Not a member yet?
Click here to register.

Forgotten your password?
Request a new one here.
You must login to post a message.

06/12/2019 3:57 PM
Just waiting on my TC4 shipment then it's roasting time

06/03/2019 11:37 AM
I rarely purchase roasted coffee. I just ordered 4 bags from Mountain Air Roaster.

05/30/2019 12:34 PM
Hi, I use a wok with a glass lid for roasting. shaking the wok is a good exercise, actually.

05/29/2019 9:08 PM

05/27/2019 1:56 AM
Kenya Nyeri Katogoto

Users Online
Guests Online: 5

Members Online: 0

Total Members: 6,013
Newest Member: frogsickle
In Memory Of Ginny

Latest Donations
Groveland Hill Roas... - 25.00
renatoa - 2.00
JitterzZ - 50.00
John Despres (Scene... - 25.00

View Thread

Who is here? 1 guest(s)
 Print Thread
Info On Centrifugal Roasters???
Terrific presentation! Thanks for finding that. I'm still puzzling over some of the terms though. Maybe "packed bed" is from the steep walls and necessarily a lot more centrifugal force packing the beans against the walls? With straight sides they would generate a thicker bed at the bottom, thinner at the top. Very interesting and a lot to think about.

I'm holding on to my prejudices. Even after seeing the roaster slide show I'm still guessing that the short fins in Chad's image are bean scoops and they could help bean circulation in both "bowl" or "packed bed" centrifugal roasters. I'm so sure I'm right I'll bet a half-cup of cold coffee that the fins are for beans, not air! Now how do we get an authoritative answer? I wonder if Probat would answer an email?

Anyway, one important point from the slide show, it looks like a homebuilder could make a wild "packed bed" roaster out of a pair of buckets! I like the idea of a narrower vertical roast chamber with the rising air flow.
11 years old... forever!
>home-built roasters and fair trade

The packed bed roaster (US Patent 5,292,005 et al) uses a stationary chamber, and the beans are circulated by high-velocity hot air fed in through angled louvres around the chamber (just like a popper!). It also uses a stationary deflector which circulates the beans from the top of the rotating bean mass back to the bottom of the chamber to ensure consistent roasting.

The bowl roaster does indeed rotate, and at such a rate that the beans move up the slope of the bowl, hit the stationary "fins" and are deflected back toward the centre of the bowl. In one description I have read, the moving bean mass is described as "toroidal" in shape, which may help you to visualise this thing in operation. The majority of the heating occurs in the "flying" portion of the bean mass.

I think you are three-fourths right, Brian. The Probat slide show of the Bowl roaster has heated air being delivered by blowing DOWN, into the center of the bowl. Yet, the image of the lid with fins clearly shows a situation where the air is being drawn UP, out of the roaster. It can't be both ways. I suspect that this lid is for a Packed Bed roaster.

What I find most interesting is that we are just now learning about these two, new roasting methods.
The lid that you see is in fact for a "bowl" roaster. Bean movement is entirely by rotation of the bowl. The fins we see are for air movement. The hot air blows downward from the big blower on top, takes a spin through the bowl then exits through two ducts in the top.

The main reason few people have heard of them is they have batch capacity's of 210-550 Kilos so they are not used in the specialty coffee industry.

Interestingly, that may change. It looks like Probat may be developing it for a smaller scale.
A life without coffee is a life not worth living.
cup-in-hand, There's a problem with your thinking, or something else is wrong with this picture. We both agree that the thing on top is a blower casing or volute (has a spiral shape). However, such blowers CANNOT run in reverse. That is, they cannot blow air out the center. The fluid always enters the center, is forced outward by the impeller vanes, and discharges tangentially. The blower cannot blow air down. Something is amis.

See also:
Dan attached the following image:

Edited by Dan on 04/24/2008 12:47 PM
Controversy is so much fun!!!

In the few numbers I've run on these techniques, there does seem to be a problem with scale on the bowl roasters. Small might not work too well. Achieving high bean-air velocities is easy in a bowl 9 feet in diameter at low spin rates. But in something the size of a normal bucket the required RPMs would be several hundred to over a thousand, with steep walls and huge forces pushing on the beans and the bucket. The basic equation for centrifugal force is proportional to the radius times the square of the rotation rate. I'm just guessing that the smallest useful size might be something like a washtub. Otherwise it might work no different than a typical small drum roaster.

