Force exerted by the expansion of freezing H2O

[BasketDaN]

How much outward force is applied when liquid H2O freezes, turning into ice?

[Bystander]

This is definitely a candidate for any FAQ that ever gets put together.

The stock answer is, "How much pressure would you like it to exert?" The solid-liquid equilibrium temperature and pressure moves from 273.16 K and five or six hundred Pa at the triple point, through 273.15 K and 0.1 MPa at the "normal" freezing point, through lower temperatures and higher pressures to the ice I- ice III- liquid triple point at ca. 250 K and 200 MPa (30,000 psi if you want more familiar units).

[russ_watters]

For a visualization of what Bystander was saying, look at the PHASE DIAGRAM (http://www.lsbu.ac.uk/water/phase.html). For comparison, atmospheric pressure is roughly 100,000pa. Notice how the line between solid and liquid is nearly vertical. As a result, it'll freeze at close to 1000x atmospheric pressure with little change in freezing temp.

[BasketDaN]

Thanks. So, just to make sure I'm interpereting that graph correctly, if I were to freeze liquid H2O at about -10 degrees C, the volume would still increase (density would decrease to .92), as long as the pressure on it was not more than 10^8 Pa?

[Michael D. Sewell]

In most practical applications, the force is limited by the compressive strength of the ice.

[Bystander]

Originally posted by BasketDaN
Thanks. So, just to make sure I'm interpereting that graph correctly, if I were to freeze liquid H2O at about -10 degrees C, the volume would still increase (density would decrease to .92), as long as the pressure on it was not more than 10^8 Pa?

Yup.

[Sariaht]

Let's say we had a small but very deep pond of salt water mixed with a substance that makes it extremely heatconducting (perhaps fine carbon) in the middle of the dessert. The infrared light from the pond could be withdrawn with a parabolid shaped foil that concentrates the light into a beem and send is away at the same time as it reflects the light comming from the sun. The salt would fall out from the water and the oxigen level of the water would rise, making it condense fast and also free of salt on top.

[BasketDaN]

How thick would a pipe have to be to be able to resist 10^8 Pa (14,500 psi) ?

[Michael D. Sewell]

Originally posted by BasketDaN
How thick would a pipe have to be to be able to resist 10^8 Pa (14,500 psi) ?

cast iron 20,000 psi
iron 50,000 psi
steel 50,000 psi and up
copper 25,000 psi and up


Stress = PD/2T

P = internal pressure(gauge)
D = inside diameter
T = pipe wall thickness

[BasketDaN]

But don't copper pipes usually burst when water freezes inside them? This freezing would usually be exerting around 14,500 psi, right? Which, by that post, would be less than what copper can handle anyway?

[Bystander]

Residential plumbing? K, L, or M? "L" is around 1/16 wall --- means you divide pipe dia. by 8, then multiply by Cu tensile strength (~25k) --- rupture ocurs at around 3000 --- that's -2 C.

[russ_watters]

Originally posted by BasketDaN
But don't copper pipes usually burst when water freezes inside them? This freezing would usually be exerting around 14,500 psi, right? Which, by that post, would be less than what copper can handle anyway? You didn't do the calculation...

[Michael D. Sewell]

Twenty years or so ago, I had maintenance contracts that included many older urban buildings in the northeastern part of the US. Frozen pipes were common in the winter. Most were copper or galvanized. Off hand I would say that approximately 20% of the time the pipes actually ruptured. Most of the broken pipes were galvanized. Many times the copper pipes were repaired promptly and they just stretched a bit. Sometimes an elbow would get pushed off, and could soldered back on. Copper tubing comes in both rigid and flexible types. Whether or not a pipe breaks after freezing is a function of how much elongation it can take before rupture occurs. You would probably be suprised at how much stretching soft copper will do before it ruptures. I think that copper tubing gets brittle with age. Russ probably can give you information on that. Cylindrical shapes always split along the longitudinal axis.
I hope this helps. -Mike

[russ_watters]

Originally posted by Michael D. Sewell
I think that copper tubing gets brittle with age. Russ probably can give you information on that. Like bending a paperclip until it breaks, repetitive stress (even if it seems small) can eventually cause metal components to fail.

[Michael D. Sewell]

Simple enough, thank you!

[turin]

Originally posted by russ_watters
Like bending a paperclip until it breaks, repetitive stress (even if it seems small) can eventually cause metal components to fail. Why is this? Does the bending induce dislocations in the lattice? This has never made sense to me (but, of course, it happens).

[Sariaht]

What is the surface tension for salt water and sweet water?

Can values of the surface tension be used to seperate salt water from sweet water?