9 replies to this topic

[Nathaniel]

This week at ALL Power Labs we've finally redesigned the fan, run a new set of tests, and found much improved static inches/H20 vac performance. (and reduced noise levels).   The goal of this new design effort is to produce a low cost fan that has the needed vac and heat tolerance needed for wood gas rigs.  Unfortunately there is not an inexpensive source for such a thing from common sources.







You need about 5" h2o vac min for a reasonable gasifier start up and run in fan. 10" is better and will let you do the whole basic range of run concern. 20" will let you do the full range for complete testing and characterization.

The ejector/venturi that comes with the GEK will do 20", but requires compressed air. This is fine for testing environments and first world run situations. It is less good for dev world and general boonies run situations. A fan is much easier and with less equipment dependencies. However, for it to work reasonably in off grid situations, it needs to be 12vdc. 120/240vAC fans don't help if you don't have ready power before you start your gasifier. Most high vac high heat tolerance fans are 120/240vAC solutions.

Thus we're trying to get to the minimum build and cost requiremets for a basic starting fan for a gasifier. We think this should have about 10" of usable vac. The original gek fan would do about 4" vac, which proved to be not enough once filter, nozzle and burner losses were added.

Standard squirrel cage fans produce about 1". Shop vacs will do 20"+, but they get fouled, are made of plastic, and generally fall apart. Proper forge combustion fans start in the $500 range, which is too expensive. We need something we can sell in the 100-200 range. For a 12vdc solution, this requires working with car/truck Fasco type fan motors, and designing the fan in relation to the available power, not the other way around. For the full range of proper high pressure/vac large fans, see cinncinnati fan here: http://www.cincinnat...03-internet.pdf


So here's what we've done so far within these constraints:

-created a fan that for the most part bolts together during production. This allows for less labor on the welding side but also for a thinner fan.

-introduced the use metallic gasketing via foil tapes to create a high temp seal.

-increased the rotor size to ten 10"DIA plus.  will likely increase it some more.

-lowered the overall width of the fan cowling to 1-3/8" and the blades to 0.7" for less bite and power consumption.

-introduced a bolt on no weld hub. for improved balance and service life.

-introduced a needle bearing and a solid motor support for other performance improvements and adjust-ability for alignment.

And the results: My word what were we waiting for?

Under a static load. that is with no air being allowed to enter the intake and just measuring the vacuume.
We achieved a 8.75"/h20 at 4100rpm. That's about 'double' the amount of static pressure from our last fan.

And have deemed this successful enough that we can ASAP (after our upcoming workshop) do one more round to finalize the design, and then offer this fan as a low cost, real solution for gasifier start up and moderate running fan.

For more pictures and general details about the fan, see here:
http://gekgasifier.p...Fan-Prototyoing

cheers!

[kenboak]

Nathaniel,

This all sounds like good progress.  

I struggled with the poor fan performance on my V2 GEK, which I think was the root of my tar problems, so this should get me going again.

Does it use the same motor - i.e. can I just buy the metalwork and swap my motor over or is this not cost effective?


Ken

[grizbach]

Great work, Nathaniel!
May I offer an idea about the rotor? Since you have access to a CNC plasma cutter. How about making the disk and vanes all in one piece? It can be made out of thinner material, there is perfect balance and no welding! You simply bend up the vanes and your ready to go! One rotor can be adjusted to the power of any motor by increasing/decreasing the bend of the vanes.  I have included a quick sketch of my thought.  It has three cutouts at the center to allow pull from both sides of the disk. So for this style, the tabs would be alternating, bent up and down.
Terry

ps. I noticed your vanes start at the hub. Have you found this to be most efficient?  I have no experience with blowers, but I think there would be less drag with the vanes just at the perimeter.