I read Dan's links on "packed-bed" roasters. But, just to feed the fire of our discussion and guessing, I saw two things that make me wonder about explanations that suggest either that the inner roasting chamber is stationary or that the beans are in any way propelled around by force of the air.

First observation is that the animation of the Probat packed-bed roaster clearly shows the drum rotating in a direction opposite to the air flow. Which is what you might expect if they wanted to maximize the bean-air relative velocity. Second, I'm not seeing how you could ever use moving air to accelerate a rotating drum or beans and not have the beans then accelerated up closer to the speed of the air, greatly reducing the relative velocity.

I'm still betting that both the bowl and the packed-bed bucket are driven. And the fins at the rim are bean scoops!

Dan, I see what you're saying about the direction of that blower. This is all very interesting. How about a field trip to the Probat factory?
11 years old... forever!
>home-built roasters and fair trade
I read the 5159764 patent description and it clearly states that in a Packed Bed design the chamber (bucket shape) does NOT rotate. My reasoning is that this causes some friction between it and the beans, so the beans have a tendency to tumble over another, aiding in mixing and even roasts. Here is the complete description:


In its simplest form, coffee roasting comprises heating a single bean to a prescribed temperature at which point chemical reactions occur that transform the bean into the desired state of pyrolysis. These reactions occur in the last part of theheating cycle. Thus, the residence time at the terminal temperature is crucial because a difference in a few seconds in heat-history can have a significant effect on the taste of the coffee.

The problem is that it is difficult to design a roaster that will roast several hundred pounds of beans at one time and to roast every bean evenly. Whether the process for heat transfer is from convection, conduction, radiation, or somecombination thereof, the heat is absorbed in the first few layers of a bean bed. Therefore, it is desirable to establish some means for equalizing bean temperature throughout the heating cycle so that when the final roasting temperatures are approached,all of the beans will be close to the same temperature during the pyrolysis process.

The aforementioned problems have been overcome to a large degree by the Apparatus and Process for Conditioning Particulate Material disclosed and claimed in our co-pending U.S. patent application, Ser. No. 07/463,557, which was filed on Jan. 11, 1990, and which is incorporated herein in its entirety by reference.

In essence, a preferred embodiment of that invention comprises a chamber for receiving a charge of particulate material such as a charge of coffee beans. The chamber has a generally circular base and an upwardly extending divergent wall defininga segment of a cone with a central axis and closed bottom. The divergent chamber wall preferably forms an included angle with respect to a horizontal plane of between to and also defines a plurality of openings in a lower portionthereof. Means are provided for inducing a mass of heated fluid generally tangentially into the chamber to rotate the coffee beans about the central axis of the chamber and for maintaining the rotating material in a relatively densely packed orcontrolled state during the heating thereof. During the rotation of the coffee beans, the chamber is stationary, i.e., it does not rotate about its central axis so that there is relative movement between the rotating material and the stationary chamber. In addition, there is also vertical and radial movement of the coffee beans with respect to the chamber. This embodiment also results in horizontal shearing within the spinning bed since beans near the chamber bottom are rotating around the chamber at ahigher angular velocity than the beans near the chamber top.

In a preferred form of the aforementioned apparatus, a second chamber is provided and disposed below the heating chamber for receiving the roasted beans and for rapidly cooling the beans in a similar manner.

Coffee roasters in accordance with our earlier invention uniformly roast batches of coffee very rapidly with an efficient use of energy. They also provide conditioning, cooling, heating and roasting apparatus which is relatively flexible,competitively priced, relatively simple in operation, free of complexity and easy to operate and maintain. In addition, such roasters occupy a relatively small area and can be rapidly converted to operate under different conditions in a job shop type ofoperation while fulfilling most of the requirements for food processing.

However, it has now been found that an improved apparatus, according to the present invention, provides all of the aforementioned advantages while adding improved flexibility. For example, the apparatus according to the present invention, canaccommodate a greater range in charge variations, types of roast and greater control over the uniformity and quality of the roast.

