Homeroasters - Home Roasting Coffee Community » BUILDING A ROASTER » Fluidbed Roaster
Who is here? 1 guest(s)
Page 1 of 9: 1234...9
 Print Thread
Bubbling Fluidized Bed Roaster Build
jkoll42
Top  #1 Print Post
Posted on 03/02/2011 10:10 AM
At the smart suggestion of Allen, I am starting a new thread since I have settled on the Bubbling bed to make searching/identification of the build easier. I am transferring the posts from the previous thread that are directly relevant to this roaster design into this one. All updates to this particular roaster will be posted here. For reference, the two original threads leading up to this design choice can be found here:

http://forum.home...post_27377
http://forum.home...post_26924
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
jkoll42
Top  #2 Print Post
Posted on 03/02/2011 10:12 AM

Quote

jkoll42 wrote:
Posted on 02/16/2011 09:18

Also wanted to relay some good news - got in touch with Sievert USA and

the will take direct orders for torch parts.

The 2941 was $30USD +/- plus shipping.

>>>>>>>>>>>>>>>
They can call direct to 815-639-1319 or 877-639-1319,
or if they want to e-mail the order - then I can call them or they can

call me for the CC #s.

Glad I could help. If there are other parts you are looking for I can

get more pages of the catalogs to you.

Becki Fairfield
Customer Service Manager
Sievert Industries, Inc.
Phone: 815-639-1319
Fax: 815-639-1320
www.sievertindustries.com
bfairfield@sievertindustries.com


Quote

jkoll42 wrote:
Posted on 02/24/2011 16:20
Care package arrive from Brown today - Thanks again Scott! Got

her loaded in the bottom of the test chamber. It looks like a missile

or something (unfortunately it would only give Gaddafi a headache).

Cant wait to start testing with the new fan
PICHERE


Quote

jkoll42 wrote:
So... I just couldn't resist giving the bubbling/boiling bed design I

laid out a try. I drilled out the base plate quick to give it a shot

with my first design and despite a bunch of holes drilled together

(have to reevaluate the area for the high density) I am really happy

with the results. Lots of bean movement, very few breaking off the bed

and the vac motor is barely above idle. VERY happy with the result for

the first shot!
[video]http://www.youtube.com/watch?v=Jw4794GeX0M[/video]


Quote

greencardigan
Posted on 02/24/2011 18:33


Looks good to me. Do you think you are using less airflow then the

previous tests? Do you have a picture of the floorplate used in this

test?


Quote

allenb
Posted on 02/24/2011 19:59


Wow! I'm very impressed with the bean movement you've achieved

especially with the fact you don't have any stalled perimeter-wall

pockets which is difficult to avoid with bubblebeds.

Definitely post your perf plate design. The one I used from Jobst

Zoellner worked very well in all areas but the perimeter which

circulated too slowly.

Are you planning on taking it to the point where it can roast? I've

been wanting to see a bubblebed actually roast coffee for quite a while

now and if you get one roasting I'll be like a kid in a candy

store!




Quote

Seedlings
Posted on 02/24/2011 21:02


Sweeeeeeeet! What's the perf pattern look like, and have you tried with

the heavier greens?

CHAD




Quote

jkoll42?
Posted on 02/25/2011 03:32

I will get the current perf pattern up this weekend. I still will do a

bunch more testing - peaberry, roasted, etc and will post them. Based

on how low the vac was turned on I am also curious how much weight I

can get to move.

And YES Allen the bubblebed will be a working roaster. My buddy has a

really good fabricator friend lined up so once the design is fully

worked out she will be be smoking.


EDIT - Scanned plate design and was going to upload but can't do so

with a pdf. Will scan to jpg tonight and put up.



Quote

jkoll42
Posted on 02/25/2011
Base Plate attempt #1 as in above video. 1/8" holes. The actual plate

didn't have the full number of holes in the high density (HD) areas.

Holes were too big for the number and area so I just kin of drilled it

out. New plate I drew up today has some mods and hopefully will test

tomorrow. Anyone with suggestions chime in!

