Change of Plans: Prototyping vs. Modeling

2011 March 17

In an earlier post about modeling versus prototyping, I had leaned toward modeling the bowl to get some initial variables figured out. As I met with machine shop owners and started hearing how much a bowl might cost, if fabricated in conventional ways, I was even more convinced that modeling would be MUCH faster and cheaper. A single bowl would cost many thousands of dollars if machined from steel and stamping it was out of the question as making the stamping dies would be even more expensive. Furthermore, the likelihood of getting it right the first time was remote, meaning that the bowl would have to be tossed out and a new, better bowl of similar expense would need to be fabricated. How many times this would need to be done to get to a final design is unknown and the expense: astronomical. Computer modeling, would be much cheaper and faster plus it is easy enough to change variables (such as bowl diameter, dishing, bowl speed, etc.) and run another set of data.

A typical stamping press for large steel panels, such as car doors.

A suggestion by Tim Eckholdt of ACME Metalworks got me thinking about alternatives. Being in the machining industry, he knew full well the cost of producing one bowl of the size I needed. He suggested that I look for some sort of stock part that could be repurposed or re-machined to meet my needs. It would be much cheaper, if such an item could be found. I started racking my brain to think of what existing items have the same shape and are made of the same material.

One item that came to mind is a wok. It has the same shape and is made of steel. I researched commercial woks and they were all too small and the steel too thin. Then, growing up on a farm, I started thinking about parts I’ve seen on the farm. Discs used for plows are of a similar shape and of thick steel. I researched the available discs and they have the right diameter, but they are too shallow of a bowl. Lastly, I thought of a component used in other assemblies that seemed promising (the actual component will remain secret until later posts). After checking out a few websites, I found that a wide range of these components are available and that the size that I thought would work best was only $250 and in stock in Los Angeles!

A stainless steel bowl of dimensions suitable for the roaster, available from stock!

This means that for a small amount of investment, I can make a full-sized prototype of the roaster bowl. Furthermore, since there are many sizes and thicknesses, I can swap out bowls until I find the best possible combination. As a visual thinker, I will be able to watch real beans move in the bowl and learn how to improve flow until it is perfect.

Since the cost of physical prototyping has gone down dramatically by repurposing existing mass produced parts, the cost of doing computer modeling has become comparatively expensive. I am excited about creating a physical prototype in the next few weeks and start moving the Roaster Project from strictly ideas and drawings into a real machine.


Comments are closed.