The improved apparatus, according to the present invention, will accommodate relatively small to relatively large charges of beans and provides a more gentle bean action which results in fewer broken beans. The apparatus can also be programmedto maintain an optimal bean circulation during the roasting operation to compensate for the green to roasted bean expansion of almost 2:1. The apparatus also includes an improved bean spill which retains most of the coffee bean's tangential velocitieswhile directing them to the bottom of the chamber. This reduces the likelihood of bean carry-over caused by slow moving beans in close proximity to the central updraft of the air leaving the bean bed and exiting the top of the chamber and is thought tobe relatively easy to fabricate. The present design may also more readily accommodate a "well"-type temperature measurement which can be used to improve temperature control during the roasting operation.
It really is a giant popper. I can't imagine how violent the airflow must be.

Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
Hi Folks

Bowl / centrifugal roaster diagram from http://www.whydon...BAT101.swf

Illustrates blowing hot air down the centre and exhausting via a sort of coaxial sleeve setup to the blower/volute.

Clear as mud!


PS I have experimented with a pound of greens, a 14-inch wok (now with a neat hole in the bottom), 2 deflectors, and a battery drill - there are real possibilities here!
Next steps: pick up the beans off the floor Grin, improve the speed control, and add a heat gun (or two) :@.
Brainiac attached the following image:
I'm now quite sure that the Centrifugal (Bowl) Roaster has a rotating chamber and the Packed Bed does not.

The gearmotor to drive the shallow bowl is clearlyl visible in Brainiac's diagram (gray). What is also clear is something I learned in a search, that the center of the bowl has a stopper that moves out of the way to dump the beans down the center of the bowl. You can see some beans being evacuated in the diagram (lower right).

I had thought about using a wok as the bowl, too. Did you know? The reason the Bowl and Packed Bed roasters are used is because they are so good at heat transfer. And heat transfer works both ways. Consequently, it is not unusual for these production machines to dump their beans into another unit that is exactly the same except they don't have a heater. They just blow ambient air over the beans to cool them quickly.

cup-in-hand is sure that this lid was for a Bowl roaster. Based on this diagram showing a coaxial airflow, I agree. Heated air enters down the center, out that large opening. Recyleced air exits behind that cowl, the donut shaped ring. Mystery solved!
Dan attached the following image:
It's just so massive! It's hard to believe that all this equipment translates to dollars well spent in the production of coffee.

Brainiac, do you have any idea how many rpms this type of wok/heatgun might require? I can't imagine it's very fast... maybe 60 rpm? I think I have a 60rpm motor laying around somewhere.

Also, can you hazard a guess at the quantaty of beans that can be circulated?

...A lot to ask from your quick experiment, I know...

HG/Wok & TO/Wok. I can't wait!
Roaster: CoffeeAir II 2# DIY air roaster
Grinder: Vintage Grindmaster 500
Brewers: Vintage Cory DCU DCL, Aeropress, Press, Osaka Titanium pourover
seedlings, Its' call economy of scale. Things get cheaper as they get larger. There is a rule of thumb that says if a something doubles in production capability its price only goes up about 1.5 times. And, the cost to operate and maintain it is cheaper, too. And, it is more efficient, to boot.
WOW! Terrific info, guys. One thing seems clear, that just about any combination of heat and coffee beans will work. That, and we have a lot of options for new roasters!!!

I have an excel spreadsheet I've been playing with that calculates various combinations of bowl sizes, RPMs, and bean speeds. I'll gladly PM/email it to anyone interested. Or you could send me some measurements and I'll plug them in.

The difficulty in giving a specific answer is that it very much depends on the exact curvature and measurements of the bowl. But I'm figuring that a good first guess would be the rotation speed necessary to balance the force of gravity at the radius of the outer rim of the bowl/wok. That's the first table on the spreadsheet.

One important point... most standard bowls will probably not work well because the the angle gets too steep at the rim, and a very high rotation speed would be needed to make the beans continue upward to the bean cycling scoops or whatever. Flatter, like a wok, is better.

Looking at some simulations, there are a great many interacting factors in these high-speed air designs. Non-intuitive factors that will probably force many good attempts to work no better than the good old HG/DB or air popper. Seems like the most common downer is that the very first beans to contact the fast air are very good at either slowing it down or speeding it up to their velocity, and much of the expected increase in efficiency can be lost.

Still working on it. These roasting concepts are really interesting!
11 years old... forever!
>home-built roasters and fair trade
Jim, Yes, this is exciting. Two new roaster designs we hadn't heard of and both could be used for homeroasting. The Packed Bed interests me the most, but the Bowl is going to be the easiest to cobble up at home. Heck, steel woks are cheap, and variable speed motors are readily available. One could just use a heat gun or two for heaters.

