Hello. I recently discovered this forum while researching my own DC system. Greatly admire Mr. Thien for his invention, for sharing it, and for maintaining such a high-caliber forum.
Having read several of the threads on this forum regarding what everyone calls "air straighteners", I just wanted to offer a link to some interesting research. The paper offers both experimental validation and design guidelines for the use of Deswirl Devices in Cyclonic DCs. I am in no way connected with the researchers; I just wanted to add to the knowledge base this site offers, as well as offer kudos to whoever brought the subject of vortex straightening to the forum. Lots of smart people here!
***
Well, I tried to post a link to the PDF file of the research paper above, but the forum software said "no thanks." If a moderator wants to correct it, here's the workaround version: www.thaiscience.info/journals/Article/Application%20of%20deswirl%20device%20in%20cyclone%20dust%20separator.pdf
The two main points I took away from the paper are: 1) a three-vaned device provides the most improvement (think peace-sign-shaped cross section) and 2) a short device is quite effective (in other words, the device doesn't have to extend very far into the outlet duct).
If this paper, or similar research has already been cited here, I apologize for the repetition. If not, I hope it is useful.
Thanks again for the information and discussion this forum offers.
Quote from: allenmck on January 07, 2014, 07:27:33 PM
Well, I tried to post a link to the PDF file of the research paper above, but the forum software said "no thanks." If a moderator wants to correct it, here's the workaround version: www.thaiscience.info/journals/Article/Application%20of%20deswirl%20device%20in%20cyclone%20dust%20separator.pdf
Sorry, that is a method to stop spammers, no links by first-time posters. I edited the link in your post and it should work now.
The # of spammers is just unbelievable. I spend all my time editing-out carp posts and removing spamming accounts.
Thanks for fixing the link. I figured it was an anti-spam technique.
And thanks again for all the hard work it takes to maintain this site.
BTW, I forgot to mention that the research paper also experimentally validates the "bellmouth" outlet modification (referred to in the paper as "streamlined entry").
Both modifications seem to already be accepted here. But I figured it was still pretty interesting that such a specialized topic had actually been the subject of published scientific research.
Very interesting paper, thanks for posting it.
One question I have concerns having a top hat before the DC. If the direction of the swirl or vortex coming out of the top hat is the same as the direction of the impeller in the DC, isn't that a good thing?
There is no question (at least, in my mind ;) ) that if the direction of the impeller is opposite of that of the vortex coming from the top hat, then energy must be wasted reversing this flow. However, if the vortex is straightened out, instead of matching the impeller, doesn't more energy have to be expended by the impeller to move it through the DC?
It seems to me that if you match the direction of the air swirl from the top hat to the DC, a straightener in between would only cause more energy to be wasted.
Quote from: revwarguy on January 08, 2014, 08:06:24 AM
Very interesting paper, thanks for posting it.
One question I have concerns having a top hat before the DC. If the direction of the swirl or vortex coming out of the top hat is the same as the direction of the impeller in the DC, isn't that a good thing?
There is no question (at least, in my mind ;) ) that if the direction of the impeller is opposite of that of the vortex coming from the top hat, then energy must be wasted reversing this flow. However, if the vortex is straightened out, instead of matching the impeller, doesn't more energy have to be expended by the impeller to move it through the DC?
It seems to me that if you match the direction of the air swirl from the top hat to the DC, a straightener in between would only cause more energy to be wasted.
The short answer is a pre-rotation of air entering a fan adversely affects performance regardless of the direction of rotation. I've posted this information many times, but here it is again. It is from a Cincinnati Fan Engineering Manual.
Duct Inlet Spin
"A major cause of reduced fan performance is an inlet duct connection that produces a spin or pre-rotation of the air entering the fan inlet. Inlet spin in the same direction of the fan wheel will reduce air volume and pressure ratings. Inlet spin in the opposite direction of the fan rotation will substantially increase the motor horsepower requirements. An ideal inlet condition is one which allows the air to enter the fan axially and evenly without spin in either direction."
Thanks, retired2. I have read it before as well, but I do not understand why "Inlet spin in the same direction of the fan wheel will reduce air volume and pressure ratings."
I can believe that the air pressure between the straightener and the impeller is lower without the straightener is lower than with it, but it seems like the pressure difference between the other sides of the impeller and the straightener would be the same with or without it. It seems to me that ultimately the mass/energy calculations here would show energy being expended by the air against the straightener (lost energy) and then more energy being expended by the impeller against the air, all to get the air moving in the same direction again.
