So 90 degree fittings are a no-no?

[prairiewood]

I'm about to go out and buy my 6" PVC. Is it vital that i use two 45 degree fittings rather than one 90 to make a right angle turn?

[Jim O'Dell]

Unless you have really long runs with lots of turns, no, probably not a killer. If you want to preserve every last bit of CFM you can, then , yeah, do it with 45's.:D My thought on this is that the CV performs so well, that you won't lose enough to make that big of a difference. But I only used 45s on my system, with a little bit of straight pipe in between. Jim.

[bababrown]

Today I bought and measured some 4" PVC fittings to get a better feel for the losses. I'll try to repeat this in 6" fittings when I can. I believe that the 6" fittings will come in at about 2/3 of the losses for 4" fittings but that remains to be shown. For 4" fittings at 4000 FPM (average) airspeed:
1) 90 degree regular (1r ells) ells: 0.69" static pressure loss,
2) 90 deg street ells (1.5r ells): 0.375" static pressure loss,
3) double 45 ells with minimum spacing: 0.56" static pressure loss.
This shows that the 45s clearly are better than regular 90s but not nearly as good as long sweep ells if you can find them in 6". The differences don't seem terribly large but realize that they add up, that they get larger with the larger airflow we want, and they are one of the few losses we can control. Also, I think a larger spacing saves a little more just by cutting the corners and saving a little straight pipe.
bababrown

[Don01]

I'm about to go out and buy my 6" PVC. Is it vital that i use two 45 degree fittings rather than one 90 to make a right angle turn?

Just be very careful. I just bought my fittings. The supplier had a short 90 which had very little radius to it. It was a very hard 90. I chose 2 45's but wonder if a nice sweep made from smooth flex pvc wouldn't be just as good.

Don

[Don01]

Today I bought and measured some 4" PVC fittings to get a better feel for the losses. I'll try to repeat this in 6" fittings when I can. I believe that the 6" fittings will come in at about 2/3 of the losses for 4" fittings but that remains to be shown. For 4" fittings at 4000 FPM (average) airspeed:
1) 90 degree regular (1r ells) ells: 0.69" static pressure loss,
2) 90 deg street ells (1.5r ells): 0.375" static pressure loss,
3) double 45 ells with minimum spacing: 0.56" static pressure loss.
This shows that the 45s clearly are better than regular 90s but not nearly as good as long sweep ells if you can find them in 6". The differences don't seem terribly large but realize that they add up, that they get larger with the larger airflow we want, and they are one of the few losses we can control. Also, I think a larger spacing saves a little more just by cutting the corners and saving a little straight pipe.
bababrown

Baba,
When using 2 45's isn't the practice to space them out with a straight run to open up the radius more than a standard 90?

Don

[bababrown]

Don,
Yes, most of us have spaced the 45s. I think it saves a little loss just by saving the extra pipe it would take if we didn't cut the corner as much.
bababrown

[Don01]

I just read bababrown's posting on 4" rigid vs flex. Based on this research, a long sweep 90 made from flex pvc would NOT be more efficient than making it from 2 - 45's.

Don

Just be very careful. I just bought my fittings. The supplier had a short 90 which had very little radius to it. It was a very hard 90. I chose 2 45's but wonder if a nice sweep made from smooth flex pvc wouldn't be just as good.

Don

[Don01]

Baba,
Is there an optimal spacing? I suspect more is better but the actual shop layout will be the limiting factor.

Don

Don,
Yes, most of us have spaced the 45s. I think it saves a little loss just by saving the extra pipe it would take if we didn't cut the corner as much.
bababrown

[bababrown]

Don,
I think the spacing between the 45s makes little difference beyond what we save in pipe by cutting the corner. If you look at two 45s spaced minimally, there is the sharp corner at the center of each 45 and then quite a bit of straight section between. So I doubt adding more pipe will decrease the loss in the resulting 90 degree fitting. We also are getting down to polishing the fuzz off the peach; it's not that exact. I think the practical problems of what you can implement take over. The other part is flex. I wouldn't use any more of it than really is necessary because it is lossy. I haven't looked much for long sweep street ells in 6" PVC but they will be better if we can find them.
bababrown

[JJohnston]

McMaster-Carr has long sweep elbows. They're not "street" ells, in that they have a female socket on both ends, and they're really (really) pricey, but if you really need one, they're available. I used 2 of them on my miter saw duct run; I needed a 180* change in direction in a small space, and 4 45s with short straight sections just wouldn't have worked. See my gallery in "Dust hoods and pickups" for a picture.

http://www.mcmaster.com/#9102k116/=3uugu9

[bababrown]

JJohnson,
I also found some thru Specified Fittings and they also are about $36 for the long 90 ell FXF. The street ells are more like $70! It's probably less demand and maybe we haven't found the right supplier(?). I've found the prices all over the map. The long fittings do look good however versus the regular 90s in which the sides of the two pipes either touch or almost do at the inside corner.
bababrown

[Don01]