The way I've figured it is that in a bucket with vertical sides, the centrifugal force needed to keep a bean suspended on the side of the spinning bucket is 1g force, not one gram, but one gravity. That's where the force of gravity is balanced. For a bucket (or wok) with slopped sides then it will be less and I think it is times the sine of the angle (use the tangent angle in curved bowls, measured at the rim).

This is true for Bowl or Packed Bed methods. However, the Bowl roaster also wants to do more than balance the beans. They need extra energy to toss the beans back into the bowl.

Of course the advantage of knowing all this is that we can build roasters BIGGER than a hot air popper!

I don't see the rpms being that high, even if you want to make sure the beans are tossed back into the center of a wok. My guess is in the 50-150 rpm range. That's not too bad, really. What does your calcs say?
Edited by Dan on 04/29/2008 2:19 AM
I am getting cranked up about these alternative high-velocity machines. These things are just too interesting not to try something myself.

The way the numbers look, the angle for the sides of a bowl/bucket to balance gravity equals centrifugal force is 45 degrees, not vertical (90 degrees). It would take an (almost) infinite angular velocity to make a bean stick to a vertical wall. But a bean might rise slowly up a vertical wall if it was being pushed from below by other beans coming off a slope. This is all Newton's fault. As the wall angle rises more than 45 degrees, the speed and force increase very quickly without giving much more vertical motion.

Probably this is best illustrated in two steps. Assume we have a straight-sided bucket. The bucket (picture attached) is about 30cm high, 40cm wide at the top rim, and I'll guess 20cm wide at the flat base. That would give a bottom radius of 10cm (0.1m), a top radius of 20cm (0.2m), and an angle of 71.6 degrees on the sides. Note that the real stainless steel bucket pictured does not have these exact measurements, but it's close.

Step 1: Put any size green coffee bean in the bottom of the bucket and power up the variable-speed motor to 164 RPM. The bean easily slides across the flat bottom and lays against the side of the bucket, but it doesn't rise up the wall. The bean is now 0.1m from the center of the bucket and all it's weight is pressing directly against the angled wall. None of it's weight is pressing on the flat bottom of the bucket, but it feels like it weighs 3.16 times heavier than when you took it out of the bag.

Step 2: Increase the speed just a little bit to overcome static wall friction. (170 RPM?) The bean will rise up the wall, slowly at first but accelerating quickly both around and upward. In somewhere around one second the bean will fly out of the bucket in some unknown direction on a low-angle trajectory tangent to the edge of the bucket rim. It's velocity will be 3.4m/s maximum, probably much less. (half?) The uncertainty is due to bean-wall sliding friction which is difficult to predict.

Friction would increase if there were a lot of little holes or slots in the wall. This would increase bean abrasion as well as increasing bean velocity nearer to the bucket rim velocity.

The problem with straight-sided containers is that they will always have this tendency to "launch" beans. At a constant RPM, the centrifugal force on each bean keeps increasing as the radius increases. Once the static friction is overcome, the beans will be quickly flung to the top rim. Increasing the wall angle may slow the bean's upward movement some, but the required increase in RPM and weight of the beans against the wall become massive. That's why the curved bowl is useful I think. Increasing the angle (curve) as the radius increases allows an even, slow moving bed of beans to slide up the sides. Maybe a two-step bucket would help, where the bottom half had a 70 degree slope and the top half had a 85 degree slope. So the top half would drag slower but be pushed up by an accumulation of bean mass from the bottom half??? Maybe it's OK having the bean mass mostly pushed up against the wall at the bottom as long as the beans keep moving to the wall and up.

Also, 2 or 3 m/s is certainly faster than drum speeds (about 0.5m/s), but air velocity in air-spout roasters is something like 15m/s. The physical bean motion would still not be enough to remove the need for a high volume blower to achieve a big increase in bean heating. In a bucket-size container, that would take something like 600 RPM. A mighty big fabrication problem in careful balancing of the parts and strength to handle imbalance in the bean mass. Bigger is at least easier in this case, higher bean speed at lower RPM. But maybe if we tried it, 3m/s would be a useful improvement in the roast profile?

