Double wall drum

[Axel]

The high airflows on a fluid bed roaster are generally thought to have less complexity, but not necessarily less body. For a drum roaster, the airflows can be much lower since the mixing of the coffee is done mechanically by the rotating drum as opposed to the air. Also, on most commercially made drum roasters, air is passed through the drum to not only roast the coffee, but to remove chaff. If chaff remains in the drum, it will burn and impart a smokey flavour to the coffee.
The heat transfer to each coffee bean within a drum is approximately 80% convection from the air, ~10% conduction from bean to bean, ~5% conduction from drum to bean, and less than 5% radiant from drum to bean. If there is no airflow at all, the other pathways dominate more and the drum needs to be a lot hotter to roast in the same time. The flavour profile will be different.
Specific flavour profiles (such as floral and high notes) depend more on the coffee type and the time/temperature profile it is subjected while roasting. All roasters will produce their own unique flavour profile and it can be difficult to replicate from one equipment type to another. In building your drum roaster, accept that it won't roast the same way as other roasters, but that doesn't mean it won't be capable of producing great roasted coffee.

Thank you ellatas
that is an explanation even I can understand. seems you have great in depth knowledge of roasting to be able to convey the idea so quickly and simply.
It sure reinforces my understanding of the roasting process as having infinite combinations of variables...

Convection
How should I figure out the air flow through the drum?
I realize that having fine control of airflow is an important feature but where should I start?
on a 24"x30" double wall drum(two 14gauge layers) rotating at about 50rpm, how much horsepower should the exhaust have available?
And more precisely.. how may CFM's(cubic feet per minute) should be a comfortable max ability?
would it be about the same with a single layer at about 11gauge?
I'm guessing that the inherent draw of hot exhaust air out the chimney/stack will do some of the work..?

Radiant
What is an optimal amount of space between the two 14gauge layers?
I've been told 1/2" can be found on some roasters, would less space(say1/4", closer to a single wall) allow for more responsive control of temperature?
I guess the question should be: if I could choose either a double wall or single wall drum, which one is better(has more relevant benefits)?
How much contact should the two layers have? just at the ends?

 

[peterjschmidt]

I still don't see any real advantages to a double wall drum that an accomplished operator would be able to provide on a single wall, and only see the disadvantages of manufacturing complexities and a slower response time in the profile.

 

[Axel]

I still don't see any real advantages to a double wall drum that an accomplished operator would be able to provide on a single wall, and only see the disadvantages of manufacturing complexities and a slower response time in the profile.

How thick should the single wall be(ideally) if the drum is about 24"ODx 30"Length?

Are any of the new probats made with single wall drums anymore?

 

[ellatas]

Axel, regarding airflow, a good place to start is knowing the pick-up velocity of chaff. This across the cross sectional area of your drum will give you a good indication of the minimum airflow. For the maximum flow, a good indicator would be to ensure you are less than the pick-up velocity of roasted whole beans as the air exits the drum. On some roasters, the air exits at the front above the discharge door then runs above the drum to a hot chaff cyclone (a so-called "Aroma Pipe"). The cross sectional area at the air discharge drives the highest air velocity and it is here you should ensure that it is always less than the pick-up velocity of the coffee.

 

[ellatas]

Double wall (or insulated) drums do take longer to warm up. However, once they are at temperature they provide a stable environment for consistent roasting. As discussed above, the primary heat transfer medium is the air which can be manipulated (in temperature and airflow) relatively quickly and does not have to wait for a drum to heat up or cool down. However, the double wall (or insulated) drums are more expensive to manufacture.

 

[ellatas]

Axel, regarding drum speeds, the RPM should be less than (30/pi).(g/r)^1/2. Too much rotation will cause coffee to ride on the drum surface throughout the rotation and consequently not be in the hot air stream. Also, most drums have counter acting flights with outer flights (mounted on the drum and furthest away from the center) pushing coffee to the door and inner flights pushing coffee away from the door. If the rpm is too high, the outer flights will dominate and push more coffee to the door. Too little rpm will also prevent the coffee from being in the hot air stream. The optimal rotation for hot air immersion is 10% to 20% less than the equation above.
Some roasters reduce their drum rotation at the end of the roast to remove the coffee from the airstream and consequently reduce the rate of temperature rise. The volume of the coffee has increased at this stage allowing for the balance of coffee moving forward and backward to be maintained at a slower rpm.

