Investigations of drilled and multi-piped impellers cavitation performance

Abstract The industry needs the rotodynamic pumps operating with ultra-low specific speed and relatively low flow rate more often. Designing of such structures on acceptable efficiency level is extremely difficult and require nonstandard approach to design as for example: drilled impeller or patented by author multi-piped impeller. Such pump elements are very easy to manufacture and operate with relatively high efficiency, but cavitation behavior is unknown. This paper focuses on experimental research in order to determine the cavitation characteristics of the drilled impellers and multi-piped impellers. The test rig was presented. Impeller models were made by means of SLS Rapid Prototyping methods. Additionally, CFD calculations were presented in order to determine static pressure distribution in the inlet sections of the investigated impellers.


Introduction
Modern world industry, especially chemical one, requires pumps that will be able to operate in range of extremely low specific speed (nq <10) with acceptable level of efficiency. Designing rotodynamic pumps operating in mentioned range of parameters is extremely difficult because of low performance [1][2][3]. Such structures must be designed in non-standard way. During the work over this problem, the idea of drilled impeller [4] was developed. The new solutions of the rotodynamic impellers were introduced as: 1. Drilled impeller with side holes [5]. 2. Multi-piped impeller [6].
The numerous researches confirmed positive operation of such structures form energy consumption point of view [7,8], but cavitation performance was unknown.
The main aim of this project was determining cavitation characteristics of the pump impellers listed above, designed in non-standard way. The special test rig was described and results of measurements were presented. Additionally, in order to interpret better the results of measurements, the CFD calculations were carried out and the static pressure at the impellers inlet passages were shown.   Impeller 1 (I1). The first one (I1) was a drilled impeller with additional side holes. Such a structure is a developing of the commonly known idea of drilled impeller presented in [4]. Basic parameters were shown in the Figure 1.
Impeller 2 (I2). The second one (I2) was multi-piped impeller, designed according to patent [6]. Such structure was described in detail in [9]. The shape and basic parameters were presented in Figure 2.
Impeller 3 (I3). Impeller 3 is a modification of Impeller 2 where the number of the channels (z) and external diameter (d 2 ) were decreased to z =4 and d 2 =130 mm respectively. The model was shown in the Figure 3.  Impeller 4 (I4). Impeller 4 is a modification of Impeller 2 concept where the pipes that create impeller passages were bended in such way in order to obtain optimal values of the inlet and outlet angles. The number of passages is z = 4, inlet angle β 1 = 50 deg., outlet angle β 2 = 30 deg.

Test rig
In order to conduct investigation, a specialized test rigpresented in Figure 5 -was designed and constructed [9]. The main element of the test rig is a pump (Figure 5b) with a special design allowing for the quick replacement of the tested impellers while maintaining the repeatability of the measurement results.
The pump is supplied from a closed tank where -depending upon the needs -one can generate overpressure or negative pressure and control the medium level and temperature.
The pump capacity can be adjusted with a ball control valve, namely MARS 88V with an IntrOM OM-1 electric drive. The measuring instruments which parameters are presented in Table 2 were used for measuring the specific values.
The characteristics of the tested impeller were measured in a fully automated way, according to the recommendations given in EN IS0 9906:2000. The measurement process is controlled by a computer and dedicated software. The test rig gives possibility of cavitation research by means of three methods: • generation vacuum in the tank, • decrease water level in the tank, • throttling at the suction pipeline.
The third method (throttling at the suction pipeline) was selected as the fastest and which generated accuracy on the acceptable level.

Result of the measurements
The basic parameters were determined according to following formulas: Pump head: Figure 5: View of the test rig [9] Hydraulic power: Pump efficiency Net Positive Suction Head

Required Net Positive Suction Head
Based on data presented in Table 1 the average measurement uncertainness were determined as follows: • for pump head ∆H = 4%, • for hydraulic power ∆P = 8%, • for NPSH ∆NPSH = 3%.
The results of the measurements are presented in the Figures 6-13.
Based on the results presented above, we can formulate the following conclusions: 1. Only impeller 2 achieved the assumed operating parameters (Table 1). 2. The maximal efficiency was reached by impeller 1 and impeller 4, respectively. 3. For the flow rate Q=4.5 m 3 /h, the best cavitation performance achieved impeller 4 (NPSHr = 8m), the worst -Impeller 1 (NPSHr=10.5 m). 4. The lowest value of NPSHr=8 m (Impeller 4) is much higher than expected for classical centrifugal impeller (NPSHr=5.5 m -value obtained from author's research).

CFD modeling
In order to better understand cavitation behavior of the investigated impellers, the CFD calculations were performed for all impellers. The main aim of this calculations was to determine the static pressure distribution and to evaluate the pressure drop. Due to this fact, analysis were done without multiphase model. General description of the numerical model, applied grid, boundary conditions and rules of CFD modeling in detail one can find in [9]. In the Figures 14-17 the static pressure distribution was presented.
Based on presented results, we can formulate following conclusions: 1. In case of Impeller 1 additional pressure drop is observed in the inlet to the first side hole. This is the reason of the worst cavitation performance.
2. The Impellers 3 and 4 have better cavitation performance than Impeller 2 ( Figure 15). This confirms assumption that impellers with lower number of passages and larger area of the inlet cross-section work better (lower pressure drop at the inlet section - Figures 16-17). 3. The Impeller 4 has lower NPSHr than Impeller 3. It means that there is optimal value of the inlet angle β 1 from cavitation performance point of view (Figure 17) [10].

Summary
Drilled impellers and multi-piped impellers are an interesting alternative for classical centrifugal impellers in the range of extremely low specific speed (nq <10). The values of the NPSHr for all investigated impellers are much higher than for classical centrifugal impellers. Among the researched impellers the best one was Impeller 4, that had 4 passages and the value of inlet angle β 1 within the optimal range [10].