Horozonal in, Horizontal out

[Bruce Seidner]

I have found one post where Phil discouraged using a horizontal outlet unless it was "off center", I don't understand what this means or where off center would be horizontally. I may have that wrong but my question is this. It seems that the main work done by the separator is accomplished by the lower baffle/slot and that without much ado debris sinks fast to the bottom through the slot in the baffle and stays put because it is protected from the turbulent spin at the top and doesn't get stirred up again to travel past the new material that is coming in, around and down.

I don't understand the dynamics of this enough to know why a horizontal exit on the opposite side of the inlet raised a above the inlet, at the same essential height of the vertical inlet as designed would not be as or more effective than a vertical outlet that then requires a elbow to a horizontal blower. All resulting in more needed head room and vertical space to fit the unit.

I am vertically challenged and am right at my joists for the top of the separator and it would really help to have it side discharged. Also I while I see that taking the outlet tube into the middle of the separator hat if low would disrupt the spin of the inlet, I wonder if just making it higher would get it out of the way. I guess I don't see how a vertical outlet in the center versus horizontal outlet at the same height would cause more debris to go to out the top. That is unless is grabbed material rotating on the wall and pulled it out before it settled below the baffle. Perhaps this could be fixed by having a top baffle that mirrored the bottom baffle only on the opposite side so that the the inlet and out let were opposite one another and the baffle slots were also opposite each other. I am using 8" duct and a large salvaged commercial blower. I would like to avoid requiring more room at the top. Making the top tall enough for two 8" ducts is already going to create a pretty tall hat. I may try this as an experiment unless there is some dumb gotcha that I am more than capable of missing.

[phil (admin)]

What do you mean by horizontal outlet?  Do you have a link to that post where I said it should be off-center?

Typically the inlet has to be horizontal, and the outlet has to be vertical.  You can put an elbow on the outlet, but that wouldn't help with your space problem.

I suppose you could build a sort of plenum with a 6" hole in the center to use as the outlet, where the plenum has a rectangular cross-section.  In that way, you might be able to get the height of the outlet down to 4" or 5".

[retired2]

Bruce,

I have never seen a cyclone separator with an outlet taken from the side wall, offset or not.  I suspect there is a good reason for that, and I have to believe the engineers who design those things have been looking for ways to improve the basic design for many, many years.  In spite of that, they all seem to follow the same basic design, i.e. horizonatal tangential inlets and on-center top outlets.

I'm sure Bernoulli's equation dictates the basic design of separators.  I'm no expert, but the simplest principle states that "as velocity increases, pressure decreases."  So, in a cyclone and Thien separator, the velocity is highest, and therefore pressure is lowest, away from the center.   Conversely, the velocity is lowest, and pressure greatest, in the center.  The center is also where the cleanest air is because centrifigal force pushes the debris to the outside.  So, any outlet that comes off horizontally, whether it be tangential or off-center, is likely to cause most, or all, of the debris to by-pass the intended separation zone.

If you are headroom limited, then Phil's suggestion to use either a short radius elbow, or build a plenum off the top center are your only two options.   And if it were me, I would find some other solution because tight bends impose the same S.P. penalty whether they are before or after the separator.

[Bruce Seidner]

I was thinking about a plenum but then it became a bit more dynamic as I thought about its function.

The Thien innovation is the slotted baffle that keeps the turbulence separate from the chips in the bottom of the bin.

What if the top of the separator had a mirrored baffle. This would make the center between the baffle a clear space and rather than merely a flat plenum over a center hole of the top of the separator, one would have a horizontal outlet from the top half of the separator.

My thoughts are that if I merely had the outlet exit from the side it might pull up some of the dust racing around the walls, hence the top baffle.

But if this makes no incremental difference then the vertical tube in the center is the eye of the storm and the cleanest place to get air out of the separator.

One additional question. How was the height of the "hat" defined. Was this empirical or is there some relationship between the CFM and the height of the "hat"?

[retired2]

I was thinking about a plenum but then it became a bit more dynamic as I thought about its function.

The Thien innovation is the slotted baffle that keeps the turbulence separate from the chips in the bottom of the bin.

What if the top of the separator had a mirrored baffle. This would make the center between the baffle a clear space and rather than merely a flat plenum over a center hole of the top of the separator, one would have a horizontal outlet from the top half of the separator.

My thoughts are that if I merely had the outlet exit from the side it might pull up some of the dust racing around the walls, hence the top baffle.

But if this makes no incremental difference then the vertical tube in the center is the eye of the storm and the cleanest place to get air out of the separator.

One additional question. How was the height of the "hat" defined. Was this empirical or is there some relationship between the CFM and the height of the "hat"?

I'm not sure I am envisioning what your are describing.  What is the point of the mirrored baffle?  What is above it?  Does the exit air have to pass through it?    Regardless of the answers to these questions, I think you are heading down a path that is going to introduce such heavy SP losses that the system will just not work satisfactorily.

Phil might want to weigh in on your last question since he has probably tested more variations than any of us, but I think it is safe to say that the height of most top hat designs is determined by the inlet size and mounting method rather than empirical testing.  Coincidentally, I think that happens to produce the optimum configuration, but that is just my opinion. 

CFM is unlikely to change much as you adjust the height of the separation chamber.  Adjustments in height will have a small effect on SP, and CFM will respond accordingly, but I doubt you could even measure the difference.  What will change more dramatically is FPM.  If you increase the height significantly, the circumferencial air speed will slow and you will start getting by-pass.  Shorten the height significanlty, and I'm not sure how you could shorten it to less than the size of the inlet, but if you could the velocity of air will increase.  That might result in better separation, but you would probably introduce some significant SP losses.

