walks like a duct, quacks like a duct, but doesn't look like a duct

[Bruce Seidner]

Sorry to take up bandwidth to obsess about this, but I anticipate my wife would be even less interested. I also have not tested her interest in a discussion of the requirements relative to electric wire gages. Why press my luck, eh?

As I explore the expense of 8" complex duct, plastic or metal, the costs get pretty unreal. Conversely, I am unrealistically cheap.

But, that said, the inlet on my blower is 8" round and horizontal to the floor. It has to connect with a cyclone outlet oriented vertically, 7" round and about 5' off the floor.

I have been looking at flexible duct and all manner of fitting and materials. I have watched YouTube tutorials on heating PVC and bending it. There are schemes to use sand and schemes to use air pressure to keep the large pipe from kinking and I have heat guns and even a large propane torch for weed eradication, ice melting, and now apparently PVC bending. But I am most comfortable with wood and keep coming back to an idea that may be aerodynamic anathema but which makes Neanderthal sense to me.

Bolting a straight square or rectangular wood vent, with an 8" round hole on one side to the face of the blower to match the height of the top of the cyclone is very appealing to me. It would take no time to use melamine shelving material right off the rack to construct such a box. It would bolt straight onto the face of the blower and go a short distance straight up. It would do nothing but create the link of suckage directly to the impeller which would whip round and right out directly into a very, very large and long fabric pup tent of a filter. I have read and believe the wages of sin associated with duct resistance in bends, transitions, turbulence, etc., coming toward a cyclone and other subtleties of duct movement science and magic. But just as I have recently learned about the physics of the dust bin and how air leakage and shape are the central dynamics and not size, all initially counterintuitive, I wonder about the ducting in this part of the system. When learning about a new content area I am almost always backassward and read the index first.

First off, such a duct from the horizontally oriented blower up to the level of the vertical oriented cyclone top would save huge coin when looking at the alternatives of fittings and pipe to achieve this same end.

Secondly, I wonder if ducting after the cyclone outlet to the blower has the same concerns as all that is writ about the dynamics of ducting from the machines to the intake of the cyclone. That seems to be where the majority of discussion and engineer brow beating occurs in the literature and forums. Here I am talking about the tail end of the system which is going to cost more than the whole 6" ASTM2729 front end of the system if I have to use 8" conventional metal or pvc duct and fittings.

Is there a similar problem using a rectangular vertical duct at this specific juncture as one gets with round to square, large to small, or short radius transitions in the ducting going into the cyclone? Or are the requirements and dynamics different after the cyclone? I see lots of pictures on professional sites using short radius 90 degree fittings after the cyclone. I don't know if this is bad design or if they know something about this end of things that I don't.

My thinking is that while transitions and ducting like this might create dreaded turbulence and resistance in ducting along a shop long intake of wye's connected to machines, that there might be different dynamics between a blower and the requirement of going straight up a few feet. The alternative is a very expensive set of fittings to make a long radius 90 degree duct up to the height I need and another set of expensive transitions and fittings 180 degrees up around and down to the inlet of the cyclone.

I can make the vent any size I want and I could even mimic a long radius duct that would be curved with flat walls.

I am open to ideas. I have no sense of how to dimension the duct but imagine something on the order of between 8"x8" or 8"x 4", etc would create the vertical duct that would rise to the occasion. Any experience or better yet, known engineering advise would be gratefully received and a holiday gift to the millions of like worried souls suffering with this problem.

[Jim Andrew]

Could you not just use some snaplock 8" pipe and a 8 to 7 reducer, and elbows? You can mix some metal in with plastic without any adverse results, just have to figure a way to connect to 6" plastic. Think I have seen some ways so just use the search feature.

[ian maybury]

Sounds like a lot of trouble Bruce for what's (over here anyway) not all that expensive if you buy spiral galvanised duct and fittings from an industrial supplier.

That said here's a guy after your own heart whose system is mostly rectangular ducts made of MDF: http://spikyfish.com/DustExtraction/ Don't know how good a solution it is/is not though.

