=1.283 for original casing compressor, is compared with M/Mdesign=1.284 for
the holed casing treatment compressor. As a supplement, the flow field at M/Mdesign=1.332 which is close to the choked point is also given for the holed casing treatment compressor. Both experiments and simulations indicate that the holed casing treatment compressor has higher pressure ratio and higher efficiency than original casing compressor on flow condition M/Mdesign=1.283.
Xu W, et al.
Sci China Tech Sci
3.1 Flows in the bleeding recirculation holes The influence of the holed casing treatment on compressor performance comes from the flow adjustment caused by the bleeding recirculation passages. When the compressor operates on varied flow conditions, the static pressure differences between the bleed port and reinjecting port are different, the mass flows and even the flow directions in the bleeding recirculation passages are also distinct. And as a result, the adjustment effects on flow field in the compressor are different. The relationship between the relative mass flow in the holes and the compressor inlet is shown in Figure 6. The y-axis denotes the relative value of the total mass flow in all holes divided by the compressor inlet mass flow. It is defined as positive if the flow direction in the hole coincides with that in the impeller passage and the flow in the holes is defined as bypass flow. Otherwise, the flow is defined as reinjected flow. The x-axis is the relative value of the compressor inlet mass flow divided by Mdesign. The mass flow in the bleeding recirculation holes changes with the compressor inlet mass flow. The relative mass flow in holes Mhole/M is linear with the relative compressor inlet mass flow M/Mdesign. When M/Mdesign is larger than 0.95, the mass flow in the holes is positive, the flow direction in the holes coincides with that in impeller passage and the flow in the holes is bypass flow. When M/Mdesign is less than 0.95, the case is contrary and the flow in the holes will be reinjected flow having reverse flow direction with that in impeller passage. If X denotes the value of x-axis, Y denotes the value of y-axis, the fitted relationship between X and Y can be got as Y=10.9148X−10.3487.
(7)
2487
September (2011) Vol.54 No.9
the hole. Figure 7(a) also indicates that the flow direction and mass flow in holes change with compressor inlet mass flow. In small flow conditions, the flow in holes is reinjected flow. While in large flowrate conditions, the flow in holes is bypass flow. The more the flow condition deviates from the design point, the larger the reinjected flow or the bypass flow will be. The bypass flow, which goes through the stationary part and the compressor, does not do work on it. This can benefit the compressor efficiency. The larger the mass flow of the bypass flow is, the larger this benefit will be. Figure 7(b) indicates that at the small flow condition the reinjected flow in holes coming from the shroud end-wall region is spiral flow because of the prerotation. The reinjected flow mixes and becomes uniform in the channel chamber. In the large flowrate condition, the bypass flow in holes is relatively stable. 3.2
Flow fields in impeller passage
The flow fields in the impeller passages of the compressors with and without the holed casing treatment in large flowrate conditions are obtained to study the influence of holed casing treatment. The studies of the following section include flow fields at 10%, 50%, 90% height of the blade and the middle surface of the blade passage. The distributions of static pressure on 10%, 50% and 90% blade height surface of compressors both with and without the holed casing treatment are shown in Figure 8. In the large flowrate condition, there is a low pressure region at the throat part of the impeller passage where the pressure is lower than the inlet pressure. The pressure near the reinjecting port is higher than that near the bleed port
The flow fields in a hole in different flow conditions are shown in Figure 7. Figure 7(a) is the absolute velocity vectors and Figure 7(b) is the absolute velocity vector lines in
Figure 6 Relative mass flow in the holes versus relative compressor inlet mass flow.
Figure 7
Flow fields in a hole.
2488
Xu W, et al.
Figure 8
Sci China Tech Sci
September (2011) Vol.54 No.9
Static pressure distributions on surfaces of 10%, 50%, and 90% blade height.
and the pressure difference can overcome the flow resistance. The inlet flow is divided into two parts. One part of the flow goes into the channel chamber becoming the bypass flow, so the flow area of the throat part is increased and the choked margin is expanded. The pressure rise of the compressor is obtained in the part from the tail of the low pressure region to the outlet of the impeller in large flowrate conditions because of the existence of the low pressure region. The bypass flow does not pass through the low pressure region and arrives directly at the tail part of the region through the stationary part. The pressure of the bypass flow is larger than that in the low pressure region and the pressure of the tail part is raised. The pressure of the outlet is then improved and the pressure
ratio of the compressor is increased. As shown in Figure 8, the static pressures in compressor with the holed casing treatment is obviously higher than that with original casing on flow condition M/Mdesign=1.283 on 10%, 50% and 90% blade height surfaces. It is particularly evident on 10% blade height surface and the throat part of 50% and 90% blade height surfaces. The static pressure at the front part of the impeller passage increases because the mass flow into impeller passage inlet decreases. The static pressure at the back part of the impeller passage is raised because of the addition of the bypass flow mentioned above. The improvement of the static pressure distribution found in the large flowrate condition is different from the situation in the small flowrate conditions. In the small flowrate condi-
Xu W, et al.
