Session: 37-07 Radial - Performance 2

Submission Number: 178193

Numerical Study on the Influence of a Volute Neck on Compressor Performance 

Volutes are an important piece of many typical centrifugal compressor flow paths, as they affect the compressor performance by collecting the discharge flow and directing it toward the outlet pipe and sometimes even performing additional flow diffusion. The volute cross section may include a neck which extends material above the minimum fluid volume radius of the volute. The neck is distinct from and should not be confused with either the tongue, or the throat of the volute. Effectively, a neck, when present, enables the extension of the rotationally periodic annular vaneless diffuser space to a radius above that of the minimum internal radius of the volute, radially elongating the available vaneless diffusion zone and allowing the swirl to be guided internally in the volute by the neck wall. Adjusting the relative radial position of the volute neck influences both how much upstream radial diffusion occurs prior to entering the volute, as well as how much change in radial position there is of the volute cross section around the 360° portion (the throat) of the volute (assuming the comparable volutes have the same outlet flow area). It was of interest to see how the relative extent of the volute neck influences the volute flow field as well as the overall compressor performance.

To perform this study, the inlet duct and impeller from the open-source NASA High-Efficiency Centrifugal Compressor (HECC) were used. A radial vaneless diffuser and volute were used downstream of the impeller. Several (7 specifically) geometric variations of the diffuser elements were analyzed, where the radial length of the diffuser was adjusted depending on the positioning of the volute neck. As the neck position was adjusted, the volute cross section form was also adjusted to ensure the preservation of the outlet flow area from the volute. Additionally, all cases maintained the same maximum (external) diameter. The extreme cases analyzed included both a volute with no neck, as well as a volute with a large neck radial position with the volute cross section centroids having a constant radial position. Six separate cases with varying volute neck positions were generated in addition to a no-neck P-type volute and their design speed-lines’ performance was analyzed using steady-state RANS CFD methods.

When comparing the resulting compressor performance across the seven cases, there clearly is a performance effect from the volute neck. Of course, in real applications, adjusting the neck for maximum performance needs to be balanced with other factors such as manufacturability.

In this analysis, the choke flow rate for each volute design speedline is slightly different. Detail review of the flow field in both the impeller passages (which are identical) and the volute cross sections indicate that the variation in choke flow is due to simultaneous choking effect (Mach 1 flow with observed shocks) in both the impeller and volute passages and may be a relatively unique situation occurring due to the specific sizing of the volute.

There is clear indication of an optimu. There appears to be an optimum in both pressure recovery and total pressure loss as the volute neck extends 31% into the overall height of the volute cross section. Just a small neck seems to have the maximum incremental benefit, with diminishing returns as the neck extends closer to the optimum. Beyond the optimum, the effect of the larger neck appears to be detrimental across nearly the entire speedline.

Overall, we see that the peak efficiency of the overall compressor can be improved from 84.6% to 86.0% by optimizing the extent of the volute neck, with a corresponding increase in overall pressure ratio of 1.9%.

The pressure recovery coefficient peaks at 0.65 for the neck extension of 31%, compared to 0.59 for the no neck (P-type) volute. Correspondingly, a P-type (no neck) volute results in a minimum pressure loss coefficient of about 0.25, while a 31% neck extension lowers the total pressure loss coefficient to a minimum of 0.21.

Presenting Author: Vlad Goldenberg SoftInWay

Presenting Author Biography: Vlad Goldenberg received his Ph.D. in Mechanical Engineering from the University of Minnesota. He has previous degrees in mechanical engineering, physics, and chemistry from the University of St.Thomas and the University of California, Berkeley. He has worked as an engineer across several industries and is presently an Engineering Manager and Innovation Lead at SoftInWay, where he leads a team of turbomachinery professionals and leads innovative projects.

Authors:

Vlad Goldenberg SoftInWay
Benjamin Conser SoftInWay, Inc