Cutting a curved piece of wood.

Hi there. Im new to the world of CNC.

It would be great if someone could help me.

Im looking at purchasing a CNC machine in order to manufacture wooden sunglasses with from laminated plywood (max 5mm thickness).

I am proficient with Adobe Illustrator and Google Sketchup, but don't have any experience with CAD software.

The pieces of wood needs to be bent into a curve before going onto the CNC router for cutting the frame. See attached picture as example of what it will look like when placed on t-slot table.

What I want to know is: Do I need to draw the sunglasses frame in CAD with the exact same curve (in 3D) as the small wood plank for it to be able to cut it correctly?

Hope someone could help me with this.

Thanks

Attachment 254374

Since you'll most likely be cutting them with a standard 3 axis routerm than no, the drawing would only need to be a 2D drawing. You'd only need it in 3D if you were using a 5 axis router and wanted the tool to remain perpendicular to the surface.

Thanks for the reply.

Will the the spindle automatically move up and down along the z-axis as it moves around the cutting line? Or do I just set the Z-height at the lowest point of the curve?

Nothing is really "automatic" about CNC work. The machines do what you tell them to do and at some point decisions are made by the programmer. You can do it either way depending on the CAM software. It's probably easier to just keep the Z height steady, but that depends also on what diameter bit you plan to use and if that bit is available with enough cutting length to do the job.

While it would be easier to cut the parts flat and bend them later, doing it the way Okker is talking about would require a toolpath that varied in Z, assuming he wants to mill a recess in the frame that would fit the lenses. And yes, in that case, the 3D curve in the CAD drawing would have to match the curve that the wood is bent to. I suppose it could all be done with 2D software if the individual pre-bent laminations were done separately and glued together later; that way the cut-outs in the upper ones could be slightly larger than the ones in the lower ones (or vice-versa) making a recess for the lenses with a ledge to seat them.

If the machine is a 3 axis machine and the tool is long enough, there is no difference between a varied Z toolpath and a fixed Z toolpath in terms of the finished part if the bit moves along identical paths. Whether the Z axis moves the cutting portion of the bit up or down, or the cutting portion is long enough to be equivalent to the stroke needed, either way you have a tool that is perfectly vertical and the result is the same provided that the toolpath is the same. In other words, you can project (verb not noun) the 3d edge geometry to a 2d drawing and it's the same toolpath along the X and Y, so again if the bit is long enough, the Z is irrelevant. The only way following the 3d surface makes sense is with a 5 axis machine where you can use the surface normals to keep the edge of the part square to the surfaces as Gerry suggested, but I doubt that he'll have a 5 axis machine or the software to program it with.

A varied Z toolpath has one advantage in that the fixture would be less likely to get compromised by being deeply cut around the part, but not having the actual part and not knowing what the extents of the 3d shape are, even that may not be any significant advantage.

That's true if you're just cutting out a part. But look at a pair of sunglasses, preferably some with a curved front. See those recesses that the lenses fit into? Those are curved, not flat. In order to make those from a curved piece of wood, he'd need to have the end of the tool follow that curve.

I can only comment on the parts he's shown us, which are just the earpieces. If there is a specific part with specific geometry, we can all be more specific about what the best way to do things is. If he's just cutting out the profile of ear pieces as he's shown in the photo, that can be done without any 3d geometry. If he needs to create a pocket for the lens to fit in, that may require a 3d toolpath or it may not. For all we know, these are just promotional sunglasses with flat lenses. Information about the lens area has not been supplied. From what I see in the photo, a 2d toolpath would be fine. With more information, my opinion may change. Just so the OP can see what I'm talking about, I've created a quick model. The long tool and the shorter tool both would follow the same path and produce the same part. If, as Andrew suggest, there is a pocket where a bottom must be present that follows the 3d contour, 3d geometry (wireframe or solids) may be required. If not, then 2d geometry is enough. Any through cut on a 3 axis router is the same whether you cut through be .001" or by 1" so long as the bit has the needed cutting edge length. If there is a bottom to the cut (not a through cut), then the Z axis is a factor.