A square duct is less efficient than a round duct, the corners choke a bit and probably won't transport chips quite as reliably - but it's pretty unlikely to be a show stopper if the duct is sized to give the correct air speed/fpm. Here's a calculator for calculating equivalent sizes (those giving the same pressure drop/ft run) in square and round: http://www.engineeringtoolbox.com/eq...ter-d_443.html It's for lowish airspeeds, but probably won't be too far off.

The square seems likely to present issues a bit earlier so far as chip transportation is concerned if in order to maximise cfm you were running a large duct size that drops the air speed on the low side of the recommended 4000fpm in vertical legs - again because of the corners.

In practice the pressure drop from bends and other fittings creating restrictions is probably far greater than anything the short straight lengths of either type of duct you are talking about will create. What's more I'm not wholly convinced (but am open to correction from somebody with hard data) that unless you go to an untypically large radius bend (most of the bends we use are around radius = 1.5xDia) that there probably isn't that much difference between the fabled two 45 deg bends versus a similar radius 90 deg.

The cyclone inlet tube and the exhaust create pressure drop in exactly the same way as the suction side of the system. Where the exhaust is different is that you can go to a large duct size without worrying about air velocity/chip transportation - in that the cyclone should by this point have dropped out everything except a very small quantity of fine dust.

Big radius bends are fine in a big workshop, but pretty tough to fit into a smaller garage type shop - and that's presuming you can get the things. They are quite a bit more expensive too.

We touched on the issue of longer/bend containing linkages between the cyclone and the blower inlet. You see various formats about, so it doesn't seem to be the case that it's a show stopper. I've never seen credible information about regarding whether or not it may have any effect on cyclone and blower performance. The extra length may not be a big deal in pressure drop terms, but it's hard to know whether or not given the pretty complex and unknowable air flows and other dynamics inside the blower and especially the cyclone what effect a given dimensional change is likely to have.

ian

[Bruce Seidner]

You sir are a sensible and considerate man. These pictures are great as are your thoughts and identification of the bottom line issues. I have a silly large 16" impeller 5hp industrial blower that I picked up surplus. There is more than enough suckage to allow whatever loss I introduce because of the contingencies of my work space.

I had copied Pentz this question and he had some sympathy for the obsessive concerns expressed and called me at my office to clarify that the box idea was problematic and that I needed to stay with duct and the principles that are relevant on the front end of the cyclone. He proposed that I turn the system round and make it a pull system and locate the blower and cyclone out of doors. I am concerned about using an aluminum impeller in a push system. So I am not sure I am going to follow in the steps of the all star dust collectors like Allen and Bill. But he introduced the idea that I do some major surgery on the cyclone itself and relocate the outlet parralel to the intake and dive right into the inner outlet of the cylcone from the side, sealing off the top. If I build a mount to raise the blower horizonatal to the cyclone and make a straight run from the blower into the cyclone my cost and complexity drop to nil. I just need to cut a hole in the side of my cyclone just ahead of the inlet which is a "dead space" for dust movement and turn this vertical outlet cyclone into a side discharge cyclone in line with my blower.. I need to buy some tin snips but that is cheap compared to a duct that would confuse a minotaur, or at least drive him to bankruptcy. I think Bill has solved the problem of "my rear end" with this advise. He thinks I am making some real heartburn by locating the bin in the ground of my crawl space but I don't have the option of putting it to the side of the house and I am not a production shop. I mostly play with sheet goods or boards that my wood jewelry store has thickness planed and made a good edge on. The bin is 55gallons and I have a portable blower that I can use to empty it when it needs emptying. I am a bit anxious about the surgery on the cyclone but I have the concept down and it is not like a gender change operation. It is merely a gastric bypass.

[Alan Schaffter]

I think Ian hit most of the points. I may be repeating what he said, but here are some of my thoughts. All ducting and fittings on both sides of the blower and cyclone affect SP and flow performance. As Ian said, however, you have the luxury of not needing velocity to keep dust in suspension at the outlet. Another important part of the ducting issue is the need to have a straight shot into the cyclone to reduce turbulence. I don't remember off-hand, but I think a minimum of 5 duct diameters of straight ducting is recommended.

