Interesting data - converging leakage curves

[Sean Maxwell]

Any idea what's causing the data to behave like this? Concrete building. Pretty tight. You think at lower levels of pressure, there are gaps that are open and air can flow through, but that at higher pressure, those gaps start to close and the leaks behave the same? I'd like to see if at very high pressures, the curves would cross each other. Or not! Let me know your thoughts.

Interesting Data

 

[Paul Carling]

It looks like you have larger openings on your depressurisation test so building elements are likely being pulled away from the seals. Larger openings mean more flow initially but the flow is turbulent which moves slower and therefore results in a lower air flow exponent. Therefore as you increase the building pressure the trend line is shallower.

 

[Sean Maxwell]

Hi Paul, that makes sense that the building leakage behaviour would be changing with pressure, but isn't an event like that more dramatically marked? For example, i've seen test results that, at some higher pressure, something blows open and so you'd get a "flap" in the data like a flap opening in real life.
Of course, not all flaps open in a sudden movement. Some lift open gradually, like a gravity-weighted damper. More pressure = more open = more flow, and the degree of open-ness is rather linear with pressure. Maybe that's what's going on with this building. Some thing or things are gradually lifting open.
I wonder if the depressurisation curve, which is higher leakage at lower pressure than the pressurisation curve, shows a thing that is flapping closed at higher pressure. Or if the pressurisation curve, which is tighter at lower pressure, shows something flapping more open at higher pressure. In other words, which of the two curves shows the building changing at higher pressure? Curious.

 

[Paul Carling]

If the results from both tests are similar but the air movement is different then you'd expect them to converge at your point of interest which is normally 50Pa. If the air movement is different because building elements are responding differently to the different flow directions then you'd expect to have a lower intercept and higher exponent on one test and a higher intercept and lower exponent on the other.

You see the same thing if you look at equipment calibration when comparing different fans of the same model. In a multiple fan setup you can identify which fan should be your main one as it's the one with the lower intercept and higher exponent as this means it gives you more flow at higher pressures meaning you can test slightly larger buildings with it than your other fans.