 

[peterjschmidt]

However, once they are at temperature they provide a stable environment for consistent roasting. As discussed above, the primary heat transfer medium is the air which can be manipulated (in temperature and airflow) relatively quickly and does not have to wait for a drum to heat up or cool down.

Which of course describes any old drum roaster.

I'm not trying to be argumentative, and am not even asking you to convince me. I just don't think the beans care if they're in a single- or double-walled drum, as long as the person operating knows how and when to add heat/decrease heat/increase airflow.

 

[Axel]

Double wall (or insulated) drums do take longer to warm up. However, once they are at temperature they provide a stable environment for consistent roasting. As discussed above, the primary heat transfer medium is the air which can be manipulated (in temperature and airflow) relatively quickly and does not have to wait for a drum to heat up or cool down. However, the double wall (or insulated) drums are more expensive to manufacture.

If it costs me $97 more in materials to make a double wall drum than a single wall, should I do it?
basically either a single 11g wall or two 14g walls.

how much more time will it take to preheat the double wall? twice as long?

does the G series from Probat feature a double wall?

 

[Axel]

Axel, regarding drum speeds, the RPM should be less than (30/pi).(g/r)^1/2. Too much rotation will cause coffee to ride on the drum surface throughout the rotation and consequently not be in the hot air stream. Also, most drums have counter acting flights with outer flights (mounted on the drum and furthest away from the center) pushing coffee to the door and inner flights pushing coffee away from the door. If the rpm is too high, the outer flights will dominate and push more coffee to the door. Too little rpm will also prevent the coffee from being in the hot air stream. The optimal rotation for hot air immersion is 10% to 20% less than the equation above.
Some roasters reduce their drum rotation at the end of the roast to remove the coffee from the airstream and consequently reduce the rate of temperature rise. The volume of the coffee has increased at this stage allowing for the balance of coffee moving forward and backward to be maintained at a slower rpm.

Great info!
thank you ellatas

making sure I use the formula correctly make take a while...
is "g" gravity and "r" rotation(RPM)?
i'm guessing "pi" is 3.14?
what is"^"?

 

[MillCityRoasters]

Great info!
thank you ellatas

making sure I use the formula correctly make take a while...
is "g" gravity and "r" rotation(RPM)?
i'm guessing "pi" is 3.14?
what is"^"?

"^" is an exponent. ^1/2 is a notation for the square root.

Among a variety of other things, I believe the equation is explained in here somewhere: http://www.nt.ntnu.no/users/skoge/prost/proceedings/aiche-2005/topical/pdffiles/T9/papers/600b.pdf

I read this a couple of years ago and my eyes are still bleeding.

 

[Axel]

"^" is an exponent. ^1/2 is a notation for the square root.

The original paper is here: http://www.nt.ntnu.no/users/skoge/prost/proceedings/aiche-2005/topical/pdffiles/T9/papers/600b.pdf

I read this a couple of years ago and my eyes are still bleeding.

Thank you Steve!
I actually remember coming across this document in a forum thread at some point in the past few days but not paying it much attention.
now it's valuable lol

 

[MillCityRoasters]

Sometimes observation is as good as calculation.

If algebra isn't your forte, you can simply watch how the beans cascade in the drum at various speeds. There's going to be a speed that most effectively fluidizes the bean mass at the height of your airflow.

Once you have decided on a specific RPM, it's easy to match a set of sprockets to that rotation.

 

[Axel]

Sometimes observation is as good as calculation.

If algebra isn't your forte, you can simply watch how the beans cascade in the drum at various speeds. There's going to be a speed that most effectively fluidizes the bean mass at the height of your airflow.

Once you have decided on a specific RPM, it's easy to match a set of sprockets to that rotation.

Algebra shouldn't be a problem now that I know the values.

Still trying to decide between single or double drum....