Many of the commercial cyclones incorporate one or more design compromises to deal with height and footprint issues. A prime example is those systems where there is a relatively sharp bend from the outlet of the blower to two filters stacked in series. First, as you know, the bend adds SP resistance and second you get less SP resistance with parallel filters in parallel. As I said in previous posts I have a push-through cyclone setup. I can not identify the source other than possibly Bill P. or vouch for its veracity, but it is supposed to result in slightly better separation. There are multiple other reasons I used this configuration however: (1) gives me more blower mounting options, (2) allows me to run a nearly straight, rectangular duct from the blower outlet to the cyclone inlet for smoother flow and eliminates the need for any SP and turbulence generating round to square duct transitions, and (3) and most important to me, it allows a trash bag to be mounted directly to the bottom of the cone for easy and relatively dust free handling and removal of the dust and chips and eliminates the need for a perfectly sealed dust drum.

Unless you plan to suck up 25' tape measures or other heavy metal objects, a properly constructed aluminum impeller, especially one that is 16", should work fine in a push-through system. Remember where the chips enter and strike the impeller- at the center, primarily on the thick impeller disk where blades are not located or if they are they their width and velocity are less. There is much less chance of damage than if chunks of wood were introduced at the blade tips. The chips and chunks are actually moving parallel to the impeller blades as they exit the blower. I have a steel impeller, but have indeed sucked up a big tape measure, a DP chuck key, and other metal parts, with absolutely no damage to the impeller.

Frankly, I would not make any mods to the cyclone and certainly would not cut a hole in it to reroute the central exhaust tube, especially if it will be a right angle bend. I may not be understanding what you are actually trying to do, but can't see how you would do that without affecting the inlet or spiral ramp if your cyclone has one. Another option that has been shown to work in low headroom situations is to tilt your cyclone. That would reduce the need to bend or angle the outlet. Remember, dust separation in a cyclone is caused by the swirling air and centrifugal force which is unaffected by cyclone orientation. Gravity only plays a small part in the dust and chips falling into the bag/bin.

[ian maybury]

I'm kind of lost Bruce, in that I'd read your post as saying you had decided on a particular cyclone and blower location, and were wondering about your duct options to get them hooked up. Locating the blower, cyclone and filters outside is a good way to save space that we mentioned before. Either that or exhausting outside all of the time if that's an option, although I gather not in your case.

If you have freedom to play with layout and location then it's down to figuring out the best layout solution - taking account of the requirements as best you know them. It makes sense too to locate your chip bin somewhere convenient for emptying

As before I doubt as the guys say that small changes (e.g. an extra bend or few feet in the connection to the blower) will sink the boat, although this is not based on experience or anything more than gut feeling and the observation that different makers run with all sorts of odd looking layouts for what seem to be packaging reasons. That said it's possible that minor differences make no difference at the gross level of separation of visible chips by the cyclone, but have quite significant effects in terms of the subtle task of dropping out more of the very fine invisible dust.

It seems for these sorts of reasons best to stick with as far as possible layouts known to be successful, meaning a decent run of unobstructed straight header into the cyclone, with the blower mounted on top of a short inlet tube. I'm not quite getting what you are saying about modification of the cyclone exhaust tube location (in that there's really nowhere else to locate it in the cyclone), but I'd be reluctant to run with anything other than the proven central and vertical positioning unless there was hard information from a reliable source that had done it before.

I know nothing about push through systems, other than that I had an issue in the past with the (steel) impeller on a bag filter system getting knocked slightly out of balance by a chunk of wood coming through, and welcomed getting away from that possibility with the (pull) cyclone system now installed.

We touched on this before I think, but don't forget that there's no point fussing about any of this if you are recirculating air back into the shop, and your filters are not of high enough specification (HEPA or similar) that they will capture whatever fines get past the cyclone....

ian

[David Kumm]

As others have stated your system is big enough to compensate for your extra restrictions. I have a cyclone positioned in the center of a long wall and can not run a straight duct into it. Using large radius 8" ells that are adjustable I make a large circle around the inlet and then branch out in two directions. The key is the direction of the pipe has to match the direction of the cyclone. My outlet pipe goes straight up a few feet, ells over to the blower and motor sitting in the attic and then exits the blower and ells another ten feet to a plenum on top of the first floor ceiling into 200 sq ft of filter area, all with 8" pipe. Not as efficient as an ideal setup but my 15" blower and 5 hp motor at 63 hz will deliver 1500 cfm to my planer 50 ft and two ells away with 8" and 7" mains. My long winded point is that if you size the pipe, fittings, and filter area correctly you will still have plenty of cfm for your application. Dave

[Bruce Seidner]

I can get a bit hyper focused on the small stuff and appreciate the big picture real world reactions that are represented by this group think. This blower is industrial surplus and it is beast. But having to refit the whole shop to accommodate a sliding table everything has gotten a bit disorganized. So point well taken, don't sweat the small stuff. The boat will continue floating.