I understand this is all theoretical, but if it has been experimentally determined that a straightener helps even when the direction is the same to any reasonable degree, then I'll add one just to see.
Quote from: revwarguy on January 08, 2014, 02:12:50 PM
Thanks, retired2. I have read it before as well, but I do not understand why "Inlet spin in the same direction of the fan wheel will reduce air volume and pressure ratings."
I can believe that the air pressure between the straightener and the impeller is lower without the straightener is lower than with it, but it seems like the pressure difference between the other sides of the impeller and the straightener would be the same with or without it. It seems to me that ultimately the mass/energy calculations here would show energy being expended by the air against the straightener (lost energy) and then more energy being expended by the impeller against the air, all to get the air moving in the same direction again.
I understand this is all theoretical, but if it has been experimentally determined that a straightener helps even when the direction is the same to any reasonable degree, then I'll add one just to see.
Unfortunately, none of the engineering documents I've seen on the subject of pre-rotation do much to explain the "why" of the phenomenon, but there are plenty of sources available on this subject and they all say the same thing, i.e. pre-rotation in either direction has an adverse effect on fan performance.
If the blower impeller spins at 1800 RPM and the air entering the blower inlet is also spinning at 1800 RPM and in the same direction, then the blower would move no air. If the air is spinning at 900 RPM the blower will move some air, but not as much as when the air entering the blower is not spinning at all. Air entering the blower that is spinning in the direction opposite that of the impeller is bad, as you would expect.
Adding a straightener will introduce some pressure drop, but it is far less significant than the benefit from increasing fan output.
Try searching Google Images for "inlet guide vane" and you will see many examples of devices that manipulate the pre-rotation of air entering fans.
Thaiscience.info seems to be overloaded at the moment :(
There was a thread regarding CW vs. CCW last year that covered this issue.
retired2 is right. I think he is the one who introduced the subject to this forum in the first place(?), which was a really insightful modification. I was gonna ask if he was an engineer by training, or just a smart civilian. :) It was also impressive that he did his own experimental testing and shared the results. IMHO advice is worth ten times as much when it comes from first-hand observation rather than "I read somewhere."
Regarding why the vortex spinning in the same direction also causes a performance hit, I read somewhere ;D (and this is a rough paraphrase) that it is due to the impellers not being able to impart as much force to the fluid. I kind of pictured it like this: when you throw a baseball TO a batter, he can crack the h--- out of it. But if you imagine a stream of baseballs coming from behind a batter, and the batter trying to hit one hard enough to speed it up, it seems like the timing of the swing would be impossible to get right. A lot of whiffs vs. a good, solid connect.
That might be total BS, but at least I could stop wondering and use the information, because the results were pretty impressive. 19% improvement (don't know if it is a 1:1 improvement in collection, or a 19% improvement in an aspect of the process). But the two mods are such easy things to do, why not give it a shot?
I'm always open to being proved wrong (kind of the point of experimenting, after all), but the explanation about the impellers not being able to move the air at all if it is spinning in the same direction doesn't seem right. That would kind of stall the whole system, right?
Anyway, Retired2 and many others have brought in a lot of useful, creative, information. And that seems to be a big motivating factor here: building your own system that is equal to, or even better than, commercial offerings. Wringing every bit of efficiency out of the system can only help. It does make you wonder how much research the commercial manufacturers have done. Or whether other considerations get priority over what seems like cheap and simple ways to push efficiency as high as possible?
When I get a chance I want to start a thread about how the super high-end DC systems (like Felder's top line) take collection to the current highest level. I'm pretty sure I know how they do it. But there are a couple of practical aspects I'm betting guys like Retired2 can figure out no problem. :)
allenmck,
I once replied to one of "vodkaman's" posts with the following response. It describes who I am.
"Vodkaman, your comments make me chuckle because they put you among a group of people in this world, me included, who have a compulsive need to understand how everything works whether we truly need to know or not, and we are always on a quest to make them work better whether they need to or not!
I don't know how many times in my life a new aquaintence has ask me "Are you an engineer?" Most of the time it is asked in a way that is not necessarily intended to be a compliment. It is more like "If you are an engineer that would explain your nutty obsessive complusive behavior and your need to fix things that ain't broke!"
I had theorized the reason straight air was always preferable was that the individual vanes on the blower wheel would always be evenly loaded. I had imagined the key was that each vane was allowed an opportunity to carry its burden on each rotation.
