Air gap causes R-value loss, why ?

[jimbob44]

I have read several comments on these posts that say that an air gap between subfloor and insulation causes an R-value loss. For example:
Case #1 - 9.5" floor joists filled up from bottom of joists to depth of 7.5" with cellulose insulation (thus giving an air gap of 2" between top of insulation and bottom of subfloor), R-value = X.
Case #2 - 7.5" floor joists filled up completely to bottom of subfloor with same 7.5" depth insulation (no air gap), R-value = more than X.
(Of course, the cellulose is going to settle after a while, but for the sake of this question, assume no settling)
Maybe there's something I'm overlooking, but I fail to see why the R-value in case #1 would be less than the R-value in case #2.
If someone believes otherwise, would you please comment as to why you think this would be?

[JRR]

I understand your point.  And if the air in the gap were kept stagnate, the overall R would probably be increased, not lessened, by the air gap.  Perhaps if some small "compartments' were set up, this would be the case.

But usually the air will be free to move about over a large area... perhaps even escape the cavity ... now we've got opportunities for real convection heat exchange.

[jraabe]

Exactly, the problem is not the air gap but the convection of air even in a sealed airspace. And most underfloor gaps are open below and around the insulation to airflown from below and even through the floor.

An air barrier such as Tyvek (if it can be kept up and intact) will go a long way towards minimizing the airflow in the joist cavity. Remember, gravity wants everthing to sag and come back to momma earth.

[glenn-k]

I have that problem myself. :-/

[jb]

LOL....