   

[Bruce Seidner]

That is enormously helpful. I really did not understand the dynamics of how this worked. I proposed a layer that detracts from the performance of the original design.

The context is that I am aghast at the tables that show the equivalent feet of duct that a 90degree fitting adds to a system. Thanks for you clarifications and clear thinking. I am back to the drawing boards.

[retired2]

That is enormously helpful. I really did not understand the dynamics of how this worked. I proposed a layer that detracts from the performance of the original design.

The context is that I am aghast at the tables that show the equivalent feet of duct that a 90degree fitting adds to a system. Thanks for you clarifications and clear thinking. I am back to the drawing boards.

Well, if you are checking tables for equivalent length of pipe, then you know how much loss a 1D bend adds compared to say a 1.5D or better yet a 2D bend.  That is the main reason my system is plumbed with metal pipe rather than plastic.  5 segment ells are readily available in metal with 1.75D radii. 

[Bruce Seidner]

Ok, if I can't go up, maybe I can go down.

I am using 8" duct and my top hat will likely be built into the cut down bottom of a 55gallon heavy plastic drum. So it will be about 24" in diameter with the height to be determined by how I build the outlet which can't go up or if it goes up it will be a sharp 90 degree elbow fitting to the blower. Using the drum will save the effort of building the top hat surround. All I will need to do is build the slotted and fitted bottom and attach the plastic top. The inlet will be mounted into the plastic hat and enter  just as it is done in a Pentz cyclone. I may even go rectangular on the inlet, we will see.

At first I thought that creating a plenum the same diameter of the top hat would be a good idea. I would just build a mdf top layer mounted above my slotted bottom and insert the whole unit into the plastic surround. My thinking was that the vertical outlet that pulls from the center of the separator would terminate at the bottom of this top dividing mdf internal piece creating a top area or plenum. Then a separate horizontal outlet would pull from this 8"x24" space and exit the side going directly to the blower inlet with no joints or bends.

I was thinking that perhaps the large diameter space in the plenum area that drew from the center of the separator would create less resistance and turbulence in the outlet from the plenum to the blower than a sharp 90degree bend from the vertical outlet of the top hat separator to the blower. While I see that I am introducing more volume I still don't see why pulling from this large plenum would not be an improvement over the dynamics of turbulence and resistance that a sharp 90 elbow would make. But my reading of the feedback is that I have not solved the problem of a sharp 90 elbow with a large diameter plenum, even with a bell mouth outlet inside the plenum and a straight run into the blower intake. But it is hard to shake this idea, it seems to work in thought, at least.

So, how about going down?

What about a 2x long radius 90 degree elbow that starts in the center of the separator, goes down through the bottom of the baffle and then out the side of the dust bin below? Dust and chips would fill in around and above the duct but since unlike the bin of a cyclone where there is dynamic tornado continuing and such a duct would interfere, the dust in the bin of a Thien separator is supposed to stay still, down, and out of the suction of the blower.

This thing is going to be hugging the floor joists and I have under 5 feet to the bench below (with concrete below the bench) to make 8" duct work. I am not looking for innovation or challenge, just a remedy.

[retired2]

Ok, if I can't go up, maybe I can go down.


So, how about going down?

What about a 2x long radius 90 degree elbow that starts in the center of the separator, goes down through the bottom of the baffle and then out the side of the dust bin below? Dust and chips would fill in around and above the duct but since unlike the bin of a cyclone where there is dynamic tornado continuing and such a duct would interfere, the dust in the bin of a Thien separator is supposed to stay still, down, and out of the suction of the blower.

This thing is going to be hugging the floor joists and I have under 5 feet to the bench below (with concrete below the bench) to make 8" duct work. I am not looking for innovation or challenge, just a remedy.

Exhausting down is about as "innovative and challenging" as it gets!  I suppose it could work, but here's the problems I see.  First with the outlet piping rounting through the dust bin, I don't know how you are going to empty the waste.  Secondly, if you've got that figured out, you have to keep all this stuff air tight.

After hearing about your blower, I wouldn't be too concerned about a tight 90 degrees off the top outlet.  You can easily overcome that loss with your blower.

[Bruce Seidner]

You sir are the voice of reason.

The fact is that there is no run over 20 feet to a machine should put capacity concerns to rest. There will be an 8" main with 3 wyes coming off it and reducing to 6" runs down to the machines. An 8" short elbow is still silly large to a guy used to 4" and 6" fittings. The blower was surplus and rusted through in places. I did not think the duct tape patches were all that attractive so I took the whole shebang apart and took the 5hp Baldor motor to a shop this morning for a once over and any maintenance it might be due for. The capacitor was looking a bit leaky and while everything is in disarray in the shop it made sense to just take care of this. I will get some sheet metal and Bondo the blower this weekend. My other project is installing a 6" riser on my 14" band saw. I also sprung for a new spring so I can tension the blade and get down to mitering up some PVC for proper 2x radius 6" elbows. I will see what room I have and I may be able to do better than a short 90 degree elbow for the top hat discharge to the blower. Heating and bending pipe over 4" is pretty hard to do effectively so it is going to be puzzle pieces time as I cut and plastic weld up my new fittings. Speaking of fitting, I found the band saw gong show recordings on this website, thank you Phil, and plan to experiment with them as I tension the new blades I will have to purchase. I went to a plastics fabricator shop last week and spent some time with a fellow who has been fabricating fittings and systems for 30 years. He has a monster band saw that he uses to cut PVC pipe and he has found that the 14-18 tpi bi-metal skip blades do the best job for the thin pipe. He uses 6tpi blades for the Schedule 80 and beyond stuff.