Build Time Estimate

Layers multiplied by what each layer costs: the beam's time scanning it, plus the recoater's time spreading the next one. $t_b = n(t_s + t_r)$. The most useful thing this equation does is not produce a number but expose a ratio.

Recoat time is fixed per layer. The plate drops, the dispenser doses powder, the blade or roller wipes it across, and that takes the same seven to twelve seconds whether the layer holds one small bracket or a full plate of forty parts. Scan time scales with the area being built; recoat time does not scale with anything. Two consequences follow and they run the whole economics of a powder-bed machine.

The first is that filling the plate is nearly free. If a lone part takes ten seconds of scanning and nine of recoating, adding thirty-nine more parts alongside it multiplies the scan time by forty and leaves the recoat time exactly where it was. The cost per part collapses. This is why service bureaux nest aggressively, and why a one-off part on a powder-bed machine is priced the way it is.

The second is that building tall and thin is the most expensive orientation there is. A slender part stood upright has many layers, each with a tiny cross-section, so the build is almost entirely recoating — the laser is idle most of the time and the machine is being paid for anyway. Laying the same part down cuts the layer count by whatever the aspect ratio is, and the arithmetic often beats the extra supports and the worse finish on the downskin faces. Often, not always: residual stress, support access and which surfaces will be machined afterwards all pull the other way, and orientation is a genuine multi-criterion decision. But the layer count is the term with the most leverage and it is the one people forget to price.

What this equation excludes is a whole shift's worth of work. Chamber purge and inert-gas fill; plate preheat and its soak; the cool-down before the door can safely be opened; powder recovery, sieving and blending; cutting the parts off the plate; stress relief; support removal and finishing. On a short build these easily exceed the build time itself, and a quotation built on machine hours alone is short by an amount that grows as the job gets smaller. It is exactly the trap of costing a machined part on cutting time — and the fix is exactly the same one: measure a real job end to end once, work out your own ratio of total elapsed time to this calculation, and apply it thereafter.