a Pipe with a fan will it help ?

[H0nDA2000]

<Yeti> weeell u guys helped me a lot here.. but there so many opinions and am lost i need some1 to recommend me using the pic i posted

 

[Archangel]

the first paint drwaing?
that would work, jut make shure the air has a way to exit the case.
and make whure the pipe has the same diameter at all times, otherwise it will work agains the airflow.

 

[Yeti]

Yeah, if you can't do a straight duct like the photo, I would go with the first drawing. Just make the duct go fan to fan so you're not covering any of the heatsink, and make sure the fans are both pulling air in from outside the case (some HSFs do blow the opposite direction).

 

[apj101]

Yeti, how would this model be effected by increasing pipe length??

 

[Yeti]

Yeti, how would this model be effected by increasing pipe length??

Not much really. A half meter, 12 cm diameter pipe will have a friction pressure drop of less than 0.5 Pa (Pascal) given a 40 CFM. Considering that normal fan will have about a 20 to 30 Pa deltaP, and that 50 cm is probably longer than anyone would need, it doesn't have much affect. Changing the diameter, on the other hand, will affect the Reynolds number which will then affect the friction factor, but I think you would have to have low flow rate with large diamters for that to make much difference.

 

[apj101]

I was more thinkink of piping the air in from outside. I have a old fan vent near my machine. Pipelength probably 1m required.

 

[Yeti]

I was more thinkink of piping the air in from outside. I have a old fan vent near my machine. Pipelength probably 1m required.

It will depend a lot on the ability of the fan (pressure rise & CFM). Most likely, though, the duct friction won't make much difference. For example, even at 100 CFM (higher CFM = higher velocities = higher friciton loss) it would take about a 10 m pipe to decrease the performance of 1 fan by 1/2.

 

[vortmax]

yeti:

we treat everything as compressable. When doing very rough analysis very steady unidirectional flows (noncurving jet streams, wind across mountains, ect...) we will treat it as incompressable to be able to hand crunch numbers.

Yes it is calculation intesive. Everything in my field is. That's why when the Crey's first came out, one of their first tasks was a numerical weather model and also why my department has several SGI boxes including a 13 node cluster.

 

[Archangel]

dont start this again... this is just to get a bit air into the case,.. nothing industrial. so no need for a perferct frictionless solution.
if it cool right its already good.

 

[Yeti]

so no need for a perferct frictionless solution.

Not trying to get a frictionless solution, just making sure the fan can handle it.

if it cool right its already good

True, but if you want to design something it helps to know before you build it if it will cool right .

 

[apj101]

dont start this again... this is just to get a bit air into the case,.. nothing industrial. so no need for a perferct frictionless solution.
if it cool right its already good.

this ranks as one of the more advanced topics discussed here, and is for the most part way above my head (and yours). But that doesn't mean we have to stiffle the thread.
And plus stuff like this is always good for the search engine "points"

 

[malikah]

Flush

If the end of the tube is flush against the top of the CPU fan and the CPU has a heat dispersion grill, then it should work fine. The best way to test it is to run the CPU for about 15 minutes and then touch the grill with your fingertip. Pretty much a no brainer...

 

[computermaineack]

Wow...the last post on this thread was over 6 months ago, I'm pretty sure the original poster has solved the problem by now.