Here's a link to a dust collector test that has been posted here before. It is a good report on dust collector performance, but it also contains a small inset on the lower left corner of the last page that might get overlooked. The inset discusses the performance improvements of a bell mouth pick up port vs square cornered pick-ups. Their testing showed improvements of more than 100cfm and a drop in SP of 1" for a bell mouth shape. That's pretty significant.
http://www.portercable.com/uploads/PCD/Documents/News/182DustCollectors.pdf
So, the question becomes, would a bell mouth outlet on a Thein separator provide some level of improved performance. It sure seems like it might be worth a try.
Has anyone done this?
HAHAHAHA You are absolutely killing me here Retired! I knew there was a reason for me to wait until today to start my build. Now you had to go and show me this? Looks like I am going to have to engineer this in my plan as well. Good find.
Quote from: Don_Z on August 04, 2011, 09:15:28 AM
HAHAHAHA You are absolutely killing me here Retired! I knew there was a reason for me to wait until today to start my build. Now you had to go and show me this? Looks like I am going to have to engineer this in my plan as well. Good find.
Don't rush into this Don. My goal is to find enough improvements that the Thein separator will work without turning on the dust collector!! Just kidding of course.
As you're thinking about how you are going to engineer a bell mouth into your design, here are some other options, but none quite as good.
(https://www.dropbox.com/s/1uicdhimdnvxzj2/FanData-BellMouth.jpg?raw=1)
Quote from: Chuck Lenz on August 04, 2011, 03:19:29 PM
Call me slow and stupid, I don't get the chart. How is there a 90% VP loss from a unflanged straight pipe ? Or even 50% VP loss with it flanged. Where did you find the chart ?
Chuck,
This is engineering data published by Cincinatti Fan. However, there is no explanation of what VP (velocity pressure) is or how it is measured. So, without that understanding, I think all we can extract from the chart is there are some real differences in performance between various shaped inlet ports, some very clearly much better than others.
Regards
Quote from: Chuck Lenz on August 04, 2011, 08:59:21 PMWHAT ? The only thing that I can extract from any of this is that your throwing out information that you have no idea about, but you still want to try to come off like you do know something about it. You have just lost all credibility and patience with me.
Chuck,
I'm sure the current Thien separator works as well as it does due to the testing of many different ideas without hard data to support the outcome. That is what this thread is all about. I threw out an idea without any personal experience to support it, but I provided two sources of data that seemed to support the idea. I made no claims whatsoever. As a matter of fact, I titled the thread with a question mark. To most people that means "I don't know, do you have an answer?"
Now, if you have no value for my posts, that is O.K. - you can simply skip over them, but please don't bother to reply if your only objective is to malign me in a public forum.
Don,
Here's an explanation why a bellmouth inlet is more efficient than a straight pipe inlet. The source of this information is online course material from the U.S. Government. Again, this explanation seems to support the idea that a bellmouth pipe might provide an improvement in performance of the Thein separator exit pipe.
Hood Entry Loss
The following discussion on hood inlets will help you understand how hood designs affect the overall pressure drop and fan requirements of a system.
When air enters a duct under suction, the airflow converges as shown in Figure 2. The area of air convergence upon entering a duct is referred to as vena contracta. The area around the mouth inlet is void of airflow due to the formation of the vena contracta.
As air passes through the vena contracta its velocity increases. After passing through the vena contracta, the airflow expands to fill the duct. As the air expands, some of the velocity pressure converts to static pressure. The hood static pressure and entry loss are related to the size of the vena contracta.
The hood geometry determines the size of the vena contracta by influencing how smoothly the airflow will enter the duct see Figure 3, [Hood Entry Loss Coefficient (Fd) for Various Duct Designs]. Figure 3 shows three different duct designs and their corresponding hood entry loss coefficients. The duct inlet design illustrated in Figure 3(c) is superior to those in parts (a) and (b) because air enters the duct more smoothly (less vena contracta effect) and air is drawn into the duct primarily from the front where the contaminated air is located.
(https://www.dropbox.com/s/l3yf63omyktemb7/fig02.gif?raw=1)
(https://www.dropbox.com/s/gkn00gc3k22jzp9/fig03.gif?raw=1)
Here's another hobby that has found an application for bellmouth pipes:
(https://www.dropbox.com/s/0kzw767izh3rav4/Carb.jpg?raw=1)
I think that the bell mouth idea is valid input and worth posting. I am assuming that unproven ideas are acceptable here. The bell mouth would definately be on my list of ideas to try. The improved efficiency might even cause more problems than it solves. Certainly the height above the baffle plate would have to be reduced.
