duct shape vs. flow resistance

[cjccmc]

I have a design in work that requires air purge (0.5 psig) be routed thru an area that requires a short portion of the duct to act as a load carrying structure. Air enters thru round 3 inch dia duct and after traveling about 8 inches and making two 90 deg turns in my piece of structural duct, it exits into another 3 inch dia duct. I tried to maximize the cross section thru my duct / box which has an irregular shape thinking it would offer lower overall resistance. One of my coworkers believes that it would be better to make the internal passage with smooth continuous surfaces even if that requires reducing the cross section area by a substantial amount, as much as 50% in some areas. Which way should I go to minimize flow resistance?

 

[katmar]

As a general rule smooth fittings are better. This is particularly true of the inside of the bend. However, increasing the flow area makes a significant difference as well.

As an example compare a smooth bend with r/D of 2 with a single weld 90 degree mitered bend. The smooth bend would have a resistance coefficient (K-value) of about 0.25 but the mitered bend would have a K-value of about 1.1. If the fittings were of the same diameter the pressure drop through the mitered bend would be 4 times that through the smooth bend.

The pressure drop through a fitting is proportional to the velocity squared, so if your cross sectional area increases by 50% the pressure drop would decrease by 1.5² (=2.25), bringing the difference between the smooth and square bends down from 4 to 4/2.25 (= 1.8). So your structural fitting may have a pressure drop of 80% more than 2 smooth 3" bends. Only you can say whether this would be important in the actual job you are doing, but I would guess this small increase in absolute terms over just one section is unlikely to be important.



Katmar Software - Uconeer 3.0

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"