Anchor Force In Thermal Expansion 1

[E. Evans]

https://files.engineering.com/getfi...-a6ee-40d0-bcd0-54a949d6eef3&file=picture.png

Hi All,
I am trying to figure out the anchor force on a pipe given thermal expansion.
with one side of the pipe anchored and the other side free to expand given a temperature increase. (see photo above)
Let’s say the pipe supports are fixed to the ground and the pipe can slide over the supports.
This generates a friction force which is:
friction force = (the weight the support holds) x (friction factor between the pipe and the support, both kinetic and static)
My question is this: does the pipe anchor feel the reaction of that friction force or does the support or both and how to look at that?

 

[XL83NL]

Draw a free body diagram and you’ll the answer. Sum of all forces has to equal zero for a static system.

Huub
- You never get what you expect, you only get what you inspect.

 

[1503-44]

The anchor force is -1 x sum(support friction loads), as long as the pipe does not buckle.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[desertfox]

Hi E Evans

The pipe support will see both the pipe dead load and the horizontal force due to friction as the pipe slides on the support as it expands, imagine the pipe support was rigidly connected to the underside of the pipe, then as the pipe expands the support would have to move with it and for that to happen the friction force generated by the floor and the base of the pipe support would have to be overcome.


“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[LittleInch]

Both.

But the anchor as you've drawn it will have some forces ftom the left hand side which will reduce the effective force.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[KevinNZ]

As drawn Fa = -ve Fb. Fa will in the opposite direction as the pipe cools.

These opens up the old question of what is the stress range from friction. Should it be 2x the calculated force in one direction?

 

[1503-44]

Fa, if it is the force from pipe to the anchor, is in the opposite direction as it is now shown.
Forces from pipe to anchor when expanding are,
<----Fa Fb----> Pipe between anchor and support is in compression.
Forces from pipe to anchor when contracting are,
---->Fa Fb<---- Pipe between anchor and support is in tension.

And Both forces will reverse their direction when cooling.

The range is 1 x the f_force in each direction.

KevinNZ I dont follow how friction force could ever be more than normal_force x (static) f_coefficient at any time.



Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[racookpe1978]

"Friction" sliding forces can be calculated. (Well, estimated. Guessed. Assumed.)
Up until the pipe gets painted. Or up until it doesn't get painted and the friction connection rusts solid against the pipe.

 

[1503-44]

Neither rust or paint bond ever held fixed any thermal force caused by more than 1°F

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[LittleInch]

It's always amused me that the piping guys use 0.3 friction factor, but for support design I would use 0.8 as a worst case corroded surface.

How many pipe supports do you see fallen over after a few years?


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

It should be lower than 0.3. That's usually a conservative number. With multiple lines on a rack, Fluor, Brown & Root, Bechtel, etc designed for uniform pipe load of 40psf x 20ft span x 0.2 f, because direction of slide amongst various lines is unknown, so it averaged out to less than 0.3 (Large lines placed individually).

Since its weak axis bending design we are talking about, using too much of a conservative value for friction has grand implications for general pipe rack design.

I've seen some bent ones, but none ever flattened from thermal loads. They might move a few inches, but then the pipe movement driving it stops and everything comes to rest. More common is seeing pipes holding up a support. Once I saw pipe holding up a compressor scrubber vessel. The scrubbers' foundation was sinking, so the anchor bolt nuts were removed. It was suspended a couple inches above the fdtn.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[LittleInch]

Multiple pipes I'll give it to you, but finding the correct CoF after a few years of operation isn't easy. Teflon pads would work. Also how much expansion has am impact and how many cycles.

To high a friction the pipe supports might jump as it expands or contacts.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

So.. pipe jumps, support pops back aligned with vertical and then there is less stress in both.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[LittleInch]

Yes, but scares the operators when it happens as they walk past. ...

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[KevinNZ]

We have had similar experiences to LittleInch. 0.3 for friction is bit low for support and foundation design. We have had whole sections of line fail because the start up friction pushed the supports over. (note our steam lines have a lot of movement)

 

[1503-44]

As you well know, steam lines always require special attention. That's never happened to me and I've always used 0.3f. Maybe someone totally forgot about friction, or line was heavy and movement affected support column eccentricity caused bending moment, or it was snagged by insulation binding? Or someone anchored in the wrong place, or didn't put in enough anchors, or they forgot to X brace between bents every 100ft. Or soil/foundations were not adequate. That's what I'd suspect. Lots can go wrong other than f factor. Losing a whole line of supports doesn't sound like a bad friction factor problem.

"Scares the operators as they walk past." Steam traps, rattlesnakes. And baboons are worse yet. They always scared me $%%##!! Man, they got huge teeth and they like showing them off.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[KevinNZ]

Fourteen, The case above was a combination of low design friction factor and poor soil.

 

[1503-44]

There you go. Unless it happened on a start up, pipe movements should have been minimal, so I think I know which one was the root cause, because when there is bad soil, it is usually very bad. One good rain or earthquake and ... quicksand. Pipe will often hold up one support, but when they all go like domino's, its either bad dirt, or a hurricane.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[racookpe1978]

-Fourteen (Petroleum)18 Oct 20 17:23
So.. pipe jumps, support pops back aligned with vertical and then there is less stress in both

True, true. But that (a relaxed case because of the pipe "jump" occurs going up in heat to some arbitrary mid-position) then becomes re-stressed when cooling back down.

Working pipe weight is affected by fluid type (steam pipes typically have a low fluid weight-per-foot, but high change-in-temperature-per-cycle) will matter 5x - 10x more when the large bore pipe is liquid-filled. The fluid weight greatly increases sideways friction load on the saddles.

Rollers? Much more likely to keep fricvtion low over the long terms.

 

[1503-44]

Not necessarily. For the relaxed, or cooling case, the movement and stresses are reversed and the jump will also probably occur in the opposite direction as well. Thus there would be no stress accumulation. The support moves backward, jump occurs, again realigned with vertical and all is well.

Friction factor does not chance with pipe temperature. Friction force is f x normal force, so of course liquid filled pipe generally will have more friction force than steam pipe, even though steam pipe may be much hotter and expand much more. That is from different normal forces, not because of different ff.

There is some evidence that f factor can vary somewhat with normal load, but those considerations are too minor and too detailed a concept to apply to every day pipe stress work. Leave those things to the watchmakers.

Rollers, slide plates, teflon, micarta, yes very good but very expensive where not needed.