Support Friction and Contraction Loops, Chilled/Refrigeration Piping

[Robster1us]

Hello all, back again with a piping thermal movement question. If I think about how a pipe might move in the thermal case of a long straight pipe between (2) anchors, with a loop to take up movement, when pipe is expanding and is resting on supports that are not frictionless, the friction forces at any two adjacent supports tend to cause the pipe in-between to bow (I think BigInch calls this "snaking") in a direction perpendicular to the direction of expansion, which then lessens the effectiveness of the expansion loop (in other words, the loop itself moves less). I am imagining some version of pushing on a semi-stiff rope.

Assuming that assertion is correct (please feel free to educate me if not), when I reverse my thinking in terms of a contracting pipe in refrigeration service, I'm now pulling on the rope, and so, assuming that the "rope" doesn't stretch much due to the frictional forces at the supports (seems like a good assumption with steel pipe), then is it reasonable to think that, most of the time, considering friction and it's effects on the functioning of a contraction loop in cold service is not anywhere close to being as important as it is in expanding pipe. In other words, barring additional assumptions, in the situation described above, the behavior of the contraction loop can be assumed that of a frictionless support. Is that right?

Some of my assumptions include: horizontal steel pipe, simply-supported every 10 feet or so (or within MSS SP-58 guidelines).

Thank you in advance for your thoughts. I enjoy the insights that the forum members always share so generously.

 

[BigInch]

Friction isn't much to consider as far as the pipe is concerned, it's more important in designing the supports. Frictional loads are a small fraction of the axial loads typically developed in the pipe itself.

Bowing would tend to distribute the lateral displacements along the pipe, rather than at the loop itself, which is why guides are used to direct the pipe expansion towards a loop without allowing much lateral displacement during the process.

In a thermal expansion, friction (theoretically) would counteract a very tiny percentage of the expansion, due to it's opposite direction when pushing back on the pipe, and likewise during a thermal contraction, friction would tend to hold back a tiny percentage of that contraction by pulling on the pipe. However friction is typically such a small percentage of total load in the pipe, that friction can be safely ignored in the pipe stress calculations, even more so, since it reduces the (absolute value of) pipe stresses in each case. But, friction may be the only lateral load on the pipe support, so it is not so easily ignored when designing the supports.

I hate Windowz 8!!!!

 

[Robster1us]

Thanks BigInch. I think those are great points.

"Bowing would tend to distribute the lateral displacements along the pipe, rather than at the loop itself, which is why guides are used to direct the pipe expansion towards a loop without allowing much lateral displacement during the process."

So what you're saying is that, regardless of hot or cold pipe, design it the same way because the refrigeration system might be down one day and you will need the guides then?

 

[BigInch]

I didn't think I said that much.

Contraction tends to keep the pipe straight. Thermally contracting pipe does not bow, so would not need guides at all for a straight run of pipe. A run in more than one direction could have lateral displacements. Thermally contracting pipe is usually in tension, expanding pipe in compression.

I hate Windowz 8!!!!

 

[Greg80]

I believe friction force (axial) is important in the design of supports.
If you have a steel on steel (friction coeff = 0.8), the friction force = 0.8 x Fy
Fy = weight + vertical load due to thermal

With a 10" pipe filled of water insulated and cladded, you can have easily friction force >2000 lbs
To reduce or eliminate this force, you can lubricate the supports and use teflon plates (teflon on teflon) ...