Air spinning in the same direction as the wheel probably would try to load one area of the blower more than others, and air spinning the opposite direction would probably do the same and also cause the wheel to have to overcome the inertia.
But I'll admit that I always sort of hoped that Retired2 would just tell us why.
;D
Quote from: phil (admin) on January 09, 2014, 08:02:22 PM
I had theorized the reason straight air was always preferable was that the individual vanes on the blower wheel would always be evenly loaded. I had imagined the key was that each vane was allowed an opportunity to carry its burden on each rotation.
Air spinning in the same direction as the wheel probably would try to load one area of the blower more than others, and air spinning the opposite direction would probably do the same and also cause the wheel to have to overcome the inertia.
But I'll admit that I always sort of hoped that Retired2 would just tell us why.
;D
Phil, do you know what a SWAG is? If not, it's an engineering acronym for an answer that lacks absolute certainty. So, here's my SWAG.
I believe the reason pre-rotation effects fan performance is simply because it changes the angle of air impingement on the impeller blades, and consequently the flow resistance. In one case that resistance is lower than the optimum fan design, and in the other case it is higher. The fan is designed and optimized for incoming air which has no spin, so any deviation from that ideal will be detrimental.
So, what do you think, Phil? Uh, before you answer, give me a break, I haven't been a practicing engineer for nearly 30 years! :)
So, absent a deswirler, is there a better side of the hat to make the inlet? Clockwise or counter clockwise?
Quote from: Jack on January 10, 2014, 08:01:41 AM
So, absent a deswirler, is there a better side of the hat to make the inlet? Clockwise or counter clockwise?
Maybe, but I can only speak to my experience with counter-rotation. My test data shows the air-straightener provided approximately 50 cfm more at a reduced amperage. Normally amperage drops when less air is being moved, but in my case I got more for less, so clearly the air straightener was justified.
I have no way of knowing what the penalty might have been if my pre-rotation had been in the same direction as the fan, but that is the direction that the experts seem to suggest is worse. Their comment is always that this pre-rotation moves the fan curve down and to the left. Simply put, that means lower cfm and less SP. They don't say by how much probably because there are too many variables.
Okay, and the commentary on the use of turning vanes at the entry to the Thien chamber, especially w/ regard to the top-entry situations? These are worth a 50% reduction in SP drop at elbows.
Likewise, what about the use of stream vortex inducers just inside the feed duct where it dumps into the Thien chamber? Tuning of this twisted jet of air would help compress it and thereby accelerate it; and it's rotation could be oriented to get it to hug the wall of the chamber at the bottom where the dust needs to be deposited?
--Cheers, Bradley
Quote from: bpotts on January 10, 2014, 05:18:33 PM
Okay, and the commentary on the use of turning vanes at the entry to the Thien chamber, especially w/ regard to the top-entry situations? These are worth a 50% reduction in SP drop at elbows.
Likewise, what about the use of stream vortex inducers just inside the feed duct where it dumps into the Thien chamber? Tuning of this twisted jet of air would help compress it and thereby accelerate it; and it's rotation could be oriented to get it to hug the wall of the chamber at the bottom where the dust needs to be deposited?
--Cheers, Bradley
I don't think air straighteners would have any benefit if used to condition the separator inlet air. That's just my opinion, but only a test with measurements would tell for sure.
I do agree with your comment about pushing the waste stream to the outside wall. That is why I have been an advocate of rectangular inlets. A good round to rectangular transition piece will put the waste stream over the drop slot where it needs to be without the aid of turning vanes. But I would not suggest "necking it down" to accelerate the air. That just adds losses to the system.
I have recently come across a problem with my air straighteners. The bin overfilled causing the waste to travel straight across the top of the baffle. Result, long strings of shavings from the lathe fouled on the straighteners and the fan inlet pipe to the filter was completely filled with this waste. It did not stop the system from sucking but did considerably reduce the effectiveness of the whole set up. Obvious symptom was fine dust reaching the DC bag because the separator had not removed it. Just shows how effective the baffle is. Moral, make sure the container does not fill more than two thirds full preferably by having a sight window in the can or some other device that makes checking this easy.
My straighteners are used baked tin cans with their bottoms removed and the sides of the can then bent in to give a series of tubes about 1 1/4" diameter. Not pretty but it improved performance considerably.
Hoping to clear my backlog of furniture projects so I can build a top hat designed to allow me to do some performance testing with various layouts.
Sorry retired2 for getting mixed up between retired and senior. I blame being very senior for my confusion. It is either that or too much beer.
