How Do You Decide the Right Depth of Cut for a Finishing Pass Versus a Roughing Pass?

Roughing depth of cut is set by what the machine, tool, and workholding can survive at maximum material removal rate; finishing depth of cut is set by what leaves the smallest, most consistent deflection and the surface finish the drawing calls for. They're solving different problems, which is why using the same axial depth for both is one of the more common ways to waste cycle time or ruin a finish.

Roughing: push until something limits you

For roughing, the goal is removing volume fast without breaking the tool, stalling the spindle, or overloading the fixture. The practical limit is usually one of three things: horsepower and torque available at the spindle, the tool's ability to evacuate chips at that depth without packing or recutting, or rigidity, either the tool's or the part's. A stubby, well-supported 3/4" end mill in a rigid setup on aluminum can often take a full-diameter-equivalent depth of cut per pass. The same tool in a thin-walled part or a long-reach holder has to back off long before horsepower becomes the limit, because deflection and chatter show up first.

There's no universal number here worth quoting. It moves with tool diameter, flute count, coating, and rigidity of the whole system.

What's consistent is the approach: start conservative, watch chip color and sound, and increase depth until you hit a real limit rather than a guessed one. Full-depth roughing with a lighter radial engagement, trochoidal or dynamic milling strategies, often lets you go deeper axially than a full-width slot would, because the tool's flutes aren't buried in material on every side at once.

Finishing: work backward from the tolerance and finish callout

Finishing depth of cut isn't about removal rate at all. It's about leaving a consistent, accurate surface. The controlling factors flip: how much deflection can you tolerate on this specific feature, and how much heat and work-hardening does the cut generate at the surface. A finishing pass that's too deep reintroduces the same deflection problems roughing has, just at a scale that shows up as dimensional error instead of a broken tool. A finishing pass that's too shallow can actually hurt too: it rubs rather than shears, generates heat, and work-hardens the surface, particularly in stainless.

A reasonable finishing depth is whatever fully re-engages the cutting edge below any burnished or work-hardened layer the roughing pass left, without asking the tool to remove enough stock to deflect measurably. In practice that's often a light, single pass sized to clean up whatever stock allowance roughing left behind evenly, which only works if roughing left a consistent allowance in the first place. An uneven roughing stock allowance forces the finishing pass to vary its actual depth of cut across the part even though the program calls it out as constant, and that inconsistency is what produces visible witness marks and dimensional drift between features.

What actually drives the decision

Ask what's the failure mode you're protecting against. Roughing failure looks like a broken tool, a stalled spindle, or a scrapped fixture. Finishing failure looks like an out-of-tolerance feature or a surface that doesn't meet the print. Set roughing depth against the first failure mode and push it until something real stops you. Set finishing depth against the second, and be conservative. A finishing pass that costs you thirty extra seconds of cycle time is cheaper than a part that fails inspection because the last cut deflected the wall.