Is there permanent size loss in metals from contraction?

[Gary Weller]

I install automatic pool covers for a living. The tracks that our cover rides on are composed of extruded anodized aluminum. These tracks come in 22' lengths. Usually there is more than one section of track, set end to end, on a given side of the pool. We drill and anchor these sections to concrete decks with NO gap between.

Several years down the road, we notice that there is now a 1/8" - 1/2" gap between the sections. These systems are installed in the heat of the Summer and we usually come back to the jobs in the same heat. If that's the case, does this mean that there is some sort of loss happening during the expansion/contraction process that happens between the seasons and hot days/cold nights?

I know that there's potential for deck heaving and expansion and contracting of the concrete, but this has been seen across the board and there's no way that so many different stones and concrete mixes have the same expansion/contraction ratio.

I guess my end-all questions are "Is there Net size-loss throughout the expansion/contraction process? If the aluminum is 10 units long at 100º, then 9 units long at 50º, but only returns to 9.9 units long when brought back to 100º, where did that .1 unit of length go? Is it gas loss that allows the aluminum atoms to get closer together?"

 

[berkeman]

I install automatic pool covers for a living. The tracks that our cover rides on are composed of extruded anodized aluminum. These tracks come in 22' lengths. Usually there is more than one section of track, set end to end, on a given side of the pool. We drill and anchor these sections to concrete decks with NO gap between.

Several years down the road, we notice that there is now a 1/8" - 1/2" gap between the sections. These systems are installed in the heat of the Summer and we usually come back to the jobs in the same heat. If that's the case, does this mean that there is some sort of loss happening during the expansion/contraction process that happens between the seasons and hot days/cold nights?

I know that there's potential for deck heaving and expansion and contracting of the concrete, but this has been seen across the board and there's no way that so many different stones and concrete mixes have the same expansion/contraction ratio.

I guess my end-all questions are "Is there Net size-loss throughout the expansion/contraction process? If the aluminum is 10 units long at 100º, then 9 units long at 50º, but only returns to 9.9 units long when brought back to 100º, where did that .1 unit of length go? Is it gas loss that allows the aluminum atoms to get closer together?"

Welcome to the PF.

Very interesting question! My initial guess is that the concrete is expanding over time, but others may have other ideas. Are these all new pools? if you install the cover hardware on a new pool, it seems plausible that the concrete structure around the pool can expand over time. But if this happens even on old pool retrofits, that would shoot that theory down...

 

[Aerospaceenginer]

If a piece of metal is heated and cooled uniformly and unconstrained, it will return to the exact original dimensions when cooled. However if the ends are constrained when heated, the metal will become shorter and fatter when cooled. The metal has yielded, as if squeezed in a press, but the volume is unchanged.

 

[berkeman]

If a piece of metal is heated and cooled uniformly and unconstrained, it will return to the exact original dimensions when cooled. However if the ends are constrained when heated, the metal will become shorter and fatter when cooled. The metal has yielded, as if squeezed in a press, but the volume is unchanged.

I yield to the better idea!

 

[Rx7man]

I think that's the right conclusion.. since there was no gap between them, and the strength of the concrete is so much higher than the aluminum, it is in fact shrinking the aluminum.. I would venture a guess that if you installed them with 1/8th gap between them, you'd probably still have a 1/8th gap later.

 

[Aerospaceenginer]

This is the principle behind "flame bending". If you heat one side of a pipe, the colder side will be a constraint. When cooled, the pipe will be bent towards the heated side. This is also called metal shrinking, but in reality a local area gets thicker and shorter. The volume is unchanged.

 

[CWatters]

You can calculate how much a length of aluminium might expand and contract as it heats up..

dl = L₀ α (t_hot - t_cold)

where
dl is the change in length (meters)
L₀ is the initial length (meters)
α is the coefficient of expansion for aluminium (22.2 * 10^-6 m/m.k)
(t_hot - t_cold) is the change in temperature (C/K)

So if you installed a 22ft/7m length when the temperature was say 20C and it was heated by the sun to say 50C (?) then the change in length works out at about 4.5mm (bit over 1/8th inch).

Do the gaps cause any problems?

If you leave an 1/8th inch gap between lengths you might stop any potential buckling/distortion of the track.

 

[Gary Weller]

You can calculate how much a length of aluminium might expand and contract as it heats up..

dl = L₀ α (t_hot - t_cold)

where
dl is the change in length (meters)
L₀ is the initial length (meters)
α is the coefficient of expansion for aluminium (22.2 * 10^-6 m/m.k)
(t_hot - t_cold) is the change in temperature (C/K)

So if you installed a 22ft/7m length when the temperature was say 20C and it was heated by the sun to say 50C (?) then the change in length works out at about 4.5mm (bit over 1/8th inch).

Do the gaps cause any problems?

If you leave an 1/8th inch gap between lengths you might stop any potential buckling/distortion of the track.

That is amazing. I'm pretty proficient in Math and Astronomy, but Physics is still somewhat new to me. This is such a fun conversation for the guys in my office right now. I've always theorized a loss in length, but my boss was skeptical. This puts Science behind it, which I always knew was there; I just never took the time to ask the right people. Thank you so much. This is great.