LED Optical Metal Sintering

This is a nucleus for development of LED-based metal sintering heads.

Introduction

Optical metal sintering provides good resolution and thus is a very desired process to be RepRapped. However, the cost of lasers powerful enough to melt metals as of 2017 is prohibitive. There's a physical workaround, however, to solve this problem. The main property of laser that makes it laser is the coherency of light. This means that two _different_ laser sources can cause an optical interference to happen between their photon fluxes.

For industrial applications such as melting, sintering and welding this is not really needed. You can actually use sunlight or LEDs and achieve the same result if you manage to output enough light of needed wavelenght.

Wavelenght

Obviously when choosing LEDs you need to consult a specter of emission of iron or whatever you're going to sinter. Take notice that every line has some width, so you don't need to match it exactly with your LEDs. However, efficiency will fall greatly with the difference rising.

Power

Around 150W of light power is needed. Effeciency of contemporary LEDs is quite high (40% as of 2012), so one will need around 600 bright 0.5W SMD diodes. Because of inevitable focusing inefficiency it is more realistic to expect around 1000 SMD LEDs. Luckily they are small and cheap.

Heat management

LEDs loose thier efficiency at high temeperature, and thus the most probable design is a full plate of LEDs (or multiple of them) with a metal heatsink glued right behind. Not really a problem since this is a metal printer which can produce efficient complex geometry sinks for itself.

Focusing

Some simulations are needed to deside how to configure LEDs in space so that they can be focused well by a single lens (otherwise it will be expensive) Cheap freeform mirroring surfaces are possible since you can print a plastic part and perform a famous silver mirror reaction on its surface.

Possible Head and Printer Design

Printhead with 600-1000 diodes and heatsinks will be quite large and proably heavy, so movable XY table and head with Z axis is foreseen instead of standard XYZ robot. Because of this instead of metal sintering, metal powder deposition may be better.