Field Trip Notes – Phoenix Analysis & Design Technologies

2010 May 28

As stated in previous posts, I believe that computer modeling of the roasting vessel is a critical step in developing a workable design. By using a computer model, I will be able to dial in the likely dimensions and shape of the bowl. I ‘ll also be able to visualize the flow of the beans within the vessel to have a better understanding of the dynamics of the process.

A commentator on this blog (Jeff Strain) suggested that I check out his firm, Phoenix Analysis & Design Technologies (PADT). They do computer modeling for all sorts of mechanical systems and fluid flow. Jeff showed me a simple version of what they could do on his laptop. I was intrigued enough with the power and flexibility of the model to schedule a visit to their offices in south Tempe. Our photojournalist, Chanelle Richardson also went on the trip.

We met with Jim Peters and Clinton Smith, who are engineers in the division that does computer modeling. As an introduction, I described the roasting process and some of the important parameters. We then discussed drum roasters and finally the basics of my design (a bowl roaster). Jim and Clinton explained some of the ways they could model the roasting vessel. We jointly decided that modeling 66,000 individual beans (approximately the maximum capacity of the roaster) flowing around within the bowl was both reasonable and likely to give the best results. They showed us how the modeling results might  be visualized and gave lots of encouragement.

PADT also does “Rapid Manufacturing”, which mean that they have the capability of converting a CAD  drawing (essential a part drafted on the computer, rather than on paper) into a finished part using various 3-D printers. Jim and Clinton took us on a tour to see how this is done. They use several different methods depending on what the final part looks like. Most rely upon lasers to “draw” the part, layer-by-layer into either a resin, which hardens, or a powdered material which melts and then hardens. Another method uses tiny blobs of molten plastic deposited in layers. The results are nothing short of phenomenal. The plastic parts can then be used directly or as the form for creating a mold.

We were shown samples of what can be done. Lots of medical devices such as individualized hip replacement components were made in titanium from the plastic parts. To demonstrate their capabilities, they also made such playthings as Rubrik’s Cubes all done in one shot with full movement of all of the pieces. It was very impressive. Rapid manufacturing is something ideally suited for a project like the Roaster Project where only one machine is being made. I could envision prototype parts being made for evaluation and mold for casting metal support structures. All-in-all it was a very enlightening and encouraging field trip.

2 Responses
  1. May 29, 2010


    I’m sure you’ve thought of this, but I’d be curious how the modelling suggests smaller quantities of beans will flow. 1/4 load, 1/2 load, and 3/4 load all seem like intruiging quantities to me.

    • May 29, 2010

      Steve, as a roaster, you know the value of being able to vary the batch size and get a consistent product. One of my goals is to be able to control enough of the variables, such as bowl rotation speed, airflow, heat input, etc. that I will be able to roast a broad range of batch sizes, as well as cacao, malted barley, and nuts. I appreciate your input as someone who roasts as a part of his business!

Comments are closed.