Impeller for 5 HP Dust Cyclone

admin

trying to fabricate a 16" impeller for a blower on my dust cyclone...v. 2.0 was made out of 11g base plate with 14g blades, full welds along blade bottoms, with a taper lock bushing welded in the middle.  Got some warping from heat but not so that it was contacting the motor mounting plate when spun by hand...ran it for a second last night and got a high pitched squeal when the edge made contact with a flish threaded insert on the inside of the motor plate.When running, the warp is sufficient that it ends up indeed contacting ,despite the 1/2" clearance from the lowest point to the faceplate...would have though that the RPM would lead the impeller to run flatter, not wobble more...Anyway, should I try using a heavier stock for the faceplate?  Am I OK using a few short beads per blade to minimize heat input to prevent warping?  Or should I skip welding all together and go with bent and riveted blades (more work).Appreciate any help...thx!
Reply:I would vote against rivets.  They would fail too quick in that environment.  Sand and stuff will eat away at the heads and create a missile.  I would use thicker backing and a more recumbent blade profile to help air flow.  Warping may be overcome by short stitches and allowing it to cool completely between welding sessions.  Just my thought, but worked on a ton of the 8 HP Billy goat blowers years ago.BobI'm spending my Kids inheritance, I dont like him that much anyway!!!!!!Enuff tools to do the job, enough sense to use em.Anybody got a spare set of kidneys?  Trade?
Reply:Thanks Bob...I think going up to 3/16 material will help a fair bit vs the 11 g too....it's only wood dust by the way, not sand or metal.  The blades are curved backward, about a 1/2" deep arc over their 6" length, and slightly angled backwards at the outer edge...I think I just halve to be more patient welding and cooling.
Reply:Here's the idea... Attached Images
Reply:Some things you have to be aware of when you are making and then spinning things, like a rotor or impeller:- warpage during fabrication;- balance!!!!- proper and adequate support (bearings);- proper and adequate design of the spinning object so that it can withstand all the forces/loads on it from spinning around.You could make something and it might not be warped, but if it is not balanced AND supported properly by adequate bearings then it will destroy itself and possibly things around it.An out-of-balance rotating object stuck onto the end of a overhanging/cantilever motor shaft and then spun up to speed could 'wobble' all over the place and :- touch things it shouldn't touch;- wobble so badly that it destroys the bearing(s) (if any) in the motor;- wobble so badly that it destroys the bearing in the motor and then bends the motor shaft.Not necessarily in that order or even as separate events (to your timescale).  A large out-of-balance and warped spinning object can DESTROY itself, the motor, the machine, and surroundings in an eyeblink.It doesn't have to be rocket science (although advanced rotor dynamics and airflow ARE parts of rocket science  ), but it certainly is at least a few steps above and beyond, tech-wise, from making say a non-moving bench leg for the shop.  The best laid schemes ... Gang oft agley ...
Reply:All well noted...given this, here is my rough approach:form bladesweld to base plate with minimal warpageweld taper lock hub into centre.use taper lock bushing to attach to test shaft, set on to bearings for balancing, slowly removing material from base using large drill bit (think dimples) until to detectable "heavy" spots can be observed.My alternative is bite the bullet and buy one for over 200 before shipping, Grizzly offering replacements for their own 5 hp cyclones...but in looking at these, I find it hard to believe I can't replicate their level of quality/tolerances to be an effective solution.
Reply:It's amazing how flexible these things can get at high rpm. Folding the bottom blade edge will stiffen the whole assembly and allow you to use thinner material. Thicker backing plate may help but it also requires more energy to run. I have some blowers that are welded, others are bolted. Clamp everything tight when welding. I'd weld a couple inches on the flange, then a couple on the back side (fold) and then finish a couple inches on the flange. I'd face the fold flange into the material as the blades tend to hold their shape better that way. You'll likely need to balance the finished assembly.Take a look at the simple balancer Steve Bedair built when building his chipper. Something like that should get you dialed in. http://bedair.org/Projects/chipper2.htmlHave you tried straitening the blower you already built?
Reply:Ok... now I see your images. Personally, I'd use straight blades. Much easier to fabricate and likely just as efficient. All the material blowers I have are straight bladed.
Reply:Would love to if I could....I'll take some pics tonight.  like I said, it came out warped, but not so badly that I'd have thought it would ever make contact...
Reply:Heard the curved blades are better for total CFM while quieting down...but you're right, flat would make life a LOT easier...great link to Bedair's site, built his radius turning attachment a ways back, knew i had the roller bearing idea in my head from somewhere...know I know where!
Reply:Originally Posted by shepdogHeard the curved blades are better for total CFM while quieting down...but you're right, flat would make life a LOT easier...great link to Bedair's site, built his radius turning attachment a ways back, knew i had the roller bearing idea in my head from somewhere...know I know where!
Reply:In reading Bill pentz' site, there were two things that affected noise and performance.   Straigt radial blades like you've drafted were effective,but noisy....angling the outer end back from the inner leading end of each blade was quieter, but sacrificed CFM.  Finally, a curved blade kept the noise down, but gained back most of the lost CFM...I like the idea that using a flange even without rivits will be stronger...what do you think about just a series of rosette welds (ie drill the flanges, and just fill the holes to wled to the baseplate....strong enough?
Reply:Originally Posted by shepdogI like the idea that using a flange even without rivits will be stronger...what do you think about just a series of rosette welds (ie drill the flanges, and just fill the holes to wled to the baseplate....strong enough?
Reply:My short test run had the missus down stairs with that special "WTF???" look she gets...translation: if it's gonna be that loud, fugheddaboudit.
Reply:Think I'll fab a straight blade one with folded flanges just to get her running...we can experiement from there if need be.
Reply:see the mower vaccum thread if you wish to see my fan similar to yours but really heavy.alsoyou can put a flat disk / ring about 1" wide at the top of the blades to stiffen it up dramatically without effecting the efficiency etc..like thishttp://www.hobartwelders.com/mboard/...=&postid=33827Last edited by dsergison; 06-29-2011 at 03:08 PM.
Reply:I think you are correct about taking your time welding , allowing material to completely cool between welds and backstitching. I guess if I were looking for something like this I would have to figure my hours to complete the project and what time I had to spend on it, ie, working hours, woulda been sittin in front of the tv anyway hours, listen to the old lady nag me hours, etc. Depending on how you value your hours , 200 might not be all that bad. Mike
Reply:I guy I know used to do production runs building fans for a company. I belive all the blades were intrmitantly welded, but I can't remember for sure, that or they weren't 100% done when I dropped by that day. I do remember him balancing the units. He had a simple bearing mounted shaft that let the fan hang vertical like the wheel on a car. He'd mark the fan for top and spin it and wait til it stopped and check if the mark was still verical. He'd do this 3 or 4 times. If it kept coming up at the same place, that was the light end and he'd do a small tack there to add weight and repeat the process. When he'd get it to the point it wouldn't come up at the same location every time, he'd set it on the "done" pile. Whether the company that finished the fans did more I can't say..No government ever voluntarily reduces itself in size. Government programs, once launched, never disappear. Actually, a government bureau is the nearest thing to eternal life we'll ever see on this earth! Ronald Reagan