Interesting input, thanks for posting.
Dave
Another place they're (bell-mouth) used is porting for low-frequency drivers (woofers).
I haven't done any testing. Maybe I will give it a shot.
Quote from: phil (admin) on August 06, 2011, 07:31:55 AM
Another place they're (bell-mouth) used is porting for low-frequency drivers (woofers).
I haven't done any testing. Maybe I will give it a shot.
Phil,
I think the other challenge this idea presents is what is the best way to create the bellmouth. I am picking one up this week at my local supplier (about $20) who is also building my round to rectangular entry transition piece. One thing he told me that concerns me a bit is the off-the-shelf bellmouths have a pretty large flare. I think he said a 5" bellmouth has a flare that is 9" or 10" O.D. It remains to be seen how that will effect internal air flow dynamics. I wish I had a way to fabricate something more in the proportions of the carburetor inlet ports posted above.
I have employed a mini bellmouth on my floorsweep by just hitting the corner with the largest roundover bit I own. However, it should be larger to get the maximum benefit, but large roundover bits are expensive and they require a large router mounted in a table.
I think I am going to fabricate my separator top with a bolt-on outlet connection so I can easily test different combinations. I don't think this idea is going to be as simple as putting a purchased bellmouth in place of a straight pipe. It will probably require some experimentation, but maybe with several of us trying different combinations we will hit on something worthwhile. Even if nothing comes of this, it is the challenge of trying that makes it fun.
Well done and thanks again Retired. Anything that helps the community is a benefit. It is the naysayers that hinder the concept of progress, so I appreciate what info you do find and contribute. I plan on adding bellmouth adapters at my machine ports using stacked mdf and then shaping by hand using my tried and true Nicholson #50.
The concept that I gathered from that article was that as long as there was a decent radius from the inlet port that it will create the vena contracta effect and therefore give an improvement. I am thinking that by opening up the factory 4" DC hole in say, a table saw and if you wanted to end up with a 6" duct connection, then by cutting it to 8" then take 2 pieces of MDF to fit the new hole, cut your 6" hole. Then either rout out the radius or use whatever your preferred method would be ( I am VERY fond of my patternmakers rasps.) continue the radius to suit your vision of a bell mouth opening.
Quote from: Don_Z on August 06, 2011, 02:31:20 PM
The concept that I gathered from that article was that as long as there was a decent radius from the inlet port that it will create the vena contracta effect and therefore give an improvement. I am thinking that by opening up the factory 4" DC hole in say, a table saw and if you wanted to end up with a 6" duct connection, then by cutting it to 8" then take 2 pieces of MDF to fit the new hole, cut your 6" hole. Then either rout out the radius or use whatever your preferred method would be ( I am VERY fond of my patternmakers rasps.) continue the radius to suit your vision of a bell mouth opening.
Don, I think what you just described actually reduces or eliminates the vena contracta, which is exactly what you want to do. And I think a filed radius on built-up MDF would work just fine since I don't think the geometry of the flare is all that critical.
However, I cringe at the thought of cutting holes in my equipment, and I'm not sure I would have the courage to open up a 4" port to an 8" hole. My cowardice would probably result in me trying to put a collar on the inside the existing port (space permitting) to give it a bellmouth entrance. But you are right, if your DC is big enough to handle 6" duct then an 8" hole with a bell mouth radius to 6" pipe is probably the ultimate.
Or there is this...
http://www.pennstateind.com/store/N-BELL06.html
Quote from: Don_Z on August 06, 2011, 03:28:26 PM
Or there is this...
http://www.pennstateind.com/store/N-BELL06.html
Don, you ain't seen nothin yet!! Take a look at these links for info on making "cheap" bellmouths. If I ever tried this in my wife's kitchen, I'll be looking for a new place to sleep.
http://forums.sohc4.net/index.php?topic=57484.0
http://forums.justcommodores.com.au/vy-holden-commodore-2002-2004/92806-cai-alien-design.html
And if you want dimensional information on the off-the-shelf models so that you can build your own, here's another link:
http://www.spiralmfg.com/hoodsf.htm
And just when you think the internet CANNOT get any dumber, you have to show me this :o? This is the stuff comedy gold is made from. My girl would have me outdoors for pulling one of those stunts. But I have to tell ya, that Spiral web site looks like a one stop shopping site for what we are trying to accomplish. I do believe I will have to inquire about pricing. Good find, thanks.
