unexplained pressure differential

December 10, 2005 10:48 PM

My old apartment had extremely leaky single-pane windows, so in the winter my roommate and I would always cover them with tightly sealed heat-shrink plastic film. Why would the plastic nearly always be bulging inward?

My intuition tells me that the plastic should sometimes bulge inward and sometimes outward, as atmospheric pressure fluctuates. I didn't ever notice the plastic flexing significantly towards the outside. Also, I might have expected the plastic to bulge outward when the winds picked up, due to Bernoulli's principle--didn't happen. What's the deal? Why was my apartment constantly behaving like a low pressure zone?

If it makes any difference, the apartment was on the lower level of a two-story building, with the windows at ground level.

posted by Galvatron to science & nature (18 comments total)

Cold air is denser then hot air. So the cold air outside was trying to push in on your toasty apartment. Think of it as a submarine collapsing on the bottom of the sea. The water is denser than the air inside.
So really, your apartment was a low pressure zone.
posted by idiotfactory at 11:00 PM on December 10, 2005

My guess is because you are heating the house, and the heat is rising and escaping at the top, through fan vents, cracks, etc., creating a suction, or low pressure inside the apartment.
posted by weapons-grade pandemonium at 11:03 PM on December 10, 2005

Cold air is denser then hot air. So the cold air outside was trying to push in on your toasty apartment.

Density doesn't have anything to do with this — pressure differentials should. And since air is pretty close to being an ideal gas, PV = nkT, i.e., pressure should increase with temperature.

Things to think about:

• What side of the building were the windows on? If it was south-facing and you got sun in the winter, I'd think that that could create a "mini-greenhouse" on the windows and end up warming the air between the windows and the plastic above the temperature of the house.
• Were there any other places in the apartment that air could escape from, especially places where it could escape only one way such as a bathroom, stove, or dryer vent? There could be a Bernoulli-like pressure gradient created by air moving over that opening.
• How well-fitted were the doors? Again, if air could escape through these gaps, then this could create a pressure gradient between your apartment and the rest of the building, as Messr. Pandemonium indicated.

posted by Johnny Assay at 11:25 PM on December 10, 2005

• There was some sun on those windows, but not a lot--and the plastic bulged inward both night and day. Also, the windows were very leaky, so a localized high pressure area between the window and plastic would have dissipated quickly.
• There may have been a bathroom vent, although it's hard for me to imagine much air escaping that way.
• The doors were poorly-fitted, and led out into a common entryway. So I could imagine a stack effect forming with the rest of the building, as weapons-grade pandemonium suggests. I wish I had taken note of the direction of airflow through the door cracks...

posted by Galvatron at 11:39 PM on December 10, 2005

I always thought it was because the window has a lot of air leaks. In my house, the older windows always bulge out the most, especially during high winds. Some of my newer windows aslo bulge, but not nearly as much.
posted by lester at 5:59 AM on December 11, 2005

The plastic is on the inside of the window, right? I suspect it's the Bernoulli effect, that the hot air rising past the windows is decreasing the outward pressure compared to the (mostly) dead air on the other side of the plastic.
posted by boaz at 7:14 AM on December 11, 2005

Are you heating with forced hot air? Perhaps the return plenum is overactive and creating low pressure in the living area. This could also decrease the efficiency of the heating system.
posted by flummox at 7:52 AM on December 11, 2005

That is an incorrect application of PV = nRT. Pressure would only increase if volume was constant, which it is not. I think that if you measured the barometric pressure indoors and outdoors you will find that it's the same, unless you are in a very tightly sealed building (i.e. not Galvatron's pad) or in a large skyscraper. Because if there really was a pressure differential, air would flow one way or the other until it was neutralized.

My theory is that the space between the window and the plastic film forms a pocket that traps air. When the wind blows outside, it "inflates" this pocket and causes it to bulge inward slightly. The surface tension / springyness of the plastic is not enough to force air back out of the cracks, so it stays slightly inflated. Remember, with such a large amount of surface area (a dozen square feet or so) even the slightest amount of positive pressure inside the pocket would cause a noticeable bulge.
posted by Rhomboid at 8:17 AM on December 11, 2005

For the same reason that shower curtains billow inwards when you are showering.

It's to do with a greater net fluid motion on one side of the curtain versus the other. The net motion is caused, as others have pointed out, by convection currents within the apartment.

You can treat air as somewhat like an a fluid. In this case, Bernoulli's equation of continuity comes into play: a1v1 = a2v2, where a is the area and v the velocity. For gases, this translates as p1v1=p2v2.

"1" is just outside the divider, "2" is just inside the divider. When I mean "just", I mean the local microenvironment.

v2>v1
==>
p1>p2

The higher pressure on one side of the divider causes it to move slightly inwards, until it attains an equilibrium where the acceleration of the concave mass balances out the force generated by the local pressure differential.

In a way, aeroplane wings use the same principle to generate an uplift force. The path of the air across the top of the wing is made longer than that across the bottom of the wing. This has the effect of increasing the net velocity of air above the wing. Therefore, the pressure just above the wing decreases. The greater pressure on the bottom half of the wing provides an uplift force.
posted by meehawl at 8:34 AM on December 11, 2005

For the same reason that shower curtains billow inwards when you are showering.

A vortex?

In a way, aeroplane wings use the same principle to generate an uplift force.

