slurry pump body code

Centrifugal Slurry Pump Vibration

Vibanalyst

Any other spectrum data type or time waveform data available? It is difficult to analyze some complex problems with velocity data only. The other thing that may be crucial..... could you provide more information about what has happened after the shaft failure on 19th September? Between 19th and 22th September? Because I find it interesting that there was another measurement taken on 22th September with more or less the same symptoms. After that 2.5x amplitude goes down and in about a month, 2x rises significantly.

CarlosQG

Hello Vibanalyst,thank you for taking the time to check what I've shared. I'm checking all work orders on this pump, and there are no entries between August and September 2023. The CMMS culture is not well established, so something could have been done and not properly recorded. A friend told me that a day before the pump shaft failure, a technician walked by and manually tried to manipulate the discharge valve. I assume he tried to open the valve fully in good faith; the pump shaft broke less than 24 hours later. My conclusion is that the pump is already working at a point far away from the BEP, and a disruption of the flow caused by the discharge valve finally ended in too much stress on the shaft. On September 19th, a new pump bearing housing and impeller was installed. The following image shows the time waveform for both bearings on the pump, and the other shows the flow trend from September 12th to the 26th.

Ralph Stewart

Does the Impeller have 3 discharge blades?If so, this might account for the 3x. What about the consistency variations of what is being pumped,(being thin and then becoming much thicker or "heavier") especially against a partially closed discharge valve? Especially the 3x, if the impeller has 3 blades. The 2.5X existence may vary also due to excessive vane-pass problems from the "thickness" of the slurry and the partially closed valve causing restrictions. Possibly the amplitudes would decrease, especially if there is a problem also with maintaining the correct consistency levels as well as the closed discharge valve. Is there a system used to automatically keep a consistent thickness at all times? The 2x might be coming from mechanical looseness somewhere, but also may be related to the "bad" valve as well as a changing, "increase and decrease" in consistency thickness.

Vibanalyst

From the information available my guess would be something happened on the 18th August during installation in terms of inducing some initial material damage. For example overheating the impeller, induced shaft eccentricity, inappropriate use of hydraulic tools/hitting, wrong part tolerances etc. I am quite sure the initial cause of a broken shaft is not in the vibration level at 8 mm/s but probably some kind of initial shaft crack.

CarlosQG

Thank you, Ralph. I'll check with operations what measurement system they have to verify slurry thickness consistency. The impeller has 5 blades. Where could the looseness come from? That's a good point. The change in consistency would make it a flow-related problem. I'm trying to wrap my head around the 2.5x harmonic. Is its origin half the Vane Pass Frequency? My next step if the vibration doesn't go down after the valve is replaced in January is to record the vibration while coasting down with a fully opened valve. I want to check if the 2x, 2.5x, and 3x harmonic are also present at different speeds. After that, I'm out of ideas.

CarlosQG

From the information available my guess would be something happened on the 18th August during installation in terms of inducing some initial material damage. For example overheating the impeller, induced shaft eccentricity, inappropriate use of hydraulic tools/hitting, wrong part tolerances etc. I am quite sure the initial cause of a broken shaft is not in the vibration level at 8 mm/s but probably some kind of initial shaft crack. Thanks Vibanalyst. That is a genuine possibility that I'm on board. A third party changed the impeller and pump casing on August 18th. I don't know about their maintenance practices. When the shaft broke in September, the pump was replaced by the maintenance crew of the mining company.

ruben

Looking at the general assembly of the equipment and the ODS videos and the way it moves and twists like an S once or twice every revolution (or so it seems), I would review the following points starting from the engine forward: 1- Alignment between pulleys, to prevent overloading of bearings and excessive wear on belts and pulleys. 2- Belt tension to avoid bends and cracks in the shaft, especially just after the pump pulley. 3- Shaft curvature, to avoid mounting bent shafts during each equipment maintenance that could eventually crack and break, as well as take away the life of the bearings. 4- Pump bench leveling, to avoid misalignments and overload on bearings when curving the structure or saddle where the bearings are mounted. 5- Alignment between the pump suction flange and the sludge tank outlet flange to avoid. If they are not aligned there will be stress in the pipes and they will overload bearings and cause the structure to twist. If they cannot be aligned correctly, consider placing an elastic tube there. As a general rule, an elastic element is also mounted on the pump discharge between the delivery flange and the delivery pipe. 6- Finally, check the balance of the pump impeller in case it has excessive wear and is unbalanced. You can take phase readings in horizontal PUMP DE and in NDE to see if it is a simple imbalance, but since many faults are mixed together it will not give you clear information, although by trying you don't lose anything.

ivibr8

Seems to me that a key issue is where you are operating on your pump curve. From the pump curve you provided, you "seem" to indicate with the red line you are operating at 750 GPM which is well off the BEP (~ 2700 to 3000 GPM) for the normal pump speed. Is this true?Operating this far left of BEP puts higher fluid pressure loads (read higher strain) on the shaft impeller end, in addition to recirculation and turbulence. I did note a "noisier" floor level on your spectral data at about the timeframe you found out from the operators that the discharge valve was malfunctioning. Is this a possible cause and effect with the observed vibration data and the broken shaft??....I dont know, but sure seems like a likely culprit to investigate.Can you tell us more about where the shaft broke (near a keyway??) and possible causes? Can you confirm where you normally run (GPM) prior to the recent issues?

ruben

7- I forgot, if possible it is VERY important to always measure the pump with the same valve opening in suction and delivery.