A lot of problem would in fact be solved by not having the dust bin be a critical component. In a push configuration I have noticed the wonderful contrast between Alan's Star Wars auto gates and elegant ducting all emptying into a simple plastic bag. I have never noticed any sparking but then I have never sucked up any screws or Lawdy Mighty, a tape measure. But a simple small volume drop box would solve that problem if I was that concerned about it.

In a push configuration I would have no concerns about vacuum or for room in the bin for the continued dynamics of the cyclone that continue down below in a pull configuration. I have to clear some space and do some experimenting with the crawl space. The ground is uneven and some spots are lower than others. I should have some time over the holidays to experiment. Also, some good holiday cheer should allow me to get over the dire loss of cfm with some suboptimal tight angles.

All good advise and perspective. Thank you. And Happy Holidays.

[jay st clair]

Hello, Just a thought on the cheap side if you have time you could build the bends by making them in sections like the way they would be made at a welding shop ie,10 degree sections screwed together sealed, and possibly sheet metal layed flat against the outer radius.

[Chris Parks]

om bends and other fittings creating restrictions is probably far greater than anything the short straight lengths of either type of duct you are talking about will create. What's more I'm not wholly convinced (but am open to correction from somebody with hard data) that unless you go to an untypically large radius bend (most of the bends we use are around radius = 1.5xDia) that there probably isn't that much difference between the fabled two 45 deg bends versus a similar radius 90 deg.

That would make an interesting experiment, I might do it sometime in the next twenty years when time becomes available. What I would contend without doing the flow test would be if you are near the fan then you may not see much of a drop, put 10 metres of duct in and the bends at the end of that and the drop should be more noticeable.I personally have tried to always utilise two 45's with a short length of pipe between them. If nothing else it looks like it should work better and makes me feel good which is the important part.

[David Kumm]

According to my cfm and SP chart a 1.5 radius 90 has twice the resistance as a 45 so that would imply no benefit from splitting although it certainly can do no harm. There are 2.5 radius ells available here that I use on the mains but I don't have numbers on the SP. It is interesting to note that for every 100' of pipe an increase in velocity of 500 fpm adds 1 sp to a 6" pipe and .6-.8 sp in an 8" pipe. That gets significant when at the top end of the fan curve. To fully utilize a 5 hp 16" system the velocity in the main runs between 5000-6000 fpm. Dave

[ian maybury]

Hi Chris. TBH i'm guesstimating and don't really know what the reality is, but not entirely without information.

In that the HVAC design tables I have (which as before are of indeterminate origin) list an R= 1.5D 90 deg bend as creating the same pressure drop as 12ft of straight spiral galvanised ducting at 4,000fpm air speed. A 45 deg bend of the same radius is listed at 6ft - so at that rate 2 x45 deg would give about the same drop as 90deg.

Don't know what the effect of a longer or shorter straight in the paired 45s would be. Taken at face value the tables make no allowance for this.

I put paired 45 deg bends right through my 160mm system too, but because of the need to get between ceiling joists to preserve head height couldn't really use more than short 3 - 400mm straights between them. Looking at them on the ground led to the thought they probably weren't going to add a whole lot. What got me thinking is that years ago I did a little work on laying out a transporation system for milk powder in a factory upgrade project. They admittedly use higher pressures and velocities than are needed in our situation, but very large radius (maybe 3ft/1m+ from memory?) bends seem to be pretty much the norm for them.

To Bruce in case I wasn't clear. What I was saying is just that purely in terms of pressure drop (subtler factors may come in to play that influence cyclone performance) the drop created by two or more 90 deg bends is by the above going to be lots more than the extra few ft of straight duct required to locate a blower a short distance from a cyclone.