The hot oil method works and was my first thought. I make fishing lures and we use this method for shaping polycarbonate. I have never tried it with anything this big though, but definately works.
I would probably make a male former and lay up some fibreglass, with some wood ribs laid in, to add some stiffness to the shape. The trick is making the former, but it is not that difficult.
Great links, I love seeing mans inginuity at work. This is what we do.
Dave
Quote from: Vodkaman on August 06, 2011, 08:54:12 PM
I would probably make a male former and lay up some fibreglass, with some wood ribs laid in, to add some stiffness to the shape. The trick is making the former, but it is not that difficult.
Dave
Dave,
I've never worked with fiberglass, but if I am envisioning the male former correctly, it sounds like something that could be turned on a lathe pretty easily.
Yes, but I do not have the luxury of a lathe. If you have the capability to turn such large blocks, you could turn a bell mouth, but it would be cheaper to buy one I guess.
Fibreglass is very easy to work with and done right, would offer a good, strong solution. Lots of Googling first, to get the feel for the process.
Dave
Quote from: Vodkaman on August 06, 2011, 09:28:08 PM
Yes, but I do not have the luxury of a lathe. If you have the capability to turn such large blocks, you could turn a bell mouth, but it would be cheaper to buy one I guess.
Dave
Well, the price was right so I picked up a 5" bell mouth this week at a local shop. I decided to post a few photos of it now because my separator project is on hold until I recover from a bad back - old age problems!
In the first photo, you can see that the heigth is only 2-5/8", so there is not much of a neck to work with. That can easily be corrected by attaching a length of straight pipe similar to the one in the second photo, but shorter.
The third photo shows the diameter of the bell is 9". It remains to be seen what will happen when a flange that large is installed in the separation chamber. The benefits of the more efficient air entry could easily be lost if that large flare upsets the dynamics of the air flow in the chamber.
I'm going to start building my separator with the premise that a bell mouth exit port does not require any changes to the established configuration for a top hat design. I will experiment with the depth of penetration of the bell mouth into the separation chamber, although I'm not sure how I will know if one position is better than the other unless there are dramatic differences. I might even try straight pipe to establish a base condition, but again I don't have a clue how to measure it for comparison purposes.
(https://www.dropbox.com/s/evqd6854yedamfg/Bellmouth-02.jpg?raw=1)
(https://www.dropbox.com/s/fckb22la83bcuic/Bellmouth-04.jpg?raw=1)
(https://www.dropbox.com/s/yu8exfl7lx9r207/Bellmouth-03.jpg?raw=1)
If you have an ammeter, you could measure the current draw of the motor. If it is pulling more amps with the bell mouth, then it is moving more air. I'm not sure if the differences would show up in current draw, but worth a check...
Or, an inexpensive weather-vane anemometer may also work the trick.
Are you intending to install this inside your separator as the exit pipe? As I believe that the science behind the bell mouth would cause enough of a disturbance inside the vortex that the separator. I am planning on using a few of these as well on my current Thien Scroll Inlet project, only my attachments will be at the machine locations and the actual impeller itself. Keep us posted on how this works out for you. My project is two phased. One is a standard "Top Hat" Thien seperator. The second, is using a Scroll Inlet method to test on which theory will yield the most benefit. The Scroll is a rather unsightly looking beast sitting on top of the trash can but I am borrowing its science of increasing air velocity while maximizing the interior separation stability and efficiency of the vortex.
Quote from: Don_Z on August 14, 2011, 10:38:13 AM
Are you intending to install this inside your separator as the exit pipe?
Yes, my plan is to start out placing the bell mouth in the separation chamber in about the same position as I would a straight pipe. Then I'll experiment with higher and lower positions to see if anything changes. Of course, if I move it all the way to the top, then it becomes an exit port much like what you plan to put on each of your pieces of equipment. That would seem like the logical place to start, but I am assuming Phil has already tested straight pipe in that position, and if it didn't work for straight pipe, I doubt it will work for a bell mouth.
I don't think there is any question that the bell mouth is the best port configuration for the machinery end of the DC system. There you are trying to suck up maximum air and dirt. In the separator, you are trying to suck up maximum air, while leaving the dirt behind.
