High velocity rectangular elbow pressure drop

[cme]

I have been assigned to an air handler replacement project serving a lab of a hospital. The system is 100% OSA, dual-duct and has approx 16,500 cfm at 60 deg supply air to the space. The new system will supply 20,000 cfm and possibly 52-54 deg air range.

At that cfm, the lab will have about 1.5 cfm/SF and the duct velocity will be approx 4000 FPM. This is like a dust collection system. The exist TSP is 7.5" and calcs indicate 3.5-4" ESP for known losses like filters, coils, etc. The new ESP may be 6" using fan laws plus losses due to duct tie-in since this unit is going on the roof.

My question is fitting losses. ASHRAE gives a constant of .12 times the velocity pressure. So that puts me at .12" SP for an elbow since my VP is unity. This doesn't seem close at all.

I have tried Carrier data books and this seems to be all in the same range.

I found a mfr (Aero Dyne) which states their PD is .17" at 2500 FPM. I can use fan laws to est SP drop at 4000 fpm.

Anybody here have any data or experience on rectangular fitting pressure drop for my situation. I'm hesitant on using turning vanes at this velocity not just for SP drop but permanent deformation or blow-out.

 

[lilliput1]

What is the dimension of the elbow (width & height), and centerline radius. Is it a "easy" bend (plane of bend parallel to the small duct dimension)or a "hard bend" (plane of bend parallel to long dimension)? Have you looked at SMACNA duct design tables/figures?

Where noise is a concern, we limit rectangular ductwork to 2000 FPM and use round or flat oval ductwork up to maximum 4000 FPM. We also provide sound attenuator (packless type in hospitals) at the supply & return air duct.

 

[cme]

The exist elbows are 2.0 Radius to CL of duct. I don't have the room for that. I'm talking abouy a hard 90 with turning vanes.

The cold deck duct is 50x16.

I'll check SMACNA tables.

 

[dpoppeli]

Have you thought of the noise issues?? I've seen HVAC trunks running at 3,500 fpm and double-wall (inside wall perforated with fiberglass fill) was the choice. As I recall elbows were 1.5-2x wide radius types probably to reduce pressure loss as well as have benefit of continuity in the noise reduction. If still forced into vanes, you could look at assemblies that are formed with dual walls with perforated material ...would increase your static loss but reduce noise as well as probably still be more efficient then a vaneless elbow.

I find SMACNA's loss calculator one of the best. I suspect there charts work pretty well too.

 

[lilliput1]

From SMACNA HVAC Systems Duct Design (note 2500 FPM max duct velocity recommended):

For Single thickness vanes w/ vane radius 2" and vane spacing 1.5", 2500 FPM, Loss Coef = 0.20
For Single thickness vanes w/ vane radius 4.5" and vane spacing 3.25", 2500 FPM, Loss Coef = 0.22

For Double thickness vanes w/ vane radius 2" and vane spacing 1.5", 2500 FPM, Loss Coef = 0.40
For Double thickness vanes w/ vane radius 2" and vane spacing 2.25", 2500 FPM, Loss Coef = 0.49
For Double thickness vanes w/ vane radius 4.5" and vane spacing 3.25", 2500 FPM, Loss Coef = 0.23

Fitting loss in. wg. = Loss Coef C x Vp
Where Vp = (FPM/4005)^2 inches wg.

 

[cme]

Thanks Lilliput1,

I just don't agree with those figures.

For Single thickness vanes w/ vane radius 4.5" and vane spacing 3.25", 2500 FPM, Loss Coef = 0.22

VP=.38 at 2500 fpm

PD=.38x.22=.09" ........ I just don't see that.

Aero Dyne vanes are max is 3000 FPM also

 

[lilliput1]

So add extra for safety. Make sure you account for all equipment pressure drops, entry loss, exit loss, transitions, convering or diverging fittings, volume dampers, fire dampers, air flow measuring stations, air terminal boxes, air terminals, room relative pressurization, duct leakage, etc. The bottom line is that the fan motor hp and fan Class must be adequate for the duty. Do not be too close, select next motor size or fan class if too close.