Trochoidal and dynamic toolpaths in hardened steel: constant engagement wins
A hard milling operation rarely fails in the middle of a straight cut. It fails in a corner.
The reason is engagement. A cutter taking a 10% radial step along a straight wall is loaded lightly and predictably. The moment that path turns an inside corner, the arc of contact jumps — on a 90° internal corner with a conventional offset path, engagement can more than double for the duration of the corner.
In mild steel the tool absorbs that. In 60 HRC D2, a CBN edge does not.
The old way and why it breaks
Traditional offset roughing follows the part profile at a constant radial depth. It is simple and it produces a predictable stock condition, but the engagement angle varies continuously with the geometry. Every concave feature is a load spike.
The usual workarounds — reducing the overall depth of cut, slowing the feed, adding corner deceleration — all work by making the whole path conservative enough to survive the worst moment in it. The tool spends 90% of its time under-loaded so that it survives the other 10%.
What trochoidal and dynamic paths do instead
Trochoidal milling replaces the straight sweep with a series of overlapping circular arcs advancing along the path. The cutter is only lightly engaged at any moment, and the engagement angle stays roughly constant regardless of the geometry it is moving through.
Dynamic or adaptive milling — the term varies by CAM vendor — is the generalisation. The software computes the actual material engagement continuously and modulates the path (and often the feed) to hold it at a target value. In open areas it looks like wide sweeping arcs; in tight corners it tightens into something close to trochoidal.
Both achieve the same thing: constant engagement. Constant engagement means constant cutting force, constant heat, constant deflection and predictable tool life.
The parameter shift that comes with it
Adopting these paths without changing the parameters wastes them. The whole point is that constant light engagement lets you change the other variables:
| Conventional offset | Dynamic / trochoidal | |
|---|---|---|
| Radial engagement (ae) | 30-50% of diameter | 5-15% of diameter |
| Axial depth (ap) | 0.5-1× diameter | 1.5-3× diameter |
| Feed per tooth | Conservative | 1.5-3× higher |
| Surface speed | Conservative | Higher |
| Flute engagement length | Short, one zone | Long, spread along the flute |
The critical one is the last. Dropping radial engagement and raising axial depth spreads wear along the whole flute length instead of concentrating it in a narrow band near the tip. In hardened steel, where the wear rate is high, that alone can multiply tool life several times over.
It also lets you use the full length of an end mill rather than the bottom 5 mm, which is a real saving on expensive solid carbide tooling.
Where it matters most in hard work
Deep pockets in hardened blocks. The classic dynamic milling case. Full axial depth, tiny radial step, high feed.
Slotting. A conventional slot means 100% radial engagement — the worst possible condition in hardened steel. Trochoidal slotting turns it into a series of light arcs and makes the operation viable at all.
Hardened material with limited machine rigidity. Lower cutting forces mean less deflection. On an older machine this can be the difference between a usable surface and chatter — though it does not substitute for genuine rigidity.
Thin walls and ribs. Lower radial force means less push-off, which means the wall comes out straight.
Where it does not help
Finishing. These are roughing and semi-finishing strategies. A finishing pass in hardened steel is a light, continuous, constant-depth pass following the final surface — usually with a ball nose or high-feed cutter, at small stepover, chosen for surface quality rather than removal rate. Trochoidal motion on a finishing pass leaves the cusp pattern you were trying to avoid.
Also: very small cutters. Below about 3 mm diameter, the arc geometry of a trochoidal path at sensible engagement produces feed rates the machine cannot actually follow, and the controller smooths the corners back into something that is no longer constant engagement.
The machine has to keep up
Dynamic paths generate a lot of short moves. A controller with a small look-ahead buffer will decelerate at every block transition, and the programmed feed never happens. Symptoms are a cycle time far longer than the CAM estimate and a surface that shows the deceleration marks.
Modern controls handle it. Older ones may need the tolerance loosened in the CAM post so fewer, longer moves are generated — at some cost in path accuracy.
What this means for your drawing
Nothing directly. Toolpath strategy is the shop’s business.
Indirectly, it is one of the clearest signals of whether a shop actually hard mills or merely owns machines that could. A shop that talks about constant engagement, radial stepdown and flute wear distribution has done this work. One that offers to “run it slower” has not.
That is a fair question to ask before placing a die block order. Send the drawing and it goes to shops where the answer is already yes.