J. Phil Thien's Projects

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? 

Quote from: bumpnstump on August 03, 2011, 07:50:19 PM
Not sure I'm smart enough to understand all of the science of this, but here's what I did:
-I built my separator using a scroll inlet, allowing for a 4" round inlet on the side, 2" offset scroll, with a 4" outlet in the top.  With pipe hooked straight up to the 4" round inlet, I was getting lots of bypass.
-Then, after reading thru some of Bill Pentz's articles, I decided to build a round-to-rectangular transition piece.  Following his instructions, I built a 12" long x 4" round to (in my case) a 2 3/4"w x 4 5/8" h rectangular transition.
-After installing and using it, there is very little-to-no bypass.
I'm not even sure what/if I did correct, only that it made a big difference.
Hope this is helpful.
Thanks,
Rick

Thanks Rick.  Your comments are indeed helpful.  They confirm that a rectangular inlet, higher than it is wide, is the probably the best inlet port.  The discussion about eccentric vs concentric transition might have been a little hard to follow since I had a bad link to my sketch.  I've corrected that and now you can see what I was talking about above.  Was your transition piece eccentric or concentric?

Also, if you don't mind, what was the inside diameter of your separator?  I'm building a 5" inlet/outlet separator and I was planning to put it on a 20 gal Brute, but Phil was a little concerned that the resulting ID might cause some bypass.  I am now planning to use a Brute 33 gal, but that makes everything a little larger than I wanted.

Thanks

Quote from: RCOX on August 03, 2011, 12:46:44 AM
Vodkaman and Retired2: I found the "noise problem" today. The duct from the impeller to the filter/catch bags was vibrating with the extra airflow. I placed a clamp across the width of said duct and it stopped. One problem solved.

I also found that my separator works better with the baffle about 1" below the inlet pipe.To do this, I had to move the baffle to 6 3/4" below the top, then I re-installed the outlet tube so it was 5" above the baffle. Noticeable improvement. Then I discovered that if I opened another gate, as was suggested, that suction reduced severely. My question, was it cfm that I lost or just velocity? With a second gate open, on the other line, I was still able to pick up the chips that I used as a test medium but at a lot slower pace. With only the 1 line open, the transfer was a lot faster and it appeared to clean the fine dust from the flex hose a lot better. All testing has been done through the following, 10' of 4" flex, blast gate, 25' pvc into the 6" wye then the separator and finally the dc.

Now, here is another observation/question. My inlet to the separator is a 4"X10"X6" end register, torpedo style. As I stated in an earlier post, placement of the baffle in relation to the inlet has a great impact on what the dust does in the separator.  I thought that if you placed more of the wide part of the baffle in front of the inlet that you should get better separation. Wrong. It appears that if the end portion of the slot is too far forward, you get more up swirl and more chips re-entering the airstream. When you place the end of the slot almost even with the beginning of the entry you get more suction and better separation, the chips/dust also has a flatter swirl in the separator. With the brute trash can I am using, you get a little flexing/collapsing of the side of the can opposite the entry point when you place the baffle with more than half of it in front of the entry. With the end of the slot even with the leading edge of the entry I will collapse the can at least a third of the diameter. In this position the baffle actually prevents the dust from rising into the incoming airstream. My question, is this normal, how it was originally designed to work or am I just trying to over analyze this whole procedure?

Enough of my rambling tonight. My brain hurts, I am going to bed.

Rcox,

With regard to cfm vs fpm, they are tied together mathmatically such that for any given size pipe a decrease in one results in a decrease in the other.  Likewise an increase in one results in an increase in the other.  Now if your comparisonis before and after a change in pipe sizes, that's a different story.

With regards to the baffle slot and the outlet pipe, there seem to be two well proven relationships.  The recommended slot position is spelled out in numerous posts.  The easiest to find is pitbull's "6" inlet/outlet how to" thread.  The most recommended outlet pipe position is to set it about 1/2 the pipe diameter below the separator top.  After taking a closer look at your inlet design, I can see why you would have a problem with the outlet recommendation.

After paying more attention to your inlet design, I really think it may be hurting the performance of your separator.  Your aspect ratio is just the opposite of what I think it should be for a rectangular inlet.  Yours is short and wide rather than tall and narrow.  That could be causing several problems.  There could be short-circuiting, i.e. waste stream getting to the outlet without being spun.  It is also likely to be causing more impingement turbulance as described in pitbull's test video, and the short body is affecting the desired position of the outlet nozzle.  I think the Brute trash can is better suited for a true top-hat design rather than a tangential internal separator.