Sci China Tech Sci
tions, the improvement of the flow fields is caused by the smaller reinjected flow. Only the shroud end-wall region of the impeller passage is influenced. In the large flowrate conditions, not only the shroud end-wall region but also the hub end-wall region is influenced because the mass flow in the holes is relatively large. Actually, the static pressure distributions in compressor with original casing at M/Mdesign= 1.283 are more like that in compressor with the holed casing treatment at M/Mdesign=1.332 which is close to the choked point. The relative velocity vectors on surfaces of 10%, 50% and 90% blade height in both compressors are shown in Figure 9. Corresponding to the distributions of static pres-
Figure 9
September (2011) Vol.54 No.9
2489
sure, the relative velocities are lower in compressor with the holed casing treatment than in compressor with original casing, especially on the throat part of the impeller passage and 10% blade height surface on flow condition M/Mdesign= 1.283. The relative velocities on 10% blade height surface in original casing compressor at M/Mdesign=1.283 are even higher than that in the holed casing treatment compressor at M/Mdesign=1.332. Figure 10 shows the relative velocity vectors on the middle S2 surface of the impeller passage. Referring to Figures 9 and 10, it can be found that there is a recirculation zone in the upper region of the compressors both with and without the holed casing treatment. The scale of the recirculation zone is smaller in the holed casing
Relative velocity vector distributions on surfaces of 10%, 50%, and 90% blade height.
2490
Xu W, et al.
Figure 10
Sci China Tech Sci
September (2011) Vol.54 No.9
Relative velocity vector distributions on the middle S2 surface of the impeller passage.
treatment compressor on 50% blade height surface as shown in Figure 9(b). This is because the bypass flow from the bleeding recirculation passages disturbs the recirculation flow and improves the flow in the impeller passage. At 90% blade height, the recirculation zone becomes greater. The improvement of the holed casing treatment on the flow in the impeller passage becomes more apparent, especially on the suction side. It can be found from Figure 9(c) that the length of the recirculation zone on suction side in compressor with original casing is about half of the blade length, while is less than 1/3 of that in compressor with the holed casing treatment. The decrease of the recirculation zone is probably caused by the bypass flow. It is known that the bypass flow is controlled by the static pressure difference between the bleed port and reinjecting port of the bleeding recirculation passages. The static pressure difference near the suction side is larger so the bypass flow near the suction side is greater and the improvement of the flow near the suction side is more significant. The recirculation zone in compressor with the holed casing treatment on flow condition M/Mdesign=1.332 is smaller than that in compressor with original casing on flow condition M/Mdesign=1.283 although the inlet mass flow is larger. This is because larger bypass flow goes through the bleeding recirculation passages on flow condition M/Mdesign=1.332 and it suppresses the flow recirculation in the shroud end-wall region further. 3.3 Flow fields in the tip clearance The static pressure distributions in the tip clearance are shown in Figure 11. It is shown that the static pressure in the tip clearance is raised by the holed casing treatment like that in the impeller passage. The parameter values of point A at the throat part and point B near the bleed port are given in Table 2. It can be found that the temperature in the low pressure region at the throat part is low and the sound velocity is low. The Mach number can easily reach 1.0. Beside the increase of the flow area at the inlet of the impeller, the bypass flow can also benefit the expansion of the choked margin. This is because the temperature and pressure of the bypass flow are higher and the velocity is lower. The Mach
number on the throat part can be decreased by the bypass flow. Table 2
Parameters of two points in the tip clearance
Parameter TA (K) pA (Pa) vA (m s−1) MaA TB (K) pB (Pa) vB (m s−1) MaB
Holed casing Holed casing Original casing (M/Mdesign=1.283) (M/Mdesign=1.284) (M/Mdesign=1.332) 250 266 253 53994 66942 56732 351 287 338 1.11 0.88 1.06 279 301 298 64975 85513 79310 319 190 214 0.95 0.55 0.62
The vorticity distributions in the tip clearance are shown in Figure 12. It is indicated that the bypass flow can reduce the vorticity in the tip clearance and inhibit the development of the vortexes in the tip clearance. This can suppress the decrease of flow area caused by the vortexes.
