Metal additive manufacturing: melt pool drifts, pores found late

The melt pool varies and leaves pores in the part

What happens. Laser power, scan speed and hatch distance are fixed in the parameter set, but the melt pool does not stay the same. Over thin walls, overhangs and narrow sections heat builds up and the pool grows; over thick sections it cools quickly and shrinks. Powder layer thickness and gas flow add their own scatter. Many machines record the melt pool with a camera or photodiode, yet the signal is only looked at after the build.

What it costs. Pores from lack of fusion or from keyholing, warped parts, scrap found only in computed tomography or after machining, and whole build plates rejected because one zone went wrong.

Why the usual fix fails. Developing a parameter set per geometry by trial takes many builds, and it holds only for that part, that powder lot and that machine. Adding supports or slowing down everywhere costs build time without removing the cause.

How a fixed-price remote diagnosis checks it. From your build logs (laser power, scan vectors, layer times, melt pool signals, gas and chamber data) and the inspection results of the same parts, I check whether the melt pool signal predicts where pores appear. The fixed-price remote diagnosis tells you whether a model-based correction of laser power per layer or per zone is realistic, and what data and machine access it would need.

Every part must come out the same, and proving it costs more than printing it

What happens. Showing what can be printed is no longer enough; customers in aerospace, medical and energy want proof that every part from every build has the same properties. In the shop, the same job on two machines, or on the same machine a month later, still gives different results.

What it costs. Witness specimens in every build, computed tomography for many parts, long qualification for each new machine or material, and customers who hesitate to move series parts to printing.

Why the usual fix fails. Freezing every parameter and inspecting more treats the symptom. The causes, such as heat build-up, powder ageing and drifting optics, keep changing underneath the frozen settings.

How a fixed-price remote diagnosis checks it. I compare build logs and in-process monitoring data across machines and builds with your inspection results and look for the process signals that separate good parts from bad ones. The fixed-price remote diagnosis answers whether those signals are strong enough to release parts with less inspection, or to correct the process during the build, and what that would take.