Powder Bed

Planning Powder Usage Across Powder Bed Builds

In powder bed fusion, most of the powder you load never becomes parts on the first pass. Here is how to plan charges, track recovery, and forecast powder buys that hold.

Powder bed processes have a brutal material arithmetic: most of the powder you load never becomes parts on the first pass. SLS and MJF builds typically pack 8 to 15 percent of chamber volume as parts; the other 85 to 92 percent is cake you must recover, sieve, and requalify before it earns anything. At 60 to 90 dollars per kilogram for PA12 and 90 to 400 for metal powders, a shop that plans charges by eyeball ties up tens of thousands of dollars in powder inventory or, worse, discovers on Thursday that it cannot charge Friday's build. Powder usage planning is what turns that arithmetic from a cash trap into a schedule you can promise against.

Plan every charge with recovery in the denominator. Powder required equals build quantity times powder per part, divided by recovery efficiency. Work it: 80 parts at 0.15 kilograms each is 12 kilograms of fused material; at a demonstrated 65 percent recovery efficiency, meaning 65 percent of what you handle comes back usable after breakout, sieving, and losses, you need to plan 18.5 kilograms of powder movement to net the build. The Powder Usage calculator runs quantity, per-part mass, and recovery efficiency in one pass. The discipline is measuring recovery from your own scale data across 10 builds, because vendor claims and reality routinely differ by 10 to 20 points.

Benchmark the physics. Part packing density: 8 to 12 percent is typical SLS, 10 to 15 percent for well-nested MJF, and metal laser powder bed runs lower because of support and spacing rules. Recovery efficiency after sieving lands at 85 to 95 percent for metals with good handling, 60 to 80 percent for polymers once you subtract cake that is too heat-aged to reuse. Sieve losses alone run 2 to 5 percent per cycle. Powder per part must include a share of the cake burden, which is why per-part powder cost on a 6 percent packed build can double versus the same part at 12 percent packing.

The levers move real money. Nesting density first: pushing a polymer build from 8 to 12 percent packing spreads the fixed cake and machine hours across 50 percent more parts and is worth more than any discount your powder vendor will ever give. Build height second: powder consumption scales with chamber height, so capping the build at the tallest nested part rather than full Z saves 10 to 30 percent of the charge. Third, recovery process: closed-loop vacuum recovery and 30-minute sieve discipline lift polymer recovery 5 to 10 points over scoop-and-bucket handling. Fourth, standardized charge recipes per platform end the 2-kilogram safety fudge every operator quietly adds.

Failure modes are consistent. Planning against virgin-only assumptions ignores that 70 to 80 percent of a healthy polymer charge is recycled powder, so purchasing over-buys 3 to 4 times actual virgin need. Untracked recovery is the opposite failure: nobody weighs cake returns, recovery quietly slides from 75 to 55 percent, and builds start short. Mixing lots without genealogy makes a bad-powder investigation impossible when elongation drops. And ignoring humidity, since PA12 above 50 percent relative humidity flows badly and jams recoaters, converts good powder into failed builds that cost far more than the powder itself.

Run it on a cadence. Per build, weigh three things: charge in, parts out, powder recovered; three scale readings, two minutes. Weekly, compute recovery efficiency and packing density by machine and review any drift beyond 5 points. Monthly, roll the demonstrated numbers into a powder purchase forecast with a 3-week buffer and reconcile physical powder inventory against the ledger; gaps above 5 percent mean unlogged losses. Quarterly, re-run charge recipes against actuals and audit the recovery station, because a torn sieve mesh or a leaking vacuum line can silently cost 2 kilograms per build for months.

World-class powder planning looks like this: packing density above 11 percent on polymers, recovery efficiency within 3 points of target every week, zero build delays for powder shortage, inventory turns of 6 or better on powder stock, and a forecast that lands within 10 percent quarter after quarter. A two-machine MJF cell hitting those numbers typically carries 15,000 to 25,000 dollars less powder inventory than an unmanaged peer and never loses a scheduled build to material. The entire system is a scale at the recovery station, a three-field log, and one weekly review.

Published 2026-07-02.