Quote from: Don_Z on August 02, 2011, 09:20:56 PM
The Rectangular port is what is connecting to the actual separator. A round inlet causes the depression/ turbulence.

I agree, but which of the two transition pieces I've drawn do you think would be better.  Both sketches go from round to rectangular where it connects to the separator.  I first thought the eccentric version might be better, but I'm no longer sure.   The reality is it might not make any difference which configuration I use.

Quote from: RCOX on July 29, 2011, 05:24:24 AM
Vodkaman: It happens with 1 gate open on each line. I did not notice it with 2 gates on the same line. Don't know if it has to do with air volume or what. It may have done it before but I never noticed it. I just thought it odd that when I opened the second line that I noticed this sound. The only difference I can explain is that the extra volume of air with 2 lines open. The problem with this thought is that this dc is designed for more air than what I have allowed it. More experiments in the works for me.

Rcox,

I just found an engineering data document on line from a prominent fan/blower manufacturer.  There are a number of duct installation guidelines that are useful.  One in particular may have significant bearing on your problem, particularly your vibration.

The guideline states that an elbow too close to the fan inlet reduces the fan's air performance due to turbulance.  If the pipe system has no turning vanes, there should be a minimum of straight pipe equal to three wheel diameters between the fan inlet and the elbow. 

When I read this guideline I immediately thought about your first photo with what looks like a short radius elbow attached directly to the fan inlet, or if not, very close to it.

Quote from: Don_Z on August 02, 2011, 05:37:51 PM
Actually the eccentric model is the most efficient in regards to the inlet. Because here is what happens, when the incoming air is introduced into the separator with a round inlet there is something you may have not noticed. Look at the videos of Pitbull's 6" inlet/outlet that he has posted that show him demonstrate his inlet. You will see a "before" video with his inlet protruding just a bit inside the separator and as he feeds it dust, you will see the turbulence as the debris is hitting the incoming air and bouncing back through the inlet. Then watch his "after" video and you will see where he tries to cut back on the inlet material, but again, during his demonstration, there is still the turbulence. The factors involved were twofold.
First, the inlet material was protruding inside the separator causing the air and debris that is being "cycloned" to smash against it.
Second, because of the round inlet being introduced into a "cyclone" environment the outside 2 corners facing the opposite side of the inlet are causing a depression in the cyclone action.
With a eccentric transition into the inlet it prepares the incoming air for the cyclone that is happening inside of the separator by removing the corners and therefore removing the depressions in the cyclone. One of the main reasons I feel that a "scroll" method would benefit in this is because it helps introduce the incoming air and debris into the cyclone with the least amount of turbulence and therefore enhancing separation. My 2 cents.

Don,

Are we talking about the same thing?  Here is sketch of what I meant by eccentric transition.  The eccentric version has the centerline for the rectangular end offset from the circular end so that the outside wall is on the same plane.  The concentric version has a single centerline with equally converging sides.

Quote from: Don_Z on August 02, 2011, 03:57:54 PM
I am thinking that the "scroll" airflow concentration will aid in reducing the turbulence inside the separator and therefore aiding the separation process and therefore reducing any scrubbing that may occur. I believe the square inlet would be the optimum choice but from what I have seen available, the typical HVAC transition that you will find at the HD or Lowes will be sufficient. I think I will start on "Project: Thien Scroll" tomorrow. The worst that can happen is I am out $30.00 in material and a day's worth of labor because it didn't work. But if it works...

Don,

I'm sure your build will work.  The question is how will we know if it worked better or not?  Even building two slightly different versions, and testing them side by side, might not reveal any differences.  Nevertheless, I'm very intersted in hearing and seeing the outcome of your effort.

Once I get started I'll post my progress as well.  I'm still leaning toward a concentric separator body with a rectangular inlet.  I will probably pay to have a transition piece fabricated so that I have full control over the shape and dimensions.  I am going to transition from 5" round to approximately 6" x 3.5".  The rectangular area is slightly larger simply to compensate for rectangular duct being less efficient than round.

The transition piece design leads to another question.  Will an eccentric transition work better than a concentric one, i.e. will it do a better job of prepping the air stream by moving the waste against the outside wall as it enters the separator?  Phil seemed to think a correctly placed small turn just prior to entry could be beneficial.  Would an eccentric transition provide that slight turn?  Unfortunately, this is another question for which there is likely no verifiable answer! 