Quote from: BernardNaish on January 22, 2014, 04:09:33 PM
Sorry retired2 for getting mixed up between retired and senior. I blame being very senior for my confusion. It is either that or too much beer.
Ii is definitely not too much beer!
Quote from: allenmck on January 07, 2014, 07:27:33 PM
The two main points I took away from the paper are: 1) a three-vaned device provides the most improvement (think peace-sign-shaped cross section) and 2) a short device is quite effective (in other words, the device doesn't have to extend very far into the outlet duct).
Took me a lot of reading threads to find this tidbit. After looking at all the tube type straighteners I thought there had to be an easier/better way. Of course this article did not test the full small tube type, but i believe I will attempt my build with the 3-vane device, and also with a "streamlined entry".
New to the forum and I am planning on building my own top hat. My problem is the whitepaper in the first entry in this thread is a broken link. Does someone have a good link for that document? It sounds like there is some good info there. Thanks.
Just wanting to check again.... Does anyone have a copy of this article? The link lists it as a PDF. I'm hoping that someone has a local copy of it when they downloaded it to read it. All references I can find to it on the web are broken links.
Thanks for your help.
http://www.researchgate.net/profile/MZ_Abdullah/publication/267700921_APPLICATION_OF_DESWIRL_DEVICE_IN_CYCLONE_DUST_SEPARATOR/links/54b1ff2a0cf28ebe92e1916e.pdf
im not sure if this is the same article but i think it might be
Thanks!. That's some deep stuff. I see where the idea of the bell mouth came from but it makes me question the use of a bunch of plastic tubes as a de-swirl device. If I read that correctly he had better results with the three bladed type then he had with the four bladed. I think that stand to reason that moving from 3 or 4 sections up to around 20 separate sections may be massive overkill. What do you guys think?
Quote from: Lance on December 29, 2015, 03:02:12 PM
Thanks!. That's some deep stuff. I see where the idea of the bell mouth came from but it makes me question the use of a bunch of plastic tubes as a de-swirl device. If I read that correctly he had better results with the three bladed type then he had with the four bladed. I think that stand to reason that moving from 3 or 4 sections up to around 20 separate sections may be massive overkill. What do you guys think?
I don't see any reason not to accept the content of that white paper. I don't recall anyone here stating the tubes are better, but I will say with certainty they are better than nothing in many cases.
I have been thinking about it and right now my plan is going to be to add a three veined piece of aluminum.
My opinion and I believe you will agree - more surface area more friction. To my thought, tubes, better than nothing - even though they contribute to a lot of area they are helping to guide the flow and somewhat overcoming themselves with better positives than negatives in the scheme of things.
Less is more. Straight vanes, really should shine - accepting the limitations of what it is they are installed in. Like water in a pipe, if it leaves to enter a rocky bottom or a variable channel, then it becomes more turbulent anyway. So into the impeller the end is a "known qualified object" unlike leaving into a specific can versus a generalization.
OK, I am going to try to explain what "I" see without the math and theory stuff. Bear with me and feel free to correct me. Some of this is feel based on experience so correct me if you see a fail.
Picture using horizontal [in relationship to the outer wall if the can is vertical] baffle plates inserted into your round or rectangular intakeS and outletS will be better. Mainly due to the pressure differentials experienced. These are exhibited due to flow changes of direction, the spiral rotational flow [caused by boundaries, geometry, friction, impeller] and the cumulative effects of friction - and thus mass not volume of air - will thus be smoothed [better smoothed - nothing is always absolute] to the OUTER wall of your can. Thus allowing the fines to gather on the outer wall areas. There is more mass in parts of the compressed stream due to flow. In a straight out centrifugal flow thee is more pressure/compression and thus mass in the outer areas - right?
