Thermal Expansion of an Oil Seal

[CTengineer2012]

All,

Question about thermal expansion of oil seals.

I have seal with a clearance between the top and bottom at the horizontal joint of .160". This measurement was taken with the seal pushed all the way to one side as to close up the left horizontal joint and measurement of .140" taken at the right horizontal joint.

Question: how do I calculate how much this seal will grow to close up during operation of the unit. If I treat it as simply linear, I get the following:

Aluminum has a thermal growth coefficient of 12.3x10^-6 in/in F. The diameter of the seal is 17". The temperature I'm going to is roughly 400F let's say. Can I use a straight calculation for this or do I need to take into account that this is a half circle?

This is what I get treating the seal as linear:

Circumfereance of the entire seal (both halves is 53.4. Half circle would be 26.7. Therefore: .0000123 * 26.7 * 400 = .131364". Is this correct or do I need to use a different formula since this is essentially a half circle?

 

[CTengineer2012]

measurement of .140" taken - this should be .160". Don't know how to edit threads.

 

[racookpe1978]

But your initial measurement was taken at room temperature, right? Or is the change in temperature = 400 F?

Thus, thermal expansion of one side of the original 0.160 gap is .0000123 * 26.7 * (400 F - 80 F)

But you need to show why the other side of the 0.160 gap is also Al, and is also at the same same change in temperature, and so why is the other side of the original 0.160 gap not also growing by the same thermal expansion?

 

[racookpe1978]

Are you sure you want an oil seal working at 400 degrees F? Seems real hot to me: Actual oil bearing temperatures are usually kept under 120 F in the systems I am used to.

 

[CTengineer2012]

Yes you are correct, I was thinking a little too hot for that area of the turbine. It is more like 120F. My question reall was; do I treat it like a straight piece of aluminum? Or do I take into account the curvature?

 

[racookpe1978]

A hollow "ring" of metal will increase in size linearly proportional to the radius of its circle at any point: Thus the ID of a ring will increase, and the OD will increase just a little bit more because the OD will be further from the center.

hence my question above: why are you apparently assuming that the ID of the two sealing surface will stay the same? Both sides of the 0.160 gap will move outward a little bit, but the gap itself will get larger by (thermal expansion of Al inch/deg F) * (0.160 inch) * (120 deg F - original deg F)

 

[CTengineer2012]

Not really worried about the ID. We do clearance checks to ensure we have sufficient clearance there. We just found an unusual gap at the horizontal split line, that's why I was asking the question.

See picture:

 

[stgrme]

It is not unusual for seal segments (half-rings or other multi-segment rings) to have gaps at the interfaces between segments. In general, these gaps are designed to prevent contact between segments. Such contact might create a larger clearance from the rotating part or create an oval clearance (wider at top or at bottom) for a half-ring.

In your particular application, you may not be looking at the complete picture. How does the seal housing expand due to the increase in oil temperature? How is the seal ring retained in its housing? Is the seal spring loaded in the radial direction?

Best of luck!

 

[racookpe1978]

Weeeeeeeeeeeeeeeelllllllllllllll, that there gap won't close from thermal expansion. (Until the metal melts.)

Seriously, something else is wrong (wrong seal sent to your plant? wrong machined diameter of the CT set at that joint? Wrong setup of the seal housing?) or at questionable to the specific person responsible for your oil seal, but you need to get your PM IMMEDIATELY to get an alert out to the vender of the seal and your CT vender to get an answer.

It "might be OK" - as just mentioned, the seal housing may be cut short at the horizontal joint while the oil seal being clamped by the upper and lower housing into position may be just fine.

If this were a Siemens machine, I'd call the district manager (or whatever your CT representative for the CT manufacturer is called in your organization) and send him/her that last photo.

 

[CTengineer2012]

The drawing for the seal shows a clearance of .065" - .070". I believe this is per side. As I mentioned earlier, the seal was pushed all the way to the left and the gap measured .160". This leaves .080" per side, which is close.

 

[EnergyMix]

Do we have the full story? How many gaps are there? Have you found the seal pushed all the way to one side before -- or to the top or bottom? If so, maybe you still need to talk to your manufacturer, because your design isn't giving you appropriate clearance.

Want to know the do's and don'ts of Eng-Tips? Read FAQ731-376.
English not your native language? Looking for some help in getting your question across to others or understanding their answers? Go to forum1529.