Mechanical seals pumps pdf

When the inside or primary seal fails, the leakage through the faces will be contained by the secondary seal until the pump can be shut down for seal replacement. The difference is in how the seals perform. It operates sealing high barrier pressure while the inboard or primary seal has clean lubricating liquid applied at differential pressure of only 20 to 30 psi.

Open navigation menu. Close suggestions Search Search. User Settings. Skip carousel. Carousel Previous. Carousel Next. What is Scribd? Explore Ebooks. Bestsellers Editors' Picks All Ebooks. Explore Audiobooks. Bestsellers Editors' Picks All audiobooks. Explore Magazines. Editors' Picks All magazines.

Explore Podcasts All podcasts. Difficulty Beginner Intermediate Advanced. Explore Documents. Session-6 - Mechanical Seals. Uploaded by Pankaj Sharma. Did you find this document useful? Is this content inappropriate? Report this Document. Flag for inappropriate content. Download now. Related titles. Carousel Previous Carousel Next.

TM Pump, Centrifugal 1. Jump to Page. Search inside document. The seal gland to the stuffing box O. Giorgiana Rosu. Andrey Pulido Barrera. Pranay Kharbe. Mahmoud Fathy. John Robert Gonzales. Matt Ferdock. Pt Prakash. Harris Jum'anianda. Sumedh Singh. Popular in Gases. Mohammad irfan. Nur Ali Said. Faizan Ahmed. Datt Nguyen. Jack Whitener. Firman Adi Susetyo. The vibration magnitude was obtained [1] and increases when the pump increases speed.

Although the reliability of the Tank 7 pumps has not been established, Fig. This report summarizes that work and provides a means to predict life for long shaft vertical pumps.

New developments were summarized throughout these reports: 1 Fatigue damage to mechanical seal faces is the primary cause of seal leakage. The cracking and spalling of the seal faces is caused by the relative vibration between the seal faces as they impact each other.

The magnitudes of forces during impact are directly affected by the cooling provided to the seal. That is, the fluid film provides damping between the faces, and the film between the faces is diminished when the cooling to the seal is reduced. Two tests on the EFP and one test on a pump later used in Tank 8 demonstrated this conclusion. In two EFP tests, no cooling was supplied to the buffer region of a double mechanical seal, and the pump was operated at its critical speed.

Seal leaks occurred within hours of operation in one test and in the other. Following the hour test, the seals were inspected and found to have several fatigue cracks caused by impacts. Under these conditions, the pump has operated in excess of 13, hours, and continues to operate.

Operating at a critical speed is expected to damage the seals [4]. Mechanical seals not equipped with a pumping ring to circulate cooling fluid were operated at a critical speed, and the rotating face was damaged within hours of operation. Operating away from this critical speed, the average life of mechanical seals was hours on pumps.

On this same style of pump, the bearings were changed from journal bearings to tilt pad bearings, and the vibrations due to whirl were reduced. The modified pumps have operated hours without seal leaks [1]. In short, replacement of the carbon journal bearings of a long-shaft vertical pump with tilt pad bearings increases the reliability of the lower mechanical seal and pump. Critical speeds causing high vibration in long- shaft pumps are attributed to both shaft bending, or flexure, and column flexure.

The light bearing loads affect the shaft stiffness and thereby reduce the critical speeds. In the vibration data, the measured shaft orbits tended to be random when whirl existed on the pumps. However, the orbits tended to settle down to a repeatable pattern when the whirl was eliminated. As an impeller moves within its volute, reaction forces set up because of the non-symmetrical static pressure difference around the periphery of the impeller. The primary reaction forces are a negative direct stiffness, and a cross-coupling stiffness.

These forces tend to be destabilizing for vertical pumps, where stability has been an issue. Along with the reaction forces, there are also excitation forces for the vibration 1X, 2X, and vane pass. Another contributor is the effect of bearing tolerances, especially in SRS pumps with carbon journals. The nominal difference between the minimum and maximum radial clearance is 0.

Whether the unbalance is due to mechanical imperfections eccentricities, tolerances, bowed shaft , or hydraulic unbalance, is not known. The vibrations recorded on pumps in the test facility and their installed location at the waste tanks are equivalent. Thus, high vibrations due to whirl or critical speeds may be reduced by selecting appropriate speeds.