-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
jkoll42
Top  #3 Print Post
Posted on 03/02/2011 10:13 AM

Quote

allenb
Posted on 02/25/2011 19:57

"jkoll42 wrote:
Base Plate attempt #1 as in above video. 1/8" holes. The actual plate

didn't have the full number of holes in the high density (HD) areas.

Holes were too big for the number and area so I just kin of drilled it

out. New plate I drew up today has some mods and hopefully will test

tomorrow. Anyone with suggestions chime in!"


In a previous post on the bubblebed Chad had asked if one might be able

to substitute wire cloth over large openings in place of the high

density perf areas. It would be interesting to try this and if it would

work it would avoid having to drill a whole lot of holes. You'd have to

do some math to try and come up with the total open area of the high

density areas and do an equivalent area in screen over the large

openings.

I'm thinking this would work but would take some experimentation.

I don't think there are any concrete rules regarding the layout of the

holes as long as you're following the basic concepts of the perf

layout.

I hope you know you will be breaking new ground here with a functioning

bubblebed roaster as I am not aware of anyone in the country doing this

as of yet except for some spinoff designs using different means to

achieve a type of rotating/pulsating bed design.

Allen


Quote

jkoll42
Posted on 02/26/2011 06:09

I will give the screen idea a shot. I have a new plate I made last

night with significantly better and more even bean movement so I will

sacrifice the old plate for the experiment. The old one looks like a

tornado and the new one looks like quicksand. Have some tweaking to do

on the new plate - some beans hanging on the edge but I have some

ideas. Now I just have to get through 4 hours in the office before

freedom!


Quote

freshbeans
Posted on 02/26/2011 07:23

Man, you get right to it!! I think quicksand will require less heat,

and give you less stripping. Sounds fantabulous! -Scott



Quote

greencardigan
02/27/2011 16:21
I've just been looking at all your new videos on youtube!

To me it looks like V1 with the previously posted perf plate has the

best movement. The others seem to have slower movement around the

perimeter at the base.

A big thanks for conducting and posting all these tests!



Quote

jkoll42
Posted on 02/27/2011 16:48

Thanks. Allen made a good suggestion to make a new thread with a

searchable title for bubble roaster so in the next couple of days I

will combine this thread and the other and start new. Would have posted

them todya, but my wireless keyboard decided to flip out and randomly

send signals for backspace so I have to grab a new one tomorrow.

Working now for some reason.

I have to work on the perimeter movement at the base. I think the first

was a too irregular in movement for my liking, but the others were a

little slow on the base. Will do some tweaks. The 6" plate is easily

pusing 2# so I mayh go that route and freeze beans.

Keyboard beinga pain in the a** again!! Arrrrggh. Bye.
jkoll42 attached the following image:
base1_1.jpg
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
jkoll42
Top  #4 Print Post
Posted on 03/02/2011 10:22 AM
[video]http://www.youtube.com/watch?v=LCFT10Hv3ek[/video]
V2 5" Plate 1.25# Load

[video]http://www.youtube.com/watch?v=MHPTcAPKy4k[/video]
V2 5" Plate 1.75#. This was just past the max this chamber would circulate without major pulsing.

[video]http://www.youtube.com/watch?v=Q5rvuEgWEAI[/video]
V2 5" Plate 2#. Overloaded chamber - past capacity

[video]http://www.youtube.com/watch?v=gd6l37KUkw0[/video]
V2 5" Plate 1# roasted beans. Could not get router control low enough to control beans. These were just store beans and were very large and light weight which made things worse.
Edited by jkoll42 on 03/02/2011 10:29 AM
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
jkoll42
Top  #5 Print Post
Posted on 03/02/2011 10:31 AM
[video]http://www.youtube.com/watch?v=hKtbULpZ5zw[/video]

Test for Chad. HD hole areas removed and replaced with a corresponding area of screening. It required a ton more air volume and produced an uncontrolled flow. It was also a pain to do - no easier than drilling. Good thought but no cigar :)
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
jkoll42
Top  #6 Print Post
Posted on 03/02/2011 10:37 AM
[video]http://www.youtube.com/watch?v=M89pZQW_aJs[/video]

6" Chamber with the same layout of the V2 plate just scaled up.

