Thread by @realscientists

A thread on Laser Powder Bed Fusion (L-PBF):

Dynamics of the process in one image (which I re-drew due to legal issues for the presentation😂). Key takeaway: this process has A TON of parameters/levers to pull on an already highly stochastic process.

#AdditiveManufacturing

https://twitter.com/realscientists/status/1202997934108921856

On powder: the goal is to get a flat, homogeneous, and tightly packed layer of spherical particles before we even start with the laser. The dynamics of powder spreading become important to understand.

(Img researchgate.net/publication/33…)

We need to control:

- Powder size distribution
- Particle morphology
- Chemistry
- Re-use of the powder

To ensure high quality metallurgy of outgoing parts.

Things get fun when the laser hits the powderbed.

In the meltpool itself, a vapor depression is formed in the top of the pool, and the fluid is recirculating through Marangoni flow. Particles are pulled into the meltpool due to surface tension.

A "vapor plume" comes off the meltpool (which must be managed by an "air knife" of laminar inert gas flow), and with it small powder particles are entrained in the plume.

Spatter particles are ejected out of the other side of the meltpool.

(img nature.com/articles/s4159…)

Researchers @argonne and @CarnegieMellon have done staggering research on the dynamics of L-PBF by using real-time synchotron imaging of the process + computational simulation.

(Also at @Livermore_Lab and @OfficialUoM)

nature.com/articles/s4159…
nature.com/articles/s4146…

@argonne @CarnegieMellon @Livermore_Lab @OfficialUoM That wobbly boi in the tweet above ⬆️ is the real-time formation of keyhole porosity in a metal part (read: not good).

What is keyhole porosity and how does it happen? Read on ⬇️.

(img iopscience.iop.org/article/10.108…)

@argonne @CarnegieMellon @Livermore_Lab @OfficialUoM PV (Power-Velocity) mapping is one way to understand this.

~ Too hot + too slow = keyhole formation
~ Too cold + too fast = lack of fusion
~ Too hot + too fast = balling (laser track balls up vs. straight line)
~ Too cold + too slow = turn the machine on 😛

(img @jul_coh)

@argonne @CarnegieMellon @Livermore_Lab @OfficialUoM @jul_coh It's critical for parameters to stay in the "conduction mode zone" where good metallurgy is formed.

So, how sensitive is your process? You'd better know if your laser power fluctuates by ±3%, speed by ±5%, that you're still inside the triangle.

(img science.sciencemag.org/content/363/64…)