Are you sure?
posted by mendel at 9:29 AM on December 11, 2005

idiotfactory is right. Even the tightest house will experience .25 air changes per hour[1]. A house built to a standard where leaky single pane windows are acceptable probably expereince 3-10 airchanges per hour. In any multistory building in a heating climate the house will experience a stack effect where the hot air in the building rises and exits the house through the upper stories and cold air enters through the lower stories. There for a pressure differential occurs on the first floor that would bow your plastic sheeting towards the inside of the house. Where the cross over point is varies depending on the air tightness of the walls, ceiling and windows and the relative placement of the windows. In a two story house that cross over point will almost always be at or above the 1st floor ceiling. This is how I knew you lived on the lower floors of the building before even reading the MI.

Wind can also cause this but it wouldn't be an always thing and it would only effect the windward side windows. Windows on the lee side would go the other way.

[1]How many times the outside air is exchanged with the inside air. A static change of less than .5 requires mechanical ventilation if you don't want problems with mould and general poor air quality.
posted by Mitheral at 10:00 AM on December 11, 2005

Are you sure

Yes. Note I did not say "the", or "all", but "an". As a first approximation, laminar fluid flow provides one of the primary uplift and inward components in the system under discussion. I also addressed the immediate local environment of the boundary. Schmidt's solution extends the analysis into the macroscopic behaviour of the entire volume enclosed by the shower, and demonstrates that a persistent, coherent vortex results. This is similar to the vortexes left behind wings, which also provide their own components to the uplift force, and the creation of which cannot be easily addressed using a laminar model.

For the purposes of Ask, a simple explanation that accounts for most of the physical behaviour is, I think most people would agree, preferable to a rather tedious finite element analysis that takes several weeks to compute, and the results of which amount to the same thing: either get a heavier curtain, or arrange a fan or stream of air blowing into the curtain to disrupt the equilibrium/disrupt the vortex.
posted by meehawl at 10:40 AM on December 11, 2005

Dang it should have previewed: weapons-grade pandemonium is also correct.

To elaborate on his answer: it's not just vents, cracks and other gross holes. Air will pass through a sheet of gyproc in easily measurable amounts.[1] Even a polyethelyne vapour barrier will alow a very small amount of air to pass through. Keep in mind though that the surface area of the walls of your house is very large and you have all sorts of joints that need to be sealed. It requires careful attention to detail and high quality fixtures(windows and doors and other things penetrating the walls) to get down to less than 1 air exchange per hour.

The stack effect in a tall well sealed building is strong enough to prevent you from opening an outward opening door if not compensated for. In more modest buildings it is sometimes the cause of CO spilling from a fireplace after the fire dies down.

During a blower door test an absolutely amazing amount of air will pass through a 1/2" pipe in one hour at .20 wc pressure.

[1] I've seen parts held to the table of CNC router with what appears to be magic. What is really happening is the table is made out of specially prepared MDF and a decent vacuum is being pulled on the other side. MDF flows enough air under these conditions to hold sheet goods to the table even when they are being pushed by the cutter head with enough force to break 3/8" carbide bits.
posted by Mitheral at 11:22 AM on December 11, 2005

I think have to reject the Bernoulli effect explanation, as I don't think there was enough air flow in the region to justify it. The stack effect, on the other hand, is plausible. It's the only explanation I've heard yet that would adequately explain the constant pressure differential.

Thanks for all the answers received so far, and for the great restraint you all have shown by not stating the obvious answer: my apartment "sucked."
posted by Galvatron at 12:44 PM on December 11, 2005

Cold air is denser then hot air. So the cold air outside was trying to push in on your toasty apartment. Think of it as a submarine collapsing on the bottom of the sea. The water is denser than the air inside.

As other people pointed out, this is totally wrong. Pressure*Volume = n*R*Temperature. where n is the number of molecules. since n = density*volume, you can change the formula to Pressure*Volume = density*volume*R*Temp, which can be simplified to Pressure = density*R*temp (R is the gas constant, btw) so you can see that the pressure increases with density and decreases with temperature.

As Rhomboid pointed out, the actual pressure differentials will normalize as much as they can if there's any way for air to get from one place to another.

--

As you know, hot gas molecules more quickly then cold gas molecules, so my personal pet theory on this is that hot gas molecules are more likely to travel through the plastic, causing more air to be on the other side. I don't know if that's true or not, it's just a guess.

I don't think it has anything to do with the gross motion of the air (like the Bernoulli Effect or a vortex or something. Air in your living room is going to be much more still then air in a shower.
posted by delmoi at 12:59 PM on December 11, 2005

i'm living in the same situation, single pane windows and plastic covering them, and i always thought it bulges inwards cos of the air circulation between the windows and the plastic..

nice info, mitheral
i'm on the 2nd floor (of 5) and no wind on the windows
posted by suni at 1:52 PM on December 11, 2005

More about stack effect in case anyone is not convinced.

A PDF on air pressure and the building envelope from the Alberta Association of Architechs is an especially good explanation of infiltration. It includes example pressure diagrams and how to calculate the stack effect.

See also: Canadian Mortgage and Housing Corporation(CMHC) on reducing stack effect effects on multi story residences, stack effect calculations for intential cooling, and CO spill from depressurization due to stack effect.

Stack effect can actually apply enough long term pressure to weaken building structure.
posted by Mitheral at 2:06 PM on December 11, 2005

Good links, Mitheral. For the geometry of that building, they suggest that the stack effect could have exerted 1-2 pounds of force on the plastic. Add in another pound or two of wind force, and I we're in the ballpark.
posted by Galvatron at 3:12 PM on December 11, 2005