[video]http://www.youtube.com/watch?v=7T6qthqUk9M[/video]

2.5# of large PNG beans in the same chamber. 2.5 seems doable but probably the limit of for a controlled bed.

I will post the plate designs tonite. Next thing I need to work on is the beans that are stuck along the bottom edge. I have some ideas so that is next on the agenda.
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
Dan
Top  #7 Print Post
Posted on 03/02/2011 10:41 AM
I don't think pulsing is an issue. To me, the 2# is your best batch size. As long as you get through mixing, then all should be well. Because coffee beans aren't small and perfectly round, a true fluid bed is not going to work very well, so I'm not surprised that a large area screen didn't work very well. That's why the successful designs "spout" a portion of the bed and then let the beans slowly return to their spout.

Besides, a true fluid bed means that the beans are always "on the heat," which I think would create baked flavors. To my mind, what makes a successful roaster is one that puts the bean "on the heat" and then "off the heat" for a while.
 
Web
seedlings
Top  #8 Print Post
Posted on 03/02/2011 11:10 AM
In every one of your GREAT videos I think you could stand to turn the blower down considerably, and very much increase efficiency and batch size. As I see the bubble bed concept, the beans shouldn't bounce up and around, but more swirl within each other. This would mean less flow, less heat, more efficiency.

It's always nice for others to take time to fail my ideas for a change instead of me doing it myself. ThumbsUp Attaboy!

CHAD
Edited by seedlings on 03/02/2011 11:11 AM
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
 
Dan
Top  #9 Print Post
Posted on 03/02/2011 11:39 AM
Chad, I noticed the overly erratic motion, too. While you need some of that, otherwise you'd get no mixing, it seemed too energetic to me.

Another idea is to straighten the flow below the perf plate. The air that is feeding it is coming from the blower, which has a distinct swirling pattern. Perhaps the heating chamber is smoothing it out as much as you need. But, if you want more, then add a flow straightener directly below the perf plate. An easy way is to bundle together some 3/4" or 1" thin-walled tubing that fits all the way across the perf plate. They only need to be about 2-3" long. That's what they do to smooth the flow in some wind tunnels.

Dan
 
Web
JimH
Top  #10 Print Post
Posted on 03/02/2011 12:25 PM
In regards to the bean caught on the edge, I wonder if borrowing an idea from poppers would work. Put in downward aimed side vents to create a slight swirl to the bean bed, moving the beans away from the wall. Also, maybe not drill the perf plate all the way to the edge, the upward lift might resist the flow to the spouting areas, while the side vents would push the beans in. Unfortunately, I haven't thought of a convenient way to fabricate it yet. BTW, I am truly enjoying this project, some great ideas so far.

Jim
 
jkoll42
Top  #11 Print Post
Posted on 03/02/2011 12:36 PM
Chad - I was hoping that the mesh would work! I think I have carpal tunnel in my wrist now after 800 or so holes :)

I understand that the bean size and inconsistand shape will prevent a true fluid bed. Most of those tests were not good in my opinion but wanted to post them so others could see how loads and plates effect the chamber. Also, I really have been fluctuating on the actual capacity of the roaster. At this point I think that the 6" RC provides better control over bean movement and more room to improve the design. I need to prevent the beans from hanging on the perimeter and the V1 5" plate, with the HD areas closer to the edge did this, but with the sacrifice of tornado-like flow. Once I get things straightened out with that, I will start to vary hole size to see how that effects things.

Aside from the hanging beans, the 6" 2# load seemed best, and once the plate is modified I should be able to turn down the air.

Dan - I agree that experimentation air smoothing is worthwhile. My friend suggested a minimum of 3x the RC diameter to smooth the air as a starting point. Another thing that may play into it is that I am using sheet metal for the plate to ease drilling. I would imagine that plate steel would further smooth the entering air. Down the road I may give some tubing in the FC a shot to smooth air for testing.
Jon
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
jkoll42
Top  #12 Print Post
Posted on 03/02/2011 12:40 PM

Quote

JimH wrote:
In regards to the bean caught on the edge, I wonder if borrowing an idea from poppers would work. Put in downward aimed side vents to create a slight swirl to the bean bed, moving the beans away from the wall. Also, maybe not drill the perf plate all the way to the edge, the upward lift might resist the flow to the spouting areas, while the side vents would push the beans in. Unfortunately, I haven't thought of a convenient way to fabricate it yet. BTW, I am truly enjoying this project, some great ideas so far.