It is not out of the question that the inefficiency of a straight mouth pipe is the very reason the Thien separator works. Wouldn't that be a gas! (pun intended)
Quote from: phil (admin) on August 13, 2011, 06:58:54 PM
If you have an ammeter, you could measure the current draw of the motor. If it is pulling more amps with the bell mouth, then it is moving more air. I'm not sure if the differences would show up in current draw, but worth a check...
Or, an inexpensive weather-vane anemometer may also work the trick.
Phil,
My bellmouth separator is not quite ready for testing, but I thought I would run some tests on my existing system to see if an ammeter can detect differences in performance. I started with extreme conditions figuring if the ammeter couldn't differentiate between those, it had no chance of detecting differences in separator setups.
All readings below are out two decimal places because that is what my ammeter is capable of (whether it's that accurate or not is another story). Also, readings bounce around about +/- 0.03 amps during any test, so the numbers below are an average.
Deadheaded 15.29 (no blast gates open)
Wide Open 16.90 (only a 5 ft. hose attached to DC inlet)
Bandsaw 15.95 (my longest equipment run)
Sander 15.76 (long run with two smaller dust ports)
Floorsweep 16.29 (short run and 5" port)
Jointer 16.19 (medium distance with large port)
Planer 16.35 (shortest distance with large hood)
The only point to all these numbers is that it shows the ammeter is detecting differences in loads for each situation, and the numbers differ in the direction expected, i.e. the smallest load moves the most air and therefore draws more amps. Conversely, the biggest load (deadheaded) moves the least air and draws the fewest amps.
So, will the ammeter detect differences in separator setup conditions? Since there is only a 1.6 amp difference between the extreme conditions above, I am skeptical. However, I intend to give it a try.
I don't own an anemometer, so I can't do any wind speed testing. I'll check ebay and Craigs list to see if I can find one really cheap. It looks like new ones start at about $25 and go to three digits.
I wish Parts-Express had a brick-and-mortar store near by. I'd love to see how 4" dust collection pipe would connect to a 4" flared subwoofer port:
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=260-403 (http://www.parts-express.com/pe/showdetl.cfm?Partnumber=260-403)
Not bad for ~$3.
Andy
Quote from: AndyF on August 27, 2011, 06:28:08 PM
I wish Parts-Express had a brick-and-mortar store near by. I'd love to see how 4" dust collection pipe would connect to a 4" flared subwoofer port:
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=260-403 (http://www.parts-express.com/pe/showdetl.cfm?Partnumber=260-403)
Not bad for ~$3.
Andy
Great find Andy. I read some of the reviews on that website and one of them said it mates up to PVC drain pipe.
At $3 I'd buy it and ask questions later!
I really thought this article in the first post was great! It really showed that some systems cannot support 6" ducting and the bell mouth fitting idea is excellent.
I am in the process of building my 1.0 DC system with a delta 760-50 with 4" pipe and will be trying some of these flares to optimize my system.
:)
Quote from: Pillguy on August 28, 2011, 06:19:03 AM
I really thought this article in the first post was great! It really showed that some systems cannot support 6" ducting and the bell mouth fitting idea is excellent.
I am in the process of building my 1.0 DC system with a delta 760-50 with 4" pipe and will be trying some of these flares to optimize my system.
:)
I purchased my 50-760 largely on the basis of that review. Here is another excellent article that became my bible as I was building my DC system.
http://americanwoodworker.com/blogs/shop/archive/2010/10/25/central-dust-collection.aspx
Retired, looks great. I have experience with anemometers and would not vouch for cheap wind or sailing versions for ductwork measurements. I suggest building a simple orifice plate with water manometer readout (coffee is easier to see). Plus it will cost you nothing but time.
http://en.wikipedia.org/wiki/Orifice_plate
Quote from: plauale on August 30, 2011, 01:08:18 AM
Retired, looks great. I have experience with anemometers and would not vouch for cheap wind or sailing versions for ductwork measurements. I suggest building a simple orifice plate with water manometer readout (coffee is easier to see). Plus it will cost you nothing but time.
http://en.wikipedia.org/wiki/Orifice_plate
Thanks for the heads up. Since my previous post I started checking the specs on some of the "affordable" anemometers. I discovered that their maximum measurement is about half the velocity of a wide open DC, so at best they would provide an inaccurate reading, and at worst would self-destruct.