Attached Files

•  fan1.JPG   27.53K   72 downloads

[jimmason]

ken, this new design will work with your old motor.  however, the the new 12vdc motors are a little stronger than the old ones so we should just send you a full new rig when this design is ready.  still, you will have even higher operation range and more control with the ejector venturi.  i'd still suggest you use that.

terry,  yes we are also considering doing the rotor as a single piece with bent up blades.  basically the method that is used to make alternator fans on cars.  currently we are just trying to get the dimensions correct so it is easier to do this welded method instead of the more difficult cad problem of the fold up blades.   well, maybe.

the tabs just out at the end won't really work.  these fans are centrifugal fans.  they work by throwing air/mass outward.  the raw push of the blade is not important.  angling the blades so they have more "lever angle" towards the output actually reduces performance.  similarly, increasing the tightness of the rotor in the housing does little to improve performance.  we've now tested all of this.  nathaniel will post the results at some point.

what matters is having some acceleration surface from center to outer edge, and have the out edge tip speed as high as possible.  you get vac performance by the combination of diameter and rpm.  diameter has more impact than rpm for some reason i'm not fully clear on yet.  if you want to details, the cincinnati fan material is very good and worth consuming with focus.  there are formulas there for all the issues of concern.

the current new fan is still not the full vac we want.  we are going for something with a stall vac over 10"h2o.  that will make it run reasonably up to 8" or so.  that covers all the relevant start up and basic run situations.

the next steps towards this will be to reduce the rotor bite so the motor will reach power equilibrium at a higher rpm.  if we can't get enough vac with this, we'll continue to increase the diameter until we do.  i think we'll likely need another inch of diameter to get where we want.

we'll post the files when we have something that is good.  it is now exceedingly easy to make.  no welding either.

jim

[Nathaniel]

We are pursuing many ideas at once... with the workshop coming up, our focus has shifted for the next few days.

Great ideas have been presented. I plan on stepping through some of them with the current fan to see if any of them further improve efficiency before stepping up to a larger size.

As usual our piles of junk/resources/treasure around here keeps giving us new ideas. Jim found a great rotor from a vacuum cleaner that predicts our direction and encompasses most of the newer ideas.

In the next week or so I will get the new rotors/housing together and tested and presented, here, with some data.

We are hot on the trail... thanks for the input. More to come.

[Nathaniel]

kenboak said:

Nathaniel,

This all sounds like good progress.  

I struggled with the poor fan performance on my V2 GEK, which I think was the root of my tar problems, so this should get me going again.

Does it use the same motor - i.e. can I just buy the metalwork and swap my motor over or is this not cost effective?


Ken

I am working to create a 'universal motor mount' that should accept your older fan. However; I am not sure which one you recieved before. If you can tell me what the model is I can more than likely find that fan and see if it will work and test it out. The idea is to make the assembly in such a way that multiple fans could work so long as it meets the RPM/watt requirements. Of course we are trying to keep the cost down, 12vdc and so on.

[carson]

how about a cast aluminum fan (cast in the backyard)? the disc and vanes could be cast as one piece. the fan shell cast in two (dimensionally correct) pieces. the disc could be 'trued up' on a lathe and should run smooth. it would be light and non-rusting.

carson

[Nathaniel]

I have been spending the last week working through ideas and various fan blade configurations. We are making progress. For now, here is a photo of some fans and the various cowlings. Many ideas lead to 'slight' increases in efficiency, but there are a few surprises.

One consistent factor is tip speed.

As we are going to the Pacific Nut Expo, fan testing is suspended, however thank-you for many of the ideas filtered down to us and we have tried variations of most of them. Keep them coming if you have them. We are slowly developing a clear picture of the design that we will be able to achieve, and produce, using our current motor.

I will be back on fan development next week. And as we narrow it down, share the data with you.

