Image of double suction impeller
Hydraulic axial thrust in double suction impellers
A standard double suction (DS) impeller has identical sides – one side is a mirror image of the other side – and has balanced wear ring diameter. The outside diameter (OD) of each ring is the same so the hydraulic axial thrust on each side is presumed to be equal and cancels each other essentially resulting in zero or negligible thrust (Fig. 1).
There are many incidents of single stage, double suction (SSDS) pumps with anti-friction (ball) thrust bearings whose bearings fail due to skidding. The ball skid is the result of the bearings being presumably unloaded due to zero thrust on DS impellers.
To prevent this failure some people think that DS impellers should have unbalanced wear rings to preload the bearings and prevent the balls from skidding (Fig. 2). Unbalanced rings refer to wear rings with different OD to change the impellers’ thrust profile; they are also called bias, offset, or stepped rings.
The idea that DS impeller has balanced thrust comes from the simple method of thrust calculations that pump companies use (Fig. 3). The calculations are based on suction pressure, differential pressure, wear ring diameter, and impeller shroud diameter. These parameters are presumed equal on both sides and cancel out each other thereby resulting in zero thrust. For this reason, the thrust of a standard DS impeller is rarely calculated.
In reality, the thrust on each side is affected by other factors such as asymmetrical flow, momentum thrust, internal flow recirculation, rotor end-float (or end-play), variations in impeller and volute castings, shaft thermal growth, clearances, etc. But their values are indeterminate so they are omitted in the calculations.
Hypothetical example
Consider a hypothetical example of a SSDS pump with an initial thrust imbalance of 200 lbs. The thrust may be towards the OB or IB end of the pump depending on the factors causing the imbalance.
Thinking that the pump has zero thrust, the engineer purposely designed the DS impeller with unbalanced wear rings to load the rotor with 200 lbs thrust towards its IB end. (The practice of balancing the thrust towards its IB end is discussed in the full version of this article.)
If the initial thrust was also in the IB direction, then the total thrust would be 400 lbs. The higher thrust load would result in reduced bearing life. But if the initial thrust was in the OB direction, then the total thrust would be zero – exactly the situation he was avoiding to prevent the bearings from failing.
This illustrates the uncertainty and high risk of using unbalanced wear rings. There have been many SSDS pumps, with unbalanced wear rings, that developed very high bearing temperature such that the unbalanced wear ring design has to be changed or abandoned.
Factors affecting the thrust of DS impellers
Indeed, the hydraulic axial load tests on several SSDS pumps, with balanced wear rings, consistently show that their thrusts are not balanced, and the imbalances are significantly high - contrary to the presumption that they are zero. Furthermore, the thrusts have no consistent direction - they may be toward the outboard or non-drive end (OB or NDE), toward the inboard or drive-end (IB or DE), or fluctuating between OB-IB, particularly when the pumps are running below their MCSF.
Factors that can cause DS impellers to have unbalanced thrust are:
· Not all DS impellers are of conventional design (Fig. 4). Many are of special service design such as impellers with rake angle discharge vane design, with angle-cut vanes, or with full center rib and staggered vanes, etc. These special design features can disturb the thrust balance of the impellers.
· Asymmetrical flow will result in thrust imbalance and imbalance of momentum thrust. (Read about the causes of asymmetrical flow in the full version of this article.)
· The hydraulic radial thrust can have unbalanced axial components depending on the volute hydraulic profile at the neck and throat areas.
· Variations in impeller and volute castings.
· Error in locating the true discharge vane centerline and its perpendicularity with the shaft centerline, during rough machining of the impeller casting. Consider this: a dynamically balanced impeller at maximum diameter would become out of balance if its diameter were cut - and would have to be rebalanced.
· Effect of rotor end-float (end-play), or shaft thermal expansion, resulting in offset impeller and volute centerlines, among others.
(Other factors are discussed in the full version of this article.)
Results from thrust load tests
A review of the hydraulic axial load tests on several SSDS pumps, with balanced wear rings, consistently show the thrusts are not balanced, and the imbalance are significantly high, contrary to the presumption that the thrusts are zero. These observations are made:
· All tests indicate significant thrust imbalance along a wide flow region within the pump performance curves.
· The lowest thrust occurs at, or near, BEP - but at a very narrow flow region.
· At some point, the thrust reverses direction – in many tests, the reversal occurred more than once.
· The thrust magnitude is sometimes erratic – it is not always a direct function of flow, or of differential head.
· The thrust may be towards OB, towards IB, or reverses direction between OB and IB. At times, the thrust reversal happened multiple times when the SSDS pumps were running below their MCSF.
Conclusion
The common method of calculating the thrust of DS impeller is unreliable. A thrust load test will give a more accurate thrust profile of the impeller, or SSDS pump.
A DS impeller does not have balanced thrust contrary to the general presumption. Pre-loading the anti-friction bearings with some amount of thrust by unbalancing the impeller ring OD can do more harm than good if the initial thrust imbalance and its direction were not known beforehand.
Attachments available in the full version of this article:
Fig. 1 – Standard double suction impeller
Fig. 2 – Impeller with unbalanced wear rings
Fig. 3 – Axial thrust calculation format
Fig. 4 – Special purpose double suction impeller
Fig. 5 – Hydraulic axial thrust curve 1
Fig. 6 – Hydraulic axial thrust curve 2
Fig. 7 – Hydraulic axial thrust curve 3
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.