Reduced vibration improves seal reliability [1, 2, and 4]. These pumps were modified for intentional, or friendly, misalignment of the shafts to reduce vibrations due to fluid instabilities in the bearings. Similar pumps without these modifications have operated hours or more. A comparison of the operating hours indicates that this attempt to correct the problem is inadequate.

The pumping rings and friendly misalignment are insufficient by themselves to improve pump reliability [2]. The bearings are immersed in water during operation with equal static pressures at both ends.

It is hypothesized, by the authors, that the lack of axial flow through the bearing can accelerates bearing wear in two ways. First, as material from the bearing is removed under normal wear, particles can get trapped inside the bearing and roll several times in the annular space before exiting.

Second, because the liquid trapped in the clearance has the potential to remain in the region for several revolutions, the likelihood of the fluid whirling inside the region is increased. The benefit of a process lubricated bearing at the hydraulic casing is also warranted. For those impossible cases, due to geometric or other constraints, the deployment of a fixed geometry design that reduces fluid damping and improves rotordynamic stability should be considered. Higher vibration magnitudes were noted at other locations along the pump, but Flange 1 was selected since data was available at that location on numerous pumps tested at SRS.

The mass of the shaft was compensated in the EFP, since this was the only pump with a one inch diameter shaft rather than a two or two and one half inch shaft. This compensation made only a negligible difference to the figures. Data from Tank 50 pumps was assumed to be identical to that for Tank 42 pumps. Both are standard slurry pumps. Stefanko Robert A. Stefanko SRS. Leishear SRS. Mechanical seals consist of two 5]. The seal failures have been corrected in most cases.

Although the smooth seal faces. One face is stationary with respect to the pump. The corrective actions were somewhat different in each case, the basic other rotates. Between the faces a fluid film evaporates as the fluid correction was the reduction of vibration or increase in cooling at the moves radially. Ideally, the film evaporates as it reaches the outer mechanical seal. The first set of cases involves horizontal pumps used surface of the seal faces, thereby preventing leakage from the pump and to circulate cooling water to a nuclear waste facility.

These are referred effectively lubricating the two surfaces. Relative vibrations between the to as Cases 1 and 2. The second set of cases involves vertical pumps two surfaces affect the fluid film, damage the faces, and decrease the used to mix nuclear waste in storage tanks. These are referred to as life of the seals. In a series of industrial applications, different Cases 3 through 5.

These vertical pumps are typically 45 feet in length techniques were used to minimize vibration, and the life of the seals and are referred to as long shaft pumps. They consist of a motor on top was shown to significantly increase. The operating speed was controlled of the tank; an impeller at the bottom of the pump submerged in the in one case, the bearing design was replaced in another case, and the waste; a drive shaft with bearings along its length connecting the stiffness of the pump was altered in still another case.

The common impeller to the shaft; a water filled pipe, or support column, corrective action in each case was a reduction in vibration. The design details of each pump are available in the references. The ID inside diameter, inches vibration data is used in different forms and formats, e. The vibration is typically measured as a rpm revolutions per minute displacement, a velocity, or an acceleration, using proximity probes, SRS Savannah River Site velomitors, and accelerometers.

Mechanical seal failures are known to have numerous causes That is, the unfiltered vibration is decomposed into several filtered Karrasik [1]. The primary considerations here are failures due to vibrations. A display of these vibrations clarifies these comments. The vibration, which are accelerated by a lack of cooling. Over the past ten Bode plot in Fig. The peak value of the unfiltered vibration is displayed for the vibrations, which were measured in two orthogonal directions, X and Y.

The waterfall plots provide details of the vibration. Each of the filtered frequencies and the maximum amplitude of each is displayed for both the X and Y directions. The filtered vibrations are displayed for a range of operating speeds in a cascade plot in Fig. Note that the operating speeds are listed to the left of the graph, the vibration amplitude is listed to the right, and the vibration frequency is listed along the bottom. At the top of the graph, several frequencies, or speeds, are identified as 1 X, 3 X, and 5 X.

These speeds are one times the operating speed, three times the operating speed, and five times the operating speed respectively. They are associated with equipment vibrations. Specifically, a 1 X vibration is typically caused by an imbalance, a misalignment of the shaft between the pump and motor, resonance, or a combination of these effects.

