While the project went slower than I had hoped during 2012, much progress was made. At the start of the year, we were still testing the prototype. The main focus was on airflow issues. Here is a photo from January:
Now we are far into Module 1, having built the bowl system with a movable carrier and the frame. We are now working on the additional parts for this module. Specifically, we are building the skirt band that goes under the bowl to prevent beans from traveling under the bowl at the end of the roasting cycle. At this point the bowl moves down 3″, briefly exposing this undesired travel path. The ring is made up of a band the same diameter as the bowl, a support ring to hold this band securely in place and bracing arms that connect this support ring to the bowl carrier.
Currently being machined are the joining plates and bracing arms. I have also designed the automated lift mechanism for the bowl and these parts are being converted to CAD files. We wish you a Happy New Year and look forward to much progress in 2013.
During the month of October, we have been checking what we have built and making modifications to improve certain details. A key assembly is the bowl hub and shaft combination. I was not entirely happy with my design and decided to either remake the hub and shaft or modify them both. After discussion with Tim at ACME Metalworks, we determined that the modified parts would work well and be as durable as new parts.
The problem with my design was that there was no positive stop for the hub relative to the shaft and this introduced too much play in the bowl when it rotated. An initial fix was putting a pin into the side of the hub and through the shaft so that the mounting nut could be tightened. It worked great, but my concern was that this fix would make access to the thrust washer virtually impossible. If it were to fail, the roaster would have to be almost completely torn apart to access this tiny part. Access and repairability are part of good design, so I decided to make a different modification: reduce the diameter of the shaft where the hub is attached, creating a shoulder and also decrease the hole diameter of the hub so that it matched the smaller shaft diameter.
A new part also arrived: the linear actuator for the bowl carrier. This device will lift or drop the bowl carrier 3″ in just a few seconds. Dropping the bowl is how we are able to evacuate the beans when the roasting process is done. There will also be a manual way of doing this using a cam and lever arrangement.
I continued drawing new parts for this module. The focus is on the bowl skirt which is a stationary 4″ tall ring that is barely below the spinning bowl. Its purpose is to prevent beans from falling under the bowl when the bowl is dropped at the end of the roasting cycle. My drawings are currently being converted to CAD drawings that will be sent to a water-jetting company for fabrication.
Due to issues within our extended family, I have not been able to focus on creative pursuits. Now that things are better, I am back on the project. Acme Metalworks is doing some modifications on existing parts to improve their performance. They have also machined the counterweights to reduce the amount of effort needed to raise and lower the bowl. We have a linear actuator on order that will be automated source for lifting and lowering the bowl (a manual cam and lever approach will be used for hand operation). I am now designing the bowl skirt and skirt supports.
This week was lucky 13 of building the roaster. It was a week of assembly, rather than making parts, which was an exciting time. Assembly is when you find out if your individual component designs actually fit together properly and if the thing works. Last week, the frame was nearly completed and this week it would be the basis for all of the assembly.
Assembly started with adding components to the frame. We added feet to be able to level the frame. Next, we installed pulleys which will guide the cables for the counterweights. The counterweights will help reduce the amount of effort it takes to lift the bowl, reducing it from 250lbs to about 30lbs.
The linear guide rods were slipped into place. Each rod is made of hardened steel and is one inch in diameter. The rods guide the carrier with great precision. Each was bolted securely into the frame. Please note my mechanics, Cindy Johnston (my talented wife) and Erich Schultz (coffee cognoscenti).
The bowl was secured with six high-strength flanged bolts. We checked the spinning of the bowl. The next step will be for the run-out and linear movement deviation to be measured with accurate and sensitive equipment so that we know that the assembly meets our specifications.
This week was spent assembling and welding the steel tube frame for the bowl and bowl carrier. The 3″ square tube sections were precisely aligned and welded into place. After this, the welds were ground flush in certain areas where parts will mate.
The next step in the process is to design and build the framework that holds the bowl carrier and the shafts that the linear bearings ride on. We decided to use 3″ square steel tubing with a 1/8″ wall thickness. It will be made of several sections welded together into a rigid unit. I gave my drawings to ACME Metalworks and they started by creating a CAD drawing.
The CAD file was then used by the machinist to start drilling and boring holes for future connections.
They also machined sleeves for certain holes that will be welded into place and ground down. The sleeves are for areas where extra rigidity is need to prevent crushing the square tube. This sleeve is for one of the linear bearing shafts. As it is bolted into place, the squeezing of the top and bottom hex nuts could warp the tube.
As the tube sections are completed, the tubes are gathered together into a kit to send to ACME’s in-house welding shop.
Meanwhile, I’ve been working with the bowl drive motor to adjust the controls and become familiar with its operation.
The roaster project will ultimately be made up of thousands of parts, each of which will be designed by me or by another designer. While it is impractical to tell the story of each part, I thought it might be interesting to tell the story of one part. In this case: part number 1B-001, the “Upper Plate” of the bowl carrier assembly. It is about 20″ square and made of 1/2″ thick machined cast aluminum.