Baba,
I appreciate your input (and everyones!). Certainly I am not interested in peach polishing, I am simply at a point where I have a ton of work ahead of me to hang all of this pipe and if there is an obvious advantage then this is the time to do it as it will not increase the work. Your fine work to date suggests there is an advantage to have the right radius, so I will apply that to my bends.
As I noted in an earlier post, the gallery entries suggest that you do what you gotta do as the CV seems to have the power to overcome most inefficiencies in piping design.
Thanks again for your input.
Don

Don,
I think the spacing between the 45s makes little difference beyond what we save in pipe by cutting the corner. If you look at two 45s spaced minimally, there is the sharp corner at the center of each 45 and then quite a bit of straight section between. So I doubt adding more pipe will decrease the loss in the resulting 90 degree fitting. We also are getting down to polishing the fuzz off the peach; it's not that exact. I think the practical problems of what you can implement take over. The other part is flex. I wouldn't use any more of it than really is necessary because it is lossy. I haven't looked much for long sweep street ells in 6" PVC but they will be better if we can find them.
bababrown

[armstrr]

its been a long time, but cant the S&D stuff be sofened and bent to make custom long curve elbows? i forget where i read it. you could just use couplers to attach the pipe. anyone have more info on that?

[Don01]

There are several references telling tthe process. You cap one end, fill the pipe with clean fine sand, cap opposite end, heat evenly, bend slightly more than you need as the pipe will spring back a bit. I haven't totally ruled out that option, but it may not be worth the trouble.
Don

its been a long time, but cant the S&D stuff be sofened and bent to make custom long curve elbows? i forget where i read it. you could just use couplers to attach the pipe. anyone have more info on that?

[dwdrury]

Just do it outside, Don, as I understand the fumes given off by heating PVC are quite toxic.

Regards,
DWD

[Don01]

Pricey?? I found a couple in the east. 36" radius for CAD $352.00 ea plus freight. Ouch..

Don

McMaster-Carr has long sweep elbows. They're not "street" ells, in that they have a female socket on both ends, and they're really (really) pricey, but if you really need one, they're available. I used 2 of them on my miter saw duct run; I needed a 180* change in direction in a small space, and 4 45s with short straight sections just wouldn't have worked. See my gallery in "Dust hoods and pickups" for a picture.

http://www.mcmaster.com/#9102k116/=3uugu9

[bababrown]

Don01 asked about spacing 45 fittings and I gave the wrong answer. In making measurements of multiple closely spaced fittings I kept getting losses that I thought were too high. I then found out about "system effect". When two fittings are spaced closely the loss is K(Loss A + Loss B) where k is 1 or greater. I improved my test setup and made more tests. The attached table shows results for individual fittings and combinations with different spacings. The main points are:
1) Use smooth fittings. Don't use short ells and tees.
2) When possible space fittings 10 pipe diameters or more.
3) Use as little flex as possible, use the smoothest you can get, and make moderate radius bends.
bababrown

[Don01]

BABA,
Are these losses really significant?
I would like to keep it to 5 or 6X for 6".
I am going to inquire about getting the factory to bend a piece of pipe for me.
Will let you know.

Don

[bababrown]

Don01,
The added losses seem quite small with smooth fittings ( if any ) but build up as the fittings get rougher. It's interesting how I got here. I put two or more fittings together so the measurement errors would be reduced. But when I added two smooth fittings in between two regular 90 degree ells the losses went down! And I repeated the decrease for a number of combinations. So then I did some research on system effect.
The added losses are significant if you use regular 90 ells and regular tees. The air flow is so disturbed that I had to switch to 10 foot test pipes on each end to get smooth flow.
If you go with 45 ells and 5 diameter spacing you won't add much loss. The loss will be about the same for the standard 45 Ys (small y shape) and 45 ells at drops if you space them 5 diameters. Most of us have put them directly together and we seem to get acceptable performance.
bababrown

[Don01]

Thanks Baba and thanks for all of your efforts in this area.

The regular S&D 90's that I looked at were not much up on an EL made by butt joining 2 pieces of pipe cut at 45 d. - a very hard 90 indeed!
I would consider 4 X 22.5 bends but the fitting are the most costly part of this adventure. No real difference in price for a 90, 45 or 22.5. Same amount of work to produce I guess. And if we want smooth walls we cannot use drain pipe and that hit the wallet pretty fast as well.

Don


Don01,
The added losses seem quite small with smooth fittings ( if any ) but build up as the fittings get rougher. It's interesting how I got here. I put two or more fittings together so the measurement errors would be reduced. But when I added two smooth fittings in between two regular 90 degree ells the losses went down! And I repeated the decrease for a number of combinations. So then I did some research on system effect.
The added losses are significant if you use regular 90 ells and regular tees. The air flow is so disturbed that I had to switch to 10 foot test pipes on each end to get smooth flow.
If you go with 45 ells and 5 diameter spacing you won't add much loss. The loss will be about the same for the standard 45 Ys (small y shape) and 45 ells at drops if you space them 5 diameters. Most of us have put them directly together and we seem to get acceptable performance.
bababrown