Good thing I didn't loose my high-school Trig and Physics books. Was I supposed to take them back?
11 years old... forever!
>home-built roasters and fair trade
Picture was dropped from previous post. Here it is I hope.

Nope. Pictures are not being attached???
dBndbit attached the following image:

Edited by dBndbit on 04/29/2008 8:57 AM
11 years old... forever!
>home-built roasters and fair trade
Jim, You do need angled walls if you want beans to rise up the wall. Beans are round enough that they will roll up that bucket wall you picture if you spin it at the right rpm.

For the sake of discussion it is quite easy to get something to not fall down a vertical wall with the correct rpm. That is whenever the centrifugal force is equal to or greater than one gravity. It is lnot an almost infinite rpm as you claim. My proof? Recall the amusement park ride called the Gravitron (aka Rotor) where people stand in vertical alcoves along the inside rim of a large wheel. The wheel spins up, the floor drops, and the passengers stay stuck to the wall. It does all this at a mind blowing velocity of, here it comes, 24rpm!


Well I'm going to the SCAA conference this weekend, and my registration confirmation arrived today. As I'm looking at the classes that I signed up for, I see that the instructor for the "Thermodynamics of Roasting" class is non other than Karl Schmidt president of Probatburns inc.. I'm going to try to ask him some questions about the packed bed and centrifugal roasters concepts, as well as stop by the Probat exhibit booth.

I'll pass along whatever I learn and if anyone has any specific questions they would like me to ask, let me know before this Friday and Ill see what I can find out.
A life without coffee is a life not worth living.
Anton, That would be great! You might ask what the problems they encounter designing and building them, and what challenges roasters face using them. That way we'll learn their negative aspects and shortcomings
Dang! SCAA... Hawaii... Double dang! Anton, thanks very much! When you talk to Herr Schmidt, make sure he's standing where he can see the beach. It's very relaxing, even hypnotic. He may reveal more inside details on how the roasters work.

Dan, I wasn't suggesting that a centrifugal roaster doesn't need some kind of slope. The question is to compare the overall bean action on a vertical wall (soup can), a single-angle straight wall (bucket), and a upward curving wall (bowl). Or some combination thereof.

Don't confuse common sense with Newtonian physics, force vectors, and rotating frames of reference. The ride you mention works because the human body offers a huge coefficient of friction when pressed against the wall. A round object on a smooth wall would have no trouble rolling down the wall on that ride.

Using the numbers from the very interesting write-up, the riders would indeed feel more than 4Gs on the ride, 45.3m/s^2! But rotating horizontally there are no forces acting to cancel the downward force of gravity (9.8m/s^2). Add the vectors and the net force on the riders is still at a downward angle, and a ball on a smooth wall will roll down. Think of the huge coefficient of friction of trying to drag four stacked bags of coffee across the kitchen floor! There are no opposing forces sideways on the bags, yet you probably couldn't move them just because of friction. However, it would be easy to roll a ball of equal weight.

Only when the ride tilts up at an angle do the centrifugal forces begin to add or subtract in the up/down direction. So that ride can tilt all the way up to 90 degrees like a Ferris wheel. Rotating at 24 RPM, at the top of the rotation 1G of gravity downward is canceled by the 4Gs of centrifugal force now pointing straight up, and the riders are still being forced upward by 3Gs. Of course, at the bottom of the ride centrifugal force and gravity point in the same direction. The forces add to make the riders feel 5Gs of downward weight.

This is all Newton's fault.
11 years old... forever!
>home-built roasters and fair trade
Jump to Forum:

Similar Threads

Thread Forum Replies Last Post
Probat Saturn Centrifugal Roaster Other Roasters 8 05/02/2019 12:37 AM
Info on DC motor power of poppers Popcorn Popper roasting 12 04/16/2019 4:20 PM
Wood Fired Roasters Other Roasters 1 04/13/2019 12:30 AM
NORCAL home roasters West HUGS 1 03/13/2019 1:53 PM
Grindmaster Model 350 - no info (?) Preparing Coffee 5 11/10/2018 12:25 AM
Homeroasters Association Logo, and all Content, Images, and Icons © 2005-2016 Homeroasters Association - Logos are the property of their respective owners.
Powered by PHP-Fusion Copyright © 2019 PHP-Fusion Inc
Released as free software without warranties under GNU Affero GPL v3
Designed with by NetriX