Jim


Are you reading my mind? I was trying th figure out a way to easily create some slight swirl but that could also be eventually fabricated (easily). Every idea got complicated other than angle drilling the perimeter holes which brings in a whole other set of trouble. I am with you on getting rid of the furthest row of holes and see if the beans will get sucked into the neg vacuum created from the updraft of the next row of holes.
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
dja
Top  #13 Print Post
Posted on 03/02/2011 4:36 PM
you have a dead spot in your roaster, that corner. If you can funnel the sides inwards some you will get rid of that dead area. I had the same thing in my roaster and the spot that was catching the beans and stopping them was only 1/8 inch wide but it put a dead spot in the flow that went completely around the roast chamber.

If you can control the heat so that it won't run away with you, you can turn the air way down.

how you want them to circulate I have no idea, I like for mine to move in a vertical swirling mass that is constandly rotating the beans.

Put something in the roaster about the same size as a coffee bean with your green something that will stand out so you can spot it easily, run the roaster with your normal load. watch and see how often you see that say white navy bean come by. In my roaster right now its about every thirty to forty five seconds. and as the beans lighten up and start to move it gets faster.
My penny's worth, inflation has hit hard in Oklahoma no more 2 cents worth.
David.
I pour Iron and roast Coffee BeansThumbsUp
If life seems normal your not going fast enough Mario Andrette
 
allenb
Top  #14 Print Post
Posted on 03/02/2011 10:24 PM
From my research on the bubblebed the goal should be to have low, gentle spouting plumes above each high density perf area and no more air flow than needed to achieve bean recirculation. As you're seeing, it's difficult to get the perimeter to drop back into the flow without a lot of agitation.

I've viewed the Nepro video more times than I'd like to admit and I don't know how they achieve a clean flow at the perimeter at the level of agitation they're at. Maybe they do have a sloped side wall just before it connects to the bottom plate. Hard to know and would not be very easy to accomplish I'm assuming.

In the video you can see the bed rotating and clearly see the stream of green entering the roast chamber traveling in an arc. I'm assuming they have a fairly high velocity convection fan at the top of the roast chamber that causes a sort of tornado affect which causes the bed to spin. Just a theory of mine. I'd sure like to see one up close in person.

Here's the link to the Nepro vid:

http://www.nepro-...asting.avi

Allen
Edited by allenb on 03/02/2011 10:31 PM
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
Unta
Top  #15 Print Post
Posted on 03/03/2011 6:40 AM
Maybe just putting a little suction on the top end of the chamber would be enough to keep things moving. Reduced pressure in the RC allows less pressure to be needed from below.
This concept is what makes my machine work..

Would be worth a try and a good reason to break out the leaf blower again..
Sean
Sean Harrington
educate.
 
Web
jkoll42
Top  #16 Print Post
Posted on 03/03/2011 6:58 AM
I too have watched the Nepro video many times... why the heck does it have to be such crappy quality!? I agree that it looks like they have a circular flow at the bottom. If that is the case, it would push the beans that are in the larger perimeter sections that are void of HP areas towards tehe upflow. The only legitimate way I can see doing it would be to have a thicker baseplate and angle drill the perimeter holes.

Going to home depot at lunch today and am going to grab a length of 8" duct. I am wondering if part of the issue is that the perimeter holes are currently so close to the side wall of the FC that there may a dead zone of air below the plate that is making things worse. Figure increasing the FC diam. would help reduce that. I also am going to switch to thin wood for testing of the baseplate. That sheet metal leaves all kinds of slag hanging off the back from the drilling, and this would also let me test the angled holes on the perimeter. Also easier to screw with as I can just putty over holes if I want to get rid of them.

