Checking Metallurgy Of Pump Impellers

Yes, you can send the impeller for a complete metallographic analysis which will look at corrosion, stress corrosion cracking, fatigue .............; or even simply do a quick scan with a hand held portable analyzer to confirm metallurgy. There are numerous companies that can perform this work.

Where abouts are you?

What's in your wastewater and pump material? Corrosive chemicals can damage an impeller as well as large sediments can . In the manufacturer's defence, the pumps are designed with a certain safety factor , so a certain amount of overstressing for a few short moments won't damage the pump that readily .

I operate a small (1.0 MGD) WWTP in south Alabama, and I treat domestic wastewater. I've had the usual problems with things like half-bricks and pieces of metal getting sucked up inside my pumps, binding, burning up motors, and bending shafts, but only with this brand of pump (ABS 1504 model; cast-iron Contrablock impeller) have I had impellers break like this, with no apparent cause other than cracking.

The most recent failure is in my shop now. You can clearly see, at the failure point, small sections of clean metal where the vane finally snapped off, surrounded by rough corroded spots, and there is a three-inch long crack clearly visible in the other side of the impeller. I'll see if I can get some good pictures tomorrow.

Modern Industries in Erie PA does nice metallurgical analyses.

What is the material being used?

How old are they?

milo

I'll look them up; thanks.

The material is, I believe, cast iron, and the units are less than six months old.

IRIS NDT could help you out as well:

http://www.irisndt.com/locations.php

Come to think of it .......

ACUREN can help you out too, and they have a branch in Alabama

http://www.acuren.com/index.php?id=17&rid=1

Ah! Someone close to home! Thank you; I'll be contacting them directly!

I think ABS has changed their manufacturing technique on these impellers. Older models of these pumps did not have this problem. It's only ones that I've bought within the past year, year and a half or so that have had impellers grenading like this.

The 1541 M105 is a 1750 RPM pump. I had someone try to sell me a pump that ran faster than what I was used to, but I told him no thanks. Speed is life if you're flying an F-16 but not if you're an electric motor. :)

I don't see any cavitation damage. The pumps that came with the plant were sized at 600 GPM at 60 feet of head. These M105s are 600 GPM at 50 feet of head, which is what I calculated it out to be. According to plant flow data, I'm actually getting about 530-540 GPM, which is just to the left of center on the pump curve. So, I don't think pump size is a problem.

Here's some shots of the impeller: Broken Impeller. Take a look and see what you think.

If you have the resources (money) it would be best to send it to a materials lab, such as ACUREN, they can provide all the "fancy" fractography, metallography, even have scanning tunneling microscopy done (if you really want to spend money and get cool pictures at the atomic level ).

I'm looking into that even as we speak. Hopefully they (or someone like them) can at least give me an idea what happened for a decent price. Unfortunately money is kinda tight right now...

Yeah I figured money would be an issue - it always is, and even more so these days.

CONSIDER YOURSELF WARNED - If you haven't had this type of work done before, and depending on the level of analysis that is required, you may want be sitting down as you read the price. You will probably be able to buy a few replacement impellors for the cost of the analysis .

How about eliminating the risk of errant bricks? By installing a screen or something?

Those places you describe as corrosion spots = cavitation foci?? That is how cavitation works, the bubbles collapse and since they are a spherical wave front collapsing they reach very high peaks and break of little bits, and then some more.

Once you have any sort of a cavity you have a place for a stress concentraion fissure to start, and it breaks off.

I am curious, Dave. What did you decide on doing? What did ACUREN say? What did they quote you for cost?

Thanks for the photos.

Apparently a piece broke off the inlet edge.

Doesn't look like cavitation. I suspect vibration induced fatigue.

Is this in an old well or a new one ?

I suspect inlet conditions causing erratic inlet eddies and pulsing of the pump.

Is the output smooth or are there pulses? Even if no pulses are evident, it could still be inducing this at the impeller eye.

In a water pump you would tend to use straightener vanes, but in sewage this is asking for a blockage.

When the well is empty can you get some photos of the installed pump and especially inlet area?

You showed good sense in sticking with a 4 pole speed pump. Obviously you have had experience of the problems of high speed pumps!

per sceptic -- "Doesn't look like cavitation. I suspect vibration induced fatigue."

You may be on to something there. That front edge of the vane looks like it has a pretty high resonant frequency; but sharp shocks such as from sources like cavitation will have those fourier components. But assuming the pump manufacturer is using the same pattern to cast his impellers what could have changed?

Here's some possibilities:

1. Stress corrosion cracking from some change in the waste water chemistry. For example are the wells that source the water experiencing salt water intrusion. (If you are located near the Gulf)

2. Increased corrosion sensitivity of the cast iron itself? Most pump manufacturers use a class 35 gray iron. This spec has a wide latitude of allowable chemistries. Typically the foundry will tailor the scrap metal charge for the melting furnace to the additional specialized requirements of the customer. Pump manufacturers are usually pretty picky about casting porosity so pump casings will pass hydro test. So efforts are made to keep aluminum out of the scrap charge. But for cast iron pumps corrosion is usually not an issue when stainless steel pumps are available for such conditions.

Now take note of the current economic conditions and the radical effect they have had on scrap metal markets. It is reasonable to suspect some change in the foundry iron composition that has not yet been noticed by the pump manufacturer. And Dave M.'s operation is the first to notice the weakened response of the hardware to corrosion.

3. There is one other remote possibility that some small process change in the machine shop (or the foundry cleaning room) is producing a stress riser in the machining or introducing some corrosion enabling contaminant into the surface of the impellers.

I realize that these suggestions are far fetched; but if Dave is having trouble getting the pump manufacturer to help then he might try banging my suggestions off their in-house engineers. I suspect that will get their attention and a bit of respect.

Ed Weldon

Hi Dave M.

I got the same problem in the other topic like you in the past. Now I'm looking for the Munsell values tables or Bridgeman Equation in order to measure ADMI of dye waste. Do you remember that topic?

I wonder if u have them now and could you show me how to get them?

My Email is huybkk47@yahoo.com

Sincerely thanks!