Burgmann seal failure

[ballpeen]

Hello all. Where I work, we have a small single package yarn dyeing machine that has a 25mm Burgmann mechanical seal that is mounted directly on the motor shaft. I'm pretty sure it is a standard carbon stationary/carbide rotator setup. The elastomer appears to be viton. The problem is that after about a few weeks of service the pump starts leaking and when I take it apart I find the same thing, the rotator is completely destroyed. It looks to me like its been overheated or that there is too much pressure on it. I've checked the running distances of the pump and I know that this thing runs really tight. Too tight from what I've seen on the pumps that I've worked on in the past. I'm quite sure that the spring on the rotator is near completely collapsed from the measurements that I have taken, when the impellor is installed on the shaft. Now this machine runs quite hot, as the liquor it circulates peaks at 275 degrees F. It could be the way its runs if they are cooling the machine down too quickly, I can't say for sure about that. Is there anything in particular that I should be looking for when I pull it apart again? It hasn't been used for a while but I was told the other day that they are thinking about starting it back up soon and that it was leaking again when it was last used.
I know the motor bearings are good and there is near zero runout. The motor shaft is up to size also. I'm beginning to think that it may be the wrong seal.
I tried to contact the dye machine vendor but I got nothing in reply. I guess I could try the Burgmann Company itself and see what that will net me. What do you guys think? Thanks.

 

[JJPellin]

It is difficult to comment with more details about the seal design. But I have a few suggestions. You indicated that the rotary was destroyed rather than the stationary. This suggests that the hard face (Tungsten Carbide, I assume) is more heavily damaged than the stationary (carbon). This tends to rule in some possible causes and rule out others. You suggested that temperature or pressure might be the issue. You description makes this sound like a rather small pumps that should not be able to produce enough pressure to fail an o-ring pusher seal, even if it is unbalanced. The primary temperature limits would be the o-rings (400 °F for Viton) or the boiling point of the liquid. But if the liquid was the limit, I would expect more damage to the carbon face rather than the hard face. The possibility of an o-ring problem opens up some interesting considerations. Depending on the configuration of the faces, badly swollen o-rings can shatter faces. Although I usually see this with a stationary face (o-ring groove or cushion seat configuration).

I would suggest the following checks:
*Verify the proper running length of the rotary unit. Burgmann should provide the correct dimension. Excessive compression on the seal cause premature failure but should not catastrophic failure of the hard face.
*Verify material compatibility. Make sure that the materials of construction of the faces and o-rings are compatible with the pumped fluid.
*Inspect the next failed seal. Ideally, ask to remove the seal as soon as it starts to drip. If you wait until it is gushing, the early evidence of the failure mode may be destroyed. Look for swollen o-rings (compare them to the new o-rings being installed). Look at the faces for classic signs of grain pull-out, blistering, abrasive damage, phonographing or plating.
*Verify operating practices. Are they starting the pump up cold? Are they running it dead-headed (zero flow) or stone-walling it (high flow)? I do not suspect thermal shock from fast heat up since the seal should be able to take this. I am more concerned about cold start with product set-up between the faces.
*Verify the condition of the machine. It sounds like you have done this already. Important measurements are shaft run-out, shaft concentricity, seat run-out (is it pressed in squarely?), axial stack-up (seal running length or compression) and impeller balance.
*Verify the seal flush piping. You didn't mention any flush, but I would normally expect some piping plan. Is there an internal port for flush from discharge back to the seal? Is there any external flush piping? Check for plugged orifices, fouled coolers, piping that is connected to the wrong ports, missing piping or components, etc.

For a shattered hard face after a short run in a hot fluid, I would be most suspicious of the following causes:
1. Cold start up with product set up between the faces.
2. Fluid incompatibility with face or o-ring materials.
3. Mechanical failure of other pump components (impeller fit, impeller balance, motor bearings.)
4. Installation error. Incorrect running length (compression), stationary face not square in the housing, seal flush piped incorrectly.

