Hi,
That's very kind of you.... Cookie Monster.
The calculation on airflow required is actually quite simple. The total heat produced by the computer is 200W, say (NB - a reasonable figure for a gaming rig with a mobster gaming card).
The exhaust air temperture is, say, 10C above input (room temperature).
Then the CFM must be the volume of air whose thermal capacity multiplied by 10C equals 200W for one minute - because CFM is Cubic Feet per minute.
The equation you need is:
Watts x1.8 = CFM x TemperatureRise in Centegrade (Air Input to Air Exhaust)
Note that this remains true whatever case you use and whatever fans you use etc.
The next issue is to determine the internal case temperature. If you design/arrange your case so that all the hot components exhaust straight out of the case the internal temperature will be much lower than the exhaust temperature. It's the internal temperature you want to control - that's the purpose of the case fans - so the potential to improve the layout and reduce the CFM required is huge. [plug] My PaQ case uses this technique -
http://paq.ltd.uk [/plug]
In an ordinary case about half of the air coming off the CPU actually exists from the case directly, and probably none of the air coming off the graphics card does. For a gaming rig, (but using only one graphics card) this means that there is about 160W going straight into the case interior, so you will need 30 CFM to keep the internal case temp down to 10C above room temperature.
And if you want to keep the temperature rise inside the case down to 5C you will need 60CFM. No way to do that quietly that I know of.
Now consider what happens if you suceed in exhausting all the CPU and graphics card hot air out of the case by means of suitable ducts. The internal heat generated will drop to about 50W and you'd need only 10CFM to maintain an internal temperature of 10C above room.
I'd prefer a 5C rise, in which case you need 20CFM. This is a good figure to aim for, as the airflow through the CPU cooler is typically 12- 15CFM and 20CFM means that some of the air is available to go out through the PSU.
The final question is how to shift this air as quietly as possible.
The key fact is this: big is beautiful. More is beautiful. Seriuosly, you can pay $30 for a fan or $2 for a fan, both 120mm. The noise levels between them for the same airflow will not vary by more than 20%.
But if you get two fans and volt them down to give the same airflow as one fan alone, the noise level will drop by 90%, typically. As a rough guide the dB produiced by any given fan changes linearly with the CFM. But dB is a log scale. Take a fan producing 30dB at 12V, then instead run two of them at 6V each to get the same total airflow (CFM proportional to fan speed or to voltage). Noise level is then 15dB each but 15dB + 15dB = 18dB, and 18DB compared to 30DB is a drop in noise by 90%, because dD is a log scale.
The same logic applies when replacing a smaller fan with a larger one, for the same CFM. The noise of the air flowing through a whole actually increases with the sixth power of the air velocity. Bigger fan, same CFM, lower velocity... go figure
So what you really want to do is figure out the airflow you need and then fit as many and as large fans as you can, and downvolt them to get the required airflow.
Peter