Diy cyclonic dirt seperator from pvc bits and a bucket

heres how i build my unit for my micro to keep my shop clean and never have to change my filter or lose vacuum pressure from loose fittings this design allows you to maintain your shop vac in stock condition
ill try to describe the parts as best as possible most should be available at your local orange box  (or blue if your like that... :)
 the theroy goes like this (as seen on dyson vacs)  get the dirt spinning really fast like a tornado
the heavier particles of dirt cant get to the top of the canister to the exhaust port  and fall thru the small opening into the waste container  (i.f. 5 gallon bucket)

addition:
the distance between the intake and the outflow is key... too close and the dirt jumps the gap  the 2" outflow pvc is pushed down into the main chamber a couple of inches creating a secondary air current up the wall across the lid and back down into the chamber pushing the dirt away from the  from the exit and back into the main wall cyclone )  if i can find a camera small enough ill video the interior ...the irony is if  i put something in there to film it its going to change the air flow
"no fair you'll change the outcome by measuring it " ...hubert farnsworth

Step 1: Parts acquirement

5 gallon bucket with air tight lid
these were about $4 make sure it has the rubber seal ring
apiece of ply is good for supporting the lid for longevity
4" pvc  had laying around
2"pvc   ditto
there was a cart of short pre cut pieces at my store so you didnt have to buy 10' sections
1 toilet floor mount adapter
4" 3"  flat adapter
2" to 4" cone adapter
4" to link with 2" t  45  not 90
4" coupler
4" to 2" flat adapter
2x 2" couplers
a tube of rtv red gasket maker
i had this from a auto store you can get this for chimneys stove to so it may be at a hardware store

Comments

[Entropy512]

Wouldn't cyclonic separation require the intake to be offset from thevertical axis instead of pointing straight at it?

It looks like just separation from gravity settling here, no cycloniceffects.  Very efficient for larger stuff like sand and PVCsawdust, but not as effective for stuff like dust.

In this configuration it isn't really any different than the design ofthe shopvac driving it.

Significantly greater separation (and probably reduced parts count)would probably be achieved by having the intake go into the edge of thebucket lid, pointing along (not at) the bucket wall, with the outletstill in the center of the bucket.

[bongodrummer]

Nice job for 30 mins! Like Entropy though I don't reallyunderstand how this creates a cyclone (more explanation?). I havefound a very useful resource on this topic to be <a href="http://www.billpentz.com/Woodworking/cyclone/BuildCyclone.cfm">http://www.billpentz.com/Woodworking/cyclone/BuildCyclone.cfm</a><br /><br />I would like to improve mine system so that I hardly ever need to changefilters, and make it more elegant... Thanks for the thought provoking'able. <br /><br /><br />

[neorazz]

if i make another one ill follow this design  

[bongodrummer]

Hay neorazz, thanks for the link - that is cool, I might have a go at something similar myself.

[vectorges]

You have some impressive results, but I am not sure why. The idea for a cyclone is to move the "dust and debris" in a spiral pattern around the outside of the container. If your intake was roughly horizontal at the outside edge of the bucket it should work better. But you can't argue with results. Good job.

[neorazz]

it is see the video

[PKM]

I suspect that a vortex may spontaneously start in the cylinder, as it's a more "stable" arrangement than turbulent flow, but that's just a stab in the dark.  Even without a vortex, it's entirely possible that the momentum of dust particles carry them downwards into the bucket while the air turns upwards- in this case the cyclone would probably just improve the filtering of smaller dust particles.

[kikiclint]

looking at the setup, it would seem that the air is really fast in the hose, but when it gets to the tube, it can slow down, like a stream suddenly getting wider, letting the particles just fall down.  That would explain the efficiency. 

[BigRedRocket]

It is entirely possible that a vortex is generated along the axis of the cylinder. I have seen another DIY dust collector that had a tangential inlet which attempted to spin the flow counter clockwise, yet it actually spun clockwise.

The asymmetry of the setup, the variable inlet conditions, etc. all play into the vortex generation.

Taping some small strings to the inside of the tube would allow you to visualize the flow. Also, adding a clear flat plate on the top, instead of the PVC, would allow you to see the strings and their movement.

Hey, but it works.

P.S. the uncertainty principle applies to quantum mechanics. Fluid mechanics can be chaotic, and this flow probably is unstable, but viewing the flow isn't going to change the flow. Drilling a small hole in the top or side, adding a light, and sticking a camera in there most likely would work great to see what is going on.

[nocode54]

I am a mechanical novice and a wood carver (strictly hand tools).  I love this concept but can't see how a vortex is created.  Can the origingator or someone else comment? 
Thanks to the poster regarding static buildup (a real concern with highly flammable, fine sawdust).  Can anyone tell me how to ground PVC?

[neorazz]

check the videos  for grounding run a piece of copper wire unshielded to a ground point and attach to pvc via a self tap screw haven't tried it yet but the principle seems sound

[trf]

I had an improvment idea. I dont know if it will work but ill throw it out there. What if you put a semi fine mesh screen to where your vaccum hose attaches to your bucket. That way, if anything does make it past the vortex, it wil hit the screen and when the vaccum is turned off this debri will fall into the bucket due to gravity?

[neorazz]

in a good system you shouldn't need any filter at all im not quite there yet

[JCambpell007]

 Great idea!
Just a word to the wise: pvc will cause a lot of static build-up andwill need to be grounded otherwise it can spark and ignite the microfine particles.  Hate to see a good idea go up in flames, so to speak.

[neorazz]

yeah i made a spark 2 inches long vacuuming up iron filings  as for flame it takes quite a bit of heat to melt this pvc  (flamethrower/ potato cannon stay tuned) and i wonder if the fuel to air ratio in a partial vacuum would even support a sustained fire now if there were a leak in your collection container  you could cause a backdraft that would be cool

[boffincentral]

You will get a static charge. However, to get enough charge to ignite will be hard. A lot of amateur woodworkers build dust collection systems using PVC pipe that have far greater CFM volumes (upwards of 1200) than a shopvac can create. I've run my (ungrounded) dust collection for hours with little effect, even in times of low humidity.

There was also an interesting episode of Mythbusters that covered static electricity and ignition as well.

I'm not saying it's impossible but that it is highly improbable.

[joebar32]

As someone who designs systems like this for moving material at up to 100tph I can tell you that this is not a true cyclone but works just fine despite it.  What you are getting here is simply the velocity drop due to the change in cross-sectional area.  Make the big pipe bigger and you'll improve efficiency.  If the upward rising velocity of a given particle is above the velocity in that cross-section, it will drop out.  If you put the inlet on a tangent to the pipe, you'll get a better cyclone. It's efficiency will be more due to loss of momentum as the particles impact and slide along the ID of the pipe, slowing down and falling to the bottom. Pipe ID has less effect on efficiency than in the first style. A tall thin rectangular inlet on the tangent would improve efficiency as well. Beware, a cyclone will have significantly more wear on the pipe wall and static generated due to the amount of particle contact it creates.

[neorazz]

100 tons per hour? for aggregates?  most of what your saying make sense  check the videos

[srilyk]

 If you cut a hole in the side of the PVC and replaced it withplexiglass (and took some care to get it pretty smooth) it should have aminimal effect on the airflow design and allow folks to see inside :D

Also, this is awesome

[neorazz]

(removed by author or community request)

[lemonie]

Interesting - add your text (above) to the Instructable (Edit button)

L

[CaseyCase]

Clever idea. I use a Thein Separator myself. http://www.cgallery.com/jpthien/cy.htm