A 2 X and 3 X component may indicate misalignment. The 5 X vibration is associated with the impeller of the pump. In this pump, there are five vanes on the impeller. Thus, the shaft is excited five times during each rotation. An examination of the figures clarifies the distinction between Figure 1: Bode Plots of Vibration Amplitude vs.

Speed, filtered and unfiltered vibrations. The Bode plot in Fig. The 1 X vibration is also shown and it dominates the vibration. In the Y- direction, the 1 X amplitude is 0. The cascade plot in Fig. Vibrations occur at one, three, and, five times the pump operating speed. Note that the 1 X speed, vibration amplitudes increase near rpm, as they do in the Bode plot.

The amplitude is measured with the scale to the right of the plot. The amplitude increases as the pump rotational speed approaches the critical speed. The critical speed is determined from the frequencies at the bottom of the plot. The maximum amplitude is reached at The critical speed affects a range of speeds.

Typically, operation within 10 percent of a critical speed is discouraged. A third type of display is a shaft displacement plot, or Lissajous plot, shown in Fig. In this example, the shaft is displaced a maximum of about 2. There function is identical, but there construction differs. The seal consists of stationary and rotating parts which contain the lapped seal faces. The rotating part is fixed to the shaft with set screws. The stationary part is mounted to the pump casing, or other fixed part of the pump.

A fluid film prevents Figure 5: Mechanical Seal Installed in an Industrial seal face contact during rotation. An appropriate fluid film thickness is Centrifugal Pump, Case 2 maintained by applied pressure to the seal. One design maintains the pressure with a spring as shown in Fig. The installed seal is shown in Fig. This type of seal is used in Cases 1 and 2. The rotating face was a soft carbon, and the stationary face was Teflon. The second design uses a metal bellows to apply the spring force to the seal faces.

Several designs were used. One of which is shown in Fig. Seals similar to this design were used in Cases 3 - 5. Both the rotating and stationary faces were silicon carbide, which is significantly harder than the carbon and Teflon materials used in Cases 1 and 2.

Although this assumption is not always valid, it provides a good starting place for an analysis. If the vibrations are higher than this you may well have a vibration problem. In this particular case, the vibrations were all below 0. The seals have not failed in the seven years since the recirculation line was installed. Both pumps used a Case 2: Reduction of vibrations to improve seal similar spring loaded seal.

Each had vibration problems of varying reliability. In this next case, high vibrations led to seal failures. Seal severity. Two different solutions were used. Added cooling was used in failures occurred every sixty days for all three pumps. The pumps called one case where the vibrations were low. A reduction in vibration was number 1, 2, and 3, were operated in parallel, and their installation is used in a second case where the vibration was high.

Identical shown in Fig. A cross-section of the pump is shown in Fig. There was no visible, external damage to the seal prior to disassembly. The two inch thick mounting plate shown under Case 1: Added cooling to reduce seal failures. In this first the pumps was later used to reduce vibration. A discussion of the case, additional cooling was provided to the single mechanical seal of vibrations is required to explain the use of the plate.

The recirculation line and the pump are shown in Fig. There were actually two separate types of seal The vibrations indicated that the pump and its original mounting failures. The original mounting structure was made of a one One type was due to the introduction of dirt into the system quarter inch steel Z — bar as shown in Fig. The motor and its following underground pipe repairs, which could have been corrected mounting had a natural frequency of 60 Hz.

To prove this, vibrations by filtering the return water of the closed loop system after repairs were were measured at several locations on the pump and motor, but the made. Typically two to four seal failures followed repairs.

Passing location shown in Fig. The vibrations measured on the motor were measured to be 0. The seal faces spalled, which created debris inches per second when the pump coupling was removed from the between the faces. The debris created circular grooves in the soft carbon motor, and 0. The grooving action created more debris, accelerating the wear the other pumps were operating.

Most of the unfiltered vibration was 60 until the seals failed. Typically, the faces were damaged too bad to see Hz, and the vibration magnitudes did not increase between seal failures.

Essentially, the faces were destroyed. Bearings were undamaged, but replaced at the time of each seal replacement. While the 0. This value is vibrations were found to be reasonable.

Vibration measurements were higher than the 0. Typically, a vibration new motors of this size by the National Electrical Manufacturers velocity level of 0.