All the best, nathaniel

[jimmason]

nathaniel's fan work is step by step eliminating all of my pet theories about how to make a fan generate more vac.  angled blades, bigger/lesser bites, curved tips, double face plates, increased tightness on intake, increased tightness of rotor to snail wall, etc etc etc are not really changing that much.  that is the big story here-- things are not changing that much.  we are maxxed at a top of 8" h2o vac or so with the current motor.

i think this might relate to the very basic and subtle fact that these fans are centripetal acceleration devices.  what these fans do is accelerate air by various means.  they are in no way pumps or air levers or linear pushing devices.   this is why things don't really change much when rotor / housing tolerance goes up or down.

if all that is at play here is centripetal acceleration of air, then its ultimate performance should be well approximated by F=ma.

in our case, force is limited by the 12vdc motor.  this motor reaches a stable max rpm when its force generation capacity is equaled by the accelerated air.  we can't change this force, like we could with a synchronous ac motor of constant rpm, but variable power.

simlarly, mass is not changeable, as air is air.

thus the only significant variable in our situation is acceleration.  

this suggests that any effort that lengthens the acceleration time will more efficiently use the available force.  diameter will certainly do this.  maybe long curved blades do this enough to do something meaningful.  

if you have non-limited force, then curved forward blades will force more acceleration for any given diameter.  but if force is limited, you'll just get to the max acceleration your force can create, with a lower rpm compensation, and that will be it.  any way you do this acceleration, as long as it is over the same period, will result in about equal end results.  

only if you can increase the efficiency of the acceleration (i.e. by increasing its time), will you get to a higher end air speed at the tips.  or that's the current theory (likely to change with more fan tests).

not sure if the efficiency of the acceleration vs drag developed during the process can be altered enough for meaningfully higher acceleration with the same amount of force.

j

[Ecnerwal]

While this is an old and theoretically dead-ended thread, as you have your new plastic blowers for furnaces coming along, for those of us in the "I'd rather cobble it up" camp, a bit more info from my vintage Buffalo fan engineering book, and some other thinking/observations.

The fan engineering book (copyright 1914/1925) provides a bunch of formulas I can get up here at some other time, and some very general principles that will make a more efficient fan - specifically, that small inlet diameter to large rotor diameter makes for an efficient centrifugal fan - if you don't want too much volume flow, then small inlet, large diameter, and not much height (or thickness depending how you think of that). Efficiency is going to be a big deal if trying to keep the power requirement down. For the seriously off-grid applications, a geared (or pulley or bicycle chain, easier to homebrew) hand (or bicycle pedal) crank comes to mind, thinking of forge blowers - if it's for a few minutes to get the flare burning clean, and then the motor starts, hand/foot operation might be just fine. I wonder if a bicycle wheel might not be a good start on a large diameter rotor with foot-drive system already attached.

The observation and thinking part which that book doesn't get into the details of is that a multi-stage fan is likely the way around too little pressure, and that's indicated in part by the fact that the shop vac (which uses a rotor of this type, typically, AFAIK) is working well (if at too much volume) That's a centrifugal fan rotor which is backed by a static vaned scroll (or stator) that directs air from the tips of stage one to the entry of stage 2, which is another rotor that either connects to another scroll and another stage, or the exit scroll.

There's also some different connotations of blade shape, as it affects horsepower drawn and volume/pressure with differing pressure or flow or fan speed. Sound reduction is not the only reason to bend blades backwards.

Finally, there are some possible efficiencies in paying attention to how the flow exits/transitions out of the fan. In particular, if it is exiting with more velocity that is needed for the application, a smooth and gradual expansion to a larger diameter will convert that velocity to useful pressure. An abrupt transition wastes much of the energy.

Trying to pull the relevant stuff from what is essentially a book for plant engineers to figure out what Buffalo Forge fan (circa 1925) they should buy is a bit tricky, but I'll pick away at it occasionally. The other side of this whole business is that a limited power budget limits what you can do, volume and pressure-wise - if you make the fan as efficient as you can make it (and at the home-brew scale that is in itself optimistic, but we can shoot for good design principles in that direction) you will hit a point where more air movement, suction, etc. simply needs more power dumped into the fan.

<edit - looks like I should have actually put this in the thread with Nathaniels stuff, as being somewhat more recent and looking at some, but not all) of this stuff.><more edit - looks like I can do that!>