Quote from: Don_Z on July 31, 2011, 07:53:01 AM
I too have been holding off on my own Tophat fabtrication until a little more data/science was introduced. There are some really fantastic builds posted on these pages but I have noticed a few things. There are quite a few that have taken some liberties in design changes to the Thien theory and not because of some new and relevant data, but more out of garage hackery (no offence). I am very interested with this introduction of the "scroll" in the conversation as that it has a history of applied mechanics and science that can back it up. Having just received my new DC unit and Wynn filters, I am not in a big hurry to start slapping something together when a little more research will give me better results with premium performance and a superior build.I have worked in the aerospace research and development industry prior to becoming a professional woodworker and rather enjoy gaining higher performance and results from my equipment and machinery rather than settle for the terrible status-quo that one must suffer from the consumer market.

Regards,
Don

Don, I think you and I are of the same mind regarding the fabrication of a separator.  However, my DC project has progressed from piping up the most offensive pieces of equipment to optimizing the DC itself - that means building a separator, adding a Wynn filter, and enclosing the DC for noise abatement.

I don't know a whole lot about cyclones, and even less about the engneering theory behind their design, but I am starting to have reservations about whether the "scroll" entry is applicable to a Thien separator.  I am even starting to worry that it might hurt the performance. 

Here's the difference that concerns me.  A cyclone has a long body that is totally concentric, either cylindrical or conical, with the exception of the top where the entry scroll makes it eccentric.  Due to the helical air flow, the air entering a cyclone quickly moves down from the eccentric part of the body to the concentric portion.   The Thien separator, by contrast, has a body length that is not much longer than the height of the inlet port. So, if the separator is built with a scroll, the air must continue to circulate in an eccentric pattern.  That just does not strike me as a more efficient design.

So, unless someone produces some data to suggest a "scroll" improves the performance of a Thien separator, I'm going to build mine with the conventional shape.  I'm still thinking about a rectangular entry port, but I'm not sure I want to spend the money to have a custom transition piece fabricated.

Quote from: bill70j on May 24, 2011, 07:53:52 AM
In industry we always design the inlets as rectangular.  We also offset the inlet from the body of the cyclone to improve efficiency.  In other words, rather than having the outer wall of the inlet ductwork in the same vertical plane as the cyclone wall, we offset it by a certain amount. This is called a scroll, which helps the dust particles stay closer to the wall and keep them from impinging on, and escaping through the outlet tube. 

The pictures below show what I am trying to say and give an idea of how a scroll will help separation efficiency.  The inlet ductwork starts with an offset, but gradually returns to the same dimension as the cyclone wall through the first 180 deg of sweep.

I built my separator with a top inlet, but am now redoing it with a rectangular side inlet to include the scroll feature.

Bill, 

I bookmarked your message some time ago when I wasn't quite ready to start my own separator.  I was intrigued by your comments regarding the rectangular inlet and the scroll feature.  I have a bunch of questions if you don't mind.

Have you started rebuilding your separator? 

Did you incorporate the scroll concept? 

Did you use a rectangular inlet?

How did you determine the aspect ratio of the rectangular inlet?

And now for the big question.  How is it working?

Thanks.

Quote from: phil (admin) on July 29, 2011, 08:11:48 AM

Quote from: retired2 on July 28, 2011, 09:06:35 PM
So, the question is, based on your experience and testing, what comments or suggestions can you make on diameter?

16" and even 17" are pretty small for 5" pipes.  If it were a typical cyclone, you'd get great separation.  But in one of my designs, I think you might get some short circuiting (where incoming debris goes right to the exit port).

Thanks Phil, you're answers are helpful even though different answers would have made my task a little easier.

Two changes I plan to make:

1) I have been debating about ditching the Delta 50-760 wheeled cart - it imposes too many design and construction limitations, and my DC will be stationary anyway.  By building my own support, I will have more freedom on the sizing of the separator.

2) I will replace my 20 gal garbage can with one that has a larger top diameter.  I hope a 30 gal can is large enough because I don't want to handle the weight of something larger.

Thanks again.   

Phil, I've searched the forum for threads discussing separator diameter, and was not able to find much specific guidance on minimum or maximum size, probably because most people simply let their container determine the diameter.

I have a Delta 50-760 and I plan to build a top hat design with 5" inlet and 5" outlet".  I might build a rectangular inlet based on the previous discussion we had.  I currently have a Brute 20 gal plastic garbage that I plan to use with the separator.  The inside diameter is 17-7/8".

My first thought was to mount the top hat right against the underside of the DC.  However, I quickly discovered that would limit my separator diameter to 16", or possibly less, due to the adjacent flange on the chip bag.

If I drop the separator down, I can build it with a diameter of about 17"(to match the 20 gal Brute).  That's not a great deal larger, but to go bigger I would need a 30 gal can.