Before a part can be designed, the basic concept of how a completed machine will work must be thought through. This is the initial sketch of how the bowl support system might work. It included the idea of a square tube frame with four shafts that would guide the movement of some sort of carrier up and down.
The idea of what a carrier might look like began to take shape in rough sketches. There would need to be upper and lower sets of linear bearings to make sure that the carrier would not have too much play or twist. This would likely take the form of an upper plate and a lower plate with some sort of connecting pieces that could stiffen the structure without adding lots of weight and excess metal.
I started to research what purchased components would be assembled with the upper plate. The bearing for the the bowl drive shaft is one such part. I needed the specification sheets for each component to understand the performance and dimensions of each, so that it would mate perfectly with the finished part.
The next step was to do a series of design study sketches for how other components I was designing would fit. In this case, I had decided not to weld the vertical supports to the plate and, instead, bolt them on. This required ‘L’ shaped stiffeners to form the 90 degree connections. These would bolt to both the vertical supports and the upper plate.
At this point, the thought process was complete and the design studies drawn. It was time to draw the actual mechanical drawing for the part. I drew it to scale and in the nomenclature used by mechanical engineers for describing parts and dimensioning. At this point refinements such as cooling fins and smoothing radii were added.
I also created a table for the coordinates of the various holes in the upper plate, including the purpose and size of each hole. Normally this would be done in CAD, but I manually calculated the rotation of holes to each arm of the ‘X’ using old school trigonometry.
I assembled the upper plate into the bowl carrier assembly to check the fit of all parts. It fit perfectly! … And that is how the design process works for each of the components in the roaster project.
This week we continued assembling parts completed by ACME Metalworks and checking for fit and other issues that might become apparent. So far, everything is in good order. We inserted the thrust bearings and tested the fit of the drive hub. The bowl is bolted to the hub, so the fit of the hub is extremely important.
We also mounted the linear bearings and fit the precision shafts that the bearings run on. The bearings are made of Frelon®, which is in that same family as Teflon®. It is self lubricating and can withstand harsh environments.
The precision shafts come already hardened, so they were sent to a grinding shop to have threads ground into them. The threaded end of each rod will be bolted into the main frame of the roasting module.
We also received the controller for the motor that will drive the bowl directly. The motor is a DC brushless motor made in the USA by Bodine Electric. This controller allows us to adjust the speed, acceleration, deceleration, and direction of the motor.
While we traveled in California, Acme Metalworks was machining many of my parts for Module 1. As a reminder, Module 1 is the roasting bowl, its carrier, and the drive system. This happens to be the heart of the roaster and the most critical module of all. The alignment of the bowl and its ability to spin without wobble and to be moved up and down very precisely is absolutely essential to the proper operation of the roaster.
Previously, the bowl had been cut and machined to the right shape. Holes were bored for the drive hub which supports the bowl and connects it to the drive shaft. The bottom curvature was accurately measured and now the hub was machined for an exact match.
The carrier for the bowl, its drive shaft, and the drive motor was completed and given to me in pieces to assemble. The carrier is machined aluminum, comprised of an upper and lower plate attached with vertical web plates. Obviously this could have been a welded assembly, but given the fact that it must be very precise (welding could introduce warpage), would be difficult to fit into a machining station after welding, and that it is a single machine, we chose to bolt it together. A production machine would probably use welding to save cost.
The upper plate also has cooling fins to help bleed off heat transmitted to the bearings from the bowl. There will be a thermocouple or RTD bolted to the plate near the bearings to monitor the temperature.
The entire carrier will move smoothly up and down approximately three inches. Four precision vertical rods will guide the carrier on special linear bearings (eight of them). Both the upper and lower plates have machined tabs to which we shall bolt the linear bearings.
We decided to take a roadtrip to California that included Los Angeles and the Central Coast. While the primary purpose of the trip was rest and relaxation, we do love coffee, so coffee stops were an important part of the trip. Several places were just so-so, but we really enjoyed two roasters.
Verve Coffee, in Santa Cruz, does a great job of roasting and preparing coffee. We had tasted their coffee in San Francisco, but had never been to their stores. We decide to visit the roasting plant and espresso bar on Seabright. The sign on the building definitely caught our eye.
The inside of the roasting plant is neatly organized around their 60’s era Probat. This one is 15kg, but they are currently having a larger machine completely refurbished. The staff made us feel very welcome and their coffee was outstanding.
A week later, on our way back to Phoenix, we stopped by Handsome Roaster in the Arts District of LA. It is a very new roastery with some very seasoned owners. The roasting plant is right behind the espresso bar and features a lovingly restored 50’s era Probat 22kg machine. In speaking with Chris Owens, co-owner and roastmaster, they do all roasting manually.
The store is very well designed and they are attempting to simplify the coffee experience without dumbing it down. They also are trying to add kindness and approachability into the experience. We thought the drinks were excellent and that the baristas were informative and welcoming. Certain rules are applied which have gotten them into some bit of controversy: there is no sweetener and nothing but whole milk is offered. We applaud their work and wish them success.