When I stick my hand in the bed, the beans are definately all suspended and fluid. I may also try increasing the hole size slightly to give me a little more airflow at lower pressure but first I really need to get the dead area fixed. If things dont go any different tonite with the larger FC and thicker plate I guess it is back to the (literal) drawing board for a plate redesign.

Sean - Thas is an interesing idea
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
seedlings
Top  #17 Print Post
Posted on 03/03/2011 8:08 AM
If you think about water boiling, there are not 3 or 4 hot spots on the pan, the entire bottom is evenly hot. I wonder if the Nepro perf pattern is, similarly, even across the bottom? Allen, have you found their patent?

The Sonofresco circulation pattern is similar, but more inconsistent:
[video]http://www.youtube.com/watch?v=N3SU4U2sBqw[/video]

With blower control (that the Sonofresco doesn't have), I don't see why an even disbursement pattern wouldn't work just peachy.

CHAD
Edited by seedlings on 03/03/2011 8:15 AM
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
 
JETROASTER
Top  #18 Print Post
Posted on 03/03/2011 8:24 AM
I'm thinking the machine being built on this thread will blow the Sono away. JMO -Scott
 
Dan
Top  #19 Print Post
Posted on 03/03/2011 8:31 AM
No pun intended, right, Scott!? :)
 
Web
jkoll42
Top  #20 Print Post
Posted on 03/03/2011 9:22 AM
Big *groan* if that was a pun :)

Scott - the patent is here:
http://www.freepa...38303.html

A true fluid bed will have a standard pattern across the bottom but you are dealing with uniform particle size. With the irregular beans, the HD/LD design of the bubbling bed like the Nepro patent allows for suspended beans with upflow to circulate the beans in a way that prevents the uneven flow like in the video above.

Of course like any good patent, the parameters given are so wide, you really can't know exactly what ratio's and sizes actually work. "The HD areas # of holes is 1.5x-3x the number in the LD areas"
Edited by jkoll42 on 03/03/2011 9:24 AM
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
JETROASTER
Top  #21 Print Post
Posted on 03/03/2011 9:43 AM
The pun was not intended....but thanks for the roasting all the same.




.....:) -Scott
 
jkoll42
Top  #22 Print Post
Posted on 03/03/2011 10:03 AM
I ought to bean you in the head for that one :(

Slapped together big bertha test setup to eliminate possible wall turbulence, increase height to smooth out air , and not have it balanced on two chunks of 2x4 and the constant tipping over.

36" tall 8" duct reduced to 6" at base to hold vac. Stuffed into vented bucket for stability with a 7" opening at the top to support bottom plate drilled for 6". Will work on plates this weekend cause I'm done for the nite!
jkoll42 attached the following image:
img_1236.jpg
Edited by jkoll42 on 03/03/2011 7:04 PM
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
allenb
Top  #23 Print Post
Posted on 03/03/2011 7:10 PM

Quote

seedlings wrote:
If you think about water boiling, there are not 3 or 4 hot spots on the pan, the entire bottom is evenly hot. I wonder if the Nepro perf pattern is, similarly, even across the bottom? Allen, have you found their patent?

The Sonofresco circulation pattern is similar, but more inconsistent:

With blower control (that the Sonofresco doesn't have), I don't see why an even disbursement pattern wouldn't work just peachy.

CHAD


The only patent I've found on Zoellner's design is for his perf plate layout. I'm certain the Nepro uses something on a theme of the HD/LD perf plate otherwise you wouldn't be able to see the 4 plume tops seen in the Nepro vid.

Allen
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
Unta
Top  #24 Print Post
Posted on 03/04/2011 12:41 AM

Quote

allenb wrote:


The only patent I've found on Zoellner's design is for his perf plate layout.
Allen

Link?
Sean Harrington
educate.
 
Web
jkoll42
Top  #25 Print Post
Posted on 03/04/2011 6:39 AM
I posted the link above - post #20
-Jon
Honey badger 1k, Bunn LPG-2E, Technivorm, Cimbali Max Hybrid, Vibiemme Double Domo V3
 
Page 1 of 9: 1234...9
Jump to Forum: