Who is here? 1 guest(s)
 Print Thread
Fluid Bed Area
drewblank
Hello everyone!

I've been diving deep into the rabbit hole of building a fluid bed roaster. I've already built one using a hot-air popcorn popper and I'm pretty happy with it.

However, I would like to build something a bit sturdier and with higher potential roast batches/better lofting/larger headspace for heating. As I'm starting to think about the design I ran into this question of fluid bed area. I've seen a lot of different designs on this forum that use the tri-clamp fittings for the roast chamber.

One thing I noticed is that there didn't seem to be any consideration for the actual diameter of the roast chamber. It seems to me there would be a direct relationship between the fluid bed area and the pressure generated. Therefore you'd want to gauge your bed size for the batch size and fan strength you're designing for.

I'm looking for somewhere around 100-200 grams per batch size, relatively low. Also, what should the shape of the roast chamber be? A flat bed like a fresh roast? A conical shape? A bowl shape?

I'm just wondering if anyone out there in the fluid bed world has thought about these design considerations.

Bonus question: has anyone considered designing fans using a 3D printer? I was wondering what that parameter space is like for increasing lofting on my popcorn popper roaster by designing a new fan and 3D printing it but I'm not even sure where to start.

Let me know, and happy roasting!
 
HarryDog
Hello Drew, what kind of popper do you have, where are the air holes?
I have used 3 different types, the one with side vents at the bottom I was able to increase the bean loft to 150g using a external 24v DC PSU, but this was starting to be a little inconsistent in evenness of the roast. Next was the vents on the bottom (Like Ikawa) and using the same PSU I could roast 200g but same issue of the evenness is starting to suffer. The third unit has a spray pattern using drilled holes on the bottom not a screen and was roasting about 200g as well but no issues with evenness in the roast. I replaced the fan with a shop blower, not as big as the yard blower. It lofts green beans with ease, tested at 227g so far (at 44%) but power to spare.
Now I don't know if my controller is all that linear but at 65% it is pushing a lot of air, I have not tried any larger charges as I'm buying beans by the pound, mostly. This is what my current roaster looks like lofting the 227g. Note the current controller in the video is at 45% just to push the beans above the SS cup.


This thread by Allenb has a interesting bean movement that's worth considering as well.
https://homeroast...ad_id=3174
Edited by HarryDog on 11/13/2022 12:28 PM
 
drewblank
I have the third type, where air is injected through perforated aluminum. I'm thinking about replacing it with stainless steel mesh because I think it will allow for better overall airflow and better lofting.

My only concern is that the heater won't be able to keep up, especially know that I'm roasting in a cold, unheated, garage. Hence building something new with more power output.
 
HarryDog
I don't have a screen version but think this would require a little more air volume to lift the beans but even a blower like what I have should lift them. I should add the glass section is 3 inches wide. The SS mug tapers down a bit.
https://www.canad...ml?loc=plp


Well I live in Canada, we have had well over 1.5 foot of snow, it dipped down to -27C, I roasted inside and run the exhaust for the roaster outside.
My heater should be about 1400 watts and I had problems at -10C. I dropped the fan to 42% then 41% as soon as it will lift the beans. Will have to bring up my charge temp some more for next roast. I turn off the furnace as it just pulls in cold air through the exhaust when it kicks in.

Will have to roast on warm days.

Was thinking of using a heat gun on my next version but maybe it won't be big enough for cold weather either? 540C on high.

Better start looking for something hotter as I would like to have heat to spare.
Edited by allenb on 11/17/2022 12:58 PM
 
allenb

Quote

drewblank wrote:

One thing I noticed is that there didn't seem to be any consideration for the actual diameter of the roast chamber. It seems to me there would be a direct relationship between the fluid bed area and the pressure generated. Therefore you'd want to gauge your bed size for the batch size and fan strength you're designing for.

I'm looking for somewhere around 100-200 grams per batch size, relatively low. Also, what should the shape of the roast chamber be? A flat bed like a fresh roast? A conical shape? A bowl shape?



I've never seen anyone post optimal fluidbed roast chamber dimensions and most of us just look at what has worked for others within our desired batch size and try and repurpose a glass or aluminum or stainless tube and go from there.
I can give you what I've seen and found to work out best for a 100-200 gram batch size. My Sirocco fluidbed made by Siemens company has a 3 15/16" or 100 mm inside diameter glass RC and works out great for it's 115 gram optimum batch capacity. It has a .75" diameter screened air inlet and the first 2.25" of glass going up from the screen is conical and then goes straight up and extends a few more inches. The bakearound glass tubes many have used are a little over 3.5" inside diameter and depending on bottom geometry, can handle anywhere from 100 to 450 grams but anything over 200 grams will be with some erratic chuffing action when the beans have fully expanded during the browning phase. This doesn't cause any negative issues but some prefer a controlled linear spout from start to finish. If using a full diameter perforated or screened bottom, you'll end up with chuffing from start to finish. You can go larger than 4" diameter for your desired charge weight and will require less blower static pressure but bottom geometry (cone angle) becomes very tricky to nail down for a good clean linear spout.

Hope this helps
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
drewblank
So I guess my question then is what are the advantages/disadvantages of the chuffing action versus the more controlled spout?

I can kind of handwave that the more erratic motion means more agitation which means more even roastings. To me, the spouting motion may be a more controlled motion but does not heat the beans as evenly.

I have zero data to back this up so I defer to those that have more experience. I guess I'm just trying to find the optimal parameters to inform design decisions.
 
renatoa
For me the most even roast comes from asymmetrical chute.
Least airflow and tuning required.
And also one of the very few allowing a realistic BT measurement.
 
HarryDog
Hi Drew, the spouting version I'm using with that fan I had it a bit high just to show it can move the 227g charge. When I roast its set to what I will call a rolling boil and the evenness is great.

I do like the asymmetrical chute that Renatoa refers to, that is shown in the Allenb link at the bottom of my first post, If you go through that you can see it in action and I think it will scale up well too. Just in case you decide to go 1 pound.

I like the tri-clamp design as you could change sections to test like different heating bases or longer/shorter viewing chamber and easily strip it down to clean it. The different parts are made to mate together as well so no struggle trying to make things fit.

Maybe we can start a chart of chamber types, motors, speed cfm/mph to give people a good area to start from?
Edited by HarryDog on 11/16/2022 5:39 PM
 
allenb

Quote

drewblank wrote:

So I guess my question then is what are the advantages/disadvantages of the chuffing action versus the more controlled spout?

I can kind of handwave that the more erratic motion means more agitation which means more even roastings. To me, the spouting motion may be a more controlled motion but does not heat the beans as evenly.

I have zero data to back this up so I defer to those that have more experience. I guess I'm just trying to find the optimal parameters to inform design decisions.


Not sure where the sources are from but I and some others here at HRO have heard over the years that coffee prefers to be in and out of the heat stream in a fluidbed as compared to all beans being immersed in the same roasting temperature throughout the roast. As far as differences between an intact, smooth spouting action where you have a clear separation between upward spout path and returning beans versus a chuffing action where there is a random pulsing bean movement, the one with chuffing action obviously has to have a return path but is intermixed with upward traveling roasting air in a random fashion. Could there be any difference in development of the coffee between the two that actually shows up in the cup? I wouldn't think so but who knows. Maybe someday, someone with both a regular spouting bed roaster and one with a flat bottom, full screen type might do some comparison tests using the same coffee and let us know if one or the other cups better.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
allenb

Quote

HarryDog wrote:

Hi Drew, the spouting version I'm using with that fan I had it a bit high just to show it can move the 227g charge. When I roast its set to what I will call a rolling boil and the evenness is great.

I do like the asymmetrical chute that Renatoa refers to, that is shown in the Allenb link at the bottom of my first post, If you go through that you can see it in action and I think it will scale up well too. Just in case you decide to go 1 pound.

I like the tri-clamp design as you could change sections to test like different heating bases or longer/shorter viewing chamber and easily strip it down to clean it. The different parts are made to mate together as well so no struggle trying to make things fit.

Maybe we can start a chart of chamber types, motors, speed cfm/mph to give people a good are to start from?


Your bean movement looks good to me, nice and smooth even at the higher flow rate.

I think your idea of putting together a spreadsheet showing dimensions and values that equate to good performance for a fluidbed is a fantastic idea. One that could be added to as an open source document. Could be a dedicated thread where everyone could discuss best practice and, as you mentioned, would be a great resource for folks just getting started with their build.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
HarryDog
Thanks Allenb, I thought the spouting version was solid and so far it has worked well. This model has actual holes drilled in the bottom plate (Not the screen version) I like the bean movement but I still don't think it's as good as that chute version you made. That looks to give you a solid controlled bean movement, like a wave above the returning beans below, that pattern looks to provide even heat to the moving beans with as little intermixing as possible.

On the spreadsheet I have 3 poppers I can add the info for and my current config that is working well. We could have a small comment section to add some info like stock motor at 20v roast 60g well at 24v it will roast 100g well. Inside and outside diameter of roaster parts or extensions used? Even things like I used a stainless steel mug for the base and the top to create an enclosed roaster for inside to vent exhaust.

Just let me know what info you want in it, we can get it started.
Edited by HarryDog on 11/16/2022 6:56 PM
 
greencardigan

Quote

allenb wrote:
Not sure where the sources are from but I and some others here at HRO have heard over the years that coffee prefers to be in and out of the heat stream in a fluidbed as compared to all beans being immersed in the same roasting temperature throughout the roast. As far as differences between an intact, smooth spouting action where you have a clear separation between upward spout path and returning beans versus a chuffing action where there is a random pulsing bean movement, the one with chuffing action obviously has to have a return path but is intermixed with upward traveling roasting air in a random fashion.


I notice that my conical spouting fluid bed roasters still have some airflow up through the returning beans. I see chaff moving up through the returning beans. But presumably the air moving up outside of the main spout is still cooler, or ends up cooler, than the main spout.
 
allenb
That would make logical sense to me. The same hot air that's forcing the central spout upward has to also be moving throughout the bean mass although at a reduced rate. I would think the temperature would probably not be much cooler than the main stream. So, with that in mind, my guess would be that the returning downward stream of beans are still being brought upward in temperature during their downward trip but at a much reduced rate of rise compared to the central stream of beans.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
HarryDog
On my roaster the inlet temp is about 35C hotter then returning beans, when first roasting on it I thought 30+ C difference something was wrong. I set the probe at about 2 beans width as not to get into the spouting stream.

Now the inlet temp was taken right at the base on the bean screen I call it.
 
allenb

Quote

HarryDog wrote:

I do like the asymmetrical chute that Renatoa refers to, that is shown in the Allenb link at the bottom of my first post, If you go through that you can see it in action and I think it will scale up well too. Just in case you decide to go 1 pound.



For anyone interested in pursuing the asymmetrical path, please see RoasterRob's excellent thread on his designs and builds. For my 1 lb fluidbed design, I utilized his existing design entirely and worked the first time with no tweaking. See the following portion of his thread:

https://homeroast...post_40101
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
drewblank
I really enjoy the details posted here!

I wanted just to show off what I'm working with:
https://photos.ap...aPJZRemFz5

This is a poplite with SSR for heater controller and a DC variable power supply for the fan. The DC supply actually lets me control the fan speed with the supply voltage. The bed is just the stock conical structure with a perforated aluminum(?) air outlet. There's also a hole in the glass for a thermocouple hooked up to artisan. I believe the roast in this video is a 100g charge. I am still very much learning how to roast with this thing. I wouldn't say I've gotten a "good" roast yet, but I'm on my way. I usually seem to end up underdeveloped I think. I'm aiming for light roasts.

I just picked up another poplite so I plan on using it to experiment with different bed designs to see if I can find any sort of systematic differences in outcomes. Will try a completely flat and asymmetric bed and post what I find. I'm also looking to use 3D prints to make a silicon gasket so I don't have to tape the glass tube to the blower/heater. I'd like to be able to friction fit the glass to the heater/blower.

I also need to design a chaff catcher. I'd love to use a 3D print, but I don't have the ability to print the material that would withstand the heat. I've seen CK in this forum use prints of high temp PLA that is then annealed. But you need a high temp hotend (up to 250C) to print. I really don't see how I could make something that isn't made of aluminum and mesh.
Edited by renatoa on 11/18/2022 1:26 AM
 
HarryDog
Hi Drew, it looks like you are off to a great start, I think you can even add more beans to the roast at that fan speed if you want to.
 
drewblank
I think I've found something that can be reconfigured as a chaff catcher.
Large tea/spice strainers! Seen here:

https://www.amazo...mp;sr=8-12

or

https://www.amazo...mp;sr=8-26

Now I just need to figure out how I can mount it to the glass RC so that it is easily removable.
Edited by allenb on 11/19/2022 11:21 AM
 
renatoa
Used the first model for exactly this purpose.
Cut the wires in the bottom and mounted there a funnel, made from some tin (from a can)
The hot air, together with chaff, will be exhausted through the funnel, the chaff will drop laterally, and the hot air will exit upward through the upper mesh.

How to mount is other story... mine was a perfect diameters match with the glass, 78mm ID, so just placed on top of glass, as a stopper. on a bottle.
Yours could require an adapter collar if diameters completely off.
~~~
Edited by renatoa on 11/20/2022 2:51 AM
 
CK
From a long time ago, this idea seen in action.

 
renatoa
Theirs is a much bigger version, imo... About 10-12cm in my eyes, if the glass is 80 mm.
Maybe a 4 inch, like this: https://www.amazo...ref=sr_1_6

At 2:08 you can see the funnel I typed above.
For funnel, look for an appropriate size sausage stuffer funnel Grin
 
drewblank
So my first attempt at making costume O-Ring gaskets was a bit of a failure. I am trying to use 3d printed molds to cast the gaskets in.

In my first attempt, the silicone (I'm using J-B Weld Red Hi-Temp RTV Silicone) just fused directly to the PLA. I bought some silicone mold release, so I will try that next.

Pic: https://imgur.com...

The mold has two halves that are taped together...when I tried to separate them the silicone just sheered in half.
Edited by renatoa on 11/27/2022 2:17 AM
 
drewblank
I just wanted to update I finally had success with getting a decent molded gasket using 3D prints and RTV silicone: the trick was using petroleum jelly to coat the mold surface. The gasket came out with no problem.

This gasket is used to friction mount the glass chamber to the heater/fan assembly. Works like a charm!

Still working on the chaff catcher assembly. Having a hard time finding a good heat resistant mount to get the tea/spice strainer I have mounted to the glass chamber.
 
Jump to Forum: