GideonTech.com Forums

  • November 19, 2019, 02:47:01 AM
  • Welcome, Guest
Please login or register.

Login with username, password and session length
Pages: 1 2 [3]   Go Down

Author Topic: Wind Tunnel  (Read 2064 times)

Fanny J. Crosby

  • Senior Modder
  • ***
  • Karma: +0/-0
  • Offline Offline
  • Posts: 577
Wind Tunnel
« Reply #20 on: June 13, 2002, 09:14:00 PM »

I am actually talking about a heatsink where all the fins are the same way. Blowing it across them... through them. It wouldnt increase turbulance much because the skinny fins won't hurt airflow much.
Logged

strongShock

  • Senior Modder
  • ***
  • Karma: +0/-0
  • Offline Offline
  • Posts: 448
    • http://www.strongshock.com
Wind Tunnel
« Reply #21 on: June 13, 2002, 11:28:00 PM »

yes, a surface normal to the flow causes stagnation at the flow axis, even with a hubless fan, which itself reduces heat transfer. ...but, the added

TURBULENCE ALWAYS INCREASES HEAT TRANSFER
(emphasize this fact above)

because it breaks up the boundary layer. This is why the good heatsinks & waterblocks have rough flow surfaces.

i am wrong in always citing Delta (I am not affilliated:), but I stand by my claim that the STATOR BLADE EXHAUST NOZZLE is not found on many fans. In fact, a few high output Delta models & Vantec Tornado fans are the only ones I see stator blades on (stator blades == stationary blades that are designed with an angle of attack & geometry which cancels the rotational momentum added by the driving rotor upstream of it). I agree though, most manufacturers make a slow speed fans, but there is an absence of low speed fans with stator blades (stators are not the typical 3-spokes you see on most fans. look very close at your Delta screamer or Tornado fan & you'll understand exactly what I'm talking about).

The funnel idea is quite complicated to detail, but I can somewhat say this guilt free:
Calculate the exit area of the fan, which is roughly:

pi*(OD^2 - ID^2)

where OD is the outer radius of the blades & ID is the radius of the center hub. Then keep the area of the duct (funnel, tube, whatever) greater than this & you'll be OK. This doesn't take into account subtly details of the flow. Generally you also need to keep the transition from the fan to the small exit area reasonably gradual. You don't want a sharp bend in the walls or anything like that. To accurately calculate the effects of geometry you need to hire an aerospace engineer, hehe.

If you use a hubless fan & consider only aerodynamics, the center lug of a heatsink would have a parabolic anulus shape (a blunted point that widens closer to & sweeps out around the CPU). But we are conducting heat, so the core or base thicknesses are limited by the need to quickly transfer heat away from the core (we still need good ole copper). Although I think the core of it is inefficient (it creates a huge stagnation area for a hubless fan), I feel that if a normal statorless fan is used, the AVC Sunflower is the geometry of choice:



We are not necessarily after creating "bluff-body" flows near our heatsinks (sorry, aero jargon for most things;). A rough surface contacting high velocity is all we want, after all, our boundary layer is only microns think.

If you get overzealous with the turbulence idea, you might end up creating large stagnation areas where you don't want them, or even create stall (very bad for heat transfer, a vortex traps heat near the metal). The AVC "accepts" the rotational flow nicely, & guides it through a path that accelerates it along the back of each bent fin, but is gentle enough to keep the flow from stagnating or stalling.

good night
Logged
]
Pages: 1 2 [3]   Go Up
 

Page created in 0.099 seconds with 19 queries.