If other straighteners are used [still better than none] they may induce fines to gather in the individual streams rather than propagate fast enough to the outer wall - perhaps not as a smoothly flowing stream but like the gulf stream in the ocean. They may be inclined to follow their own individual streams on the inside of the entire combined streams rather than fling to the outer wall pushing through the other pressure/velocity barriers. Thus requiring more time to fall through an all ready flowing and "somewhat laminar" flow stream wall. So some "air streams or flows" may contain more mass than others and there is a boundary that is harder to penetrate for the lighter loads. The best analogy I can think of is an "air curtain" such as used in warehouses and places like Home Depot to keep either cold air out or hot air out [works for a lot of bugs too]. They work because the form a "boundary wall" of fast air that is hard to penetrate. If you allow your fines to enter the inside diameter of your "can" stream by creating a curtain wall of several separate and differing velocities you have created [if you can] fast laminar flow air tubes. Laminar flow is talked about but very hard to achieve in these situations. These streams will induce spirals in each of the other individual streams boundaries due to the change in air speed as they round the can - also friction between the "tubes of air" merging once they exit the straighteners and thus cause spiraling and turbulence within the individual streams and the stream as a unit. Thus these could create a pressure and speed gradient boundary they must overcome to get to the outer wall. Why significant? Depends on the turbulence, the guide, the time-to-distance factor in traveling to their destination and the time to fall out of the stream due to mostly - gravity. So the gravity acceleration/velocity factor over distance - trajectories come to play too you would think - would you not?
So using horizontal "flow straighteners" in your intake stream should - theoretically - allow the flows to be more "even" across their horizontal gradient while they travel through the horizontal plates and some of the exit boundaries. This should allow the fines to better penetrate a pressure and speed boundary they already find themselves in - no air curtain is present to overcome as they are already IN the curtain.
This may prove useful EXCEPT in the impeller intake where some form of Extended "Y" allows for proper loading of the impeller by presenting somewhat even flow to all of the blade centers - someone else mentioned this before. Of course we can then talk about the off center fan housing - radius of the impeller versus off center radius of the housing and the pressure fields generated about the housing - Is the air pushed or pressured through the fan outlet or both [both most are active]. "My" biggest concern about flow straighteners is that they have TIME [the old trajectory thingy for me] to smooth the flow, TIME for all fines to migrate out before additional turbulence or frictional component is encountered, before they end the limited travel distance and end up around the inside of the can and attempt to enter another intake stream - effectively pushing them into the exhaust due to less pressure/flow to move out and more pressure/flow encouragement to move in to the exit. This is why [I assume] most have followed the guidelines Phil has found through experimentation [are you really a Physicist Phil?] for tube below/above base plate rules. Once air can bleed out of the stream and head for the exit [exhaust opening or fan intake on a top hat] the fines will no longer wish to run to the outside fighting the stream, it's pressure/velocity front, and air flow headed to the exit [analogy of the air curtain once again seems to somewhat apply] - they will follow the exit flow - path of least resistance.
I hope that somewhat makes sense and apologize if I wasted time. I may be just full of it? Sorry for the long windiness but I do not know how else to cover it. Per all my posts, depends on you config and your scale, your "significance factor" as yours is built. Wow, I may have to limit my self from now on. It's like writing a technical paper for me, I like to somewhat cover the subject from my viewpoint with as little misconception as possible.
Moving on, this leads to some experimenting I have done with the exit pipe position and configuration. That's another item and still in test.
Here's a spiral design I am tinkering with on the bench - it is in pieces right at the moment - yes, friction is somewhat "area compensated", but as above stated, straightening flow also reduces friction and turbulence. So where does it end?
I am previewing my post but the picture jpg will not post here (?) and I do not have an external account I can use for forums at this time. I will work on that.
FYI - if you wish me to stop and be quiet - I can do that too. No offense taken.
I note that the paper concluded that a 3-vane setup with "streamlined" input (called a "bellmouth" in this forum) and a "central body" ("Type E1" in its Figure 3 (https://www.dropbox.com/s/wzy76wk4pmrxrzn/ApplicationOfDeswirlDeviceInCycloneDustSeparator_AbdullahHusainFraser_2003_Figure3.jpg?dl=0)) was superior to 3-vane without central body (Type C1) because it completely killed the "swirl velocity component". The "central body" is a cylinder, small in diameter compared to the output tube, axially centered in the output tube, with a rounded leading/entry/downward-pointing end and long-tapered trailing/exit/upward-pointing end -- similar to the nacelle of modern wind turbines. BTW, I was able to find the paper/PDF with a web search for its title, "Application Of Deswirl Device In Cyclone Dust Separator".
I agree about his finding with the Type E1 (3-vaned Bellmouth and central body) being his best found method of stopping swirl velocity but I would also question the loss of air volume by decreasing the overall size of the pipe by removing the center portion. Is the extra bit of swirl cancelation with the loss of volume in the exit pipe of the separator.
I found the part in the white paper that said good things about the three vane design. Has anyone found a reference as to how long the vanes should be? If I can't find any documentation, I think I'm going to try vane length equal to duct diameter. Also, do you think there would be any advantage to putting vanes both at the output of the separator and at the input to the impeller housing?