4 Conclusions Experiments and numerical simulations indicate that the holed casing treatment can improve effectively the stall and choked margin of the centrifugal compressor. It can also improve the pressure ratios and efficiencies of the compressor in large flowrate conditions obviously. By analysis on the flow fields in compressors with and without the holed casing treatment at ndesign, the following conclusions can be obtained: (1) In the large flowrate case, there is a low pressure region on the throat part of the impeller passage where the pressure is lower than the inlet pressure. Then the pressure difference between the reinjecting port and the bleed port will overcome the flow resistance on the hole and drive the bypass flows from the reinjecting port directly to the tail part of the low pressure region, which means the flow area at the inlet of the impeller increases as well as the flowrate. (2) As we know, the relative velocity is higher and the
Xu W, et al.
Figure 11
Sci China Tech Sci
September (2011) Vol.54 No.9
2491
Static pressure distributions in tip clearance.
Figure 12
Vorticity distributions in tip clearance.
temperature is lower at the throat part near shroud end-wall, which will make the Mach number reach 1.0 easily, thus choking status appear. However, the existence of the bypass flow increases both temperature and the sound velocity, then the Mach number decreases and the choked margin will be expanded. Besides, the bypass flow would inhibit the development of the vortexes in the tip clearance and suppress the decrease of the flow area caused by the vortexes, which will raise the working range at large flowrates. (3) In large flowrate conditions, the pressure rise of the compressor is obtained in the part from the tail of the low pressure region to the outlet of the impeller. The bypass flow can increase the pressure at the tail part of the low pressure region, and then make the pressure at the outlet rise. (4) Since the bypass flow goes through the stationary part, the impeller does not do any work on it. This implies that the impeller saves a part of energy and increases the compressor efficiency. Besides, the bypass flow breaks into the recirculation zone near the shroud end-wall and would improve the flow pattern in the upper region especially the tip
clearance zone of the impeller, which would help to increase the efficiency. This work was supported by the National Natural Science Foundation of China (Grant No. 50776056) and the High Technology Research and Development Program of China (“863” Program) (Grant No. 2009AA05Z201). 1 2 3 4
5 6 7
Pampreen R C. Compressor Surge and Stall. Norwich, Vermont: Concepts ETI, Inc., 1993 Cumpsty N A. Compressor Aerodynamics. Malabar, Florida: Krieger Publishing Company, 2004 Hartmann M J, Benser W A, Hauser C H. Fan and compressor technology. NASA SP-259, 1970 Lu X G, Chu W L, Zhu J Q, et al. A review of studies on casing treatment of axial-flow compressor (in Chinese). Adv Mech, 2006, 36(2): 222–232 Hathaway M D. Self-recirculating casing treatment concept for enhanced compressor performance. ASME Paper, GT-2002-30368, 2002 Strazisar A J, Bright M M, Thorp S, et al. Compressor stall control through endwall recirculation. ASME Paper, GT2004-54295, 2004 Beheshti B H, Teixeira J A, Ivey P C, et al. Parametric study of tip clearance-casing treatment on performance and stability of a transonic axial compressor. J Turbomach, 2004, 126(4): 527–535
2492 8
9
10 11
Xu W, et al.
Sci China Tech Sci
Huang X D, Chen H X, Shi K, et al. An analysis of the circumferential grooves casing treatment for transonic compressor flow. Sci China Phys Mech Astron, 2010, 53(2): 353–359 Gao P, Chu W L, Yang Y. Numerical analysis for the unsteady flow and casing treatment in a low speed centrifugal compressor (in Chinese). J Propul Technol, 2007, 28(1): 32–35 Xiao J, Xu W, Gu C G, et al. Self-recirculating casing treatment for a radial compressor. Chinese J Mechl Eng, 2009, 22(4): 567–573 Ishida M, Surana T, Ueki H, et al. Suppression of unstable flow at small flow rates in a centrifugal blower by controlling tip leakage flow and reverse flow. J Turbomach, 2005, 127(1): 76–83
September (2011) Vol.54 No.9
12 13 14
15
Yin J, Li P, Pees S. Optimization of turbocharger ported shroud compressor stages. ASME Paper, GT2009-59248, 2009 Fisher F B. Application of map width enhancement devices to turbocharger compressor stages. SAE paper, No. 880794, 1988 Wang T, Xu W, Gu C G, et al. A new type of casing treatment with self-adaptive for a centrifugal compressor. ASME Paper, GT-201023457, 2010 Xu W, Wang T, GU C G. Analysis of centrifugal compressor with holed casing treatment at the small flow condition (in Chinese). In: Chinese Society of Engineering Thermophysics Conference. Nanjing: Chinese Society of Engineering Thermophysics, 2010
