Gene Vogel
EASA Pump and Vibration Specialist
When motors are installed on top of vertical pumps, high vibration is a common problem. The source of the problem can be a mechanical issue with the pump, motor or coupling, or it can be hydraulic forces from the pump. Often structural issues involving resonance amplify the vibration. An understanding of the nature of this style pump and the various forces is essential to diagnosing and correcting vibration problems on vertical pump motors.
There are quite a number of configurations of vertical pumps. Submersible pumps fall into this general category. This discussion, however, will omit submersibles and focus on those pumps that are surface mounted where the motor is bolted to a pedestal on top of the pump. See Figure 1. This is the style that most commonly exhibits high vibration conditions. An important contributing condition is resonance, and specifically “reed frequency” resonance. But an understanding of the vibratory forces is important also.
The most common vibratory force from rotating machinery is mass unbalance. That is certainly true for vertical pump motors. Accurate balancing of the motor rotor in a balancing machine will not eliminate all of the unbalance sources for this style motor. Often the top thrust bearing is mounted on a carrier hub, which is a clearance fit to the shaft and is locked in place by a shaft nut. This allows some variance of the rotating axis in relationship to the principal axis of inertia, creating unbalance.
Then, on hollow shaft designs, there is a massive hub that mounts to the top of the rotor. The hub is also subject to unbalance, eccentric mounting, and it may also have moving parts for the anti-reversing ratchet. This hollow shaft design also has the pump shaft fitted through the motor shaft and bolted to the top hub. If the pump shaft and retaining hardware are slightly eccentric, unbalance results.
Coupling alignment
A second vibratory force is coupling alignment. For the hollow shaft design the coupling is simply the motor shaft bolted to that top hub. The only flexible element in this drive train is the pump shaft, so if there is any offset or angular misalignment of the pump shaft and hub, significant forces result.
An alternate design is for the pump shaft and motor to be coupled at the bottom of the motor. Here too, often the coupling is not a flexible style. It is a solid coupling supporting the weight and thrust of the pump. The opportunity for vibratory forces from misalignment is much greater than for conventional horizontal mounted flexible couplings.
Mechanical action of pump shaft, impeller
A third source of vibratory force is the mechanical action of the pump shaft and impeller. This style pump shaft is very long and slender, and susceptible to twisting and bending. Most commonly, the pump shaft is stabilized by journal type guide bearings at intervals down the pump casing. The clearance on these bearing is subject to variation and the shaft can whip or “dance” in the casing, especially in very deep pumps.
Hydraulic action of fluid
The fourth significant vibratory force is the hydraulic action of the fluid being pumped as it moves through the pump casing and out the discharge. The discharge from most of these pumps is horizontal, so the fluid must make a ninety degree turn, usually at the top of the pump, just below the motor. Flow turbulence at this location can be a strong exciting force.
Several of these forces are also potentially present for horizontal mounted pumps and motors, and those base structures are designed to withstand the forces. However, the long moment arm created by the tall motor mounted on top of the pump flange makes it difficult to design rigidity into the structure. Thus resonance frequencies may be low enough to coincide with any of these vibratory forces, and high vibration results. So a discussion of resonance is in order.
Resonant frequencies
First, all machine structures have resonant frequencies. Ideally, those frequencies are well above any exciting force. Design engineers work hard to achieve that condition. The resonant frequency results from the mass and stiffness characteristics of the structure. Adding stiffness raises the resonant frequency; adding mass lowers it. The strings of a guitar are a good analogy: Tightening a string stiffens it, raising the tone; the thicker strings have more mass and produce the lower tones. Another example which simulates the vertical pump motor is a simple metal ruler. Hold the ruler on one end and strike it, and it will ring at its natural, or reed, frequency.
The similarity to the vertical pump motor, held on one end and free on the other, is obvious. The motor represents a large mass relative to the structure which tends to lower the resonant frequency. The stiffness of the supporting mounting flange is affected by the foundation, and especially by how well the pump is bolted to that foundation. Now the pump that hangs below the mounting flange also is a significant mass and has resonant frequencies, as does the discharge piping, also attached to the mounting flange.
When you look at the machine as a whole – pump, motor and discharge piping, the flange to which they all attach looks pretty small and weak. So if you wanted to make music, the vertical pump has some real possibilities. Unfortunately, that music is what most facilities call the sound of trouble.
The question about why such a machine is vibrating severely really should be: Why wouldn’t it vibrate severely? First, it would have to have enough stiffness in that mounting flange to raise the resonant frequencies well above operating speed and other exciting forces. Second, the mechanical alignment and balance would have to be adjusted to within reasonably tight tolerances. And third, the hydraulic flow through the pump and piping would have to be smooth and not turbulent. Any of those factors can cause high vibration.
Conduct basic frequency analysis
Analyzing vibration on these machines to determine the contributing factors is fairly straightforward, and involves phase analysis and operating deflection shape (ODS). First, however, a basic frequency analysis should be conducted with a portable vibration analyzer. In most cases, a single vibration frequency will be dominant; often that frequency is rotating speed. A simple process of elimination comes into play here: If the dominant frequency is rotating speed, the problem is not hydraulic forces from the pump. If the dominant frequency is not rotating speed, then neither unbalance nor misalignment is a source. A simple bump test should also be conducted to identify the resonant frequencies of the mounted motor.
By far, the most common occurrence is vibration at 1 x rpm, so let’s look further at that condition. Phase analysis will confirm that the vibration is at 1 x rpm, and the amplitude and phase, taken at various locations on the motor and pump flange, will show a “picture” of the vibratory motion. The process of using amplitude and phase vectors to visualize the motion is the basis of an ODS analysis. See Figure 2. While computer software can animate an ODS, a simple pencil sketch with the vectors plotted will often provide adequate information. Amplitude and phase can also be used to attempt to reduce the vibration by trim balancing.
The coupling hub or fan at the top of the motor will provide a convenient location for trim balancing weights. This is also an important diagnostic step. If very small weights cause significant changes in the amplitude and phase vector, then a resonant condition is confirmed. If reducing the vibration in one direction, say the North-South direction, causes an increase in the East-West direction, then unbalance is not the primary exciting force, and alignment or a warped pump shaft should be suspected.
Consider other possibilities
When vibration is not at rotating speed, the process of elimination needs to consider a number of other possible forces. While these are not included in this article, basic vibration analysis techniques can help to isolate the most likely causes. While the focus of attention is the motor, the pump, perhaps unseen below the surface, can generate vibratory forces that will be reflected in the motor vibration.
The seemingly simple case of vertical pump motor vibration can take the vibration technician on an analytical journey. By recognizing that resonance, pump generated forces and the interaction between the motor and the pump shaft can all be contributing factors, the technician can approach the problem objectively and use the process of elimination to find solutions to the problem. Those who approach the problem unprepared may find only frustration.
Print