John Socha-Leialoha
September 2003

O-Scale First Prototype

N Scale First Prototype

I recently went camping with my two brothers and their families. I don't normally like camping, especially far from trains. So it was with great reluctance that I agreed to go along. But the promise of great weather and company made it impossible to refuse, especially since it came from my wife.

Despite the weather (cold and wet), we did have some fun, particularly our 6-year old son. My favorite part, of course, was taking photographs of microwave towers since I'm currently working on a model for Mark Bridgwater's Feather River Route that will appear in the city of Sacramento.

Inspiration

On the drive back I got to thinking. Wouldn't it be nice to build an N-scale picnic table out of etched brass. I figured I could build it with just five parts to make it easy to assemble, and being etched, it could have the appearance of individual boards for the table top and bench seats.

I also happen to have recently purchased a small table-top CNC milling machine (a Roland MDX-20) in partnership with another modeler (Hans Spiller). These milling machines are pretty amazing: you make drawings on the computer and the milling machine simply cuts them out of materials like wood or brass. And you can get remarkably small cutting tools, such as 0.010" or 0.005" diameter cutters. However, as I was soon to learn, cutting brass with such small cuters is very difficult without breaking the cutter. My plan, you see, was to use the 0.010" diameter cutter to cut the brass so it would look almost exactly like a photo-etched part. I'd done this previously with the top section of the microwave tower I'm working on and it worked quite well, although it did take a long time.

The top tower section, by the way, is an N-scale model and is about 1-1/2" tall.

Research

The next day I got on the web. My favorite tool for research is Google. They have a set of tabs at the top, and one of these tabs is "Images." I clicked on this tab and entered "picnic table" into the search box. This provided over 5,000 images that were mostly of picnic tables. Exploring these soon lead to links with actual plans for a picnic table, including all the dimensions. This research project was much simpler than many projects since building a picnic table, as it turns out, is a common do-it-yourself project.

Creating the Drawings

My software package of choice for drafting is Microsoft Visio, which really isn't a drafting package, but it actually works quite well for that purpose. I do all my drawings in actual feet and inches. The final size of the image is determined by the scale of the page. So for the O-scale prototype of the picnic table, I set the page scale to 1:48.

Laser cutters, and also my milling machine software, allow you to cut at different depths, such as all the way through and half way through. I used the half-way setting to create notches in the table top and bench seats for the legs. These notches make this kit far easier to build since the tabs provide the precision and alignment.

You mark the different depths using different color. Here black indicates that the line should be cut all the way through. The red rectangles indicate that an area of wood should be removed, but at a different depth. I set black to cut all the way through, and red to cut half way through.

Proof of Concept Prototype

My first attempt, once I had the drawings finished, was to mill a "kit" in brass on my CNC milling machine. I started with a 1/32" diameter cutter to remove most of the material, and then I decided to "optimize" the second pass with a 0.010" diameter cutter so it wouldn't have to cut as much material. This is when I broke my last 0.010" cutter--it broke while trying to cut the slots between the boards on the top of the table. Hence, I ended up with a prototype model with solid slabs for the table and bench tops.

The result told me that an N-scale picnic table in etched brass would work, and that trying to simulate etching of brass with a very small cutter was just asking to break the cutter.

That's when I decided to shift both to a larger scale, and also to basswood, although I later learned that aircraft-quality birch plywood is stronger and cuts cleaner than basswood in a milling machine. The photo at the top of this page shows the basswood version of the table in O scale. The O-scale "kit" contains just nine parts, all assembled by gluing them into alignment slots. It took me about 5 minutes to assemble this "kit".

Once I received some replacement 0.010" diameter cutters, I decided to create an N-scale prototype of a laser cut kit. You can see the result on the right. The lines don't look as crisp as you'd expect from a laser cutter because there is some small sawdust in the cuts left by the milling machine, and which the vacuum wasn't strong enough to remove. I later used a small hobby knife to remove the sawdust.

A Trial Run on a Laser Cutter

Now for the next problem. I don't own a laser cutter. They're quite a bit more expensive than the CNC milling machine, which by itself was a difficult purchase to justify since this is a hobby for me. Fortunately, a little bit of searching on Google turned up a company less than four miles from my house.

Before doing a test run, I redesigned the kit a little. When assembling the first N-scale prototype, one of the diagonal braces went flying, and I couldn't find it, so I added extra parts to the drawings in case you break or lose a part. For this test I decided to create an HO-scale version of the kit.

There are some common file formats among people who do laser cutting and brass etching, and CorelDraw is one of those formats. I didn't own a copy of CorelDraw, but then I discovered that you can buy a previous version, now called CorelDraw Essentials, for just $79, so I purchased a copy. I then copy and pasted the drawings from Visio to CorelDraw. When sent to the laser, it produced the results shown here.

The alignment grooves, which are the thick black lines in the laser-cut part, are cut using less than full laser power so the laser doesn't cut all the way through. The picture shown here was "scribed" at 30% of full power.

You may also notice that one of the diagonal braces fell out. This happened because the braces were spaced too close to each other. I corrected this error in the drawings. I also discovered that the groves in the underside of the bench tops were in the wrong location, so I also fixed this.

Production Run

The final production process took a lot longer than I expected. First I spent a few hours with the laser people getting the settings just right, and figuring out the exact process that would be most efficient for getting drawings from my computer to their laser cutter. This, of course, wasn't cheap.

Once we had all this done, I created a Corel Draw file that was 12" by 24"--the size of the birch plywood I purchased for the production run. I placed multiple copies of the kits onto this page so the laser would cut out an entire run of kits at once. A few hours later I returned and picked up the finished kits.

Designing the packing and instructions also took quite a bit of time. I used Corel Draw to design and layout the packaging, and then exported the finished result to a PDF file and wrote it to a CD. With this CD in hand, I headed off to Kinkos, where they printed pages onto thick card stock. Since there were multiple packages per printed page, we had to cut out each package and then score and bend over the top. Finally, we put one kit into a bag and stapled everything together, resulting in a finished product. I ended up working three very long days during the production of our first batch of kits. If I were to do the calculate, this wouldn't even add up to minimum wage for all my time invested. The learning curve was somewhat steep, so future kits should be easier.


Copyright © 2003-2008 New Rail Models, LLC
Last modified: 01/17/08