DMLS vs CNC for a Metal Part: Routing Features, Batch Size and Post-Processing in the RFQ

Rule out any DMLS quote that promises tight tolerance or smooth finish as-built

A DMLS quote that offers tight tolerance and a smooth finish straight off the build plate is not comparable to a CNC quote, because the process does not deliver either as-built. Standard DMLS tolerance is around ±0.1 mm, and critical surfaces are then post-machined below ±0.05 mm³¹⁵.

CNC holds ±0.05 mm on 3-axis and ±0.025 mm on 5-axis¹⁵. Surface finish separates the routes by roughly an order of magnitude.

As-built DMLS lands at Ra 4–16 µm¹⁵, with one source putting metal AM at Ra 5–15 µm²⁹, against Ra 1.6–3.2 µm on 3-axis and Ra 0.8–1.6 µm on 5-axis CNC¹⁵.

DMLS roughness also depends on geometry and build orientation, so a single number on a drawing does not describe the whole part¹. That is why the tight features are reached by oversizing the printed area and CNC-machining the critical interfaces afterwards²⁰³.

When you compare quotes, price that post-machining step into the DMLS route from the start. A DMLS price without it is a price for a part that will not hold your tolerances.

Split the part: which features can only be DMLS, which are better machined

Treat the routing as a feature decision, not a whole-part one. Internal channels, undercuts and lattice structures are where DMLS is capable and CNC is not, or is limited by tool access¹⁵²⁹.

Part consolidation — building several components as one — is the other DMLS lever, because it removes assembly steps³⁰. The complement is on the same drawing.

Mating surfaces, bores and seating faces are what CNC finishes to functional tolerance, and a hybrid build-print-then-machine plan is described as standard practice in tooling, medical device prototyping and aerospace weight-reduction work¹⁵.

So mark up the drawing before you send it out: which surfaces carry a tolerance or a sealing function, and which volumes exist only because the geometry is internal. The first group is a machining requirement whatever route the body takes; the second is the reason to print at all.

Check whether batch size has already moved the answer to CNC

DMLS economics rest on complexity and part consolidation, while CNC economics improve with volume³⁰. One comparison puts DMLS cost advantage at "complex metal only" and CNC advantage from 20–50+ parts¹⁵.

Another reports CNC cost dropping significantly at 100 parts²⁹. Those are directional statements from supplier-tier sources, not a crossover curve.

What they tell you is that a simple part at a few dozen pieces or more is a CNC conversation, and a complex consolidated part at low volume is a DMLS one. The sources do not give a part count at which the two meet.

When a DMLS quote arrives, check that it separates per-build cost from per-part cost. Build time, stress relief and HIP are incurred per build plate, while support removal, CNC post-machining and surface treatment are incurred per part²⁶.

A single per-part figure that hides those lines cannot be compared with a CNC price.

What the DMLS RFQ must specify as line items

DMLS carries a multi-step post-processing chain, and CNC is described as needing minimal post-processing²⁶²⁹¹⁵. Each step you leave out of the RFQ shifts its cost and its risk to you after award, so ask for them as separate lines rather than one finished-part price.

• ✓Support removal, quoted per part
• ✓Stress relief, quoted per build
• ✓HIP and heat treatment, quoted per build, with the cycle named
• ✓CNC machining of the critical interfaces, quoted per part, with the surfaces listed
• ✓Surface treatment, quoted per part, with the target finish stated
• ✓Material certification for the alloy, tied to the HIP and heat-treatment record

Decide whether the part can be accepted as load-bearing

Do not accept an as-built DMLS part as equivalent to wrought stock. For specific alloys such as Ti-6Al-4V and AlSi10Mg, DMLS can approach wrought properties when a validated HIP cycle and post-HIP heat treatment are applied.

And as-built properties without documented post-processing must not be assumed equivalent¹². Sources disagree on how far that goes.

One describes DMLS/SLM parts as isotropic with strength comparable to wrought and porosity below 0.2%, while noting lower fatigue strength from surface roughness and internal porosity⁷. Another puts metal AM as anisotropic at typically 80–95% of wrought properties²⁹.

The gap is partly alloy and post-processing, which is why the RFQ has to name the alloy, the HIP and heat-treatment cycle, and the certification you expect back¹²¹⁵. If the part carries structural load, that specification is the condition of acceptance, not a later refinement.

Check the schedule before committing to DMLS

CNC lead time is generally shorter than DMLS for comparable parts, and DMLS is penalised by printing plus the post-processing chain³⁰. One turning service quotes 3–7 days for small batches with 24–48 h expedite⁵; a prototype comparison puts CNC at 5–14 business days and metal AM at 7–14 days²⁹.

If your programme needs parts inside a week, the DMLS route is unlikely to fit once support removal, stress relief and HIP are queued behind the build. Ask the DMLS supplier for the build, post-processing and finishing stages as separate dates, because no source in this set quantifies build time or queue time.

Where the sources disagree, and what to demand before award

Two disagreements matter at award: the DMLS tolerance figure and the material-property claim. Neither is settled by picking a side — the tolerance spread reflects different supplier standards, and the property spread reflects alloy and post-processing.

Resolve both by making the supplier state their basis in the quote.

Before you request a quote

• Ask the supplier for model-specific evidence and record the condition that would change your buying decision.
• Ask the supplier for model-specific evidence and record the condition that would change your buying decision.
• Ask the supplier for model-specific evidence and record the condition that would change your buying decision.
• Ask the supplier for model-specific evidence and record the condition that would change your buying decision.
• Ask the supplier for model-specific evidence and record the condition that would change your buying decision.
• Ask the supplier for model-specific evidence and record the condition that would change your buying decision.
• Ask the supplier for model-specific evidence and record the condition that would change your buying decision.