CNC Making 2D and 3D Objects: How Toolpaths Become Parts

• 1
  Key takeaways
• 2
  What CNC making 2D and 3D objects actually means
• 3
  Cutter geometry sets the real limit on 3D surfaces
• 4
  Why 3-axis and 5-axis produce different 3D objects
• 5
  Workholding is where most 2D and 3D plans break
• 6
  Where tolerance and surface finish diverge
• 7
  From CAD model to finished 2D and 3D part
• 8
  Choosing the right approach for 2D and 3D work

Subtractive machining starts from a solid block and removes material until the remaining shape matches the CAD model. The machine never sees the model. It sees coordinates. A CAM system converts the model into toolpath moves, then a post-processor turns those moves into G-code the controller can run.

The word 2D describes a toolpath that stays at a single Z height. The cutter plunges to depth, follows a closed contour, lifts, and moves on. Think of a mounting plate with bolt holes and an outer outline. Every feature can be described by an XY position plus a depth value.

3D means the tool tip changes height continuously while it moves in XY. That is what produces a curved surface: a turbine blade fillet, an aerodynamic fairing, a lens housing. The controller interpolates three axes at once, so the cut is a swept surface rather than a flat floor.

The gap between them is not a matter of quality. It is a matter of geometry. A 2D profile is a line extruded downward. A 3D surface is a set of points with no constant height. Different tools, different stepovers, different fixtures.

• 1
  2DConstant Z, closed contours, through-holes, slots.
• 2
  2.5DMultiple Z levels with vertical walls between them.
• 3
  3DContinuously varying Z, curved or freeform surfaces.

A flat end mill leaves a flat floor. That is why it dominates 2D work. Push it across a curved surface and it gouges the part, because the corner of the flute digs in. Curved surfaces are cut with a ball nose tool, where the tip radius equals the tool radius and contact is a single point.

The trade-off is speed. A ball nose tool with a Ø6 mm tip cutting a shallow slope leaves scallops between passes. The height of those scallops depends on the stepover and the surface curvature. Tighten the stepover and the finish improves, but cycle time rises almost linearly. This is the single biggest cost driver in 3D machining.

For a typical aluminum part we run stepovers of 0.1–0.3 mm on finishing passes to hold Ra 0.8–1.6 μm. Where a finer Ra 0.2–0.8 μm is called for, the part usually goes to a separate finishing operation or a polishing step rather than a slower toolpath, because the toolpath cost outruns the benefit.

A bull nose cutter sits between the two. It has a small corner radius, so it can rough a curved region and still leave a reasonable floor. Many shops use it for the semi-finish pass and switch to ball nose only for the final pass. That keeps the tool count low without sacrificing the surface.

On a 3-axis machine the part is fixed and the tool moves in X, Y and Z. Every surface must be reachable from above, or the part has to be re-fixtured. Each re-fixture adds setup time and, more importantly, adds stack-up error. Two setups can easily cost more tolerance than the machining itself.

A 5-axis machine tilts the tool or the table, so the cutter can approach a face from an angle. That removes most re-fixturing. It also lets the tool shank clear the part on deep cavities, which means a shorter, stiffer cutter can be used. Short tools chatter less. Less chatter means a better surface and longer tool life.

The catch is programming. Simultaneous 5-axis toolpaths need collision checking between holder, tool and part. A path that looks clean in the CAM preview can still crash if the holder clearance is 2 mm and the stock was modeled 3 mm oversize. We verify with the actual stock model, not the nominal one.

For parts with holes on several faces, or pockets deeper than three times the cutter diameter, 5-axis usually wins on total cost even though the hourly rate is higher. For flat plates with through-features, 3-axis is faster and cheaper. That decision belongs in the DFM review, not after the first article.

A thin plate is the classic problem. Clamp it on the edges and the middle deflects under cutting force, so the floor of a pocket comes out dished. Clamp it hard and it springs back after unclamping, and the flatness is gone. Vacuum chucks and low-melt fixturing solve most of this, but they need to be planned before the stock is cut.

For a 3D contour on a long part, we often leave a sacrificial tab and cut it off in a second operation. The tab keeps the part rigid through the finishing pass. Removing it adds one setup, but it prevents the chatter marks that would otherwise need hand polishing.

The 4,000 mm maximum processing size on our large travel machines, 4,000 × 400 × 150 mm, is a machine limit, not a fixturing guarantee. A part that long still needs support along its length or it will sag. We look at the length-to-thickness ratio during quoting and tell you if the part needs a different approach.

Rotary tables change the math. A Ø400 mm table lets a part be machined on four sides in one setup, which is often the cheapest way to get true position between faces. It is not the same as simultaneous 5-axis. Indexed 4-axis work is faster to program and usually enough for prismatic parts.

Dimensional tolerance and surface finish are separate specifications and they behave differently. A 2D profile can hold ±0.005 mm on a well-supported part with a rigid setup. The same profile on a thin wall will move, and no amount of machine accuracy fixes that. The material moves, not the machine.

Surface finish on a 3D surface is a function of stepover, tool condition and spindle speed. A worn tool leaves a darker, rougher patch that is easy to see after anodizing, because the oxide grows differently on a torn surface. If the part will be anodized, it is worth specifying the finish requirement before machining, not after.

Sharp internal corners are the most common drawing error we see. A pocket with a 0.5 mm internal corner radius cannot be cut with a Ø10 mm end mill. The tool radius is the floor. Either the corner radius goes up to match an available cutter, or the corner is drilled out first and the rest is milled.

Depth-to-diameter ratio has a similar hard limit. A pocket 60 mm deep with a 6 mm cutter is a 10:1 ratio. That needs a reduced-neck tool, which deflects. We would rather open the pocket, drill a pilot, or split the depth across two setups than fight the ratio.