Put another way - I ended up using a 90 deg bend to hook up the dust system to my band saw. Scary to think that the bend accounts for as much pressure drop as the run of ducting right across the width of the shop to get to it.

The other point that's maybe worth repeating is that while a larger diameter blower will due to its higher tip speed for a given rpm normally perform better at higher pressure drops (up to 12 or 13in WG - see table 5 well down this page of Bill Ps http://billpentz.com/woodworking/cyclone/DCBasics.cfm) that what you get is still determined by the fan curve. (a graph showing the CFM delivered across the full range of operating pressure drops)

Meaning that no matter how hairy chested your nice big 5HP blower looks that the CFM delivered (and power drawn) will still tail off in accordance with the curve as the total pressure drop moves upwards out of its design range. i.e. too much ducting length, too small a duct and too many bends and other restrictions can strangle it too.

ian

[Alan Schaffter]

This whole discussion makes it worthwhile with a fixed system to do your best to figure out how to vent outside with a short run. My garage level push-through vents straight up to a second story shop level dual filter plenum which is just a few inches above the cyclone in these old construction pics. It now has a diverter so I can bypass the filters and vent straight up another 3' through the roof when the weather is mild. I guess I can't do much better, unless I eliminate the cyclone.



Simple truck tire inner tube coupling- stretches to any size and reduces the transmission of vibrations and noise, and often free!!





The level of discussions and knowledge sure has improved over the early days when Bill was repeatedly beat up on these forums!

[Chris Parks]

According to my cfm and SP chart a 1.5 radius 90 has twice the resistance as a 45 so that would imply no benefit from splitting although it certainly can do no harm.

We need someone to do a flow test and stop guessing. If flow drops over a length of duct then any impedance should be magnified? If a real world test says that two 45's make no difference then we have been led up the garden path so to speak.

[ian maybury]

I've not been around long enough to have a good historical perspective Alan, but maybe the possibility of a rational and largely engineering based discussion on a public forum on the topic is a measure of the scale of Bill Ps achievement. My feeling is that we probably owe him quite a lot in terms of the body of reliable baseline data he got out there, of the approach to system design he evolved, and of what he did to raise our consciousness of dust issues. Mine anyway - I was all set to waste €1,000s on another dodgy dust system (having struggled for years with a basic bag filter unit) until I was steered towards reading his stuff.

From what I've seen he quite apart from getting a huge amount of work done took some significant risks in tackling the cosy concensus - by speaking out against the then status quo.

It'd be nice if we could mark this in some way.

It seems unlikely to be the final word on the topic given that the need for 5HP blowers and equivalent ducting sizes on garage size shops is at least in part to do with the less than ideal hooding/dust collection arrangements on current woodworking machinery, but it's hard to underestimate the value of placing a fact based design framework in the hands of the public. Knowledge is power....

My understanding is that he's at this stage struggling with deteriorating health, and finding it harder to undertake hands on work than was the case.

I'm guessing perhaps Chris that the 2X 45deg thing is not so much untrue, as that it may not be that big a deal at small radii, and that as a rule of thumb it has its origins in large shop situations where larger bend radii get used. I should say that if I was building my system again next week that barring hard information I'd still use doubled 45s.

The length of the duct feeding a bend probably doesn't influence the situation on bends too much. In that the CFM flowing from one end of a line to the other barring duct leakage remains pretty much constant. The length (along with bends etc) does influence the total pressure drop in that line though, which in turn determines the point on the fan curve (the graph of pressure drop versus CFM for that blower) at which the blower will be operating - and hence the CFM the blower delivers. Which is what flows through the line.

Air at our low pressures (up to about 25in WG) is for this sort of purpose normally regarded as incompressible. i.e. it behaves more like water in a pipe than a gas. The flow rate/fpm will speed up to get past a reduction in cross sectional area, but it will slow down again afterwards in response in (inverse) proportion to whatever changes in duct cross section it meets. i.e. CFM = duct cross section in square ft X air speed in ft/min. It won't compress the way the air at higher pressures in say your shop air system does to get through a tiny nozzle with relatively low pressure loss though.

The pressure drop in the bend or a restriction is presumably the result of the energy lost in making the speed changes involved in either event happen.

Some testing would be useful all right...

ian