Quote from: retired2 on August 30, 2011, 07:34:43 AM
Quote from: plauale on August 30, 2011, 01:08:18 AM
Retired, looks great. I have experience with anemometers and would not vouch for cheap wind or sailing versions for ductwork measurements. I suggest building a simple orifice plate with water manometer readout (coffee is easier to see). Plus it will cost you nothing but time.
http://en.wikipedia.org/wiki/Orifice_plate
Thanks for the heads up. Since my previous post I started checking the specs on some of the "affordable" anemometers. I discovered that their maximum measurement is about half the velocity of a wide open DC, so at best they would provide an inaccurate reading, and at worst would self-destruct.
I use a Kestrel. I often up-size the pipe to the next size (so 2-1/2" gets adapted to a length of 4", a 4" would get adapted to a length of 6", etc.). I take my measurements at the end of the up-sized pipe. The Kestrel can handle something nuts like 7800-fpm. But switching to a larger section of pipe makes other tests I perform easier, too.
http://www.profblairandassociates.com/pdfs/RET_Bellmouth_Sept.pdf
Contains a CFD analysis of Bell openings that applies to this thread.
Quote from: rmitchell62 on December 18, 2012, 12:52:21 AM
http://www.profblairandassociates.com/pdfs/RET_Bellmouth_Sept.pdf
Contains a CFD analysis of Bell openings that applies to this thread.
Interesting article, even though it made my head hurt trying to understand it. What was interesting was the writer claims a mass flow rate improvement of 16% for a simple radius bell-mouth vs. a straight pipe. I looked back at the data I published when testing my separator build, and it shows a 13% improvement. Given the lack of sophistication in my testing procedures that is pretty close.
The article goes on to say that a bell-mouth with an elliptical radius would yield another 4% improvement. Unfortunately, I doubt there is a source anywhere for off the shelf elliptical radiused bell-mouths. So, that 4% will probably elude us wood butchers forever.
Never the less, the article provides more verification that all top hat separators should be built with a commercially available simple radius bell mouth on the outlet pipe - and even at the pick-up ports where possible.
Regards,
Retired2
@retired2
One other thing I believe was in this article was a description of the optimal ratio of dimensions (but I've been reading alot). If the pipe diameter is x then bell rim diameter is 2X and optimal length of pipe is also X. So if I plan an 8" wide pipe outlet (inlet to fan/vacuum) then the pipe should be 8" long. I haven't found any research on the distance below the bell, to the baffle. I know yours is about 3" or so but that seemed really low to me. I'm going to try 8" above the bell and >1x below. We will see.
Maybe we should bite the bullet and do a computational analysis of the design. Any current University Engineering students lurking?
.
Rob
Quote from: rmitchell62 on December 20, 2012, 05:46:16 AM
@retired2
One other thing I believe was in this article was a description of the optimal ratio of dimensions (but I've been reading alot). If the pipe diameter is x then bell rim diameter is 2X and optimal length of pipe is also X. So if I plan an 8" wide pipe outlet (inlet to fan/vacuum) then the pipe should be 8" long. I haven't found any research on the distance below the bell, to the baffle. I know yours is about 3" or so but that seemed really low to me. I'm going to try 8" above the bell and >1x below. We will see.
Maybe we should bite the bullet and do a computational analysis of the design. Any current University Engineering students lurking?
.
Rob
It makes sense that there is an optimal neck length for a bell mouth, but it never occurred to me at the time, so mine is arbitrary in length. Also, my air straightener probably extends into the optimal neck area, so that may also have some unknown effect on performance.
With regard to bell mouth position relative to the baffle, my air flow was highest when the bell mouth was closest to the baffle. That always puzzled me because that is not what I would have expectecd. I did not leave it there simply because Phil's testing suggests the position to be 1/2 the pipe diameter.
It's been 30 years since I've done any serious engineering, and that was fluid flow, so any analysis beyond elementary basics is over my head! The real question is whether it is worth it? I don't doubt that a few more cfm's could be squeezed out of the typical separator, but I suspect most DIY dust collection systems have other "low hanging fruit" that would produce bigger gains.
Youtube DIY link on how to make a flared port tube:
http://www.youtube.com/watch?v=fzdnlBkG2mI
Not sure about HSE practices (electrics and H2O!) but the technique looks fairly simple. Might give this a go on my 150mm (6in) connector pipe between separator and impeller housing.
Any suggestions on how to square off a large diameter pipe? I received a rough hand sawn length of pipe which needs squaring first. I'm even struggling to draw a straight line around the pipe so I can cut and file to the line.
Thanks.
if u need to draw a line around a pipe , get a long lenght of paper with a straight edge and wrap it @ the pipe at least twice and keep the st edge straight is you go around the pipe then just mark the line and cut