So, the question is, based on your experience and testing, what comments or suggestions can you make on diameter?

Quote from: RCOX on July 28, 2011, 04:53:01 PM

Retired2: the black cheap flex that Chuck is referring to makes the clear premium flex smooth as a babies bottom. I have some of both and there is a very noticeable difference. To give you a comparison the black flex I bought was about $23.00 for 20' while the clear was about $30.00 for 10'. I will admit you get what you pay for with this stuff.

Glad I never bought any of the cheap black stuff!! 

Here's a revealing quote from Wynn Environmental's web-site regarding their clear flex hose.  "All of our hose is what the industry refers to as "Smooth Bore Flexible Hose".  We are not big fans of this term, because it is misleading,.."

Quote from: Chuck Lenz on July 28, 2011, 12:41:27 PM

Quote from: retired2 on July 28, 2011, 09:07:46 AM

Rcox, published friction losses for flex hose states that it is 3 to 4 times greater than smooth pipe.  So, to put that in perspective, if you have 5 foot of flex hose, it is the equivlent of 15 or 20 feet of smooth pipe. 




Hope this helps.  Regards.

Retired, you don't say if thats the generic black plastic inexpensive DC flex, or smooth interior wall flex, I'm guessing that those losses are from the inexpensive DC flex.

Chuck, none of the documents where I found information regarding line losses for flex hose make any distinction between type of hose.  I don't own any "generic inexpensive" black hose, so I don't know how it is different in construction or feel from the clear hose I own. 

My clear hose was purchased from three sources: Wynn Environmental, Oneida Air, and Rockler.  They are all constructed the same way except for the gauge of the plastic, the gauge of the wire, and the direction of rotation of the helix.  I would not say any of them are smooth.  Yes, the plastic is smooth to the touch, but they all have an inherent rib that is produced by the wire helix.  My understanding is that it is this rib that produces the turbulance and resulting line losses.

The length of gray hose that came with my Delta DC is the closest thing I've seen to a flat interior, but again I wouldn't call it smooth because it also has a spiraling groove on about a 1/2" pitch.  I've tried stretching my clear hose using weights on a 90 degree day, but it has a memory, and if left unrestrained it will return to nearly its original length. 

I suspect the line loss differences between all these hose types is pretty similar, because none of them have a smooth flat inside wall. 

Quote from: RCOX on July 27, 2011, 08:46:17 PM
I have bought 2 wyes from Lowes and they do not have the tabs in them. The short leg is fitted into the opening and spot welded. The only thing I did not like is the crimped end is on the wrong end for dust collection. I bought a crimping tool for hvac duct years ago and now use it to crimp the other pipe on the end I want. You lose almost nothing in diameter if you are careful and it eliminates a possibility of a jam.

Rcox, I have a crimping tool that I use only when I have to.  It is a good crimper, but like most of the hand held crimpers, it produces deeper crimps and a steeper taper on the end of the pipe.  However small, that taper produces more turbulance and losses than a machine crimped pipe end.  As the title of this thread suggests, I go a bit overboard trying to keep the inside of my system as smooth as possible.

Quote from: RCOX on July 28, 2011, 05:15:50 AM
No, I haven't forgotten it. I have not decided how to raise it and still be able to reach the filter flappers. Be nice if my Powermatic had flapper motors like the JDS portable cyclones do. May not be an issue with the separator but have really had to use them when using the drum sander. Was also curious if using a short run of flex would increase or decrease restriction. It would make it easier to smooth out the bends with the flex but would I just be trading 1 issue for another. Right now, I have a couple of projects and a small honeydo list that needs to get done. The projects will give a real world test of my setup instead of just sucking up a drum of dust, dumping it , make some changes and suck it up again.

Interested on input on the flex idea. I can make do with about 5' and could really smooth out the turns.

Rcox, published friction losses for flex hose states that it is 3 to 4 times greater than smooth pipe.  So, to put that in perspective, if you have 5 foot of flex hose, it is the equivlent of 15 or 20 feet of smooth pipe. 

I have not found any data specific to bends, but I would expect the same rule to apply.  For example, a 6" 1.5D 90 degree bend is the equivalent of 12 feet of straight pipe.  That same bend using flex hose would be the equivalent of 36' to 48' of straight pipe.

And remember if you are considering 1D bends, all the numbers in the above examples double again.  So, if you are unable to find larger radius ells locally, just use two 90 degree 1D bends.  Rotate the segments to make each one a 45 degree bend and then connect them together.  That gives you a large radius 90 degree bend for only a small additional investment.

Hope this helps.  Regards.