Hi guys, welcome back to another prototyping session with Ravi. In today’s blog, we’re making a topographical map of Mt. Everest. This was a proficient-level assignment – it was definitely challenging but super rewarding in the end.
The process wasn’t too hard but debugging the process was an amazing learning experience. First, I started with a 6”x3.25”x1.38” block of wood. After measuring its dimensions, I was ready to create my toolpaths. Here, I used VCarve Pro instead of Carbide Create as the instructions indicated. After importing the everest.stl file and scaling it to my block of wood, I started creating the toolpaths. I used three tools – (1) 1/4” end mill, (2) 1/8” ball mill, and (3) 1/16” ball mill. I used these three tools to get finer finishes each time. At first, I created all the .gcodes separately before realizing that I could have it all in one file and just have Carbide Motion (the program Shapeoko Pro uses) prompt tool changes. Here are some toolpath rendering images:
After creating the .gcode, I first attempted to make Everest using the Nomad Desktop CNC which didn’t work at all. The 1/4” end mill was just too aggressive or wasn’t spinning fast enough to cut the wood. I set up the job using just the 1/8” end mill for roughing and realized the estimated time was ~3.5 hrs.
After Dr. W recommended that I set it up on the Shapeoko Pro, I realized in hindsight that this was a far more robust and superior approach. The setup was super similar (if not easier) and I would definitely recommend it over the Desktop Nomad anytime. Anyway, here are some reflections from my v1 process. I think the dimensions (all axes) are super important because I noticed the mills were cutting into the clamps that were holding the wood down. To speed up the process, especially at the finishing stages, I definitely increased the feed rate A LOT. It seemed like the machine could handle it so I did it. I didn’t do this for ¼” end mill roughing because there’s a chance it could jam up because it already happened once. Here’s what the v1 product looked like – I think it came out pretty good with minimal post-processing (like Dremel sanding). After I cut the next block since I need two parts, I will also stain them.
In my v2 approach, here are the changes I’m making – (1) accurate XYZ dimensions (w/ caliper), (2) ensuring bits are tightened & have sufficient length to prevent collet interference, (3) smaller stepover rate (8% of bit diameter). The last change is crucial – I had a misunderstanding where I thought adding smaller bit finishing toolpaths would help increase the resolution but it was recommended that I use a smaller stepover rate (like a lawnmower). After looking at VCarve forums, people like to use 3-8% of bit diameter – I will be using 8% so I don’t need to stay in the OEDK for ~4-5 hours straight. The default is ~45-50% so using an 8% stepover rate is a huge change. Also, an artificial way to decrease time is to increase feed rate but I think a higher spindle RPM is needed. But if it’s too high, the bit can overheat and can potentially lower end product resolution.
Another thing I did was probe with the measuring tool instead of the ¼” drill bit. This might’ve helped me gain some accuracy, especially with determining path boundaries since I’m using the full face of my wood stock. This is less significant if you intend to vector/contour a .svg file since only part of the wood is used. Since I used the full dimensions of my wood block, precision was key here. With the changes above, here’s the final result (v2 is on the right):
It looks way better than v1 even though I didn’t use a 1/16” ball mill at the end. Changing the stepover rate was crucial to achieving a higher resolution – another thing I’d try is reducing the plunge rate. I think it’s like the same parameter except in the z-direction.
Here’s both v1 (left) and v2 (right) after some post-processing (added some boiled linseed oil – the only finish I could find) – overall, this was a pretty fun and enlightening process!
The oil highlights the difference between both pieces (v1, v2) – here are key toolpath settings in VCarvePro you should try if you’d like to try and replicate this in a reasonable time.
|
V1 (worse ~ 2 hrs) |
** V2 (better ~ 2 hrs) |
|
1. 1/4” end mill a. 40-50% stepover 2. 1/8” ball mill a. 40-50% stepover 3. 1/16” ball mill a. 40-50% stepover |
1. 1/4” end mill a. 40-50% stepover 2. 1/8” ball mill a. ** 8% stepover |
- don’t be afraid to increase the feed rate (double it even) for FINISHING ONLY – will save you lots of time (change in carbide motion)
- use default settings from carbide tools import file – provided on Canvas or refer to Shapeoko Pro feed and speed charts for pine wood.
COST BREAKDOWN
| Cost | Price | Source | Quantity | Total |
| 1.38” thick wood stock | ~$3/plank | lowes | ~1 plank | $3 |
| overhead – machine wear | ~ $50 | estimate | ~ | $50 |
| operator cost (me) | $20 | me, the operator | 10 hrs. | $200 |
| TOTAL: | ~$253 |