 

[ballpeen]

Well JJ. I am certain that there is no seal flush, unless it is internal to the pump by design, which again as far as I can see does not exist. I really don't feel that there is any liquor between the seal faces at startup as long as the faces are running together. As far as the elastomer, the o-rings look as good as the day they were installed. I know that the bearings are good because the motor was recently sent out and the stator was rewound and bearings were replaced. Now getting to the installation, that could be a problem as the instruction book that came with the machine is a little sketchy, (we got the machine used)....... Compression length is where I feel the problem lies. As stated before, from the measurements I've taken it looks to me like the spring behind the rotator is fully compressed. Usually its around 2/3 of the spring.
On this particular setup its a large single coil spring behind a silicon or tungsten carbide rotator and a carbon stationary.
I've been on the Burgmann site and am working on getting some more info. so I can send it to them and see if this is an incorrect seal for this application. I personally feel that this is the crux of the problem here.
One problem here is that the pump is made by one company and the seal by another and the machine by still another. And the machine manufacturer is of little help, as I tried last Fall to contact them.

 

[JJPellin]

You mention that it is a single spring design. This configuration tends to be less sensitive to compression than a multiple spring design. But it is very important that it is not bottomed out. You feel that starting up cold is not the problem. But there must be fluid between the seal faces in order for it to work. Even in a static condition, fluid will be drawn between the faces by capilary action. If the fluid sets up when cold, I would still feel that this may be the problem. Excessive compression might cause rapid wear of the carbon face. But starting up with hardened product between the faces can destroy the faces.
If you are not sure whether the seal design is good for your application, you can change to a different design or manufacturer easily. All of the seal manufacturer's have web sites where you can select a seal based on temperature, product and size. Small seals of this type tend to be inexpensive and readily available off the shelf. It doesn't matter that the seal company didn't make the pump. Except for Flowserve, this is rarely the case. In fact, as a policy, I don't purchase seals from pump manufacturers in any case. They don't tend to be good at making and installing seals. Purchase a new seal that will fit in your pump and get instructions on proper installation and setting dimensions.
Otherwise, it sounds like you are asking all the right questions and making the right checks.

 

[ballpeen]

Thanks JJ for your interest in my post. You seem to be knowledgeable in this area based on what you have replied with in your posts. One thing that I did want to add that I neglected to tell you in my last reply was the condition of the rotator. In the two times I've pulled the pump out for repair the rotator is really busted up. Its looked like someone took a hammer to it and started hitting it and cracking the tungsten carbide, there was actual pieces broken away from it. Wouldn't that indicate a too quick of a cooling effect of the metal or would you say that the overcompression of the loading spring or a combination of the two were in play here?
I think that on this application, a carbon rotator with a tungsten carbide stationary would be a better choice. If there is enough room for it...... I'm pretty sure that the stationary is not a standard VDMA. As we have tried to purchase an aftermarket for another set of pumps with similar mechanical seals from Chesterton and they couldn't come up with anything to fit in the hole. The Rep. said if there was enough call for it, the Company would setup and make the seals but, the cost doesn't merit doing it. So were basically stuck buying OEM's for the smaller pumps.

 

[JJPellin]

When I see a broken TC mating ring, I tend to think of one of the following:

Mishandling - Installation or removal practices often result in broken hard faces.
Product set up between the faces - In a coker charge pump (asphalt) we used to see this a lot. The fractures started at the grooves for the anti-rotation pins on the back of the face. The carbon face is softer, but less brittle and did not crack up. I don't know how Burgmann drives their face, but there is some sort of groove or pin or something that drives it.

I do not tend to think of overcompression or thermal shock. I don't know what your service is, but I have never seen a service that could cool a seal off so fast that the hard face shattered. The face is probably no interference fit into any metal part, but floats on the o-ring. It might be possible to shock this face enough to crack it, but I have never seen that. And overcompression should result in a breakdown of the fluid film resulting in accelerated carbon wear. If you were running two hard faces, one of the two might be broken in this case. But if the carbon is not ground down to the nub, I would not expect overcompression. The one exception to this would be if the overcompression resulted in the seal totally bottomed out and then the face might crack during assembly. But this would fall into my first category as mishandling. I would expect the mechanic to be able to tell that he cracked the face during assembly. And I would expect the seal to leak catestrophically as soon as it was flooded before it even ran.